From 8b8f3e666fc6c2f33acb7ec6c0c66946307baddb Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Wed, 19 Aug 2020 14:06:36 +0200 Subject: net: Use generic ns_common::count Switch over network namespaces to use the newly introduced common lifetime counter. Network namespaces have an additional counter named "passive". This counter does not guarantee that the network namespace is not already de-initialized and so isn't concerned with the actual lifetime of the network namespace; only the "count" counter is. So the latter is moved into struct ns_common. Currently every namespace type has its own lifetime counter which is stored in the specific namespace struct. The lifetime counters are used identically for all namespaces types. Namespaces may of course have additional unrelated counters and these are not altered. This introduces a common lifetime counter into struct ns_common. The ns_common struct encompasses information that all namespaces share. That should include the lifetime counter since its common for all of them. It also allows us to unify the type of the counters across all namespaces. Most of them use refcount_t but one uses atomic_t and at least one uses kref. Especially the last one doesn't make much sense since it's just a wrapper around refcount_t since 2016 and actually complicates cleanup operations by having to use container_of() to cast the correct namespace struct out of struct ns_common. Having the lifetime counter for the namespaces in one place reduces maintenance cost. Not just because after switching all namespaces over we will have removed more code than we added but also because the logic is more easily understandable and we indicate to the user that the basic lifetime requirements for all namespaces are currently identical. Signed-off-by: Kirill Tkhai Reviewed-by: Kees Cook Acked-by: Christian Brauner [christian.brauner@ubuntu.com: rewrite commit] Link: https://lore.kernel.org/r/159644977635.604812.1319877322927063560.stgit@localhost.localdomain Signed-off-by: Christian Brauner --- include/net/net_namespace.h | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 2ee5901bec7a..cb4b33d7834b 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -60,9 +60,6 @@ struct net { refcount_t passive; /* To decide when the network * namespace should be freed. */ - refcount_t count; /* To decided when the network - * namespace should be shut down. - */ spinlock_t rules_mod_lock; unsigned int dev_unreg_count; @@ -245,7 +242,7 @@ void __put_net(struct net *net); static inline struct net *get_net(struct net *net) { - refcount_inc(&net->count); + refcount_inc(&net->ns.count); return net; } @@ -256,14 +253,14 @@ static inline struct net *maybe_get_net(struct net *net) * exists. If the reference count is zero this * function fails and returns NULL. */ - if (!refcount_inc_not_zero(&net->count)) + if (!refcount_inc_not_zero(&net->ns.count)) net = NULL; return net; } static inline void put_net(struct net *net) { - if (refcount_dec_and_test(&net->count)) + if (refcount_dec_and_test(&net->ns.count)) __put_net(net); } @@ -275,7 +272,7 @@ int net_eq(const struct net *net1, const struct net *net2) static inline int check_net(const struct net *net) { - return refcount_read(&net->count) != 0; + return refcount_read(&net->ns.count) != 0; } void net_drop_ns(void *); -- cgit v1.2.3 From e5e1a4bc916d29958c3b587354293738fcb984d7 Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Tue, 27 Oct 2020 13:32:01 +0100 Subject: xsk: Fix possible memory leak at socket close MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix a possible memory leak at xsk socket close that is caused by the refcounting of the umem object being wrong. The reference count of the umem was decremented only after the pool had been freed. Note that if the buffer pool is destroyed, it is important that the umem is destroyed after the pool, otherwise the umem would disappear while the driver is still running. And as the buffer pool needs to be destroyed in a work queue, the umem is also (if its refcount reaches zero) destroyed after the buffer pool in that same work queue. What was missing is that the refcount also needs to be decremented when the pool is not freed and when the pool has not even been created. The first case happens when the refcount of the pool is higher than 1, i.e. it is still being used by some other socket using the same device and queue id. In this case, it is safe to decrement the refcount of the umem outside of the work queue as the umem will never be freed because the refcount of the umem is always greater than or equal to the refcount of the buffer pool. The second case is if the buffer pool has not been created yet, i.e. the socket was closed before it was bound but after the umem was created. In this case, it is safe to destroy the umem outside of the work queue, since there is no pool that can use it by definition. Fixes: 1c1efc2af158 ("xsk: Create and free buffer pool independently from umem") Reported-by: syzbot+eb71df123dc2be2c1456@syzkaller.appspotmail.com Signed-off-by: Magnus Karlsson Signed-off-by: Daniel Borkmann Acked-by: Björn Töpel Link: https://lore.kernel.org/bpf/1603801921-2712-1-git-send-email-magnus.karlsson@gmail.com --- include/net/xsk_buff_pool.h | 2 +- net/xdp/xsk.c | 3 ++- net/xdp/xsk_buff_pool.c | 7 +++++-- 3 files changed, 8 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/xsk_buff_pool.h b/include/net/xsk_buff_pool.h index 0140d086dc84..01755b838c74 100644 --- a/include/net/xsk_buff_pool.h +++ b/include/net/xsk_buff_pool.h @@ -86,7 +86,7 @@ int xp_assign_dev_shared(struct xsk_buff_pool *pool, struct xdp_umem *umem, void xp_destroy(struct xsk_buff_pool *pool); void xp_release(struct xdp_buff_xsk *xskb); void xp_get_pool(struct xsk_buff_pool *pool); -void xp_put_pool(struct xsk_buff_pool *pool); +bool xp_put_pool(struct xsk_buff_pool *pool); void xp_clear_dev(struct xsk_buff_pool *pool); void xp_add_xsk(struct xsk_buff_pool *pool, struct xdp_sock *xs); void xp_del_xsk(struct xsk_buff_pool *pool, struct xdp_sock *xs); diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index b71a32eeae65..cfbec3989a76 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -1146,7 +1146,8 @@ static void xsk_destruct(struct sock *sk) if (!sock_flag(sk, SOCK_DEAD)) return; - xp_put_pool(xs->pool); + if (!xp_put_pool(xs->pool)) + xdp_put_umem(xs->umem); sk_refcnt_debug_dec(sk); } diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 64c9e55d4d4e..8a3bf4e1318e 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -251,15 +251,18 @@ void xp_get_pool(struct xsk_buff_pool *pool) refcount_inc(&pool->users); } -void xp_put_pool(struct xsk_buff_pool *pool) +bool xp_put_pool(struct xsk_buff_pool *pool) { if (!pool) - return; + return false; if (refcount_dec_and_test(&pool->users)) { INIT_WORK(&pool->work, xp_release_deferred); schedule_work(&pool->work); + return true; } + + return false; } static struct xsk_dma_map *xp_find_dma_map(struct xsk_buff_pool *pool) -- cgit v1.2.3 From f54ec58fee837ec847cb8b50593e81bfaa46107f Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 27 Oct 2020 21:12:12 +0100 Subject: wimax: move out to staging There are no known users of this driver as of October 2020, and it will be removed unless someone turns out to still need it in future releases. According to https://en.wikipedia.org/wiki/List_of_WiMAX_networks, there have been many public wimax networks, but it appears that many of these have migrated to LTE or discontinued their service altogether. As most PCs and phones lack WiMAX hardware support, the remaining networks tend to use standalone routers. These almost certainly run Linux, but not a modern kernel or the mainline wimax driver stack. NetworkManager appears to have dropped userspace support in 2015 https://bugzilla.gnome.org/show_bug.cgi?id=747846, the www.linuxwimax.org site had already shut down earlier. WiMax is apparently still being deployed on airport campus networks ("AeroMACS"), but in a frequency band that was not supported by the old Intel 2400m (used in Sandy Bridge laptops and earlier), which is the only driver using the kernel's wimax stack. Move all files into drivers/staging/wimax, including the uapi header files and documentation, to make it easier to remove it when it gets to that. Only minimal changes are made to the source files, in order to make it possible to port patches across the move. Also remove the MAINTAINERS entry that refers to a broken mailing list and website. Acked-by: Jakub Kicinski Acked-by: Greg Kroah-Hartman Acked-By: Inaky Perez-Gonzalez Acked-by: Johannes Berg Suggested-by: Inaky Perez-Gonzalez Signed-off-by: Arnd Bergmann --- Documentation/admin-guide/index.rst | 1 - Documentation/admin-guide/wimax/i2400m.rst | 283 ---- Documentation/admin-guide/wimax/index.rst | 19 - Documentation/admin-guide/wimax/wimax.rst | 89 -- Documentation/networking/kapi.rst | 21 - .../translations/zh_CN/admin-guide/index.rst | 1 - MAINTAINERS | 22 - drivers/net/Kconfig | 2 - drivers/net/Makefile | 1 - drivers/net/wimax/Kconfig | 18 - drivers/net/wimax/Makefile | 2 - drivers/net/wimax/i2400m/Kconfig | 37 - drivers/net/wimax/i2400m/Makefile | 23 - drivers/net/wimax/i2400m/control.c | 1434 ----------------- drivers/net/wimax/i2400m/debug-levels.h | 32 - drivers/net/wimax/i2400m/debugfs.c | 253 --- drivers/net/wimax/i2400m/driver.c | 1002 ------------ drivers/net/wimax/i2400m/fw.c | 1653 -------------------- drivers/net/wimax/i2400m/i2400m-usb.h | 275 ---- drivers/net/wimax/i2400m/i2400m.h | 970 ------------ drivers/net/wimax/i2400m/netdev.c | 603 ------- drivers/net/wimax/i2400m/op-rfkill.c | 196 --- drivers/net/wimax/i2400m/rx.c | 1395 ----------------- drivers/net/wimax/i2400m/sysfs.c | 65 - drivers/net/wimax/i2400m/tx.c | 1011 ------------ drivers/net/wimax/i2400m/usb-debug-levels.h | 28 - drivers/net/wimax/i2400m/usb-fw.c | 365 ----- drivers/net/wimax/i2400m/usb-notif.c | 258 --- drivers/net/wimax/i2400m/usb-rx.c | 462 ------ drivers/net/wimax/i2400m/usb-tx.c | 273 ---- drivers/net/wimax/i2400m/usb.c | 764 --------- drivers/staging/Kconfig | 2 + drivers/staging/Makefile | 1 + drivers/staging/wimax/Documentation/i2400m.rst | 283 ++++ drivers/staging/wimax/Documentation/index.rst | 19 + drivers/staging/wimax/Documentation/wimax.rst | 89 ++ drivers/staging/wimax/Kconfig | 46 + drivers/staging/wimax/Makefile | 15 + drivers/staging/wimax/TODO | 18 + drivers/staging/wimax/debug-levels.h | 29 + drivers/staging/wimax/debugfs.c | 38 + drivers/staging/wimax/i2400m/Kconfig | 37 + drivers/staging/wimax/i2400m/Makefile | 23 + drivers/staging/wimax/i2400m/control.c | 1434 +++++++++++++++++ drivers/staging/wimax/i2400m/debug-levels.h | 32 + drivers/staging/wimax/i2400m/debugfs.c | 253 +++ drivers/staging/wimax/i2400m/driver.c | 1002 ++++++++++++ drivers/staging/wimax/i2400m/fw.c | 1653 ++++++++++++++++++++ drivers/staging/wimax/i2400m/i2400m-usb.h | 275 ++++ drivers/staging/wimax/i2400m/i2400m.h | 970 ++++++++++++ drivers/staging/wimax/i2400m/linux-wimax-i2400m.h | 572 +++++++ drivers/staging/wimax/i2400m/netdev.c | 603 +++++++ drivers/staging/wimax/i2400m/op-rfkill.c | 196 +++ drivers/staging/wimax/i2400m/rx.c | 1395 +++++++++++++++++ drivers/staging/wimax/i2400m/sysfs.c | 65 + drivers/staging/wimax/i2400m/tx.c | 1011 ++++++++++++ drivers/staging/wimax/i2400m/usb-debug-levels.h | 28 + drivers/staging/wimax/i2400m/usb-fw.c | 365 +++++ drivers/staging/wimax/i2400m/usb-notif.c | 258 +++ drivers/staging/wimax/i2400m/usb-rx.c | 462 ++++++ drivers/staging/wimax/i2400m/usb-tx.c | 273 ++++ drivers/staging/wimax/i2400m/usb.c | 764 +++++++++ drivers/staging/wimax/id-table.c | 130 ++ drivers/staging/wimax/linux-wimax-debug.h | 491 ++++++ drivers/staging/wimax/linux-wimax.h | 239 +++ drivers/staging/wimax/net-wimax.h | 503 ++++++ drivers/staging/wimax/op-msg.c | 391 +++++ drivers/staging/wimax/op-reset.c | 108 ++ drivers/staging/wimax/op-rfkill.c | 431 +++++ drivers/staging/wimax/op-state-get.c | 52 + drivers/staging/wimax/stack.c | 616 ++++++++ drivers/staging/wimax/wimax-internal.h | 85 + include/linux/wimax/debug.h | 491 ------ include/net/wimax.h | 503 ------ include/uapi/linux/wimax.h | 239 --- include/uapi/linux/wimax/i2400m.h | 572 ------- net/Kconfig | 2 - net/Makefile | 1 - net/wimax/Kconfig | 40 - net/wimax/Makefile | 13 - net/wimax/debug-levels.h | 29 - net/wimax/debugfs.c | 38 - net/wimax/id-table.c | 130 -- net/wimax/op-msg.c | 391 ----- net/wimax/op-reset.c | 108 -- net/wimax/op-rfkill.c | 431 ----- net/wimax/op-state-get.c | 52 - net/wimax/stack.c | 616 -------- net/wimax/wimax-internal.h | 85 - 89 files changed, 15257 insertions(+), 15299 deletions(-) delete mode 100644 Documentation/admin-guide/wimax/i2400m.rst delete mode 100644 Documentation/admin-guide/wimax/index.rst delete mode 100644 Documentation/admin-guide/wimax/wimax.rst delete mode 100644 drivers/net/wimax/Kconfig delete mode 100644 drivers/net/wimax/Makefile delete mode 100644 drivers/net/wimax/i2400m/Kconfig delete mode 100644 drivers/net/wimax/i2400m/Makefile delete mode 100644 drivers/net/wimax/i2400m/control.c delete mode 100644 drivers/net/wimax/i2400m/debug-levels.h delete mode 100644 drivers/net/wimax/i2400m/debugfs.c delete mode 100644 drivers/net/wimax/i2400m/driver.c delete mode 100644 drivers/net/wimax/i2400m/fw.c delete mode 100644 drivers/net/wimax/i2400m/i2400m-usb.h delete mode 100644 drivers/net/wimax/i2400m/i2400m.h delete mode 100644 drivers/net/wimax/i2400m/netdev.c delete mode 100644 drivers/net/wimax/i2400m/op-rfkill.c delete mode 100644 drivers/net/wimax/i2400m/rx.c delete mode 100644 drivers/net/wimax/i2400m/sysfs.c delete mode 100644 drivers/net/wimax/i2400m/tx.c delete mode 100644 drivers/net/wimax/i2400m/usb-debug-levels.h delete mode 100644 drivers/net/wimax/i2400m/usb-fw.c delete mode 100644 drivers/net/wimax/i2400m/usb-notif.c delete mode 100644 drivers/net/wimax/i2400m/usb-rx.c delete mode 100644 drivers/net/wimax/i2400m/usb-tx.c delete mode 100644 drivers/net/wimax/i2400m/usb.c create mode 100644 drivers/staging/wimax/Documentation/i2400m.rst create mode 100644 drivers/staging/wimax/Documentation/index.rst create mode 100644 drivers/staging/wimax/Documentation/wimax.rst create mode 100644 drivers/staging/wimax/Kconfig create mode 100644 drivers/staging/wimax/Makefile create mode 100644 drivers/staging/wimax/TODO create mode 100644 drivers/staging/wimax/debug-levels.h create mode 100644 drivers/staging/wimax/debugfs.c create mode 100644 drivers/staging/wimax/i2400m/Kconfig create mode 100644 drivers/staging/wimax/i2400m/Makefile create mode 100644 drivers/staging/wimax/i2400m/control.c create mode 100644 drivers/staging/wimax/i2400m/debug-levels.h create mode 100644 drivers/staging/wimax/i2400m/debugfs.c create mode 100644 drivers/staging/wimax/i2400m/driver.c create mode 100644 drivers/staging/wimax/i2400m/fw.c create mode 100644 drivers/staging/wimax/i2400m/i2400m-usb.h create mode 100644 drivers/staging/wimax/i2400m/i2400m.h create mode 100644 drivers/staging/wimax/i2400m/linux-wimax-i2400m.h create mode 100644 drivers/staging/wimax/i2400m/netdev.c create mode 100644 drivers/staging/wimax/i2400m/op-rfkill.c create mode 100644 drivers/staging/wimax/i2400m/rx.c create mode 100644 drivers/staging/wimax/i2400m/sysfs.c create mode 100644 drivers/staging/wimax/i2400m/tx.c create mode 100644 drivers/staging/wimax/i2400m/usb-debug-levels.h create mode 100644 drivers/staging/wimax/i2400m/usb-fw.c create mode 100644 drivers/staging/wimax/i2400m/usb-notif.c create mode 100644 drivers/staging/wimax/i2400m/usb-rx.c create mode 100644 drivers/staging/wimax/i2400m/usb-tx.c create mode 100644 drivers/staging/wimax/i2400m/usb.c create mode 100644 drivers/staging/wimax/id-table.c create mode 100644 drivers/staging/wimax/linux-wimax-debug.h create mode 100644 drivers/staging/wimax/linux-wimax.h create mode 100644 drivers/staging/wimax/net-wimax.h create mode 100644 drivers/staging/wimax/op-msg.c create mode 100644 drivers/staging/wimax/op-reset.c create mode 100644 drivers/staging/wimax/op-rfkill.c create mode 100644 drivers/staging/wimax/op-state-get.c create mode 100644 drivers/staging/wimax/stack.c create mode 100644 drivers/staging/wimax/wimax-internal.h delete mode 100644 include/linux/wimax/debug.h delete mode 100644 include/net/wimax.h delete mode 100644 include/uapi/linux/wimax.h delete mode 100644 include/uapi/linux/wimax/i2400m.h delete mode 100644 net/wimax/Kconfig delete mode 100644 net/wimax/Makefile delete mode 100644 net/wimax/debug-levels.h delete mode 100644 net/wimax/debugfs.c delete mode 100644 net/wimax/id-table.c delete mode 100644 net/wimax/op-msg.c delete mode 100644 net/wimax/op-reset.c delete mode 100644 net/wimax/op-rfkill.c delete mode 100644 net/wimax/op-state-get.c delete mode 100644 net/wimax/stack.c delete mode 100644 net/wimax/wimax-internal.h (limited to 'include/net') diff --git a/Documentation/admin-guide/index.rst b/Documentation/admin-guide/index.rst index ed1cf94ea50c..d53986a424c4 100644 --- a/Documentation/admin-guide/index.rst +++ b/Documentation/admin-guide/index.rst @@ -115,7 +115,6 @@ configure specific aspects of kernel behavior to your liking. unicode vga-softcursor video-output - wimax/index xfs .. only:: subproject and html diff --git a/Documentation/admin-guide/wimax/i2400m.rst b/Documentation/admin-guide/wimax/i2400m.rst deleted file mode 100644 index 194388c0c351..000000000000 --- a/Documentation/admin-guide/wimax/i2400m.rst +++ /dev/null @@ -1,283 +0,0 @@ -.. include:: - -==================================================== -Driver for the Intel Wireless Wimax Connection 2400m -==================================================== - -:Copyright: |copy| 2008 Intel Corporation < linux-wimax@intel.com > - - This provides a driver for the Intel Wireless WiMAX Connection 2400m - and a basic Linux kernel WiMAX stack. - -1. Requirements -=============== - - * Linux installation with Linux kernel 2.6.22 or newer (if building - from a separate tree) - * Intel i2400m Echo Peak or Baxter Peak; this includes the Intel - Wireless WiMAX/WiFi Link 5x50 series. - * build tools: - - + Linux kernel development package for the target kernel; to - build against your currently running kernel, you need to have - the kernel development package corresponding to the running - image installed (usually if your kernel is named - linux-VERSION, the development package is called - linux-dev-VERSION or linux-headers-VERSION). - + GNU C Compiler, make - -2. Compilation and installation -=============================== - -2.1. Compilation of the drivers included in the kernel ------------------------------------------------------- - - Configure the kernel; to enable the WiMAX drivers select Drivers > - Networking Drivers > WiMAX device support. Enable all of them as - modules (easier). - - If USB or SDIO are not enabled in the kernel configuration, the options - to build the i2400m USB or SDIO drivers will not show. Enable said - subsystems and go back to the WiMAX menu to enable the drivers. - - Compile and install your kernel as usual. - -2.2. Compilation of the drivers distributed as an standalone module -------------------------------------------------------------------- - - To compile:: - - $ cd source/directory - $ make - - Once built you can load and unload using the provided load.sh script; - load.sh will load the modules, load.sh u will unload them. - - To install in the default kernel directories (and enable auto loading - when the device is plugged):: - - $ make install - $ depmod -a - - If your kernel development files are located in a non standard - directory or if you want to build for a kernel that is not the - currently running one, set KDIR to the right location:: - - $ make KDIR=/path/to/kernel/dev/tree - - For more information, please contact linux-wimax@intel.com. - -3. Installing the firmware --------------------------- - - The firmware can be obtained from http://linuxwimax.org or might have - been supplied with your hardware. - - It has to be installed in the target system:: - - $ cp FIRMWAREFILE.sbcf /lib/firmware/i2400m-fw-BUSTYPE-1.3.sbcf - - * NOTE: if your firmware came in an .rpm or .deb file, just install - it as normal, with the rpm (rpm -i FIRMWARE.rpm) or dpkg - (dpkg -i FIRMWARE.deb) commands. No further action is needed. - * BUSTYPE will be usb or sdio, depending on the hardware you have. - Each hardware type comes with its own firmware and will not work - with other types. - -4. Design -========= - - This package contains two major parts: a WiMAX kernel stack and a - driver for the Intel i2400m. - - The WiMAX stack is designed to provide for common WiMAX control - services to current and future WiMAX devices from any vendor; please - see README.wimax for details. - - The i2400m kernel driver is broken up in two main parts: the bus - generic driver and the bus-specific drivers. The bus generic driver - forms the drivercore and contain no knowledge of the actual method we - use to connect to the device. The bus specific drivers are just the - glue to connect the bus-generic driver and the device. Currently only - USB and SDIO are supported. See drivers/net/wimax/i2400m/i2400m.h for - more information. - - The bus generic driver is logically broken up in two parts: OS-glue and - hardware-glue. The OS-glue interfaces with Linux. The hardware-glue - interfaces with the device on using an interface provided by the - bus-specific driver. The reason for this breakup is to be able to - easily reuse the hardware-glue to write drivers for other OSes; note - the hardware glue part is written as a native Linux driver; no - abstraction layers are used, so to port to another OS, the Linux kernel - API calls should be replaced with the target OS's. - -5. Usage -======== - - To load the driver, follow the instructions in the install section; - once the driver is loaded, plug in the device (unless it is permanently - plugged in). The driver will enumerate the device, upload the firmware - and output messages in the kernel log (dmesg, /var/log/messages or - /var/log/kern.log) such as:: - - ... - i2400m_usb 5-4:1.0: firmware interface version 8.0.0 - i2400m_usb 5-4:1.0: WiMAX interface wmx0 (00:1d:e1:01:94:2c) ready - - At this point the device is ready to work. - - Current versions require the Intel WiMAX Network Service in userspace - to make things work. See the network service's README for instructions - on how to scan, connect and disconnect. - -5.1. Module parameters ----------------------- - - Module parameters can be set at kernel or module load time or by - echoing values:: - - $ echo VALUE > /sys/module/MODULENAME/parameters/PARAMETERNAME - - To make changes permanent, for example, for the i2400m module, you can - also create a file named /etc/modprobe.d/i2400m containing:: - - options i2400m idle_mode_disabled=1 - - To find which parameters are supported by a module, run:: - - $ modinfo path/to/module.ko - - During kernel bootup (if the driver is linked in the kernel), specify - the following to the kernel command line:: - - i2400m.PARAMETER=VALUE - -5.1.1. i2400m: idle_mode_disabled -^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - - The i2400m module supports a parameter to disable idle mode. This - parameter, once set, will take effect only when the device is - reinitialized by the driver (eg: following a reset or a reconnect). - -5.2. Debug operations: debugfs entries --------------------------------------- - - The driver will register debugfs entries that allow the user to tweak - debug settings. There are three main container directories where - entries are placed, which correspond to the three blocks a i2400m WiMAX - driver has: - - * /sys/kernel/debug/wimax:DEVNAME/ for the generic WiMAX stack - controls - * /sys/kernel/debug/wimax:DEVNAME/i2400m for the i2400m generic - driver controls - * /sys/kernel/debug/wimax:DEVNAME/i2400m-usb (or -sdio) for the - bus-specific i2400m-usb or i2400m-sdio controls). - - Of course, if debugfs is mounted in a directory other than - /sys/kernel/debug, those paths will change. - -5.2.1. Increasing debug output -^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - - The files named *dl_* indicate knobs for controlling the debug output - of different submodules:: - - # find /sys/kernel/debug/wimax\:wmx0 -name \*dl_\* - /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_tx - /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_rx - /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_notif - /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_fw - /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_usb - /sys/kernel/debug/wimax:wmx0/i2400m/dl_tx - /sys/kernel/debug/wimax:wmx0/i2400m/dl_rx - /sys/kernel/debug/wimax:wmx0/i2400m/dl_rfkill - /sys/kernel/debug/wimax:wmx0/i2400m/dl_netdev - /sys/kernel/debug/wimax:wmx0/i2400m/dl_fw - /sys/kernel/debug/wimax:wmx0/i2400m/dl_debugfs - /sys/kernel/debug/wimax:wmx0/i2400m/dl_driver - /sys/kernel/debug/wimax:wmx0/i2400m/dl_control - /sys/kernel/debug/wimax:wmx0/wimax_dl_stack - /sys/kernel/debug/wimax:wmx0/wimax_dl_op_rfkill - /sys/kernel/debug/wimax:wmx0/wimax_dl_op_reset - /sys/kernel/debug/wimax:wmx0/wimax_dl_op_msg - /sys/kernel/debug/wimax:wmx0/wimax_dl_id_table - /sys/kernel/debug/wimax:wmx0/wimax_dl_debugfs - - By reading the file you can obtain the current value of said debug - level; by writing to it, you can set it. - - To increase the debug level of, for example, the i2400m's generic TX - engine, just write:: - - $ echo 3 > /sys/kernel/debug/wimax:wmx0/i2400m/dl_tx - - Increasing numbers yield increasing debug information; for details of - what is printed and the available levels, check the source. The code - uses 0 for disabled and increasing values until 8. - -5.2.2. RX and TX statistics -^^^^^^^^^^^^^^^^^^^^^^^^^^^ - - The i2400m/rx_stats and i2400m/tx_stats provide statistics about the - data reception/delivery from the device:: - - $ cat /sys/kernel/debug/wimax:wmx0/i2400m/rx_stats - 45 1 3 34 3104 48 480 - - The numbers reported are: - - * packets/RX-buffer: total, min, max - * RX-buffers: total RX buffers received, accumulated RX buffer size - in bytes, min size received, max size received - - Thus, to find the average buffer size received, divide accumulated - RX-buffer / total RX-buffers. - - To clear the statistics back to 0, write anything to the rx_stats file:: - - $ echo 1 > /sys/kernel/debug/wimax:wmx0/i2400m_rx_stats - - Likewise for TX. - - Note the packets this debug file refers to are not network packet, but - packets in the sense of the device-specific protocol for communication - to the host. See drivers/net/wimax/i2400m/tx.c. - -5.2.3. Tracing messages received from user space -^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - - To echo messages received from user space into the trace pipe that the - i2400m driver creates, set the debug file i2400m/trace_msg_from_user to - 1:: - - $ echo 1 > /sys/kernel/debug/wimax:wmx0/i2400m/trace_msg_from_user - -5.2.4. Performing a device reset -^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - - By writing a 0, a 1 or a 2 to the file - /sys/kernel/debug/wimax:wmx0/reset, the driver performs a warm (without - disconnecting from the bus), cold (disconnecting from the bus) or bus - (bus specific) reset on the device. - -5.2.5. Asking the device to enter power saving mode -^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - - By writing any value to the /sys/kernel/debug/wimax:wmx0 file, the - device will attempt to enter power saving mode. - -6. Troubleshooting -================== - -6.1. Driver complains about ``i2400m-fw-usb-1.2.sbcf: request failed`` ----------------------------------------------------------------------- - - If upon connecting the device, the following is output in the kernel - log:: - - i2400m_usb 5-4:1.0: fw i2400m-fw-usb-1.3.sbcf: request failed: -2 - - This means that the driver cannot locate the firmware file named - /lib/firmware/i2400m-fw-usb-1.2.sbcf. Check that the file is present in - the right location. diff --git a/Documentation/admin-guide/wimax/index.rst b/Documentation/admin-guide/wimax/index.rst deleted file mode 100644 index fdf7c1f99ff5..000000000000 --- a/Documentation/admin-guide/wimax/index.rst +++ /dev/null @@ -1,19 +0,0 @@ -.. SPDX-License-Identifier: GPL-2.0 - -=============== -WiMAX subsystem -=============== - -.. toctree:: - :maxdepth: 2 - - wimax - - i2400m - -.. only:: subproject and html - - Indices - ======= - - * :ref:`genindex` diff --git a/Documentation/admin-guide/wimax/wimax.rst b/Documentation/admin-guide/wimax/wimax.rst deleted file mode 100644 index 817ee8ba2732..000000000000 --- a/Documentation/admin-guide/wimax/wimax.rst +++ /dev/null @@ -1,89 +0,0 @@ -.. include:: - -======================== -Linux kernel WiMAX stack -======================== - -:Copyright: |copy| 2008 Intel Corporation < linux-wimax@intel.com > - - This provides a basic Linux kernel WiMAX stack to provide a common - control API for WiMAX devices, usable from kernel and user space. - -1. Design -========= - - The WiMAX stack is designed to provide for common WiMAX control - services to current and future WiMAX devices from any vendor. - - Because currently there is only one and we don't know what would be the - common services, the APIs it currently provides are very minimal. - However, it is done in such a way that it is easily extensible to - accommodate future requirements. - - The stack works by embedding a struct wimax_dev in your device's - control structures. This provides a set of callbacks that the WiMAX - stack will call in order to implement control operations requested by - the user. As well, the stack provides API functions that the driver - calls to notify about changes of state in the device. - - The stack exports the API calls needed to control the device to user - space using generic netlink as a marshalling mechanism. You can access - them using your own code or use the wrappers provided for your - convenience in libwimax (in the wimax-tools package). - - For detailed information on the stack, please see - include/linux/wimax.h. - -2. Usage -======== - - For usage in a driver (registration, API, etc) please refer to the - instructions in the header file include/linux/wimax.h. - - When a device is registered with the WiMAX stack, a set of debugfs - files will appear in /sys/kernel/debug/wimax:wmxX can tweak for - control. - -2.1. Obtaining debug information: debugfs entries -------------------------------------------------- - - The WiMAX stack is compiled, by default, with debug messages that can - be used to diagnose issues. By default, said messages are disabled. - - The drivers will register debugfs entries that allow the user to tweak - debug settings. - - Each driver, when registering with the stack, will cause a debugfs - directory named wimax:DEVICENAME to be created; optionally, it might - create more subentries below it. - -2.1.1. Increasing debug output ------------------------------- - - The files named *dl_* indicate knobs for controlling the debug output - of different submodules of the WiMAX stack:: - - # find /sys/kernel/debug/wimax\:wmx0 -name \*dl_\* - /sys/kernel/debug/wimax:wmx0/wimax_dl_stack - /sys/kernel/debug/wimax:wmx0/wimax_dl_op_rfkill - /sys/kernel/debug/wimax:wmx0/wimax_dl_op_reset - /sys/kernel/debug/wimax:wmx0/wimax_dl_op_msg - /sys/kernel/debug/wimax:wmx0/wimax_dl_id_table - /sys/kernel/debug/wimax:wmx0/wimax_dl_debugfs - /sys/kernel/debug/wimax:wmx0/.... # other driver specific files - - NOTE: - Of course, if debugfs is mounted in a directory other than - /sys/kernel/debug, those paths will change. - - By reading the file you can obtain the current value of said debug - level; by writing to it, you can set it. - - To increase the debug level of, for example, the id-table submodule, - just write: - - $ echo 3 > /sys/kernel/debug/wimax:wmx0/wimax_dl_id_table - - Increasing numbers yield increasing debug information; for details of - what is printed and the available levels, check the source. The code - uses 0 for disabled and increasing values until 8. diff --git a/Documentation/networking/kapi.rst b/Documentation/networking/kapi.rst index d198fa5eaacd..ea55f462cefa 100644 --- a/Documentation/networking/kapi.rst +++ b/Documentation/networking/kapi.rst @@ -83,27 +83,6 @@ SUN RPC subsystem .. kernel-doc:: net/sunrpc/clnt.c :export: -WiMAX ------ - -.. kernel-doc:: net/wimax/op-msg.c - :export: - -.. kernel-doc:: net/wimax/op-reset.c - :export: - -.. kernel-doc:: net/wimax/op-rfkill.c - :export: - -.. kernel-doc:: net/wimax/stack.c - :export: - -.. kernel-doc:: include/net/wimax.h - :internal: - -.. kernel-doc:: include/uapi/linux/wimax.h - :internal: - Network device support ====================== diff --git a/Documentation/translations/zh_CN/admin-guide/index.rst b/Documentation/translations/zh_CN/admin-guide/index.rst index ed5ab7e37f38..48bbd3ebad48 100644 --- a/Documentation/translations/zh_CN/admin-guide/index.rst +++ b/Documentation/translations/zh_CN/admin-guide/index.rst @@ -114,7 +114,6 @@ Todolist: unicode vga-softcursor video-output - wimax/index xfs .. only:: subproject and html diff --git a/MAINTAINERS b/MAINTAINERS index e73636b75f29..17f5571788c9 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9106,16 +9106,6 @@ W: https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi T: git git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/iwlwifi.git F: drivers/net/wireless/intel/iwlwifi/ -INTEL WIRELESS WIMAX CONNECTION 2400 -M: Inaky Perez-Gonzalez -M: linux-wimax@intel.com -L: wimax@linuxwimax.org (subscribers-only) -S: Supported -W: http://linuxwimax.org -F: Documentation/admin-guide/wimax/i2400m.rst -F: drivers/net/wimax/i2400m/ -F: include/uapi/linux/wimax/i2400m.h - INTEL WMI SLIM BOOTLOADER (SBL) FIRMWARE UPDATE DRIVER M: Jithu Joseph R: Maurice Ma @@ -18907,18 +18897,6 @@ S: Supported W: https://wireless.wiki.kernel.org/en/users/Drivers/wil6210 F: drivers/net/wireless/ath/wil6210/ -WIMAX STACK -M: Inaky Perez-Gonzalez -M: linux-wimax@intel.com -L: wimax@linuxwimax.org (subscribers-only) -S: Supported -W: http://linuxwimax.org -F: Documentation/admin-guide/wimax/wimax.rst -F: include/linux/wimax/debug.h -F: include/net/wimax.h -F: include/uapi/linux/wimax.h -F: net/wimax/ - WINBOND CIR DRIVER M: David Härdeman S: Maintained diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index c3dbe64e628e..c0af2dc8b938 100644 --- a/drivers/net/Kconfig +++ b/drivers/net/Kconfig @@ -489,8 +489,6 @@ source "drivers/net/usb/Kconfig" source "drivers/net/wireless/Kconfig" -source "drivers/net/wimax/Kconfig" - source "drivers/net/wan/Kconfig" source "drivers/net/ieee802154/Kconfig" diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 72e18d505d1a..b27e8633c305 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -66,7 +66,6 @@ obj-$(CONFIG_NET_SB1000) += sb1000.o obj-$(CONFIG_SUNGEM_PHY) += sungem_phy.o obj-$(CONFIG_WAN) += wan/ obj-$(CONFIG_WLAN) += wireless/ -obj-$(CONFIG_WIMAX) += wimax/ obj-$(CONFIG_IEEE802154) += ieee802154/ obj-$(CONFIG_VMXNET3) += vmxnet3/ diff --git a/drivers/net/wimax/Kconfig b/drivers/net/wimax/Kconfig deleted file mode 100644 index 2249e3d77a76..000000000000 --- a/drivers/net/wimax/Kconfig +++ /dev/null @@ -1,18 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only -# -# WiMAX LAN device drivers configuration -# - - -comment "Enable WiMAX (Networking options) to see the WiMAX drivers" - depends on WIMAX = n - -if WIMAX - -menu "WiMAX Wireless Broadband devices" - -source "drivers/net/wimax/i2400m/Kconfig" - -endmenu - -endif diff --git a/drivers/net/wimax/Makefile b/drivers/net/wimax/Makefile deleted file mode 100644 index b4575bacf994..000000000000 --- a/drivers/net/wimax/Makefile +++ /dev/null @@ -1,2 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only -obj-$(CONFIG_WIMAX_I2400M) += i2400m/ diff --git a/drivers/net/wimax/i2400m/Kconfig b/drivers/net/wimax/i2400m/Kconfig deleted file mode 100644 index 843b905a26a3..000000000000 --- a/drivers/net/wimax/i2400m/Kconfig +++ /dev/null @@ -1,37 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only - -config WIMAX_I2400M - tristate - depends on WIMAX - select FW_LOADER - -comment "Enable USB support to see WiMAX USB drivers" - depends on USB = n - -config WIMAX_I2400M_USB - tristate "Intel Wireless WiMAX Connection 2400 over USB (including 5x50)" - depends on WIMAX && USB - select WIMAX_I2400M - help - Select if you have a device based on the Intel WiMAX - Connection 2400 over USB (like any of the Intel Wireless - WiMAX/WiFi Link 5x50 series). - - If unsure, it is safe to select M (module). - -config WIMAX_I2400M_DEBUG_LEVEL - int "WiMAX i2400m debug level" - depends on WIMAX_I2400M - default 8 - help - - Select the maximum debug verbosity level to be compiled into - the WiMAX i2400m driver code. - - By default, this is disabled at runtime and can be - selectively enabled at runtime for different parts of the - code using the sysfs debug-levels file. - - If set at zero, this will compile out all the debug code. - - It is recommended that it is left at 8. diff --git a/drivers/net/wimax/i2400m/Makefile b/drivers/net/wimax/i2400m/Makefile deleted file mode 100644 index b1db1eff0648..000000000000 --- a/drivers/net/wimax/i2400m/Makefile +++ /dev/null @@ -1,23 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0 - -obj-$(CONFIG_WIMAX_I2400M) += i2400m.o -obj-$(CONFIG_WIMAX_I2400M_USB) += i2400m-usb.o - -i2400m-y := \ - control.o \ - driver.o \ - fw.o \ - op-rfkill.o \ - sysfs.o \ - netdev.o \ - tx.o \ - rx.o - -i2400m-$(CONFIG_DEBUG_FS) += debugfs.o - -i2400m-usb-y := \ - usb-fw.o \ - usb-notif.o \ - usb-tx.o \ - usb-rx.o \ - usb.o diff --git a/drivers/net/wimax/i2400m/control.c b/drivers/net/wimax/i2400m/control.c deleted file mode 100644 index 8df98757d901..000000000000 --- a/drivers/net/wimax/i2400m/control.c +++ /dev/null @@ -1,1434 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * Miscellaneous control functions for managing the device - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Inaky Perez-Gonzalez - * - Initial implementation - * - * This is a collection of functions used to control the device (plus - * a few helpers). - * - * There are utilities for handling TLV buffers, hooks on the device's - * reports to act on device changes of state [i2400m_report_hook()], - * on acks to commands [i2400m_msg_ack_hook()], a helper for sending - * commands to the device and blocking until a reply arrives - * [i2400m_msg_to_dev()], a few high level commands for manipulating - * the device state, powersving mode and configuration plus the - * routines to setup the device once communication is stablished with - * it [i2400m_dev_initialize()]. - * - * ROADMAP - * - * i2400m_dev_initialize() Called by i2400m_dev_start() - * i2400m_set_init_config() - * i2400m_cmd_get_state() - * i2400m_dev_shutdown() Called by i2400m_dev_stop() - * i2400m_reset() - * - * i2400m_{cmd,get,set}_*() - * i2400m_msg_to_dev() - * i2400m_msg_check_status() - * - * i2400m_report_hook() Called on reception of an event - * i2400m_report_state_hook() - * i2400m_tlv_buffer_walk() - * i2400m_tlv_match() - * i2400m_report_tlv_system_state() - * i2400m_report_tlv_rf_switches_status() - * i2400m_report_tlv_media_status() - * i2400m_cmd_enter_powersave() - * - * i2400m_msg_ack_hook() Called on reception of a reply to a - * command, get or set - */ - -#include -#include "i2400m.h" -#include -#include -#include -#include -#include - - -#define D_SUBMODULE control -#include "debug-levels.h" - -static int i2400m_idle_mode_disabled;/* 0 (idle mode enabled) by default */ -module_param_named(idle_mode_disabled, i2400m_idle_mode_disabled, int, 0644); -MODULE_PARM_DESC(idle_mode_disabled, - "If true, the device will not enable idle mode negotiation " - "with the base station (when connected) to save power."); - -/* 0 (power saving enabled) by default */ -static int i2400m_power_save_disabled; -module_param_named(power_save_disabled, i2400m_power_save_disabled, int, 0644); -MODULE_PARM_DESC(power_save_disabled, - "If true, the driver will not tell the device to enter " - "power saving mode when it reports it is ready for it. " - "False by default (so the device is told to do power " - "saving)."); - -static int i2400m_passive_mode; /* 0 (passive mode disabled) by default */ -module_param_named(passive_mode, i2400m_passive_mode, int, 0644); -MODULE_PARM_DESC(passive_mode, - "If true, the driver will not do any device setup " - "and leave it up to user space, who must be properly " - "setup."); - - -/* - * Return if a TLV is of a give type and size - * - * @tlv_hdr: pointer to the TLV - * @tlv_type: type of the TLV we are looking for - * @tlv_size: expected size of the TLV we are looking for (if -1, - * don't check the size). This includes the header - * Returns: 0 if the TLV matches - * < 0 if it doesn't match at all - * > 0 total TLV + payload size, if the type matches, but not - * the size - */ -static -ssize_t i2400m_tlv_match(const struct i2400m_tlv_hdr *tlv, - enum i2400m_tlv tlv_type, ssize_t tlv_size) -{ - if (le16_to_cpu(tlv->type) != tlv_type) /* Not our type? skip */ - return -1; - if (tlv_size != -1 - && le16_to_cpu(tlv->length) + sizeof(*tlv) != tlv_size) { - size_t size = le16_to_cpu(tlv->length) + sizeof(*tlv); - printk(KERN_WARNING "W: tlv type 0x%x mismatched because of " - "size (got %zu vs %zd expected)\n", - tlv_type, size, tlv_size); - return size; - } - return 0; -} - - -/* - * Given a buffer of TLVs, iterate over them - * - * @i2400m: device instance - * @tlv_buf: pointer to the beginning of the TLV buffer - * @buf_size: buffer size in bytes - * @tlv_pos: seek position; this is assumed to be a pointer returned - * by i2400m_tlv_buffer_walk() [and thus, validated]. The - * TLV returned will be the one following this one. - * - * Usage: - * - * tlv_itr = NULL; - * while (tlv_itr = i2400m_tlv_buffer_walk(i2400m, buf, size, tlv_itr)) { - * ... - * // Do stuff with tlv_itr, DON'T MODIFY IT - * ... - * } - */ -static -const struct i2400m_tlv_hdr *i2400m_tlv_buffer_walk( - struct i2400m *i2400m, - const void *tlv_buf, size_t buf_size, - const struct i2400m_tlv_hdr *tlv_pos) -{ - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_tlv_hdr *tlv_top = tlv_buf + buf_size; - size_t offset, length, avail_size; - unsigned type; - - if (tlv_pos == NULL) /* Take the first one? */ - tlv_pos = tlv_buf; - else /* Nope, the next one */ - tlv_pos = (void *) tlv_pos - + le16_to_cpu(tlv_pos->length) + sizeof(*tlv_pos); - if (tlv_pos == tlv_top) { /* buffer done */ - tlv_pos = NULL; - goto error_beyond_end; - } - if (tlv_pos > tlv_top) { - tlv_pos = NULL; - WARN_ON(1); - goto error_beyond_end; - } - offset = (void *) tlv_pos - (void *) tlv_buf; - avail_size = buf_size - offset; - if (avail_size < sizeof(*tlv_pos)) { - dev_err(dev, "HW BUG? tlv_buf %p [%zu bytes], tlv @%zu: " - "short header\n", tlv_buf, buf_size, offset); - goto error_short_header; - } - type = le16_to_cpu(tlv_pos->type); - length = le16_to_cpu(tlv_pos->length); - if (avail_size < sizeof(*tlv_pos) + length) { - dev_err(dev, "HW BUG? tlv_buf %p [%zu bytes], " - "tlv type 0x%04x @%zu: " - "short data (%zu bytes vs %zu needed)\n", - tlv_buf, buf_size, type, offset, avail_size, - sizeof(*tlv_pos) + length); - goto error_short_header; - } -error_short_header: -error_beyond_end: - return tlv_pos; -} - - -/* - * Find a TLV in a buffer of sequential TLVs - * - * @i2400m: device descriptor - * @tlv_hdr: pointer to the first TLV in the sequence - * @size: size of the buffer in bytes; all TLVs are assumed to fit - * fully in the buffer (otherwise we'll complain). - * @tlv_type: type of the TLV we are looking for - * @tlv_size: expected size of the TLV we are looking for (if -1, - * don't check the size). This includes the header - * - * Returns: NULL if the TLV is not found, otherwise a pointer to - * it. If the sizes don't match, an error is printed and NULL - * returned. - */ -static -const struct i2400m_tlv_hdr *i2400m_tlv_find( - struct i2400m *i2400m, - const struct i2400m_tlv_hdr *tlv_hdr, size_t size, - enum i2400m_tlv tlv_type, ssize_t tlv_size) -{ - ssize_t match; - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_tlv_hdr *tlv = NULL; - while ((tlv = i2400m_tlv_buffer_walk(i2400m, tlv_hdr, size, tlv))) { - match = i2400m_tlv_match(tlv, tlv_type, tlv_size); - if (match == 0) /* found it :) */ - break; - if (match > 0) - dev_warn(dev, "TLV type 0x%04x found with size " - "mismatch (%zu vs %zd needed)\n", - tlv_type, match, tlv_size); - } - return tlv; -} - - -static const struct -{ - char *msg; - int errno; -} ms_to_errno[I2400M_MS_MAX] = { - [I2400M_MS_DONE_OK] = { "", 0 }, - [I2400M_MS_DONE_IN_PROGRESS] = { "", 0 }, - [I2400M_MS_INVALID_OP] = { "invalid opcode", -ENOSYS }, - [I2400M_MS_BAD_STATE] = { "invalid state", -EILSEQ }, - [I2400M_MS_ILLEGAL_VALUE] = { "illegal value", -EINVAL }, - [I2400M_MS_MISSING_PARAMS] = { "missing parameters", -ENOMSG }, - [I2400M_MS_VERSION_ERROR] = { "bad version", -EIO }, - [I2400M_MS_ACCESSIBILITY_ERROR] = { "accesibility error", -EIO }, - [I2400M_MS_BUSY] = { "busy", -EBUSY }, - [I2400M_MS_CORRUPTED_TLV] = { "corrupted TLV", -EILSEQ }, - [I2400M_MS_UNINITIALIZED] = { "uninitialized", -EILSEQ }, - [I2400M_MS_UNKNOWN_ERROR] = { "unknown error", -EIO }, - [I2400M_MS_PRODUCTION_ERROR] = { "production error", -EIO }, - [I2400M_MS_NO_RF] = { "no RF", -EIO }, - [I2400M_MS_NOT_READY_FOR_POWERSAVE] = - { "not ready for powersave", -EACCES }, - [I2400M_MS_THERMAL_CRITICAL] = { "thermal critical", -EL3HLT }, -}; - - -/* - * i2400m_msg_check_status - translate a message's status code - * - * @i2400m: device descriptor - * @l3l4_hdr: message header - * @strbuf: buffer to place a formatted error message (unless NULL). - * @strbuf_size: max amount of available space; larger messages will - * be truncated. - * - * Returns: errno code corresponding to the status code in @l3l4_hdr - * and a message in @strbuf describing the error. - */ -int i2400m_msg_check_status(const struct i2400m_l3l4_hdr *l3l4_hdr, - char *strbuf, size_t strbuf_size) -{ - int result; - enum i2400m_ms status = le16_to_cpu(l3l4_hdr->status); - const char *str; - - if (status == 0) - return 0; - if (status >= ARRAY_SIZE(ms_to_errno)) { - str = "unknown status code"; - result = -EBADR; - } else { - str = ms_to_errno[status].msg; - result = ms_to_errno[status].errno; - } - if (strbuf) - snprintf(strbuf, strbuf_size, "%s (%d)", str, status); - return result; -} - - -/* - * Act on a TLV System State reported by the device - * - * @i2400m: device descriptor - * @ss: validated System State TLV - */ -static -void i2400m_report_tlv_system_state(struct i2400m *i2400m, - const struct i2400m_tlv_system_state *ss) -{ - struct device *dev = i2400m_dev(i2400m); - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - enum i2400m_system_state i2400m_state = le32_to_cpu(ss->state); - - d_fnstart(3, dev, "(i2400m %p ss %p [%u])\n", i2400m, ss, i2400m_state); - - if (i2400m->state != i2400m_state) { - i2400m->state = i2400m_state; - wake_up_all(&i2400m->state_wq); - } - switch (i2400m_state) { - case I2400M_SS_UNINITIALIZED: - case I2400M_SS_INIT: - case I2400M_SS_CONFIG: - case I2400M_SS_PRODUCTION: - wimax_state_change(wimax_dev, WIMAX_ST_UNINITIALIZED); - break; - - case I2400M_SS_RF_OFF: - case I2400M_SS_RF_SHUTDOWN: - wimax_state_change(wimax_dev, WIMAX_ST_RADIO_OFF); - break; - - case I2400M_SS_READY: - case I2400M_SS_STANDBY: - case I2400M_SS_SLEEPACTIVE: - wimax_state_change(wimax_dev, WIMAX_ST_READY); - break; - - case I2400M_SS_CONNECTING: - case I2400M_SS_WIMAX_CONNECTED: - wimax_state_change(wimax_dev, WIMAX_ST_READY); - break; - - case I2400M_SS_SCAN: - case I2400M_SS_OUT_OF_ZONE: - wimax_state_change(wimax_dev, WIMAX_ST_SCANNING); - break; - - case I2400M_SS_IDLE: - d_printf(1, dev, "entering BS-negotiated idle mode\n"); - fallthrough; - case I2400M_SS_DISCONNECTING: - case I2400M_SS_DATA_PATH_CONNECTED: - wimax_state_change(wimax_dev, WIMAX_ST_CONNECTED); - break; - - default: - /* Huh? just in case, shut it down */ - dev_err(dev, "HW BUG? unknown state %u: shutting down\n", - i2400m_state); - i2400m_reset(i2400m, I2400M_RT_WARM); - break; - } - d_fnend(3, dev, "(i2400m %p ss %p [%u]) = void\n", - i2400m, ss, i2400m_state); -} - - -/* - * Parse and act on a TLV Media Status sent by the device - * - * @i2400m: device descriptor - * @ms: validated Media Status TLV - * - * This will set the carrier up on down based on the device's link - * report. This is done asides of what the WiMAX stack does based on - * the device's state as sometimes we need to do a link-renew (the BS - * wants us to renew a DHCP lease, for example). - * - * In fact, doc says that every time we get a link-up, we should do a - * DHCP negotiation... - */ -static -void i2400m_report_tlv_media_status(struct i2400m *i2400m, - const struct i2400m_tlv_media_status *ms) -{ - struct device *dev = i2400m_dev(i2400m); - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - struct net_device *net_dev = wimax_dev->net_dev; - enum i2400m_media_status status = le32_to_cpu(ms->media_status); - - d_fnstart(3, dev, "(i2400m %p ms %p [%u])\n", i2400m, ms, status); - - switch (status) { - case I2400M_MEDIA_STATUS_LINK_UP: - netif_carrier_on(net_dev); - break; - case I2400M_MEDIA_STATUS_LINK_DOWN: - netif_carrier_off(net_dev); - break; - /* - * This is the network telling us we need to retrain the DHCP - * lease -- so far, we are trusting the WiMAX Network Service - * in user space to pick this up and poke the DHCP client. - */ - case I2400M_MEDIA_STATUS_LINK_RENEW: - netif_carrier_on(net_dev); - break; - default: - dev_err(dev, "HW BUG? unknown media status %u\n", - status); - } - d_fnend(3, dev, "(i2400m %p ms %p [%u]) = void\n", - i2400m, ms, status); -} - - -/* - * Process a TLV from a 'state report' - * - * @i2400m: device descriptor - * @tlv: pointer to the TLV header; it has been already validated for - * consistent size. - * @tag: for error messages - * - * Act on the TLVs from a 'state report'. - */ -static -void i2400m_report_state_parse_tlv(struct i2400m *i2400m, - const struct i2400m_tlv_hdr *tlv, - const char *tag) -{ - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_tlv_media_status *ms; - const struct i2400m_tlv_system_state *ss; - const struct i2400m_tlv_rf_switches_status *rfss; - - if (0 == i2400m_tlv_match(tlv, I2400M_TLV_SYSTEM_STATE, sizeof(*ss))) { - ss = container_of(tlv, typeof(*ss), hdr); - d_printf(2, dev, "%s: system state TLV " - "found (0x%04x), state 0x%08x\n", - tag, I2400M_TLV_SYSTEM_STATE, - le32_to_cpu(ss->state)); - i2400m_report_tlv_system_state(i2400m, ss); - } - if (0 == i2400m_tlv_match(tlv, I2400M_TLV_RF_STATUS, sizeof(*rfss))) { - rfss = container_of(tlv, typeof(*rfss), hdr); - d_printf(2, dev, "%s: RF status TLV " - "found (0x%04x), sw 0x%02x hw 0x%02x\n", - tag, I2400M_TLV_RF_STATUS, - le32_to_cpu(rfss->sw_rf_switch), - le32_to_cpu(rfss->hw_rf_switch)); - i2400m_report_tlv_rf_switches_status(i2400m, rfss); - } - if (0 == i2400m_tlv_match(tlv, I2400M_TLV_MEDIA_STATUS, sizeof(*ms))) { - ms = container_of(tlv, typeof(*ms), hdr); - d_printf(2, dev, "%s: Media Status TLV: %u\n", - tag, le32_to_cpu(ms->media_status)); - i2400m_report_tlv_media_status(i2400m, ms); - } -} - - -/* - * Parse a 'state report' and extract information - * - * @i2400m: device descriptor - * @l3l4_hdr: pointer to message; it has been already validated for - * consistent size. - * @size: size of the message (header + payload). The header length - * declaration is assumed to be congruent with @size (as in - * sizeof(*l3l4_hdr) + l3l4_hdr->length == size) - * - * Walk over the TLVs in a report state and act on them. - */ -static -void i2400m_report_state_hook(struct i2400m *i2400m, - const struct i2400m_l3l4_hdr *l3l4_hdr, - size_t size, const char *tag) -{ - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_tlv_hdr *tlv; - size_t tlv_size = le16_to_cpu(l3l4_hdr->length); - - d_fnstart(4, dev, "(i2400m %p, l3l4_hdr %p, size %zu, %s)\n", - i2400m, l3l4_hdr, size, tag); - tlv = NULL; - - while ((tlv = i2400m_tlv_buffer_walk(i2400m, &l3l4_hdr->pl, - tlv_size, tlv))) - i2400m_report_state_parse_tlv(i2400m, tlv, tag); - d_fnend(4, dev, "(i2400m %p, l3l4_hdr %p, size %zu, %s) = void\n", - i2400m, l3l4_hdr, size, tag); -} - - -/* - * i2400m_report_hook - (maybe) act on a report - * - * @i2400m: device descriptor - * @l3l4_hdr: pointer to message; it has been already validated for - * consistent size. - * @size: size of the message (header + payload). The header length - * declaration is assumed to be congruent with @size (as in - * sizeof(*l3l4_hdr) + l3l4_hdr->length == size) - * - * Extract information we might need (like carrien on/off) from a - * device report. - */ -void i2400m_report_hook(struct i2400m *i2400m, - const struct i2400m_l3l4_hdr *l3l4_hdr, size_t size) -{ - struct device *dev = i2400m_dev(i2400m); - unsigned msg_type; - - d_fnstart(3, dev, "(i2400m %p l3l4_hdr %p size %zu)\n", - i2400m, l3l4_hdr, size); - /* Chew on the message, we might need some information from - * here */ - msg_type = le16_to_cpu(l3l4_hdr->type); - switch (msg_type) { - case I2400M_MT_REPORT_STATE: /* carrier detection... */ - i2400m_report_state_hook(i2400m, - l3l4_hdr, size, "REPORT STATE"); - break; - /* If the device is ready for power save, then ask it to do - * it. */ - case I2400M_MT_REPORT_POWERSAVE_READY: /* zzzzz */ - if (l3l4_hdr->status == cpu_to_le16(I2400M_MS_DONE_OK)) { - if (i2400m_power_save_disabled) - d_printf(1, dev, "ready for powersave, " - "not requesting (disabled by module " - "parameter)\n"); - else { - d_printf(1, dev, "ready for powersave, " - "requesting\n"); - i2400m_cmd_enter_powersave(i2400m); - } - } - break; - } - d_fnend(3, dev, "(i2400m %p l3l4_hdr %p size %zu) = void\n", - i2400m, l3l4_hdr, size); -} - - -/* - * i2400m_msg_ack_hook - process cmd/set/get ack for internal status - * - * @i2400m: device descriptor - * @l3l4_hdr: pointer to message; it has been already validated for - * consistent size. - * @size: size of the message - * - * Extract information we might need from acks to commands and act on - * it. This is akin to i2400m_report_hook(). Note most of this - * processing should be done in the function that calls the - * command. This is here for some cases where it can't happen... - */ -static void i2400m_msg_ack_hook(struct i2400m *i2400m, - const struct i2400m_l3l4_hdr *l3l4_hdr, - size_t size) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - unsigned int ack_type; - char strerr[32]; - - /* Chew on the message, we might need some information from - * here */ - ack_type = le16_to_cpu(l3l4_hdr->type); - switch (ack_type) { - case I2400M_MT_CMD_ENTER_POWERSAVE: - /* This is just left here for the sake of example, as - * the processing is done somewhere else. */ - if (0) { - result = i2400m_msg_check_status( - l3l4_hdr, strerr, sizeof(strerr)); - if (result >= 0) - d_printf(1, dev, "ready for power save: %zd\n", - size); - } - break; - } -} - - -/* - * i2400m_msg_size_check() - verify message size and header are congruent - * - * It is ok if the total message size is larger than the expected - * size, as there can be padding. - */ -int i2400m_msg_size_check(struct i2400m *i2400m, - const struct i2400m_l3l4_hdr *l3l4_hdr, - size_t msg_size) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - size_t expected_size; - d_fnstart(4, dev, "(i2400m %p l3l4_hdr %p msg_size %zu)\n", - i2400m, l3l4_hdr, msg_size); - if (msg_size < sizeof(*l3l4_hdr)) { - dev_err(dev, "bad size for message header " - "(expected at least %zu, got %zu)\n", - (size_t) sizeof(*l3l4_hdr), msg_size); - result = -EIO; - goto error_hdr_size; - } - expected_size = le16_to_cpu(l3l4_hdr->length) + sizeof(*l3l4_hdr); - if (msg_size < expected_size) { - dev_err(dev, "bad size for message code 0x%04x (expected %zu, " - "got %zu)\n", le16_to_cpu(l3l4_hdr->type), - expected_size, msg_size); - result = -EIO; - } else - result = 0; -error_hdr_size: - d_fnend(4, dev, - "(i2400m %p l3l4_hdr %p msg_size %zu) = %d\n", - i2400m, l3l4_hdr, msg_size, result); - return result; -} - - - -/* - * Cancel a wait for a command ACK - * - * @i2400m: device descriptor - * @code: [negative] errno code to cancel with (don't use - * -EINPROGRESS) - * - * If there is an ack already filled out, free it. - */ -void i2400m_msg_to_dev_cancel_wait(struct i2400m *i2400m, int code) -{ - struct sk_buff *ack_skb; - unsigned long flags; - - spin_lock_irqsave(&i2400m->rx_lock, flags); - ack_skb = i2400m->ack_skb; - if (ack_skb && !IS_ERR(ack_skb)) - kfree_skb(ack_skb); - i2400m->ack_skb = ERR_PTR(code); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); -} - - -/** - * i2400m_msg_to_dev - Send a control message to the device and get a response - * - * @i2400m: device descriptor - * - * @buf: pointer to the buffer containing the message to be sent; it - * has to start with a &struct i2400M_l3l4_hdr and then - * followed by the payload. Once this function returns, the - * buffer can be reused. - * - * @buf_len: buffer size - * - * Returns: - * - * Pointer to skb containing the ack message. You need to check the - * pointer with IS_ERR(), as it might be an error code. Error codes - * could happen because: - * - * - the message wasn't formatted correctly - * - couldn't send the message - * - failed waiting for a response - * - the ack message wasn't formatted correctly - * - * The returned skb has been allocated with wimax_msg_to_user_alloc(), - * it contains the response in a netlink attribute and is ready to be - * passed up to user space with wimax_msg_to_user_send(). To access - * the payload and its length, use wimax_msg_{data,len}() on the skb. - * - * The skb has to be freed with kfree_skb() once done. - * - * Description: - * - * This function delivers a message/command to the device and waits - * for an ack to be received. The format is described in - * linux/wimax/i2400m.h. In summary, a command/get/set is followed by an - * ack. - * - * This function will not check the ack status, that's left up to the - * caller. Once done with the ack skb, it has to be kfree_skb()ed. - * - * The i2400m handles only one message at the same time, thus we need - * the mutex to exclude other players. - * - * We write the message and then wait for an answer to come back. The - * RX path intercepts control messages and handles them in - * i2400m_rx_ctl(). Reports (notifications) are (maybe) processed - * locally and then forwarded (as needed) to user space on the WiMAX - * stack message pipe. Acks are saved and passed back to us through an - * skb in i2400m->ack_skb which is ready to be given to generic - * netlink if need be. - */ -struct sk_buff *i2400m_msg_to_dev(struct i2400m *i2400m, - const void *buf, size_t buf_len) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_l3l4_hdr *msg_l3l4_hdr; - struct sk_buff *ack_skb; - const struct i2400m_l3l4_hdr *ack_l3l4_hdr; - size_t ack_len; - int ack_timeout; - unsigned msg_type; - unsigned long flags; - - d_fnstart(3, dev, "(i2400m %p buf %p len %zu)\n", - i2400m, buf, buf_len); - - rmb(); /* Make sure we see what i2400m_dev_reset_handle() */ - if (i2400m->boot_mode) - return ERR_PTR(-EL3RST); - - msg_l3l4_hdr = buf; - /* Check msg & payload consistency */ - result = i2400m_msg_size_check(i2400m, msg_l3l4_hdr, buf_len); - if (result < 0) - goto error_bad_msg; - msg_type = le16_to_cpu(msg_l3l4_hdr->type); - d_printf(1, dev, "CMD/GET/SET 0x%04x %zu bytes\n", - msg_type, buf_len); - d_dump(2, dev, buf, buf_len); - - /* Setup the completion, ack_skb ("we are waiting") and send - * the message to the device */ - mutex_lock(&i2400m->msg_mutex); - spin_lock_irqsave(&i2400m->rx_lock, flags); - i2400m->ack_skb = ERR_PTR(-EINPROGRESS); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - init_completion(&i2400m->msg_completion); - result = i2400m_tx(i2400m, buf, buf_len, I2400M_PT_CTRL); - if (result < 0) { - dev_err(dev, "can't send message 0x%04x: %d\n", - le16_to_cpu(msg_l3l4_hdr->type), result); - goto error_tx; - } - - /* Some commands take longer to execute because of crypto ops, - * so we give them some more leeway on timeout */ - switch (msg_type) { - case I2400M_MT_GET_TLS_OPERATION_RESULT: - case I2400M_MT_CMD_SEND_EAP_RESPONSE: - ack_timeout = 5 * HZ; - break; - default: - ack_timeout = HZ; - } - - if (unlikely(i2400m->trace_msg_from_user)) - wimax_msg(&i2400m->wimax_dev, "echo", buf, buf_len, GFP_KERNEL); - /* The RX path in rx.c will put any response for this message - * in i2400m->ack_skb and wake us up. If we cancel the wait, - * we need to change the value of i2400m->ack_skb to something - * not -EINPROGRESS so RX knows there is no one waiting. */ - result = wait_for_completion_interruptible_timeout( - &i2400m->msg_completion, ack_timeout); - if (result == 0) { - dev_err(dev, "timeout waiting for reply to message 0x%04x\n", - msg_type); - result = -ETIMEDOUT; - i2400m_msg_to_dev_cancel_wait(i2400m, result); - goto error_wait_for_completion; - } else if (result < 0) { - dev_err(dev, "error waiting for reply to message 0x%04x: %d\n", - msg_type, result); - i2400m_msg_to_dev_cancel_wait(i2400m, result); - goto error_wait_for_completion; - } - - /* Pull out the ack data from i2400m->ack_skb -- see if it is - * an error and act accordingly */ - spin_lock_irqsave(&i2400m->rx_lock, flags); - ack_skb = i2400m->ack_skb; - if (IS_ERR(ack_skb)) - result = PTR_ERR(ack_skb); - else - result = 0; - i2400m->ack_skb = NULL; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - if (result < 0) - goto error_ack_status; - ack_l3l4_hdr = wimax_msg_data_len(ack_skb, &ack_len); - - /* Check the ack and deliver it if it is ok */ - if (unlikely(i2400m->trace_msg_from_user)) - wimax_msg(&i2400m->wimax_dev, "echo", - ack_l3l4_hdr, ack_len, GFP_KERNEL); - result = i2400m_msg_size_check(i2400m, ack_l3l4_hdr, ack_len); - if (result < 0) { - dev_err(dev, "HW BUG? reply to message 0x%04x: %d\n", - msg_type, result); - goto error_bad_ack_len; - } - if (msg_type != le16_to_cpu(ack_l3l4_hdr->type)) { - dev_err(dev, "HW BUG? bad reply 0x%04x to message 0x%04x\n", - le16_to_cpu(ack_l3l4_hdr->type), msg_type); - result = -EIO; - goto error_bad_ack_type; - } - i2400m_msg_ack_hook(i2400m, ack_l3l4_hdr, ack_len); - mutex_unlock(&i2400m->msg_mutex); - d_fnend(3, dev, "(i2400m %p buf %p len %zu) = %p\n", - i2400m, buf, buf_len, ack_skb); - return ack_skb; - -error_bad_ack_type: -error_bad_ack_len: - kfree_skb(ack_skb); -error_ack_status: -error_wait_for_completion: -error_tx: - mutex_unlock(&i2400m->msg_mutex); -error_bad_msg: - d_fnend(3, dev, "(i2400m %p buf %p len %zu) = %d\n", - i2400m, buf, buf_len, result); - return ERR_PTR(result); -} - - -/* - * Definitions for the Enter Power Save command - * - * The Enter Power Save command requests the device to go into power - * saving mode. The device will ack or nak the command depending on it - * being ready for it. If it acks, we tell the USB subsystem to - * - * As well, the device might request to go into power saving mode by - * sending a report (REPORT_POWERSAVE_READY), in which case, we issue - * this command. The hookups in the RX coder allow - */ -enum { - I2400M_WAKEUP_ENABLED = 0x01, - I2400M_WAKEUP_DISABLED = 0x02, - I2400M_TLV_TYPE_WAKEUP_MODE = 144, -}; - -struct i2400m_cmd_enter_power_save { - struct i2400m_l3l4_hdr hdr; - struct i2400m_tlv_hdr tlv; - __le32 val; -} __packed; - - -/* - * Request entering power save - * - * This command is (mainly) executed when the device indicates that it - * is ready to go into powersave mode via a REPORT_POWERSAVE_READY. - */ -int i2400m_cmd_enter_powersave(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct i2400m_cmd_enter_power_save *cmd; - char strerr[32]; - - result = -ENOMEM; - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->hdr.type = cpu_to_le16(I2400M_MT_CMD_ENTER_POWERSAVE); - cmd->hdr.length = cpu_to_le16(sizeof(*cmd) - sizeof(cmd->hdr)); - cmd->hdr.version = cpu_to_le16(I2400M_L3L4_VERSION); - cmd->tlv.type = cpu_to_le16(I2400M_TLV_TYPE_WAKEUP_MODE); - cmd->tlv.length = cpu_to_le16(sizeof(cmd->val)); - cmd->val = cpu_to_le32(I2400M_WAKEUP_ENABLED); - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - result = PTR_ERR(ack_skb); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'Enter power save' command: %d\n", - result); - goto error_msg_to_dev; - } - result = i2400m_msg_check_status(wimax_msg_data(ack_skb), - strerr, sizeof(strerr)); - if (result == -EACCES) - d_printf(1, dev, "Cannot enter power save mode\n"); - else if (result < 0) - dev_err(dev, "'Enter power save' (0x%04x) command failed: " - "%d - %s\n", I2400M_MT_CMD_ENTER_POWERSAVE, - result, strerr); - else - d_printf(1, dev, "device ready to power save\n"); - kfree_skb(ack_skb); -error_msg_to_dev: - kfree(cmd); -error_alloc: - return result; -} -EXPORT_SYMBOL_GPL(i2400m_cmd_enter_powersave); - - -/* - * Definitions for getting device information - */ -enum { - I2400M_TLV_DETAILED_DEVICE_INFO = 140 -}; - -/** - * i2400m_get_device_info - Query the device for detailed device information - * - * @i2400m: device descriptor - * - * Returns: an skb whose skb->data points to a 'struct - * i2400m_tlv_detailed_device_info'. When done, kfree_skb() it. The - * skb is *guaranteed* to contain the whole TLV data structure. - * - * On error, IS_ERR(skb) is true and ERR_PTR(skb) is the error - * code. - */ -struct sk_buff *i2400m_get_device_info(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct i2400m_l3l4_hdr *cmd; - const struct i2400m_l3l4_hdr *ack; - size_t ack_len; - const struct i2400m_tlv_hdr *tlv; - const struct i2400m_tlv_detailed_device_info *ddi; - char strerr[32]; - - ack_skb = ERR_PTR(-ENOMEM); - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->type = cpu_to_le16(I2400M_MT_GET_DEVICE_INFO); - cmd->length = 0; - cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'get device info' command: %ld\n", - PTR_ERR(ack_skb)); - goto error_msg_to_dev; - } - ack = wimax_msg_data_len(ack_skb, &ack_len); - result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); - if (result < 0) { - dev_err(dev, "'get device info' (0x%04x) command failed: " - "%d - %s\n", I2400M_MT_GET_DEVICE_INFO, result, - strerr); - goto error_cmd_failed; - } - tlv = i2400m_tlv_find(i2400m, ack->pl, ack_len - sizeof(*ack), - I2400M_TLV_DETAILED_DEVICE_INFO, sizeof(*ddi)); - if (tlv == NULL) { - dev_err(dev, "GET DEVICE INFO: " - "detailed device info TLV not found (0x%04x)\n", - I2400M_TLV_DETAILED_DEVICE_INFO); - result = -EIO; - goto error_no_tlv; - } - skb_pull(ack_skb, (void *) tlv - (void *) ack_skb->data); -error_msg_to_dev: - kfree(cmd); -error_alloc: - return ack_skb; - -error_no_tlv: -error_cmd_failed: - kfree_skb(ack_skb); - kfree(cmd); - return ERR_PTR(result); -} - - -/* Firmware interface versions we support */ -enum { - I2400M_HDIv_MAJOR = 9, - I2400M_HDIv_MINOR = 1, - I2400M_HDIv_MINOR_2 = 2, -}; - - -/** - * i2400m_firmware_check - check firmware versions are compatible with - * the driver - * - * @i2400m: device descriptor - * - * Returns: 0 if ok, < 0 errno code an error and a message in the - * kernel log. - * - * Long function, but quite simple; first chunk launches the command - * and double checks the reply for the right TLV. Then we process the - * TLV (where the meat is). - * - * Once we process the TLV that gives us the firmware's interface - * version, we encode it and save it in i2400m->fw_version for future - * reference. - */ -int i2400m_firmware_check(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct i2400m_l3l4_hdr *cmd; - const struct i2400m_l3l4_hdr *ack; - size_t ack_len; - const struct i2400m_tlv_hdr *tlv; - const struct i2400m_tlv_l4_message_versions *l4mv; - char strerr[32]; - unsigned major, minor, branch; - - result = -ENOMEM; - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->type = cpu_to_le16(I2400M_MT_GET_LM_VERSION); - cmd->length = 0; - cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - if (IS_ERR(ack_skb)) { - result = PTR_ERR(ack_skb); - dev_err(dev, "Failed to issue 'get lm version' command: %-d\n", - result); - goto error_msg_to_dev; - } - ack = wimax_msg_data_len(ack_skb, &ack_len); - result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); - if (result < 0) { - dev_err(dev, "'get lm version' (0x%04x) command failed: " - "%d - %s\n", I2400M_MT_GET_LM_VERSION, result, - strerr); - goto error_cmd_failed; - } - tlv = i2400m_tlv_find(i2400m, ack->pl, ack_len - sizeof(*ack), - I2400M_TLV_L4_MESSAGE_VERSIONS, sizeof(*l4mv)); - if (tlv == NULL) { - dev_err(dev, "get lm version: TLV not found (0x%04x)\n", - I2400M_TLV_L4_MESSAGE_VERSIONS); - result = -EIO; - goto error_no_tlv; - } - l4mv = container_of(tlv, typeof(*l4mv), hdr); - major = le16_to_cpu(l4mv->major); - minor = le16_to_cpu(l4mv->minor); - branch = le16_to_cpu(l4mv->branch); - result = -EINVAL; - if (major != I2400M_HDIv_MAJOR) { - dev_err(dev, "unsupported major fw version " - "%u.%u.%u\n", major, minor, branch); - goto error_bad_major; - } - result = 0; - if (minor > I2400M_HDIv_MINOR_2 || minor < I2400M_HDIv_MINOR) - dev_warn(dev, "untested minor fw version %u.%u.%u\n", - major, minor, branch); - /* Yes, we ignore the branch -- we don't have to track it */ - i2400m->fw_version = major << 16 | minor; - dev_info(dev, "firmware interface version %u.%u.%u\n", - major, minor, branch); -error_bad_major: -error_no_tlv: -error_cmd_failed: - kfree_skb(ack_skb); -error_msg_to_dev: - kfree(cmd); -error_alloc: - return result; -} - - -/* - * Send an DoExitIdle command to the device to ask it to go out of - * basestation-idle mode. - * - * @i2400m: device descriptor - * - * This starts a renegotiation with the basestation that might involve - * another crypto handshake with user space. - * - * Returns: 0 if ok, < 0 errno code on error. - */ -int i2400m_cmd_exit_idle(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct i2400m_l3l4_hdr *cmd; - char strerr[32]; - - result = -ENOMEM; - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->type = cpu_to_le16(I2400M_MT_CMD_EXIT_IDLE); - cmd->length = 0; - cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - result = PTR_ERR(ack_skb); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'exit idle' command: %d\n", - result); - goto error_msg_to_dev; - } - result = i2400m_msg_check_status(wimax_msg_data(ack_skb), - strerr, sizeof(strerr)); - kfree_skb(ack_skb); -error_msg_to_dev: - kfree(cmd); -error_alloc: - return result; - -} - - -/* - * Query the device for its state, update the WiMAX stack's idea of it - * - * @i2400m: device descriptor - * - * Returns: 0 if ok, < 0 errno code on error. - * - * Executes a 'Get State' command and parses the returned - * TLVs. - * - * Because this is almost identical to a 'Report State', we use - * i2400m_report_state_hook() to parse the answer. This will set the - * carrier state, as well as the RF Kill switches state. - */ -static int i2400m_cmd_get_state(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct i2400m_l3l4_hdr *cmd; - const struct i2400m_l3l4_hdr *ack; - size_t ack_len; - char strerr[32]; - - result = -ENOMEM; - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->type = cpu_to_le16(I2400M_MT_GET_STATE); - cmd->length = 0; - cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'get state' command: %ld\n", - PTR_ERR(ack_skb)); - result = PTR_ERR(ack_skb); - goto error_msg_to_dev; - } - ack = wimax_msg_data_len(ack_skb, &ack_len); - result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); - if (result < 0) { - dev_err(dev, "'get state' (0x%04x) command failed: " - "%d - %s\n", I2400M_MT_GET_STATE, result, strerr); - goto error_cmd_failed; - } - i2400m_report_state_hook(i2400m, ack, ack_len - sizeof(*ack), - "GET STATE"); - result = 0; - kfree_skb(ack_skb); -error_cmd_failed: -error_msg_to_dev: - kfree(cmd); -error_alloc: - return result; -} - -/** - * Set basic configuration settings - * - * @i2400m: device descriptor - * @args: array of pointers to the TLV headers to send for - * configuration (each followed by its payload). - * TLV headers and payloads must be properly initialized, with the - * right endianess (LE). - * @arg_size: number of pointers in the @args array - */ -static int i2400m_set_init_config(struct i2400m *i2400m, - const struct i2400m_tlv_hdr **arg, - size_t args) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct i2400m_l3l4_hdr *cmd; - char strerr[32]; - unsigned argc, argsize, tlv_size; - const struct i2400m_tlv_hdr *tlv_hdr; - void *buf, *itr; - - d_fnstart(3, dev, "(i2400m %p arg %p args %zu)\n", i2400m, arg, args); - result = 0; - if (args == 0) - goto none; - /* Compute the size of all the TLVs, so we can alloc a - * contiguous command block to copy them. */ - argsize = 0; - for (argc = 0; argc < args; argc++) { - tlv_hdr = arg[argc]; - argsize += sizeof(*tlv_hdr) + le16_to_cpu(tlv_hdr->length); - } - WARN_ON(argc >= 9); /* As per hw spec */ - - /* Alloc the space for the command and TLVs*/ - result = -ENOMEM; - buf = kzalloc(sizeof(*cmd) + argsize, GFP_KERNEL); - if (buf == NULL) - goto error_alloc; - cmd = buf; - cmd->type = cpu_to_le16(I2400M_MT_SET_INIT_CONFIG); - cmd->length = cpu_to_le16(argsize); - cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); - - /* Copy the TLVs */ - itr = buf + sizeof(*cmd); - for (argc = 0; argc < args; argc++) { - tlv_hdr = arg[argc]; - tlv_size = sizeof(*tlv_hdr) + le16_to_cpu(tlv_hdr->length); - memcpy(itr, tlv_hdr, tlv_size); - itr += tlv_size; - } - - /* Send the message! */ - ack_skb = i2400m_msg_to_dev(i2400m, buf, sizeof(*cmd) + argsize); - result = PTR_ERR(ack_skb); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'init config' command: %d\n", - result); - - goto error_msg_to_dev; - } - result = i2400m_msg_check_status(wimax_msg_data(ack_skb), - strerr, sizeof(strerr)); - if (result < 0) - dev_err(dev, "'init config' (0x%04x) command failed: %d - %s\n", - I2400M_MT_SET_INIT_CONFIG, result, strerr); - kfree_skb(ack_skb); -error_msg_to_dev: - kfree(buf); -error_alloc: -none: - d_fnend(3, dev, "(i2400m %p arg %p args %zu) = %d\n", - i2400m, arg, args, result); - return result; - -} - -/** - * i2400m_set_idle_timeout - Set the device's idle mode timeout - * - * @i2400m: i2400m device descriptor - * - * @msecs: milliseconds for the timeout to enter idle mode. Between - * 100 to 300000 (5m); 0 to disable. In increments of 100. - * - * After this @msecs of the link being idle (no data being sent or - * received), the device will negotiate with the basestation entering - * idle mode for saving power. The connection is maintained, but - * getting out of it (done in tx.c) will require some negotiation, - * possible crypto re-handshake and a possible DHCP re-lease. - * - * Only available if fw_version >= 0x00090002. - * - * Returns: 0 if ok, < 0 errno code on error. - */ -int i2400m_set_idle_timeout(struct i2400m *i2400m, unsigned msecs) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct { - struct i2400m_l3l4_hdr hdr; - struct i2400m_tlv_config_idle_timeout cit; - } *cmd; - const struct i2400m_l3l4_hdr *ack; - size_t ack_len; - char strerr[32]; - - result = -ENOSYS; - if (i2400m_le_v1_3(i2400m)) - goto error_alloc; - result = -ENOMEM; - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->hdr.type = cpu_to_le16(I2400M_MT_GET_STATE); - cmd->hdr.length = cpu_to_le16(sizeof(*cmd) - sizeof(cmd->hdr)); - cmd->hdr.version = cpu_to_le16(I2400M_L3L4_VERSION); - - cmd->cit.hdr.type = - cpu_to_le16(I2400M_TLV_CONFIG_IDLE_TIMEOUT); - cmd->cit.hdr.length = cpu_to_le16(sizeof(cmd->cit.timeout)); - cmd->cit.timeout = cpu_to_le32(msecs); - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'set idle timeout' command: " - "%ld\n", PTR_ERR(ack_skb)); - result = PTR_ERR(ack_skb); - goto error_msg_to_dev; - } - ack = wimax_msg_data_len(ack_skb, &ack_len); - result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); - if (result < 0) { - dev_err(dev, "'set idle timeout' (0x%04x) command failed: " - "%d - %s\n", I2400M_MT_GET_STATE, result, strerr); - goto error_cmd_failed; - } - result = 0; - kfree_skb(ack_skb); -error_cmd_failed: -error_msg_to_dev: - kfree(cmd); -error_alloc: - return result; -} - - -/** - * i2400m_dev_initialize - Initialize the device once communications are ready - * - * @i2400m: device descriptor - * - * Returns: 0 if ok, < 0 errno code on error. - * - * Configures the device to work the way we like it. - * - * At the point of this call, the device is registered with the WiMAX - * and netdev stacks, firmware is uploaded and we can talk to the - * device normally. - */ -int i2400m_dev_initialize(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct i2400m_tlv_config_idle_parameters idle_params; - struct i2400m_tlv_config_idle_timeout idle_timeout; - struct i2400m_tlv_config_d2h_data_format df; - struct i2400m_tlv_config_dl_host_reorder dlhr; - const struct i2400m_tlv_hdr *args[9]; - unsigned argc = 0; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - if (i2400m_passive_mode) - goto out_passive; - /* Disable idle mode? (enabled by default) */ - if (i2400m_idle_mode_disabled) { - if (i2400m_le_v1_3(i2400m)) { - idle_params.hdr.type = - cpu_to_le16(I2400M_TLV_CONFIG_IDLE_PARAMETERS); - idle_params.hdr.length = cpu_to_le16( - sizeof(idle_params) - sizeof(idle_params.hdr)); - idle_params.idle_timeout = 0; - idle_params.idle_paging_interval = 0; - args[argc++] = &idle_params.hdr; - } else { - idle_timeout.hdr.type = - cpu_to_le16(I2400M_TLV_CONFIG_IDLE_TIMEOUT); - idle_timeout.hdr.length = cpu_to_le16( - sizeof(idle_timeout) - sizeof(idle_timeout.hdr)); - idle_timeout.timeout = 0; - args[argc++] = &idle_timeout.hdr; - } - } - if (i2400m_ge_v1_4(i2400m)) { - /* Enable extended RX data format? */ - df.hdr.type = - cpu_to_le16(I2400M_TLV_CONFIG_D2H_DATA_FORMAT); - df.hdr.length = cpu_to_le16( - sizeof(df) - sizeof(df.hdr)); - df.format = 1; - args[argc++] = &df.hdr; - - /* Enable RX data reordering? - * (switch flipped in rx.c:i2400m_rx_setup() after fw upload) */ - if (i2400m->rx_reorder) { - dlhr.hdr.type = - cpu_to_le16(I2400M_TLV_CONFIG_DL_HOST_REORDER); - dlhr.hdr.length = cpu_to_le16( - sizeof(dlhr) - sizeof(dlhr.hdr)); - dlhr.reorder = 1; - args[argc++] = &dlhr.hdr; - } - } - result = i2400m_set_init_config(i2400m, args, argc); - if (result < 0) - goto error; -out_passive: - /* - * Update state: Here it just calls a get state; parsing the - * result (System State TLV and RF Status TLV [done in the rx - * path hooks]) will set the hardware and software RF-Kill - * status. - */ - result = i2400m_cmd_get_state(i2400m); -error: - if (result < 0) - dev_err(dev, "failed to initialize the device: %d\n", result); - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; -} - - -/** - * i2400m_dev_shutdown - Shutdown a running device - * - * @i2400m: device descriptor - * - * Release resources acquired during the running of the device; in - * theory, should also tell the device to go to sleep, switch off the - * radio, all that, but at this point, in most cases (driver - * disconnection, reset handling) we can't even talk to the device. - */ -void i2400m_dev_shutdown(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} diff --git a/drivers/net/wimax/i2400m/debug-levels.h b/drivers/net/wimax/i2400m/debug-levels.h deleted file mode 100644 index 00942bb1489b..000000000000 --- a/drivers/net/wimax/i2400m/debug-levels.h +++ /dev/null @@ -1,32 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Intel Wireless WiMAX Connection 2400m - * Debug levels control file for the i2400m module - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - */ -#ifndef __debug_levels__h__ -#define __debug_levels__h__ - -/* Maximum compile and run time debug level for all submodules */ -#define D_MODULENAME i2400m -#define D_MASTER CONFIG_WIMAX_I2400M_DEBUG_LEVEL - -#include - -/* List of all the enabled modules */ -enum d_module { - D_SUBMODULE_DECLARE(control), - D_SUBMODULE_DECLARE(driver), - D_SUBMODULE_DECLARE(debugfs), - D_SUBMODULE_DECLARE(fw), - D_SUBMODULE_DECLARE(netdev), - D_SUBMODULE_DECLARE(rfkill), - D_SUBMODULE_DECLARE(rx), - D_SUBMODULE_DECLARE(sysfs), - D_SUBMODULE_DECLARE(tx), -}; - - -#endif /* #ifndef __debug_levels__h__ */ diff --git a/drivers/net/wimax/i2400m/debugfs.c b/drivers/net/wimax/i2400m/debugfs.c deleted file mode 100644 index 1c640b41ea4c..000000000000 --- a/drivers/net/wimax/i2400m/debugfs.c +++ /dev/null @@ -1,253 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Intel Wireless WiMAX Connection 2400m - * Debugfs interfaces to manipulate driver and device information - * - * Copyright (C) 2007 Intel Corporation - * Inaky Perez-Gonzalez - */ - -#include -#include -#include -#include -#include -#include -#include "i2400m.h" - - -#define D_SUBMODULE debugfs -#include "debug-levels.h" - -static -int debugfs_netdev_queue_stopped_get(void *data, u64 *val) -{ - struct i2400m *i2400m = data; - *val = netif_queue_stopped(i2400m->wimax_dev.net_dev); - return 0; -} -DEFINE_DEBUGFS_ATTRIBUTE(fops_netdev_queue_stopped, - debugfs_netdev_queue_stopped_get, - NULL, "%llu\n"); - -/* - * We don't allow partial reads of this file, as then the reader would - * get weirdly confused data as it is updated. - * - * So or you read it all or nothing; if you try to read with an offset - * != 0, we consider you are done reading. - */ -static -ssize_t i2400m_rx_stats_read(struct file *filp, char __user *buffer, - size_t count, loff_t *ppos) -{ - struct i2400m *i2400m = filp->private_data; - char buf[128]; - unsigned long flags; - - if (*ppos != 0) - return 0; - if (count < sizeof(buf)) - return -ENOSPC; - spin_lock_irqsave(&i2400m->rx_lock, flags); - snprintf(buf, sizeof(buf), "%u %u %u %u %u %u %u\n", - i2400m->rx_pl_num, i2400m->rx_pl_min, - i2400m->rx_pl_max, i2400m->rx_num, - i2400m->rx_size_acc, - i2400m->rx_size_min, i2400m->rx_size_max); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - return simple_read_from_buffer(buffer, count, ppos, buf, strlen(buf)); -} - - -/* Any write clears the stats */ -static -ssize_t i2400m_rx_stats_write(struct file *filp, const char __user *buffer, - size_t count, loff_t *ppos) -{ - struct i2400m *i2400m = filp->private_data; - unsigned long flags; - - spin_lock_irqsave(&i2400m->rx_lock, flags); - i2400m->rx_pl_num = 0; - i2400m->rx_pl_max = 0; - i2400m->rx_pl_min = UINT_MAX; - i2400m->rx_num = 0; - i2400m->rx_size_acc = 0; - i2400m->rx_size_min = UINT_MAX; - i2400m->rx_size_max = 0; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - return count; -} - -static -const struct file_operations i2400m_rx_stats_fops = { - .owner = THIS_MODULE, - .open = simple_open, - .read = i2400m_rx_stats_read, - .write = i2400m_rx_stats_write, - .llseek = default_llseek, -}; - - -/* See i2400m_rx_stats_read() */ -static -ssize_t i2400m_tx_stats_read(struct file *filp, char __user *buffer, - size_t count, loff_t *ppos) -{ - struct i2400m *i2400m = filp->private_data; - char buf[128]; - unsigned long flags; - - if (*ppos != 0) - return 0; - if (count < sizeof(buf)) - return -ENOSPC; - spin_lock_irqsave(&i2400m->tx_lock, flags); - snprintf(buf, sizeof(buf), "%u %u %u %u %u %u %u\n", - i2400m->tx_pl_num, i2400m->tx_pl_min, - i2400m->tx_pl_max, i2400m->tx_num, - i2400m->tx_size_acc, - i2400m->tx_size_min, i2400m->tx_size_max); - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - return simple_read_from_buffer(buffer, count, ppos, buf, strlen(buf)); -} - -/* Any write clears the stats */ -static -ssize_t i2400m_tx_stats_write(struct file *filp, const char __user *buffer, - size_t count, loff_t *ppos) -{ - struct i2400m *i2400m = filp->private_data; - unsigned long flags; - - spin_lock_irqsave(&i2400m->tx_lock, flags); - i2400m->tx_pl_num = 0; - i2400m->tx_pl_max = 0; - i2400m->tx_pl_min = UINT_MAX; - i2400m->tx_num = 0; - i2400m->tx_size_acc = 0; - i2400m->tx_size_min = UINT_MAX; - i2400m->tx_size_max = 0; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - return count; -} - -static -const struct file_operations i2400m_tx_stats_fops = { - .owner = THIS_MODULE, - .open = simple_open, - .read = i2400m_tx_stats_read, - .write = i2400m_tx_stats_write, - .llseek = default_llseek, -}; - - -/* Write 1 to ask the device to go into suspend */ -static -int debugfs_i2400m_suspend_set(void *data, u64 val) -{ - int result; - struct i2400m *i2400m = data; - result = i2400m_cmd_enter_powersave(i2400m); - if (result >= 0) - result = 0; - return result; -} -DEFINE_DEBUGFS_ATTRIBUTE(fops_i2400m_suspend, - NULL, debugfs_i2400m_suspend_set, - "%llu\n"); - -/* - * Reset the device - * - * Write 0 to ask the device to soft reset, 1 to cold reset, 2 to bus - * reset (as defined by enum i2400m_reset_type). - */ -static -int debugfs_i2400m_reset_set(void *data, u64 val) -{ - int result; - struct i2400m *i2400m = data; - enum i2400m_reset_type rt = val; - switch(rt) { - case I2400M_RT_WARM: - case I2400M_RT_COLD: - case I2400M_RT_BUS: - result = i2400m_reset(i2400m, rt); - if (result >= 0) - result = 0; - break; - default: - result = -EINVAL; - } - return result; -} -DEFINE_DEBUGFS_ATTRIBUTE(fops_i2400m_reset, - NULL, debugfs_i2400m_reset_set, - "%llu\n"); - -void i2400m_debugfs_add(struct i2400m *i2400m) -{ - struct dentry *dentry = i2400m->wimax_dev.debugfs_dentry; - - dentry = debugfs_create_dir("i2400m", dentry); - i2400m->debugfs_dentry = dentry; - - d_level_register_debugfs("dl_", control, dentry); - d_level_register_debugfs("dl_", driver, dentry); - d_level_register_debugfs("dl_", debugfs, dentry); - d_level_register_debugfs("dl_", fw, dentry); - d_level_register_debugfs("dl_", netdev, dentry); - d_level_register_debugfs("dl_", rfkill, dentry); - d_level_register_debugfs("dl_", rx, dentry); - d_level_register_debugfs("dl_", tx, dentry); - - debugfs_create_size_t("tx_in", 0400, dentry, &i2400m->tx_in); - debugfs_create_size_t("tx_out", 0400, dentry, &i2400m->tx_out); - debugfs_create_u32("state", 0600, dentry, &i2400m->state); - - /* - * Trace received messages from user space - * - * In order to tap the bidirectional message stream in the - * 'msg' pipe, user space can read from the 'msg' pipe; - * however, due to limitations in libnl, we can't know what - * the different applications are sending down to the kernel. - * - * So we have this hack where the driver will echo any message - * received on the msg pipe from user space [through a call to - * wimax_dev->op_msg_from_user() into - * i2400m_op_msg_from_user()] into the 'trace' pipe that this - * driver creates. - * - * So then, reading from both the 'trace' and 'msg' pipes in - * user space will provide a full dump of the traffic. - * - * Write 1 to activate, 0 to clear. - * - * It is not really very atomic, but it is also not too - * critical. - */ - debugfs_create_u8("trace_msg_from_user", 0600, dentry, - &i2400m->trace_msg_from_user); - - debugfs_create_file("netdev_queue_stopped", 0400, dentry, i2400m, - &fops_netdev_queue_stopped); - - debugfs_create_file("rx_stats", 0600, dentry, i2400m, - &i2400m_rx_stats_fops); - - debugfs_create_file("tx_stats", 0600, dentry, i2400m, - &i2400m_tx_stats_fops); - - debugfs_create_file("suspend", 0200, dentry, i2400m, - &fops_i2400m_suspend); - - debugfs_create_file("reset", 0200, dentry, i2400m, &fops_i2400m_reset); -} - -void i2400m_debugfs_rm(struct i2400m *i2400m) -{ - debugfs_remove_recursive(i2400m->debugfs_dentry); -} diff --git a/drivers/net/wimax/i2400m/driver.c b/drivers/net/wimax/i2400m/driver.c deleted file mode 100644 index ecb3fccca603..000000000000 --- a/drivers/net/wimax/i2400m/driver.c +++ /dev/null @@ -1,1002 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Intel Wireless WiMAX Connection 2400m - * Generic probe/disconnect, reset and message passing - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - * - * See i2400m.h for driver documentation. This contains helpers for - * the driver model glue [_setup()/_release()], handling device resets - * [_dev_reset_handle()], and the backends for the WiMAX stack ops - * reset [_op_reset()] and message from user [_op_msg_from_user()]. - * - * ROADMAP: - * - * i2400m_op_msg_from_user() - * i2400m_msg_to_dev() - * wimax_msg_to_user_send() - * - * i2400m_op_reset() - * i240m->bus_reset() - * - * i2400m_dev_reset_handle() - * __i2400m_dev_reset_handle() - * __i2400m_dev_stop() - * __i2400m_dev_start() - * - * i2400m_setup() - * i2400m->bus_setup() - * i2400m_bootrom_init() - * register_netdev() - * wimax_dev_add() - * i2400m_dev_start() - * __i2400m_dev_start() - * i2400m_dev_bootstrap() - * i2400m_tx_setup() - * i2400m->bus_dev_start() - * i2400m_firmware_check() - * i2400m_check_mac_addr() - * - * i2400m_release() - * i2400m_dev_stop() - * __i2400m_dev_stop() - * i2400m_dev_shutdown() - * i2400m->bus_dev_stop() - * i2400m_tx_release() - * i2400m->bus_release() - * wimax_dev_rm() - * unregister_netdev() - */ -#include "i2400m.h" -#include -#include -#include -#include -#include -#include - -#define D_SUBMODULE driver -#include "debug-levels.h" - - -static char i2400m_debug_params[128]; -module_param_string(debug, i2400m_debug_params, sizeof(i2400m_debug_params), - 0644); -MODULE_PARM_DESC(debug, - "String of space-separated NAME:VALUE pairs, where NAMEs " - "are the different debug submodules and VALUE are the " - "initial debug value to set."); - -static char i2400m_barkers_params[128]; -module_param_string(barkers, i2400m_barkers_params, - sizeof(i2400m_barkers_params), 0644); -MODULE_PARM_DESC(barkers, - "String of comma-separated 32-bit values; each is " - "recognized as the value the device sends as a reboot " - "signal; values are appended to a list--setting one value " - "as zero cleans the existing list and starts a new one."); - -/* - * WiMAX stack operation: relay a message from user space - * - * @wimax_dev: device descriptor - * @pipe_name: named pipe the message is for - * @msg_buf: pointer to the message bytes - * @msg_len: length of the buffer - * @genl_info: passed by the generic netlink layer - * - * The WiMAX stack will call this function when a message was received - * from user space. - * - * For the i2400m, this is an L3L4 message, as specified in - * include/linux/wimax/i2400m.h, and thus prefixed with a 'struct - * i2400m_l3l4_hdr'. Driver (and device) expect the messages to be - * coded in Little Endian. - * - * This function just verifies that the header declaration and the - * payload are consistent and then deals with it, either forwarding it - * to the device or procesing it locally. - * - * In the i2400m, messages are basically commands that will carry an - * ack, so we use i2400m_msg_to_dev() and then deliver the ack back to - * user space. The rx.c code might intercept the response and use it - * to update the driver's state, but then it will pass it on so it can - * be relayed back to user space. - * - * Note that asynchronous events from the device are processed and - * sent to user space in rx.c. - */ -static -int i2400m_op_msg_from_user(struct wimax_dev *wimax_dev, - const char *pipe_name, - const void *msg_buf, size_t msg_len, - const struct genl_info *genl_info) -{ - int result; - struct i2400m *i2400m = wimax_dev_to_i2400m(wimax_dev); - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - - d_fnstart(4, dev, "(wimax_dev %p [i2400m %p] msg_buf %p " - "msg_len %zu genl_info %p)\n", wimax_dev, i2400m, - msg_buf, msg_len, genl_info); - ack_skb = i2400m_msg_to_dev(i2400m, msg_buf, msg_len); - result = PTR_ERR(ack_skb); - if (IS_ERR(ack_skb)) - goto error_msg_to_dev; - result = wimax_msg_send(&i2400m->wimax_dev, ack_skb); -error_msg_to_dev: - d_fnend(4, dev, "(wimax_dev %p [i2400m %p] msg_buf %p msg_len %zu " - "genl_info %p) = %d\n", wimax_dev, i2400m, msg_buf, msg_len, - genl_info, result); - return result; -} - - -/* - * Context to wait for a reset to finalize - */ -struct i2400m_reset_ctx { - struct completion completion; - int result; -}; - - -/* - * WiMAX stack operation: reset a device - * - * @wimax_dev: device descriptor - * - * See the documentation for wimax_reset() and wimax_dev->op_reset for - * the requirements of this function. The WiMAX stack guarantees - * serialization on calls to this function. - * - * Do a warm reset on the device; if it fails, resort to a cold reset - * and return -ENODEV. On successful warm reset, we need to block - * until it is complete. - * - * The bus-driver implementation of reset takes care of falling back - * to cold reset if warm fails. - */ -static -int i2400m_op_reset(struct wimax_dev *wimax_dev) -{ - int result; - struct i2400m *i2400m = wimax_dev_to_i2400m(wimax_dev); - struct device *dev = i2400m_dev(i2400m); - struct i2400m_reset_ctx ctx = { - .completion = COMPLETION_INITIALIZER_ONSTACK(ctx.completion), - .result = 0, - }; - - d_fnstart(4, dev, "(wimax_dev %p)\n", wimax_dev); - mutex_lock(&i2400m->init_mutex); - i2400m->reset_ctx = &ctx; - mutex_unlock(&i2400m->init_mutex); - result = i2400m_reset(i2400m, I2400M_RT_WARM); - if (result < 0) - goto out; - result = wait_for_completion_timeout(&ctx.completion, 4*HZ); - if (result == 0) - result = -ETIMEDOUT; - else if (result > 0) - result = ctx.result; - /* if result < 0, pass it on */ - mutex_lock(&i2400m->init_mutex); - i2400m->reset_ctx = NULL; - mutex_unlock(&i2400m->init_mutex); -out: - d_fnend(4, dev, "(wimax_dev %p) = %d\n", wimax_dev, result); - return result; -} - - -/* - * Check the MAC address we got from boot mode is ok - * - * @i2400m: device descriptor - * - * Returns: 0 if ok, < 0 errno code on error. - */ -static -int i2400m_check_mac_addr(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *skb; - const struct i2400m_tlv_detailed_device_info *ddi; - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - skb = i2400m_get_device_info(i2400m); - if (IS_ERR(skb)) { - result = PTR_ERR(skb); - dev_err(dev, "Cannot verify MAC address, error reading: %d\n", - result); - goto error; - } - /* Extract MAC address */ - ddi = (void *) skb->data; - BUILD_BUG_ON(ETH_ALEN != sizeof(ddi->mac_address)); - d_printf(2, dev, "GET DEVICE INFO: mac addr %pM\n", - ddi->mac_address); - if (!memcmp(net_dev->perm_addr, ddi->mac_address, - sizeof(ddi->mac_address))) - goto ok; - dev_warn(dev, "warning: device reports a different MAC address " - "to that of boot mode's\n"); - dev_warn(dev, "device reports %pM\n", ddi->mac_address); - dev_warn(dev, "boot mode reported %pM\n", net_dev->perm_addr); - if (is_zero_ether_addr(ddi->mac_address)) - dev_err(dev, "device reports an invalid MAC address, " - "not updating\n"); - else { - dev_warn(dev, "updating MAC address\n"); - net_dev->addr_len = ETH_ALEN; - memcpy(net_dev->perm_addr, ddi->mac_address, ETH_ALEN); - memcpy(net_dev->dev_addr, ddi->mac_address, ETH_ALEN); - } -ok: - result = 0; - kfree_skb(skb); -error: - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; -} - - -/** - * __i2400m_dev_start - Bring up driver communication with the device - * - * @i2400m: device descriptor - * @flags: boot mode flags - * - * Returns: 0 if ok, < 0 errno code on error. - * - * Uploads firmware and brings up all the resources needed to be able - * to communicate with the device. - * - * The workqueue has to be setup early, at least before RX handling - * (it's only real user for now) so it can process reports as they - * arrive. We also want to destroy it if we retry, to make sure it is - * flushed...easier like this. - * - * TX needs to be setup before the bus-specific code (otherwise on - * shutdown, the bus-tx code could try to access it). - */ -static -int __i2400m_dev_start(struct i2400m *i2400m, enum i2400m_bri flags) -{ - int result; - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - struct net_device *net_dev = wimax_dev->net_dev; - struct device *dev = i2400m_dev(i2400m); - int times = i2400m->bus_bm_retries; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); -retry: - result = i2400m_dev_bootstrap(i2400m, flags); - if (result < 0) { - dev_err(dev, "cannot bootstrap device: %d\n", result); - goto error_bootstrap; - } - result = i2400m_tx_setup(i2400m); - if (result < 0) - goto error_tx_setup; - result = i2400m_rx_setup(i2400m); - if (result < 0) - goto error_rx_setup; - i2400m->work_queue = create_singlethread_workqueue(wimax_dev->name); - if (i2400m->work_queue == NULL) { - result = -ENOMEM; - dev_err(dev, "cannot create workqueue\n"); - goto error_create_workqueue; - } - if (i2400m->bus_dev_start) { - result = i2400m->bus_dev_start(i2400m); - if (result < 0) - goto error_bus_dev_start; - } - i2400m->ready = 1; - wmb(); /* see i2400m->ready's documentation */ - /* process pending reports from the device */ - queue_work(i2400m->work_queue, &i2400m->rx_report_ws); - result = i2400m_firmware_check(i2400m); /* fw versions ok? */ - if (result < 0) - goto error_fw_check; - /* At this point is ok to send commands to the device */ - result = i2400m_check_mac_addr(i2400m); - if (result < 0) - goto error_check_mac_addr; - result = i2400m_dev_initialize(i2400m); - if (result < 0) - goto error_dev_initialize; - - /* We don't want any additional unwanted error recovery triggered - * from any other context so if anything went wrong before we come - * here, let's keep i2400m->error_recovery untouched and leave it to - * dev_reset_handle(). See dev_reset_handle(). */ - - atomic_dec(&i2400m->error_recovery); - /* Every thing works so far, ok, now we are ready to - * take error recovery if it's required. */ - - /* At this point, reports will come for the device and set it - * to the right state if it is different than UNINITIALIZED */ - d_fnend(3, dev, "(net_dev %p [i2400m %p]) = %d\n", - net_dev, i2400m, result); - return result; - -error_dev_initialize: -error_check_mac_addr: -error_fw_check: - i2400m->ready = 0; - wmb(); /* see i2400m->ready's documentation */ - flush_workqueue(i2400m->work_queue); - if (i2400m->bus_dev_stop) - i2400m->bus_dev_stop(i2400m); -error_bus_dev_start: - destroy_workqueue(i2400m->work_queue); -error_create_workqueue: - i2400m_rx_release(i2400m); -error_rx_setup: - i2400m_tx_release(i2400m); -error_tx_setup: -error_bootstrap: - if (result == -EL3RST && times-- > 0) { - flags = I2400M_BRI_SOFT|I2400M_BRI_MAC_REINIT; - goto retry; - } - d_fnend(3, dev, "(net_dev %p [i2400m %p]) = %d\n", - net_dev, i2400m, result); - return result; -} - - -static -int i2400m_dev_start(struct i2400m *i2400m, enum i2400m_bri bm_flags) -{ - int result = 0; - mutex_lock(&i2400m->init_mutex); /* Well, start the device */ - if (i2400m->updown == 0) { - result = __i2400m_dev_start(i2400m, bm_flags); - if (result >= 0) { - i2400m->updown = 1; - i2400m->alive = 1; - wmb();/* see i2400m->updown and i2400m->alive's doc */ - } - } - mutex_unlock(&i2400m->init_mutex); - return result; -} - - -/** - * i2400m_dev_stop - Tear down driver communication with the device - * - * @i2400m: device descriptor - * - * Returns: 0 if ok, < 0 errno code on error. - * - * Releases all the resources allocated to communicate with the - * device. Note we cannot destroy the workqueue earlier as until RX is - * fully destroyed, it could still try to schedule jobs. - */ -static -void __i2400m_dev_stop(struct i2400m *i2400m) -{ - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - wimax_state_change(wimax_dev, __WIMAX_ST_QUIESCING); - i2400m_msg_to_dev_cancel_wait(i2400m, -EL3RST); - complete(&i2400m->msg_completion); - i2400m_net_wake_stop(i2400m); - i2400m_dev_shutdown(i2400m); - /* - * Make sure no report hooks are running *before* we stop the - * communication infrastructure with the device. - */ - i2400m->ready = 0; /* nobody can queue work anymore */ - wmb(); /* see i2400m->ready's documentation */ - flush_workqueue(i2400m->work_queue); - - if (i2400m->bus_dev_stop) - i2400m->bus_dev_stop(i2400m); - destroy_workqueue(i2400m->work_queue); - i2400m_rx_release(i2400m); - i2400m_tx_release(i2400m); - wimax_state_change(wimax_dev, WIMAX_ST_DOWN); - d_fnend(3, dev, "(i2400m %p) = 0\n", i2400m); -} - - -/* - * Watch out -- we only need to stop if there is a need for it. The - * device could have reset itself and failed to come up again (see - * _i2400m_dev_reset_handle()). - */ -static -void i2400m_dev_stop(struct i2400m *i2400m) -{ - mutex_lock(&i2400m->init_mutex); - if (i2400m->updown) { - __i2400m_dev_stop(i2400m); - i2400m->updown = 0; - i2400m->alive = 0; - wmb(); /* see i2400m->updown and i2400m->alive's doc */ - } - mutex_unlock(&i2400m->init_mutex); -} - - -/* - * Listen to PM events to cache the firmware before suspend/hibernation - * - * When the device comes out of suspend, it might go into reset and - * firmware has to be uploaded again. At resume, most of the times, we - * can't load firmware images from disk, so we need to cache it. - * - * i2400m_fw_cache() will allocate a kobject and attach the firmware - * to it; that way we don't have to worry too much about the fw loader - * hitting a race condition. - * - * Note: modus operandi stolen from the Orinoco driver; thx. - */ -static -int i2400m_pm_notifier(struct notifier_block *notifier, - unsigned long pm_event, - void *unused) -{ - struct i2400m *i2400m = - container_of(notifier, struct i2400m, pm_notifier); - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p pm_event %lx)\n", i2400m, pm_event); - switch (pm_event) { - case PM_HIBERNATION_PREPARE: - case PM_SUSPEND_PREPARE: - i2400m_fw_cache(i2400m); - break; - case PM_POST_RESTORE: - /* Restore from hibernation failed. We need to clean - * up in exactly the same way, so fall through. */ - case PM_POST_HIBERNATION: - case PM_POST_SUSPEND: - i2400m_fw_uncache(i2400m); - break; - - case PM_RESTORE_PREPARE: - default: - break; - } - d_fnend(3, dev, "(i2400m %p pm_event %lx) = void\n", i2400m, pm_event); - return NOTIFY_DONE; -} - - -/* - * pre-reset is called before a device is going on reset - * - * This has to be followed by a call to i2400m_post_reset(), otherwise - * bad things might happen. - */ -int i2400m_pre_reset(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - d_printf(1, dev, "pre-reset shut down\n"); - - mutex_lock(&i2400m->init_mutex); - if (i2400m->updown) { - netif_tx_disable(i2400m->wimax_dev.net_dev); - __i2400m_dev_stop(i2400m); - /* down't set updown to zero -- this way - * post_reset can restore properly */ - } - mutex_unlock(&i2400m->init_mutex); - if (i2400m->bus_release) - i2400m->bus_release(i2400m); - d_fnend(3, dev, "(i2400m %p) = 0\n", i2400m); - return 0; -} -EXPORT_SYMBOL_GPL(i2400m_pre_reset); - - -/* - * Restore device state after a reset - * - * Do the work needed after a device reset to bring it up to the same - * state as it was before the reset. - * - * NOTE: this requires i2400m->init_mutex taken - */ -int i2400m_post_reset(struct i2400m *i2400m) -{ - int result = 0; - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - d_printf(1, dev, "post-reset start\n"); - if (i2400m->bus_setup) { - result = i2400m->bus_setup(i2400m); - if (result < 0) { - dev_err(dev, "bus-specific setup failed: %d\n", - result); - goto error_bus_setup; - } - } - mutex_lock(&i2400m->init_mutex); - if (i2400m->updown) { - result = __i2400m_dev_start( - i2400m, I2400M_BRI_SOFT | I2400M_BRI_MAC_REINIT); - if (result < 0) - goto error_dev_start; - } - mutex_unlock(&i2400m->init_mutex); - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; - -error_dev_start: - if (i2400m->bus_release) - i2400m->bus_release(i2400m); - /* even if the device was up, it could not be recovered, so we - * mark it as down. */ - i2400m->updown = 0; - wmb(); /* see i2400m->updown's documentation */ - mutex_unlock(&i2400m->init_mutex); -error_bus_setup: - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; -} -EXPORT_SYMBOL_GPL(i2400m_post_reset); - - -/* - * The device has rebooted; fix up the device and the driver - * - * Tear down the driver communication with the device, reload the - * firmware and reinitialize the communication with the device. - * - * If someone calls a reset when the device's firmware is down, in - * theory we won't see it because we are not listening. However, just - * in case, leave the code to handle it. - * - * If there is a reset context, use it; this means someone is waiting - * for us to tell him when the reset operation is complete and the - * device is ready to rock again. - * - * NOTE: if we are in the process of bringing up or down the - * communication with the device [running i2400m_dev_start() or - * _stop()], don't do anything, let it fail and handle it. - * - * This function is ran always in a thread context - * - * This function gets passed, as payload to i2400m_work() a 'const - * char *' ptr with a "reason" why the reset happened (for messages). - */ -static -void __i2400m_dev_reset_handle(struct work_struct *ws) -{ - struct i2400m *i2400m = container_of(ws, struct i2400m, reset_ws); - const char *reason = i2400m->reset_reason; - struct device *dev = i2400m_dev(i2400m); - struct i2400m_reset_ctx *ctx = i2400m->reset_ctx; - int result; - - d_fnstart(3, dev, "(ws %p i2400m %p reason %s)\n", ws, i2400m, reason); - - i2400m->boot_mode = 1; - wmb(); /* Make sure i2400m_msg_to_dev() sees boot_mode */ - - result = 0; - if (mutex_trylock(&i2400m->init_mutex) == 0) { - /* We are still in i2400m_dev_start() [let it fail] or - * i2400m_dev_stop() [we are shutting down anyway, so - * ignore it] or we are resetting somewhere else. */ - dev_err(dev, "device rebooted somewhere else?\n"); - i2400m_msg_to_dev_cancel_wait(i2400m, -EL3RST); - complete(&i2400m->msg_completion); - goto out; - } - - dev_err(dev, "%s: reinitializing driver\n", reason); - rmb(); - if (i2400m->updown) { - __i2400m_dev_stop(i2400m); - i2400m->updown = 0; - wmb(); /* see i2400m->updown's documentation */ - } - - if (i2400m->alive) { - result = __i2400m_dev_start(i2400m, - I2400M_BRI_SOFT | I2400M_BRI_MAC_REINIT); - if (result < 0) { - dev_err(dev, "%s: cannot start the device: %d\n", - reason, result); - result = -EUCLEAN; - if (atomic_read(&i2400m->bus_reset_retries) - >= I2400M_BUS_RESET_RETRIES) { - result = -ENODEV; - dev_err(dev, "tried too many times to " - "reset the device, giving up\n"); - } - } - } - - if (i2400m->reset_ctx) { - ctx->result = result; - complete(&ctx->completion); - } - mutex_unlock(&i2400m->init_mutex); - if (result == -EUCLEAN) { - /* - * We come here because the reset during operational mode - * wasn't successfully done and need to proceed to a bus - * reset. For the dev_reset_handle() to be able to handle - * the reset event later properly, we restore boot_mode back - * to the state before previous reset. ie: just like we are - * issuing the bus reset for the first time - */ - i2400m->boot_mode = 0; - wmb(); - - atomic_inc(&i2400m->bus_reset_retries); - /* ops, need to clean up [w/ init_mutex not held] */ - result = i2400m_reset(i2400m, I2400M_RT_BUS); - if (result >= 0) - result = -ENODEV; - } else { - rmb(); - if (i2400m->alive) { - /* great, we expect the device state up and - * dev_start() actually brings the device state up */ - i2400m->updown = 1; - wmb(); - atomic_set(&i2400m->bus_reset_retries, 0); - } - } -out: - d_fnend(3, dev, "(ws %p i2400m %p reason %s) = void\n", - ws, i2400m, reason); -} - - -/** - * i2400m_dev_reset_handle - Handle a device's reset in a thread context - * - * Schedule a device reset handling out on a thread context, so it - * is safe to call from atomic context. We can't use the i2400m's - * queue as we are going to destroy it and reinitialize it as part of - * the driver bringup/bringup process. - * - * See __i2400m_dev_reset_handle() for details; that takes care of - * reinitializing the driver to handle the reset, calling into the - * bus-specific functions ops as needed. - */ -int i2400m_dev_reset_handle(struct i2400m *i2400m, const char *reason) -{ - i2400m->reset_reason = reason; - return schedule_work(&i2400m->reset_ws); -} -EXPORT_SYMBOL_GPL(i2400m_dev_reset_handle); - - - /* - * The actual work of error recovery. - * - * The current implementation of error recovery is to trigger a bus reset. - */ -static -void __i2400m_error_recovery(struct work_struct *ws) -{ - struct i2400m *i2400m = container_of(ws, struct i2400m, recovery_ws); - - i2400m_reset(i2400m, I2400M_RT_BUS); -} - -/* - * Schedule a work struct for error recovery. - * - * The intention of error recovery is to bring back the device to some - * known state whenever TX sees -110 (-ETIMEOUT) on copying the data to - * the device. The TX failure could mean a device bus stuck, so the current - * error recovery implementation is to trigger a bus reset to the device - * and hopefully it can bring back the device. - * - * The actual work of error recovery has to be in a thread context because - * it is kicked off in the TX thread (i2400ms->tx_workqueue) which is to be - * destroyed by the error recovery mechanism (currently a bus reset). - * - * Also, there may be already a queue of TX works that all hit - * the -ETIMEOUT error condition because the device is stuck already. - * Since bus reset is used as the error recovery mechanism and we don't - * want consecutive bus resets simply because the multiple TX works - * in the queue all hit the same device erratum, the flag "error_recovery" - * is introduced for preventing unwanted consecutive bus resets. - * - * Error recovery shall only be invoked again if previous one was completed. - * The flag error_recovery is set when error recovery mechanism is scheduled, - * and is checked when we need to schedule another error recovery. If it is - * in place already, then we shouldn't schedule another one. - */ -void i2400m_error_recovery(struct i2400m *i2400m) -{ - if (atomic_add_return(1, &i2400m->error_recovery) == 1) - schedule_work(&i2400m->recovery_ws); - else - atomic_dec(&i2400m->error_recovery); -} -EXPORT_SYMBOL_GPL(i2400m_error_recovery); - -/* - * Alloc the command and ack buffers for boot mode - * - * Get the buffers needed to deal with boot mode messages. - */ -static -int i2400m_bm_buf_alloc(struct i2400m *i2400m) -{ - i2400m->bm_cmd_buf = kzalloc(I2400M_BM_CMD_BUF_SIZE, GFP_KERNEL); - if (i2400m->bm_cmd_buf == NULL) - goto error_bm_cmd_kzalloc; - i2400m->bm_ack_buf = kzalloc(I2400M_BM_ACK_BUF_SIZE, GFP_KERNEL); - if (i2400m->bm_ack_buf == NULL) - goto error_bm_ack_buf_kzalloc; - return 0; - -error_bm_ack_buf_kzalloc: - kfree(i2400m->bm_cmd_buf); -error_bm_cmd_kzalloc: - return -ENOMEM; -} - - -/* - * Free boot mode command and ack buffers. - */ -static -void i2400m_bm_buf_free(struct i2400m *i2400m) -{ - kfree(i2400m->bm_ack_buf); - kfree(i2400m->bm_cmd_buf); -} - - -/** - * i2400m_init - Initialize a 'struct i2400m' from all zeroes - * - * This is a bus-generic API call. - */ -void i2400m_init(struct i2400m *i2400m) -{ - wimax_dev_init(&i2400m->wimax_dev); - - i2400m->boot_mode = 1; - i2400m->rx_reorder = 1; - init_waitqueue_head(&i2400m->state_wq); - - spin_lock_init(&i2400m->tx_lock); - i2400m->tx_pl_min = UINT_MAX; - i2400m->tx_size_min = UINT_MAX; - - spin_lock_init(&i2400m->rx_lock); - i2400m->rx_pl_min = UINT_MAX; - i2400m->rx_size_min = UINT_MAX; - INIT_LIST_HEAD(&i2400m->rx_reports); - INIT_WORK(&i2400m->rx_report_ws, i2400m_report_hook_work); - - mutex_init(&i2400m->msg_mutex); - init_completion(&i2400m->msg_completion); - - mutex_init(&i2400m->init_mutex); - /* wake_tx_ws is initialized in i2400m_tx_setup() */ - - INIT_WORK(&i2400m->reset_ws, __i2400m_dev_reset_handle); - INIT_WORK(&i2400m->recovery_ws, __i2400m_error_recovery); - - atomic_set(&i2400m->bus_reset_retries, 0); - - i2400m->alive = 0; - - /* initialize error_recovery to 1 for denoting we - * are not yet ready to take any error recovery */ - atomic_set(&i2400m->error_recovery, 1); -} -EXPORT_SYMBOL_GPL(i2400m_init); - - -int i2400m_reset(struct i2400m *i2400m, enum i2400m_reset_type rt) -{ - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - - /* - * Make sure we stop TXs and down the carrier before - * resetting; this is needed to avoid things like - * i2400m_wake_tx() scheduling stuff in parallel. - */ - if (net_dev->reg_state == NETREG_REGISTERED) { - netif_tx_disable(net_dev); - netif_carrier_off(net_dev); - } - return i2400m->bus_reset(i2400m, rt); -} -EXPORT_SYMBOL_GPL(i2400m_reset); - - -/** - * i2400m_setup - bus-generic setup function for the i2400m device - * - * @i2400m: device descriptor (bus-specific parts have been initialized) - * - * Returns: 0 if ok, < 0 errno code on error. - * - * Sets up basic device comunication infrastructure, boots the ROM to - * read the MAC address, registers with the WiMAX and network stacks - * and then brings up the device. - */ -int i2400m_setup(struct i2400m *i2400m, enum i2400m_bri bm_flags) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - - snprintf(wimax_dev->name, sizeof(wimax_dev->name), - "i2400m-%s:%s", dev->bus->name, dev_name(dev)); - - result = i2400m_bm_buf_alloc(i2400m); - if (result < 0) { - dev_err(dev, "cannot allocate bootmode scratch buffers\n"); - goto error_bm_buf_alloc; - } - - if (i2400m->bus_setup) { - result = i2400m->bus_setup(i2400m); - if (result < 0) { - dev_err(dev, "bus-specific setup failed: %d\n", - result); - goto error_bus_setup; - } - } - - result = i2400m_bootrom_init(i2400m, bm_flags); - if (result < 0) { - dev_err(dev, "read mac addr: bootrom init " - "failed: %d\n", result); - goto error_bootrom_init; - } - result = i2400m_read_mac_addr(i2400m); - if (result < 0) - goto error_read_mac_addr; - eth_random_addr(i2400m->src_mac_addr); - - i2400m->pm_notifier.notifier_call = i2400m_pm_notifier; - register_pm_notifier(&i2400m->pm_notifier); - - result = register_netdev(net_dev); /* Okey dokey, bring it up */ - if (result < 0) { - dev_err(dev, "cannot register i2400m network device: %d\n", - result); - goto error_register_netdev; - } - netif_carrier_off(net_dev); - - i2400m->wimax_dev.op_msg_from_user = i2400m_op_msg_from_user; - i2400m->wimax_dev.op_rfkill_sw_toggle = i2400m_op_rfkill_sw_toggle; - i2400m->wimax_dev.op_reset = i2400m_op_reset; - - result = wimax_dev_add(&i2400m->wimax_dev, net_dev); - if (result < 0) - goto error_wimax_dev_add; - - /* Now setup all that requires a registered net and wimax device. */ - result = sysfs_create_group(&net_dev->dev.kobj, &i2400m_dev_attr_group); - if (result < 0) { - dev_err(dev, "cannot setup i2400m's sysfs: %d\n", result); - goto error_sysfs_setup; - } - - i2400m_debugfs_add(i2400m); - - result = i2400m_dev_start(i2400m, bm_flags); - if (result < 0) - goto error_dev_start; - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; - -error_dev_start: - i2400m_debugfs_rm(i2400m); - sysfs_remove_group(&i2400m->wimax_dev.net_dev->dev.kobj, - &i2400m_dev_attr_group); -error_sysfs_setup: - wimax_dev_rm(&i2400m->wimax_dev); -error_wimax_dev_add: - unregister_netdev(net_dev); -error_register_netdev: - unregister_pm_notifier(&i2400m->pm_notifier); -error_read_mac_addr: -error_bootrom_init: - if (i2400m->bus_release) - i2400m->bus_release(i2400m); -error_bus_setup: - i2400m_bm_buf_free(i2400m); -error_bm_buf_alloc: - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; -} -EXPORT_SYMBOL_GPL(i2400m_setup); - - -/** - * i2400m_release - release the bus-generic driver resources - * - * Sends a disconnect message and undoes any setup done by i2400m_setup() - */ -void i2400m_release(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - netif_stop_queue(i2400m->wimax_dev.net_dev); - - i2400m_dev_stop(i2400m); - - cancel_work_sync(&i2400m->reset_ws); - cancel_work_sync(&i2400m->recovery_ws); - - i2400m_debugfs_rm(i2400m); - sysfs_remove_group(&i2400m->wimax_dev.net_dev->dev.kobj, - &i2400m_dev_attr_group); - wimax_dev_rm(&i2400m->wimax_dev); - unregister_netdev(i2400m->wimax_dev.net_dev); - unregister_pm_notifier(&i2400m->pm_notifier); - if (i2400m->bus_release) - i2400m->bus_release(i2400m); - i2400m_bm_buf_free(i2400m); - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} -EXPORT_SYMBOL_GPL(i2400m_release); - - -/* - * Debug levels control; see debug.h - */ -struct d_level D_LEVEL[] = { - D_SUBMODULE_DEFINE(control), - D_SUBMODULE_DEFINE(driver), - D_SUBMODULE_DEFINE(debugfs), - D_SUBMODULE_DEFINE(fw), - D_SUBMODULE_DEFINE(netdev), - D_SUBMODULE_DEFINE(rfkill), - D_SUBMODULE_DEFINE(rx), - D_SUBMODULE_DEFINE(sysfs), - D_SUBMODULE_DEFINE(tx), -}; -size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); - - -static -int __init i2400m_driver_init(void) -{ - d_parse_params(D_LEVEL, D_LEVEL_SIZE, i2400m_debug_params, - "i2400m.debug"); - return i2400m_barker_db_init(i2400m_barkers_params); -} -module_init(i2400m_driver_init); - -static -void __exit i2400m_driver_exit(void) -{ - i2400m_barker_db_exit(); -} -module_exit(i2400m_driver_exit); - -MODULE_AUTHOR("Intel Corporation "); -MODULE_DESCRIPTION("Intel 2400M WiMAX networking bus-generic driver"); -MODULE_LICENSE("GPL"); diff --git a/drivers/net/wimax/i2400m/fw.c b/drivers/net/wimax/i2400m/fw.c deleted file mode 100644 index 6c9a41bff2e0..000000000000 --- a/drivers/net/wimax/i2400m/fw.c +++ /dev/null @@ -1,1653 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * Firmware uploader - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * Inaky Perez-Gonzalez - * - Initial implementation - * - * - * THE PROCEDURE - * - * The 2400m and derived devices work in two modes: boot-mode or - * normal mode. In boot mode we can execute only a handful of commands - * targeted at uploading the firmware and launching it. - * - * The 2400m enters boot mode when it is first connected to the - * system, when it crashes and when you ask it to reboot. There are - * two submodes of the boot mode: signed and non-signed. Signed takes - * firmwares signed with a certain private key, non-signed takes any - * firmware. Normal hardware takes only signed firmware. - * - * On boot mode, in USB, we write to the device using the bulk out - * endpoint and read from it in the notification endpoint. - * - * Upon entrance to boot mode, the device sends (preceded with a few - * zero length packets (ZLPs) on the notification endpoint in USB) a - * reboot barker (4 le32 words with the same value). We ack it by - * sending the same barker to the device. The device acks with a - * reboot ack barker (4 le32 words with value I2400M_ACK_BARKER) and - * then is fully booted. At this point we can upload the firmware. - * - * Note that different iterations of the device and EEPROM - * configurations will send different [re]boot barkers; these are - * collected in i2400m_barker_db along with the firmware - * characteristics they require. - * - * This process is accomplished by the i2400m_bootrom_init() - * function. All the device interaction happens through the - * i2400m_bm_cmd() [boot mode command]. Special return values will - * indicate if the device did reset during the process. - * - * After this, we read the MAC address and then (if needed) - * reinitialize the device. We need to read it ahead of time because - * in the future, we might not upload the firmware until userspace - * 'ifconfig up's the device. - * - * We can then upload the firmware file. The file is composed of a BCF - * header (basic data, keys and signatures) and a list of write - * commands and payloads. Optionally more BCF headers might follow the - * main payload. We first upload the header [i2400m_dnload_init()] and - * then pass the commands and payloads verbatim to the i2400m_bm_cmd() - * function [i2400m_dnload_bcf()]. Then we tell the device to jump to - * the new firmware [i2400m_dnload_finalize()]. - * - * Once firmware is uploaded, we are good to go :) - * - * When we don't know in which mode we are, we first try by sending a - * warm reset request that will take us to boot-mode. If we time out - * waiting for a reboot barker, that means maybe we are already in - * boot mode, so we send a reboot barker. - * - * COMMAND EXECUTION - * - * This code (and process) is single threaded; for executing commands, - * we post a URB to the notification endpoint, post the command, wait - * for data on the notification buffer. We don't need to worry about - * others as we know we are the only ones in there. - * - * BACKEND IMPLEMENTATION - * - * This code is bus-generic; the bus-specific driver provides back end - * implementations to send a boot mode command to the device and to - * read an acknolwedgement from it (or an asynchronous notification) - * from it. - * - * FIRMWARE LOADING - * - * Note that in some cases, we can't just load a firmware file (for - * example, when resuming). For that, we might cache the firmware - * file. Thus, when doing the bootstrap, if there is a cache firmware - * file, it is used; if not, loading from disk is attempted. - * - * ROADMAP - * - * i2400m_barker_db_init Called by i2400m_driver_init() - * i2400m_barker_db_add - * - * i2400m_barker_db_exit Called by i2400m_driver_exit() - * - * i2400m_dev_bootstrap Called by __i2400m_dev_start() - * request_firmware - * i2400m_fw_bootstrap - * i2400m_fw_check - * i2400m_fw_hdr_check - * i2400m_fw_dnload - * release_firmware - * - * i2400m_fw_dnload - * i2400m_bootrom_init - * i2400m_bm_cmd - * i2400m_reset - * i2400m_dnload_init - * i2400m_dnload_init_signed - * i2400m_dnload_init_nonsigned - * i2400m_download_chunk - * i2400m_bm_cmd - * i2400m_dnload_bcf - * i2400m_bm_cmd - * i2400m_dnload_finalize - * i2400m_bm_cmd - * - * i2400m_bm_cmd - * i2400m->bus_bm_cmd_send() - * i2400m->bus_bm_wait_for_ack - * __i2400m_bm_ack_verify - * i2400m_is_boot_barker - * - * i2400m_bm_cmd_prepare Used by bus-drivers to prep - * commands before sending - * - * i2400m_pm_notifier Called on Power Management events - * i2400m_fw_cache - * i2400m_fw_uncache - */ -#include -#include -#include -#include -#include -#include "i2400m.h" - - -#define D_SUBMODULE fw -#include "debug-levels.h" - - -static const __le32 i2400m_ACK_BARKER[4] = { - cpu_to_le32(I2400M_ACK_BARKER), - cpu_to_le32(I2400M_ACK_BARKER), - cpu_to_le32(I2400M_ACK_BARKER), - cpu_to_le32(I2400M_ACK_BARKER) -}; - - -/** - * Prepare a boot-mode command for delivery - * - * @cmd: pointer to bootrom header to prepare - * - * Computes checksum if so needed. After calling this function, DO NOT - * modify the command or header as the checksum won't work anymore. - * - * We do it from here because some times we cannot do it in the - * original context the command was sent (it is a const), so when we - * copy it to our staging buffer, we add the checksum there. - */ -void i2400m_bm_cmd_prepare(struct i2400m_bootrom_header *cmd) -{ - if (i2400m_brh_get_use_checksum(cmd)) { - int i; - u32 checksum = 0; - const u32 *checksum_ptr = (void *) cmd->payload; - for (i = 0; i < cmd->data_size / 4; i++) - checksum += cpu_to_le32(*checksum_ptr++); - checksum += cmd->command + cmd->target_addr + cmd->data_size; - cmd->block_checksum = cpu_to_le32(checksum); - } -} -EXPORT_SYMBOL_GPL(i2400m_bm_cmd_prepare); - - -/* - * Database of known barkers. - * - * A barker is what the device sends indicating he is ready to be - * bootloaded. Different versions of the device will send different - * barkers. Depending on the barker, it might mean the device wants - * some kind of firmware or the other. - */ -static struct i2400m_barker_db { - __le32 data[4]; -} *i2400m_barker_db; -static size_t i2400m_barker_db_used, i2400m_barker_db_size; - - -static -int i2400m_zrealloc_2x(void **ptr, size_t *_count, size_t el_size, - gfp_t gfp_flags) -{ - size_t old_count = *_count, - new_count = old_count ? 2 * old_count : 2, - old_size = el_size * old_count, - new_size = el_size * new_count; - void *nptr = krealloc(*ptr, new_size, gfp_flags); - if (nptr) { - /* zero the other half or the whole thing if old_count - * was zero */ - if (old_size == 0) - memset(nptr, 0, new_size); - else - memset(nptr + old_size, 0, old_size); - *_count = new_count; - *ptr = nptr; - return 0; - } else - return -ENOMEM; -} - - -/* - * Add a barker to the database - * - * This cannot used outside of this module and only at at module_init - * time. This is to avoid the need to do locking. - */ -static -int i2400m_barker_db_add(u32 barker_id) -{ - int result; - - struct i2400m_barker_db *barker; - if (i2400m_barker_db_used >= i2400m_barker_db_size) { - result = i2400m_zrealloc_2x( - (void **) &i2400m_barker_db, &i2400m_barker_db_size, - sizeof(i2400m_barker_db[0]), GFP_KERNEL); - if (result < 0) - return result; - } - barker = i2400m_barker_db + i2400m_barker_db_used++; - barker->data[0] = le32_to_cpu(barker_id); - barker->data[1] = le32_to_cpu(barker_id); - barker->data[2] = le32_to_cpu(barker_id); - barker->data[3] = le32_to_cpu(barker_id); - return 0; -} - - -void i2400m_barker_db_exit(void) -{ - kfree(i2400m_barker_db); - i2400m_barker_db = NULL; - i2400m_barker_db_size = 0; - i2400m_barker_db_used = 0; -} - - -/* - * Helper function to add all the known stable barkers to the barker - * database. - */ -static -int i2400m_barker_db_known_barkers(void) -{ - int result; - - result = i2400m_barker_db_add(I2400M_NBOOT_BARKER); - if (result < 0) - goto error_add; - result = i2400m_barker_db_add(I2400M_SBOOT_BARKER); - if (result < 0) - goto error_add; - result = i2400m_barker_db_add(I2400M_SBOOT_BARKER_6050); - if (result < 0) - goto error_add; -error_add: - return result; -} - - -/* - * Initialize the barker database - * - * This can only be used from the module_init function for this - * module; this is to avoid the need to do locking. - * - * @options: command line argument with extra barkers to - * recognize. This is a comma-separated list of 32-bit hex - * numbers. They are appended to the existing list. Setting 0 - * cleans the existing list and starts a new one. - */ -int i2400m_barker_db_init(const char *_options) -{ - int result; - char *options = NULL, *options_orig, *token; - - i2400m_barker_db = NULL; - i2400m_barker_db_size = 0; - i2400m_barker_db_used = 0; - - result = i2400m_barker_db_known_barkers(); - if (result < 0) - goto error_add; - /* parse command line options from i2400m.barkers */ - if (_options != NULL) { - unsigned barker; - - options_orig = kstrdup(_options, GFP_KERNEL); - if (options_orig == NULL) { - result = -ENOMEM; - goto error_parse; - } - options = options_orig; - - while ((token = strsep(&options, ",")) != NULL) { - if (*token == '\0') /* eat joint commas */ - continue; - if (sscanf(token, "%x", &barker) != 1 - || barker > 0xffffffff) { - printk(KERN_ERR "%s: can't recognize " - "i2400m.barkers value '%s' as " - "a 32-bit number\n", - __func__, token); - result = -EINVAL; - goto error_parse; - } - if (barker == 0) { - /* clean list and start new */ - i2400m_barker_db_exit(); - continue; - } - result = i2400m_barker_db_add(barker); - if (result < 0) - goto error_parse_add; - } - kfree(options_orig); - } - return 0; - -error_parse_add: -error_parse: - kfree(options_orig); -error_add: - kfree(i2400m_barker_db); - return result; -} - - -/* - * Recognize a boot barker - * - * @buf: buffer where the boot barker. - * @buf_size: size of the buffer (has to be 16 bytes). It is passed - * here so the function can check it for the caller. - * - * Note that as a side effect, upon identifying the obtained boot - * barker, this function will set i2400m->barker to point to the right - * barker database entry. Subsequent calls to the function will result - * in verifying that the same type of boot barker is returned when the - * device [re]boots (as long as the same device instance is used). - * - * Return: 0 if @buf matches a known boot barker. -ENOENT if the - * buffer in @buf doesn't match any boot barker in the database or - * -EILSEQ if the buffer doesn't have the right size. - */ -int i2400m_is_boot_barker(struct i2400m *i2400m, - const void *buf, size_t buf_size) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct i2400m_barker_db *barker; - int i; - - result = -ENOENT; - if (buf_size != sizeof(i2400m_barker_db[i].data)) - return result; - - /* Short circuit if we have already discovered the barker - * associated with the device. */ - if (i2400m->barker && - !memcmp(buf, i2400m->barker, sizeof(i2400m->barker->data))) - return 0; - - for (i = 0; i < i2400m_barker_db_used; i++) { - barker = &i2400m_barker_db[i]; - BUILD_BUG_ON(sizeof(barker->data) != 16); - if (memcmp(buf, barker->data, sizeof(barker->data))) - continue; - - if (i2400m->barker == NULL) { - i2400m->barker = barker; - d_printf(1, dev, "boot barker set to #%u/%08x\n", - i, le32_to_cpu(barker->data[0])); - if (barker->data[0] == le32_to_cpu(I2400M_NBOOT_BARKER)) - i2400m->sboot = 0; - else - i2400m->sboot = 1; - } else if (i2400m->barker != barker) { - dev_err(dev, "HW inconsistency: device " - "reports a different boot barker " - "than set (from %08x to %08x)\n", - le32_to_cpu(i2400m->barker->data[0]), - le32_to_cpu(barker->data[0])); - result = -EIO; - } else - d_printf(2, dev, "boot barker confirmed #%u/%08x\n", - i, le32_to_cpu(barker->data[0])); - result = 0; - break; - } - return result; -} -EXPORT_SYMBOL_GPL(i2400m_is_boot_barker); - - -/* - * Verify the ack data received - * - * Given a reply to a boot mode command, chew it and verify everything - * is ok. - * - * @opcode: opcode which generated this ack. For error messages. - * @ack: pointer to ack data we received - * @ack_size: size of that data buffer - * @flags: I2400M_BM_CMD_* flags we called the command with. - * - * Way too long function -- maybe it should be further split - */ -static -ssize_t __i2400m_bm_ack_verify(struct i2400m *i2400m, int opcode, - struct i2400m_bootrom_header *ack, - size_t ack_size, int flags) -{ - ssize_t result = -ENOMEM; - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(8, dev, "(i2400m %p opcode %d ack %p size %zu)\n", - i2400m, opcode, ack, ack_size); - if (ack_size < sizeof(*ack)) { - result = -EIO; - dev_err(dev, "boot-mode cmd %d: HW BUG? notification didn't " - "return enough data (%zu bytes vs %zu expected)\n", - opcode, ack_size, sizeof(*ack)); - goto error_ack_short; - } - result = i2400m_is_boot_barker(i2400m, ack, ack_size); - if (result >= 0) { - result = -ERESTARTSYS; - d_printf(6, dev, "boot-mode cmd %d: HW boot barker\n", opcode); - goto error_reboot; - } - if (ack_size == sizeof(i2400m_ACK_BARKER) - && memcmp(ack, i2400m_ACK_BARKER, sizeof(*ack)) == 0) { - result = -EISCONN; - d_printf(3, dev, "boot-mode cmd %d: HW reboot ack barker\n", - opcode); - goto error_reboot_ack; - } - result = 0; - if (flags & I2400M_BM_CMD_RAW) - goto out_raw; - ack->data_size = le32_to_cpu(ack->data_size); - ack->target_addr = le32_to_cpu(ack->target_addr); - ack->block_checksum = le32_to_cpu(ack->block_checksum); - d_printf(5, dev, "boot-mode cmd %d: notification for opcode %u " - "response %u csum %u rr %u da %u\n", - opcode, i2400m_brh_get_opcode(ack), - i2400m_brh_get_response(ack), - i2400m_brh_get_use_checksum(ack), - i2400m_brh_get_response_required(ack), - i2400m_brh_get_direct_access(ack)); - result = -EIO; - if (i2400m_brh_get_signature(ack) != 0xcbbc) { - dev_err(dev, "boot-mode cmd %d: HW BUG? wrong signature " - "0x%04x\n", opcode, i2400m_brh_get_signature(ack)); - goto error_ack_signature; - } - if (opcode != -1 && opcode != i2400m_brh_get_opcode(ack)) { - dev_err(dev, "boot-mode cmd %d: HW BUG? " - "received response for opcode %u, expected %u\n", - opcode, i2400m_brh_get_opcode(ack), opcode); - goto error_ack_opcode; - } - if (i2400m_brh_get_response(ack) != 0) { /* failed? */ - dev_err(dev, "boot-mode cmd %d: error; hw response %u\n", - opcode, i2400m_brh_get_response(ack)); - goto error_ack_failed; - } - if (ack_size < ack->data_size + sizeof(*ack)) { - dev_err(dev, "boot-mode cmd %d: SW BUG " - "driver provided only %zu bytes for %zu bytes " - "of data\n", opcode, ack_size, - (size_t) le32_to_cpu(ack->data_size) + sizeof(*ack)); - goto error_ack_short_buffer; - } - result = ack_size; - /* Don't you love this stack of empty targets? Well, I don't - * either, but it helps track exactly who comes in here and - * why :) */ -error_ack_short_buffer: -error_ack_failed: -error_ack_opcode: -error_ack_signature: -out_raw: -error_reboot_ack: -error_reboot: -error_ack_short: - d_fnend(8, dev, "(i2400m %p opcode %d ack %p size %zu) = %d\n", - i2400m, opcode, ack, ack_size, (int) result); - return result; -} - - -/** - * i2400m_bm_cmd - Execute a boot mode command - * - * @cmd: buffer containing the command data (pointing at the header). - * This data can be ANYWHERE (for USB, we will copy it to an - * specific buffer). Make sure everything is in proper little - * endian. - * - * A raw buffer can be also sent, just cast it and set flags to - * I2400M_BM_CMD_RAW. - * - * This function will generate a checksum for you if the - * checksum bit in the command is set (unless I2400M_BM_CMD_RAW - * is set). - * - * You can use the i2400m->bm_cmd_buf to stage your commands and - * send them. - * - * If NULL, no command is sent (we just wait for an ack). - * - * @cmd_size: size of the command. Will be auto padded to the - * bus-specific drivers padding requirements. - * - * @ack: buffer where to place the acknowledgement. If it is a regular - * command response, all fields will be returned with the right, - * native endianess. - * - * You *cannot* use i2400m->bm_ack_buf for this buffer. - * - * @ack_size: size of @ack, 16 aligned; you need to provide at least - * sizeof(*ack) bytes and then enough to contain the return data - * from the command - * - * @flags: see I2400M_BM_CMD_* above. - * - * @returns: bytes received by the notification; if < 0, an errno code - * denoting an error or: - * - * -ERESTARTSYS The device has rebooted - * - * Executes a boot-mode command and waits for a response, doing basic - * validation on it; if a zero length response is received, it retries - * waiting for a response until a non-zero one is received (timing out - * after %I2400M_BOOT_RETRIES retries). - */ -static -ssize_t i2400m_bm_cmd(struct i2400m *i2400m, - const struct i2400m_bootrom_header *cmd, size_t cmd_size, - struct i2400m_bootrom_header *ack, size_t ack_size, - int flags) -{ - ssize_t result = -ENOMEM, rx_bytes; - struct device *dev = i2400m_dev(i2400m); - int opcode = cmd == NULL ? -1 : i2400m_brh_get_opcode(cmd); - - d_fnstart(6, dev, "(i2400m %p cmd %p size %zu ack %p size %zu)\n", - i2400m, cmd, cmd_size, ack, ack_size); - BUG_ON(ack_size < sizeof(*ack)); - BUG_ON(i2400m->boot_mode == 0); - - if (cmd != NULL) { /* send the command */ - result = i2400m->bus_bm_cmd_send(i2400m, cmd, cmd_size, flags); - if (result < 0) - goto error_cmd_send; - if ((flags & I2400M_BM_CMD_RAW) == 0) - d_printf(5, dev, - "boot-mode cmd %d csum %u rr %u da %u: " - "addr 0x%04x size %u block csum 0x%04x\n", - opcode, i2400m_brh_get_use_checksum(cmd), - i2400m_brh_get_response_required(cmd), - i2400m_brh_get_direct_access(cmd), - cmd->target_addr, cmd->data_size, - cmd->block_checksum); - } - result = i2400m->bus_bm_wait_for_ack(i2400m, ack, ack_size); - if (result < 0) { - dev_err(dev, "boot-mode cmd %d: error waiting for an ack: %d\n", - opcode, (int) result); /* bah, %zd doesn't work */ - goto error_wait_for_ack; - } - rx_bytes = result; - /* verify the ack and read more if necessary [result is the - * final amount of bytes we get in the ack] */ - result = __i2400m_bm_ack_verify(i2400m, opcode, ack, ack_size, flags); - if (result < 0) - goto error_bad_ack; - /* Don't you love this stack of empty targets? Well, I don't - * either, but it helps track exactly who comes in here and - * why :) */ - result = rx_bytes; -error_bad_ack: -error_wait_for_ack: -error_cmd_send: - d_fnend(6, dev, "(i2400m %p cmd %p size %zu ack %p size %zu) = %d\n", - i2400m, cmd, cmd_size, ack, ack_size, (int) result); - return result; -} - - -/** - * i2400m_download_chunk - write a single chunk of data to the device's memory - * - * @i2400m: device descriptor - * @buf: the buffer to write - * @buf_len: length of the buffer to write - * @addr: address in the device memory space - * @direct: bootrom write mode - * @do_csum: should a checksum validation be performed - */ -static int i2400m_download_chunk(struct i2400m *i2400m, const void *chunk, - size_t __chunk_len, unsigned long addr, - unsigned int direct, unsigned int do_csum) -{ - int ret; - size_t chunk_len = ALIGN(__chunk_len, I2400M_PL_ALIGN); - struct device *dev = i2400m_dev(i2400m); - struct { - struct i2400m_bootrom_header cmd; - u8 cmd_payload[]; - } __packed *buf; - struct i2400m_bootrom_header ack; - - d_fnstart(5, dev, "(i2400m %p chunk %p __chunk_len %zu addr 0x%08lx " - "direct %u do_csum %u)\n", i2400m, chunk, __chunk_len, - addr, direct, do_csum); - buf = i2400m->bm_cmd_buf; - memcpy(buf->cmd_payload, chunk, __chunk_len); - memset(buf->cmd_payload + __chunk_len, 0xad, chunk_len - __chunk_len); - - buf->cmd.command = i2400m_brh_command(I2400M_BRH_WRITE, - __chunk_len & 0x3 ? 0 : do_csum, - __chunk_len & 0xf ? 0 : direct); - buf->cmd.target_addr = cpu_to_le32(addr); - buf->cmd.data_size = cpu_to_le32(__chunk_len); - ret = i2400m_bm_cmd(i2400m, &buf->cmd, sizeof(buf->cmd) + chunk_len, - &ack, sizeof(ack), 0); - if (ret >= 0) - ret = 0; - d_fnend(5, dev, "(i2400m %p chunk %p __chunk_len %zu addr 0x%08lx " - "direct %u do_csum %u) = %d\n", i2400m, chunk, __chunk_len, - addr, direct, do_csum, ret); - return ret; -} - - -/* - * Download a BCF file's sections to the device - * - * @i2400m: device descriptor - * @bcf: pointer to firmware data (first header followed by the - * payloads). Assumed verified and consistent. - * @bcf_len: length (in bytes) of the @bcf buffer. - * - * Returns: < 0 errno code on error or the offset to the jump instruction. - * - * Given a BCF file, downloads each section (a command and a payload) - * to the device's address space. Actually, it just executes each - * command i the BCF file. - * - * The section size has to be aligned to 4 bytes AND the padding has - * to be taken from the firmware file, as the signature takes it into - * account. - */ -static -ssize_t i2400m_dnload_bcf(struct i2400m *i2400m, - const struct i2400m_bcf_hdr *bcf, size_t bcf_len) -{ - ssize_t ret; - struct device *dev = i2400m_dev(i2400m); - size_t offset, /* iterator offset */ - data_size, /* Size of the data payload */ - section_size, /* Size of the whole section (cmd + payload) */ - section = 1; - const struct i2400m_bootrom_header *bh; - struct i2400m_bootrom_header ack; - - d_fnstart(3, dev, "(i2400m %p bcf %p bcf_len %zu)\n", - i2400m, bcf, bcf_len); - /* Iterate over the command blocks in the BCF file that start - * after the header */ - offset = le32_to_cpu(bcf->header_len) * sizeof(u32); - while (1) { /* start sending the file */ - bh = (void *) bcf + offset; - data_size = le32_to_cpu(bh->data_size); - section_size = ALIGN(sizeof(*bh) + data_size, 4); - d_printf(7, dev, - "downloading section #%zu (@%zu %zu B) to 0x%08x\n", - section, offset, sizeof(*bh) + data_size, - le32_to_cpu(bh->target_addr)); - /* - * We look for JUMP cmd from the bootmode header, - * either I2400M_BRH_SIGNED_JUMP for secure boot - * or I2400M_BRH_JUMP for unsecure boot, the last chunk - * should be the bootmode header with JUMP cmd. - */ - if (i2400m_brh_get_opcode(bh) == I2400M_BRH_SIGNED_JUMP || - i2400m_brh_get_opcode(bh) == I2400M_BRH_JUMP) { - d_printf(5, dev, "jump found @%zu\n", offset); - break; - } - if (offset + section_size > bcf_len) { - dev_err(dev, "fw %s: bad section #%zu, " - "end (@%zu) beyond EOF (@%zu)\n", - i2400m->fw_name, section, - offset + section_size, bcf_len); - ret = -EINVAL; - goto error_section_beyond_eof; - } - __i2400m_msleep(20); - ret = i2400m_bm_cmd(i2400m, bh, section_size, - &ack, sizeof(ack), I2400M_BM_CMD_RAW); - if (ret < 0) { - dev_err(dev, "fw %s: section #%zu (@%zu %zu B) " - "failed %d\n", i2400m->fw_name, section, - offset, sizeof(*bh) + data_size, (int) ret); - goto error_send; - } - offset += section_size; - section++; - } - ret = offset; -error_section_beyond_eof: -error_send: - d_fnend(3, dev, "(i2400m %p bcf %p bcf_len %zu) = %d\n", - i2400m, bcf, bcf_len, (int) ret); - return ret; -} - - -/* - * Indicate if the device emitted a reboot barker that indicates - * "signed boot" - */ -static -unsigned i2400m_boot_is_signed(struct i2400m *i2400m) -{ - return likely(i2400m->sboot); -} - - -/* - * Do the final steps of uploading firmware - * - * @bcf_hdr: BCF header we are actually using - * @bcf: pointer to the firmware image (which matches the first header - * that is followed by the actual payloads). - * @offset: [byte] offset into @bcf for the command we need to send. - * - * Depending on the boot mode (signed vs non-signed), different - * actions need to be taken. - */ -static -int i2400m_dnload_finalize(struct i2400m *i2400m, - const struct i2400m_bcf_hdr *bcf_hdr, - const struct i2400m_bcf_hdr *bcf, size_t offset) -{ - int ret = 0; - struct device *dev = i2400m_dev(i2400m); - struct i2400m_bootrom_header *cmd, ack; - struct { - struct i2400m_bootrom_header cmd; - u8 cmd_pl[0]; - } __packed *cmd_buf; - size_t signature_block_offset, signature_block_size; - - d_fnstart(3, dev, "offset %zu\n", offset); - cmd = (void *) bcf + offset; - if (i2400m_boot_is_signed(i2400m) == 0) { - struct i2400m_bootrom_header jump_ack; - d_printf(1, dev, "unsecure boot, jumping to 0x%08x\n", - le32_to_cpu(cmd->target_addr)); - cmd_buf = i2400m->bm_cmd_buf; - memcpy(&cmd_buf->cmd, cmd, sizeof(*cmd)); - cmd = &cmd_buf->cmd; - /* now cmd points to the actual bootrom_header in cmd_buf */ - i2400m_brh_set_opcode(cmd, I2400M_BRH_JUMP); - cmd->data_size = 0; - ret = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), - &jump_ack, sizeof(jump_ack), 0); - } else { - d_printf(1, dev, "secure boot, jumping to 0x%08x\n", - le32_to_cpu(cmd->target_addr)); - cmd_buf = i2400m->bm_cmd_buf; - memcpy(&cmd_buf->cmd, cmd, sizeof(*cmd)); - signature_block_offset = - sizeof(*bcf_hdr) - + le32_to_cpu(bcf_hdr->key_size) * sizeof(u32) - + le32_to_cpu(bcf_hdr->exponent_size) * sizeof(u32); - signature_block_size = - le32_to_cpu(bcf_hdr->modulus_size) * sizeof(u32); - memcpy(cmd_buf->cmd_pl, - (void *) bcf_hdr + signature_block_offset, - signature_block_size); - ret = i2400m_bm_cmd(i2400m, &cmd_buf->cmd, - sizeof(cmd_buf->cmd) + signature_block_size, - &ack, sizeof(ack), I2400M_BM_CMD_RAW); - } - d_fnend(3, dev, "returning %d\n", ret); - return ret; -} - - -/** - * i2400m_bootrom_init - Reboots a powered device into boot mode - * - * @i2400m: device descriptor - * @flags: - * I2400M_BRI_SOFT: a reboot barker has been seen - * already, so don't wait for it. - * - * I2400M_BRI_NO_REBOOT: Don't send a reboot command, but wait - * for a reboot barker notification. This is a one shot; if - * the state machine needs to send a reboot command it will. - * - * Returns: - * - * < 0 errno code on error, 0 if ok. - * - * Description: - * - * Tries hard enough to put the device in boot-mode. There are two - * main phases to this: - * - * a. (1) send a reboot command and (2) get a reboot barker - * - * b. (1) echo/ack the reboot sending the reboot barker back and (2) - * getting an ack barker in return - * - * We want to skip (a) in some cases [soft]. The state machine is - * horrible, but it is basically: on each phase, send what has to be - * sent (if any), wait for the answer and act on the answer. We might - * have to backtrack and retry, so we keep a max tries counter for - * that. - * - * It sucks because we don't know ahead of time which is going to be - * the reboot barker (the device might send different ones depending - * on its EEPROM config) and once the device reboots and waits for the - * echo/ack reboot barker being sent back, it doesn't understand - * anything else. So we can be left at the point where we don't know - * what to send to it -- cold reset and bus reset seem to have little - * effect. So the function iterates (in this case) through all the - * known barkers and tries them all until an ACK is - * received. Otherwise, it gives up. - * - * If we get a timeout after sending a warm reset, we do it again. - */ -int i2400m_bootrom_init(struct i2400m *i2400m, enum i2400m_bri flags) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct i2400m_bootrom_header *cmd; - struct i2400m_bootrom_header ack; - int count = i2400m->bus_bm_retries; - int ack_timeout_cnt = 1; - unsigned i; - - BUILD_BUG_ON(sizeof(*cmd) != sizeof(i2400m_barker_db[0].data)); - BUILD_BUG_ON(sizeof(ack) != sizeof(i2400m_ACK_BARKER)); - - d_fnstart(4, dev, "(i2400m %p flags 0x%08x)\n", i2400m, flags); - result = -ENOMEM; - cmd = i2400m->bm_cmd_buf; - if (flags & I2400M_BRI_SOFT) - goto do_reboot_ack; -do_reboot: - ack_timeout_cnt = 1; - if (--count < 0) - goto error_timeout; - d_printf(4, dev, "device reboot: reboot command [%d # left]\n", - count); - if ((flags & I2400M_BRI_NO_REBOOT) == 0) - i2400m_reset(i2400m, I2400M_RT_WARM); - result = i2400m_bm_cmd(i2400m, NULL, 0, &ack, sizeof(ack), - I2400M_BM_CMD_RAW); - flags &= ~I2400M_BRI_NO_REBOOT; - switch (result) { - case -ERESTARTSYS: - /* - * at this point, i2400m_bm_cmd(), through - * __i2400m_bm_ack_process(), has updated - * i2400m->barker and we are good to go. - */ - d_printf(4, dev, "device reboot: got reboot barker\n"); - break; - case -EISCONN: /* we don't know how it got here...but we follow it */ - d_printf(4, dev, "device reboot: got ack barker - whatever\n"); - goto do_reboot; - case -ETIMEDOUT: - /* - * Device has timed out, we might be in boot mode - * already and expecting an ack; if we don't know what - * the barker is, we just send them all. Cold reset - * and bus reset don't work. Beats me. - */ - if (i2400m->barker != NULL) { - dev_err(dev, "device boot: reboot barker timed out, " - "trying (set) %08x echo/ack\n", - le32_to_cpu(i2400m->barker->data[0])); - goto do_reboot_ack; - } - for (i = 0; i < i2400m_barker_db_used; i++) { - struct i2400m_barker_db *barker = &i2400m_barker_db[i]; - memcpy(cmd, barker->data, sizeof(barker->data)); - result = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), - &ack, sizeof(ack), - I2400M_BM_CMD_RAW); - if (result == -EISCONN) { - dev_warn(dev, "device boot: got ack barker " - "after sending echo/ack barker " - "#%d/%08x; rebooting j.i.c.\n", - i, le32_to_cpu(barker->data[0])); - flags &= ~I2400M_BRI_NO_REBOOT; - goto do_reboot; - } - } - dev_err(dev, "device boot: tried all the echo/acks, could " - "not get device to respond; giving up"); - result = -ESHUTDOWN; - case -EPROTO: - case -ESHUTDOWN: /* dev is gone */ - case -EINTR: /* user cancelled */ - goto error_dev_gone; - default: - dev_err(dev, "device reboot: error %d while waiting " - "for reboot barker - rebooting\n", result); - d_dump(1, dev, &ack, result); - goto do_reboot; - } - /* At this point we ack back with 4 REBOOT barkers and expect - * 4 ACK barkers. This is ugly, as we send a raw command -- - * hence the cast. _bm_cmd() will catch the reboot ack - * notification and report it as -EISCONN. */ -do_reboot_ack: - d_printf(4, dev, "device reboot ack: sending ack [%d # left]\n", count); - memcpy(cmd, i2400m->barker->data, sizeof(i2400m->barker->data)); - result = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), - &ack, sizeof(ack), I2400M_BM_CMD_RAW); - switch (result) { - case -ERESTARTSYS: - d_printf(4, dev, "reboot ack: got reboot barker - retrying\n"); - if (--count < 0) - goto error_timeout; - goto do_reboot_ack; - case -EISCONN: - d_printf(4, dev, "reboot ack: got ack barker - good\n"); - break; - case -ETIMEDOUT: /* no response, maybe it is the other type? */ - if (ack_timeout_cnt-- < 0) { - d_printf(4, dev, "reboot ack timedout: retrying\n"); - goto do_reboot_ack; - } else { - dev_err(dev, "reboot ack timedout too long: " - "trying reboot\n"); - goto do_reboot; - } - break; - case -EPROTO: - case -ESHUTDOWN: /* dev is gone */ - goto error_dev_gone; - default: - dev_err(dev, "device reboot ack: error %d while waiting for " - "reboot ack barker - rebooting\n", result); - goto do_reboot; - } - d_printf(2, dev, "device reboot ack: got ack barker - boot done\n"); - result = 0; -exit_timeout: -error_dev_gone: - d_fnend(4, dev, "(i2400m %p flags 0x%08x) = %d\n", - i2400m, flags, result); - return result; - -error_timeout: - dev_err(dev, "Timed out waiting for reboot ack\n"); - result = -ETIMEDOUT; - goto exit_timeout; -} - - -/* - * Read the MAC addr - * - * The position this function reads is fixed in device memory and - * always available, even without firmware. - * - * Note we specify we want to read only six bytes, but provide space - * for 16, as we always get it rounded up. - */ -int i2400m_read_mac_addr(struct i2400m *i2400m) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - struct i2400m_bootrom_header *cmd; - struct { - struct i2400m_bootrom_header ack; - u8 ack_pl[16]; - } __packed ack_buf; - - d_fnstart(5, dev, "(i2400m %p)\n", i2400m); - cmd = i2400m->bm_cmd_buf; - cmd->command = i2400m_brh_command(I2400M_BRH_READ, 0, 1); - cmd->target_addr = cpu_to_le32(0x00203fe8); - cmd->data_size = cpu_to_le32(6); - result = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), - &ack_buf.ack, sizeof(ack_buf), 0); - if (result < 0) { - dev_err(dev, "BM: read mac addr failed: %d\n", result); - goto error_read_mac; - } - d_printf(2, dev, "mac addr is %pM\n", ack_buf.ack_pl); - if (i2400m->bus_bm_mac_addr_impaired == 1) { - ack_buf.ack_pl[0] = 0x00; - ack_buf.ack_pl[1] = 0x16; - ack_buf.ack_pl[2] = 0xd3; - get_random_bytes(&ack_buf.ack_pl[3], 3); - dev_err(dev, "BM is MAC addr impaired, faking MAC addr to " - "mac addr is %pM\n", ack_buf.ack_pl); - result = 0; - } - net_dev->addr_len = ETH_ALEN; - memcpy(net_dev->dev_addr, ack_buf.ack_pl, ETH_ALEN); -error_read_mac: - d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; -} - - -/* - * Initialize a non signed boot - * - * This implies sending some magic values to the device's memory. Note - * we convert the values to little endian in the same array - * declaration. - */ -static -int i2400m_dnload_init_nonsigned(struct i2400m *i2400m) -{ - unsigned i = 0; - int ret = 0; - struct device *dev = i2400m_dev(i2400m); - d_fnstart(5, dev, "(i2400m %p)\n", i2400m); - if (i2400m->bus_bm_pokes_table) { - while (i2400m->bus_bm_pokes_table[i].address) { - ret = i2400m_download_chunk( - i2400m, - &i2400m->bus_bm_pokes_table[i].data, - sizeof(i2400m->bus_bm_pokes_table[i].data), - i2400m->bus_bm_pokes_table[i].address, 1, 1); - if (ret < 0) - break; - i++; - } - } - d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, ret); - return ret; -} - - -/* - * Initialize the signed boot process - * - * @i2400m: device descriptor - * - * @bcf_hdr: pointer to the firmware header; assumes it is fully in - * memory (it has gone through basic validation). - * - * Returns: 0 if ok, < 0 errno code on error, -ERESTARTSYS if the hw - * rebooted. - * - * This writes the firmware BCF header to the device using the - * HASH_PAYLOAD_ONLY command. - */ -static -int i2400m_dnload_init_signed(struct i2400m *i2400m, - const struct i2400m_bcf_hdr *bcf_hdr) -{ - int ret; - struct device *dev = i2400m_dev(i2400m); - struct { - struct i2400m_bootrom_header cmd; - struct i2400m_bcf_hdr cmd_pl; - } __packed *cmd_buf; - struct i2400m_bootrom_header ack; - - d_fnstart(5, dev, "(i2400m %p bcf_hdr %p)\n", i2400m, bcf_hdr); - cmd_buf = i2400m->bm_cmd_buf; - cmd_buf->cmd.command = - i2400m_brh_command(I2400M_BRH_HASH_PAYLOAD_ONLY, 0, 0); - cmd_buf->cmd.target_addr = 0; - cmd_buf->cmd.data_size = cpu_to_le32(sizeof(cmd_buf->cmd_pl)); - memcpy(&cmd_buf->cmd_pl, bcf_hdr, sizeof(*bcf_hdr)); - ret = i2400m_bm_cmd(i2400m, &cmd_buf->cmd, sizeof(*cmd_buf), - &ack, sizeof(ack), 0); - if (ret >= 0) - ret = 0; - d_fnend(5, dev, "(i2400m %p bcf_hdr %p) = %d\n", i2400m, bcf_hdr, ret); - return ret; -} - - -/* - * Initialize the firmware download at the device size - * - * Multiplex to the one that matters based on the device's mode - * (signed or non-signed). - */ -static -int i2400m_dnload_init(struct i2400m *i2400m, - const struct i2400m_bcf_hdr *bcf_hdr) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - - if (i2400m_boot_is_signed(i2400m)) { - d_printf(1, dev, "signed boot\n"); - result = i2400m_dnload_init_signed(i2400m, bcf_hdr); - if (result == -ERESTARTSYS) - return result; - if (result < 0) - dev_err(dev, "firmware %s: signed boot download " - "initialization failed: %d\n", - i2400m->fw_name, result); - } else { - /* non-signed boot process without pokes */ - d_printf(1, dev, "non-signed boot\n"); - result = i2400m_dnload_init_nonsigned(i2400m); - if (result == -ERESTARTSYS) - return result; - if (result < 0) - dev_err(dev, "firmware %s: non-signed download " - "initialization failed: %d\n", - i2400m->fw_name, result); - } - return result; -} - - -/* - * Run consistency tests on the firmware file and load up headers - * - * Check for the firmware being made for the i2400m device, - * etc...These checks are mostly informative, as the device will make - * them too; but the driver's response is more informative on what - * went wrong. - * - * This will also look at all the headers present on the firmware - * file, and update i2400m->fw_bcf_hdr to point to them. - */ -static -int i2400m_fw_hdr_check(struct i2400m *i2400m, - const struct i2400m_bcf_hdr *bcf_hdr, - size_t index, size_t offset) -{ - struct device *dev = i2400m_dev(i2400m); - - unsigned module_type, header_len, major_version, minor_version, - module_id, module_vendor, date, size; - - module_type = le32_to_cpu(bcf_hdr->module_type); - header_len = sizeof(u32) * le32_to_cpu(bcf_hdr->header_len); - major_version = (le32_to_cpu(bcf_hdr->header_version) & 0xffff0000) - >> 16; - minor_version = le32_to_cpu(bcf_hdr->header_version) & 0x0000ffff; - module_id = le32_to_cpu(bcf_hdr->module_id); - module_vendor = le32_to_cpu(bcf_hdr->module_vendor); - date = le32_to_cpu(bcf_hdr->date); - size = sizeof(u32) * le32_to_cpu(bcf_hdr->size); - - d_printf(1, dev, "firmware %s #%zd@%08zx: BCF header " - "type:vendor:id 0x%x:%x:%x v%u.%u (%u/%u B) built %08x\n", - i2400m->fw_name, index, offset, - module_type, module_vendor, module_id, - major_version, minor_version, header_len, size, date); - - /* Hard errors */ - if (major_version != 1) { - dev_err(dev, "firmware %s #%zd@%08zx: major header version " - "v%u.%u not supported\n", - i2400m->fw_name, index, offset, - major_version, minor_version); - return -EBADF; - } - - if (module_type != 6) { /* built for the right hardware? */ - dev_err(dev, "firmware %s #%zd@%08zx: unexpected module " - "type 0x%x; aborting\n", - i2400m->fw_name, index, offset, - module_type); - return -EBADF; - } - - if (module_vendor != 0x8086) { - dev_err(dev, "firmware %s #%zd@%08zx: unexpected module " - "vendor 0x%x; aborting\n", - i2400m->fw_name, index, offset, module_vendor); - return -EBADF; - } - - if (date < 0x20080300) - dev_warn(dev, "firmware %s #%zd@%08zx: build date %08x " - "too old; unsupported\n", - i2400m->fw_name, index, offset, date); - return 0; -} - - -/* - * Run consistency tests on the firmware file and load up headers - * - * Check for the firmware being made for the i2400m device, - * etc...These checks are mostly informative, as the device will make - * them too; but the driver's response is more informative on what - * went wrong. - * - * This will also look at all the headers present on the firmware - * file, and update i2400m->fw_hdrs to point to them. - */ -static -int i2400m_fw_check(struct i2400m *i2400m, const void *bcf, size_t bcf_size) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - size_t headers = 0; - const struct i2400m_bcf_hdr *bcf_hdr; - const void *itr, *next, *top; - size_t slots = 0, used_slots = 0; - - for (itr = bcf, top = itr + bcf_size; - itr < top; - headers++, itr = next) { - size_t leftover, offset, header_len, size; - - leftover = top - itr; - offset = itr - bcf; - if (leftover <= sizeof(*bcf_hdr)) { - dev_err(dev, "firmware %s: %zu B left at @%zx, " - "not enough for BCF header\n", - i2400m->fw_name, leftover, offset); - break; - } - bcf_hdr = itr; - /* Only the first header is supposed to be followed by - * payload */ - header_len = sizeof(u32) * le32_to_cpu(bcf_hdr->header_len); - size = sizeof(u32) * le32_to_cpu(bcf_hdr->size); - if (headers == 0) - next = itr + size; - else - next = itr + header_len; - - result = i2400m_fw_hdr_check(i2400m, bcf_hdr, headers, offset); - if (result < 0) - continue; - if (used_slots + 1 >= slots) { - /* +1 -> we need to account for the one we'll - * occupy and at least an extra one for - * always being NULL */ - result = i2400m_zrealloc_2x( - (void **) &i2400m->fw_hdrs, &slots, - sizeof(i2400m->fw_hdrs[0]), - GFP_KERNEL); - if (result < 0) - goto error_zrealloc; - } - i2400m->fw_hdrs[used_slots] = bcf_hdr; - used_slots++; - } - if (headers == 0) { - dev_err(dev, "firmware %s: no usable headers found\n", - i2400m->fw_name); - result = -EBADF; - } else - result = 0; -error_zrealloc: - return result; -} - - -/* - * Match a barker to a BCF header module ID - * - * The device sends a barker which tells the firmware loader which - * header in the BCF file has to be used. This does the matching. - */ -static -unsigned i2400m_bcf_hdr_match(struct i2400m *i2400m, - const struct i2400m_bcf_hdr *bcf_hdr) -{ - u32 barker = le32_to_cpu(i2400m->barker->data[0]) - & 0x7fffffff; - u32 module_id = le32_to_cpu(bcf_hdr->module_id) - & 0x7fffffff; /* high bit used for something else */ - - /* special case for 5x50 */ - if (barker == I2400M_SBOOT_BARKER && module_id == 0) - return 1; - if (module_id == barker) - return 1; - return 0; -} - -static -const struct i2400m_bcf_hdr *i2400m_bcf_hdr_find(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_bcf_hdr **bcf_itr, *bcf_hdr; - unsigned i = 0; - u32 barker = le32_to_cpu(i2400m->barker->data[0]); - - d_printf(2, dev, "finding BCF header for barker %08x\n", barker); - if (barker == I2400M_NBOOT_BARKER) { - bcf_hdr = i2400m->fw_hdrs[0]; - d_printf(1, dev, "using BCF header #%u/%08x for non-signed " - "barker\n", 0, le32_to_cpu(bcf_hdr->module_id)); - return bcf_hdr; - } - for (bcf_itr = i2400m->fw_hdrs; *bcf_itr != NULL; bcf_itr++, i++) { - bcf_hdr = *bcf_itr; - if (i2400m_bcf_hdr_match(i2400m, bcf_hdr)) { - d_printf(1, dev, "hit on BCF hdr #%u/%08x\n", - i, le32_to_cpu(bcf_hdr->module_id)); - return bcf_hdr; - } else - d_printf(1, dev, "miss on BCF hdr #%u/%08x\n", - i, le32_to_cpu(bcf_hdr->module_id)); - } - dev_err(dev, "cannot find a matching BCF header for barker %08x\n", - barker); - return NULL; -} - - -/* - * Download the firmware to the device - * - * @i2400m: device descriptor - * @bcf: pointer to loaded (and minimally verified for consistency) - * firmware - * @bcf_size: size of the @bcf buffer (header plus payloads) - * - * The process for doing this is described in this file's header. - * - * Note we only reinitialize boot-mode if the flags say so. Some hw - * iterations need it, some don't. In any case, if we loop, we always - * need to reinitialize the boot room, hence the flags modification. - */ -static -int i2400m_fw_dnload(struct i2400m *i2400m, const struct i2400m_bcf_hdr *bcf, - size_t fw_size, enum i2400m_bri flags) -{ - int ret = 0; - struct device *dev = i2400m_dev(i2400m); - int count = i2400m->bus_bm_retries; - const struct i2400m_bcf_hdr *bcf_hdr; - size_t bcf_size; - - d_fnstart(5, dev, "(i2400m %p bcf %p fw size %zu)\n", - i2400m, bcf, fw_size); - i2400m->boot_mode = 1; - wmb(); /* Make sure other readers see it */ -hw_reboot: - if (count-- == 0) { - ret = -ERESTARTSYS; - dev_err(dev, "device rebooted too many times, aborting\n"); - goto error_too_many_reboots; - } - if (flags & I2400M_BRI_MAC_REINIT) { - ret = i2400m_bootrom_init(i2400m, flags); - if (ret < 0) { - dev_err(dev, "bootrom init failed: %d\n", ret); - goto error_bootrom_init; - } - } - flags |= I2400M_BRI_MAC_REINIT; - - /* - * Initialize the download, push the bytes to the device and - * then jump to the new firmware. Note @ret is passed with the - * offset of the jump instruction to _dnload_finalize() - * - * Note we need to use the BCF header in the firmware image - * that matches the barker that the device sent when it - * rebooted, so it has to be passed along. - */ - ret = -EBADF; - bcf_hdr = i2400m_bcf_hdr_find(i2400m); - if (bcf_hdr == NULL) - goto error_bcf_hdr_find; - - ret = i2400m_dnload_init(i2400m, bcf_hdr); - if (ret == -ERESTARTSYS) - goto error_dev_rebooted; - if (ret < 0) - goto error_dnload_init; - - /* - * bcf_size refers to one header size plus the fw sections size - * indicated by the header,ie. if there are other extended headers - * at the tail, they are not counted - */ - bcf_size = sizeof(u32) * le32_to_cpu(bcf_hdr->size); - ret = i2400m_dnload_bcf(i2400m, bcf, bcf_size); - if (ret == -ERESTARTSYS) - goto error_dev_rebooted; - if (ret < 0) { - dev_err(dev, "fw %s: download failed: %d\n", - i2400m->fw_name, ret); - goto error_dnload_bcf; - } - - ret = i2400m_dnload_finalize(i2400m, bcf_hdr, bcf, ret); - if (ret == -ERESTARTSYS) - goto error_dev_rebooted; - if (ret < 0) { - dev_err(dev, "fw %s: " - "download finalization failed: %d\n", - i2400m->fw_name, ret); - goto error_dnload_finalize; - } - - d_printf(2, dev, "fw %s successfully uploaded\n", - i2400m->fw_name); - i2400m->boot_mode = 0; - wmb(); /* Make sure i2400m_msg_to_dev() sees boot_mode */ -error_dnload_finalize: -error_dnload_bcf: -error_dnload_init: -error_bcf_hdr_find: -error_bootrom_init: -error_too_many_reboots: - d_fnend(5, dev, "(i2400m %p bcf %p size %zu) = %d\n", - i2400m, bcf, fw_size, ret); - return ret; - -error_dev_rebooted: - dev_err(dev, "device rebooted, %d tries left\n", count); - /* we got the notification already, no need to wait for it again */ - flags |= I2400M_BRI_SOFT; - goto hw_reboot; -} - -static -int i2400m_fw_bootstrap(struct i2400m *i2400m, const struct firmware *fw, - enum i2400m_bri flags) -{ - int ret; - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_bcf_hdr *bcf; /* Firmware data */ - - d_fnstart(5, dev, "(i2400m %p)\n", i2400m); - bcf = (void *) fw->data; - ret = i2400m_fw_check(i2400m, bcf, fw->size); - if (ret >= 0) - ret = i2400m_fw_dnload(i2400m, bcf, fw->size, flags); - if (ret < 0) - dev_err(dev, "%s: cannot use: %d, skipping\n", - i2400m->fw_name, ret); - kfree(i2400m->fw_hdrs); - i2400m->fw_hdrs = NULL; - d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, ret); - return ret; -} - - -/* Refcounted container for firmware data */ -struct i2400m_fw { - struct kref kref; - const struct firmware *fw; -}; - - -static -void i2400m_fw_destroy(struct kref *kref) -{ - struct i2400m_fw *i2400m_fw = - container_of(kref, struct i2400m_fw, kref); - release_firmware(i2400m_fw->fw); - kfree(i2400m_fw); -} - - -static -struct i2400m_fw *i2400m_fw_get(struct i2400m_fw *i2400m_fw) -{ - if (i2400m_fw != NULL && i2400m_fw != (void *) ~0) - kref_get(&i2400m_fw->kref); - return i2400m_fw; -} - - -static -void i2400m_fw_put(struct i2400m_fw *i2400m_fw) -{ - kref_put(&i2400m_fw->kref, i2400m_fw_destroy); -} - - -/** - * i2400m_dev_bootstrap - Bring the device to a known state and upload firmware - * - * @i2400m: device descriptor - * - * Returns: >= 0 if ok, < 0 errno code on error. - * - * This sets up the firmware upload environment, loads the firmware - * file from disk, verifies and then calls the firmware upload process - * per se. - * - * Can be called either from probe, or after a warm reset. Can not be - * called from within an interrupt. All the flow in this code is - * single-threade; all I/Os are synchronous. - */ -int i2400m_dev_bootstrap(struct i2400m *i2400m, enum i2400m_bri flags) -{ - int ret, itr; - struct device *dev = i2400m_dev(i2400m); - struct i2400m_fw *i2400m_fw; - const struct firmware *fw; - const char *fw_name; - - d_fnstart(5, dev, "(i2400m %p)\n", i2400m); - - ret = -ENODEV; - spin_lock(&i2400m->rx_lock); - i2400m_fw = i2400m_fw_get(i2400m->fw_cached); - spin_unlock(&i2400m->rx_lock); - if (i2400m_fw == (void *) ~0) { - dev_err(dev, "can't load firmware now!"); - goto out; - } else if (i2400m_fw != NULL) { - dev_info(dev, "firmware %s: loading from cache\n", - i2400m->fw_name); - ret = i2400m_fw_bootstrap(i2400m, i2400m_fw->fw, flags); - i2400m_fw_put(i2400m_fw); - goto out; - } - - /* Load firmware files to memory. */ - for (itr = 0, ret = -ENOENT; ; itr++) { - fw_name = i2400m->bus_fw_names[itr]; - if (fw_name == NULL) { - dev_err(dev, "Could not find a usable firmware image\n"); - break; - } - d_printf(1, dev, "trying firmware %s (%d)\n", fw_name, itr); - ret = request_firmware(&fw, fw_name, dev); - if (ret < 0) { - dev_err(dev, "fw %s: cannot load file: %d\n", - fw_name, ret); - continue; - } - i2400m->fw_name = fw_name; - ret = i2400m_fw_bootstrap(i2400m, fw, flags); - release_firmware(fw); - if (ret >= 0) /* firmware loaded successfully */ - break; - i2400m->fw_name = NULL; - } -out: - d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, ret); - return ret; -} -EXPORT_SYMBOL_GPL(i2400m_dev_bootstrap); - - -void i2400m_fw_cache(struct i2400m *i2400m) -{ - int result; - struct i2400m_fw *i2400m_fw; - struct device *dev = i2400m_dev(i2400m); - - /* if there is anything there, free it -- now, this'd be weird */ - spin_lock(&i2400m->rx_lock); - i2400m_fw = i2400m->fw_cached; - spin_unlock(&i2400m->rx_lock); - if (i2400m_fw != NULL && i2400m_fw != (void *) ~0) { - i2400m_fw_put(i2400m_fw); - WARN(1, "%s:%u: still cached fw still present?\n", - __func__, __LINE__); - } - - if (i2400m->fw_name == NULL) { - dev_err(dev, "firmware n/a: can't cache\n"); - i2400m_fw = (void *) ~0; - goto out; - } - - i2400m_fw = kzalloc(sizeof(*i2400m_fw), GFP_ATOMIC); - if (i2400m_fw == NULL) - goto out; - kref_init(&i2400m_fw->kref); - result = request_firmware(&i2400m_fw->fw, i2400m->fw_name, dev); - if (result < 0) { - dev_err(dev, "firmware %s: failed to cache: %d\n", - i2400m->fw_name, result); - kfree(i2400m_fw); - i2400m_fw = (void *) ~0; - } else - dev_info(dev, "firmware %s: cached\n", i2400m->fw_name); -out: - spin_lock(&i2400m->rx_lock); - i2400m->fw_cached = i2400m_fw; - spin_unlock(&i2400m->rx_lock); -} - - -void i2400m_fw_uncache(struct i2400m *i2400m) -{ - struct i2400m_fw *i2400m_fw; - - spin_lock(&i2400m->rx_lock); - i2400m_fw = i2400m->fw_cached; - i2400m->fw_cached = NULL; - spin_unlock(&i2400m->rx_lock); - - if (i2400m_fw != NULL && i2400m_fw != (void *) ~0) - i2400m_fw_put(i2400m_fw); -} - diff --git a/drivers/net/wimax/i2400m/i2400m-usb.h b/drivers/net/wimax/i2400m/i2400m-usb.h deleted file mode 100644 index eff4f464a23e..000000000000 --- a/drivers/net/wimax/i2400m/i2400m-usb.h +++ /dev/null @@ -1,275 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * USB-specific i2400m driver definitions - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Inaky Perez-Gonzalez - * Yanir Lubetkin - * - Initial implementation - * - * - * This driver implements the bus-specific part of the i2400m for - * USB. Check i2400m.h for a generic driver description. - * - * ARCHITECTURE - * - * This driver listens to notifications sent from the notification - * endpoint (in usb-notif.c); when data is ready to read, the code in - * there schedules a read from the device (usb-rx.c) and then passes - * the data to the generic RX code (rx.c). - * - * When the generic driver needs to send data (network or control), it - * queues up in the TX FIFO (tx.c) and that will notify the driver - * through the i2400m->bus_tx_kick() callback - * (usb-tx.c:i2400mu_bus_tx_kick) which will send the items in the - * FIFO queue. - * - * This driver, as well, implements the USB-specific ops for the generic - * driver to be able to setup/teardown communication with the device - * [i2400m_bus_dev_start() and i2400m_bus_dev_stop()], reseting the - * device [i2400m_bus_reset()] and performing firmware upload - * [i2400m_bus_bm_cmd() and i2400_bus_bm_wait_for_ack()]. - */ - -#ifndef __I2400M_USB_H__ -#define __I2400M_USB_H__ - -#include "i2400m.h" -#include - - -/* - * Error Density Count: cheapo error density (over time) counter - * - * Originally by Reinette Chatre - * - * Embed an 'struct edc' somewhere. Each time there is a soft or - * retryable error, call edc_inc() and check if the error top - * watermark has been reached. - */ -enum { - EDC_MAX_ERRORS = 10, - EDC_ERROR_TIMEFRAME = HZ, -}; - -/* error density counter */ -struct edc { - unsigned long timestart; - u16 errorcount; -}; - -struct i2400m_endpoint_cfg { - unsigned char bulk_out; - unsigned char notification; - unsigned char reset_cold; - unsigned char bulk_in; -}; - -static inline void edc_init(struct edc *edc) -{ - edc->timestart = jiffies; -} - -/** - * edc_inc - report a soft error and check if we are over the watermark - * - * @edc: pointer to error density counter. - * @max_err: maximum number of errors we can accept over the timeframe - * @timeframe: length of the timeframe (in jiffies). - * - * Returns: !0 1 if maximum acceptable errors per timeframe has been - * exceeded. 0 otherwise. - * - * This is way to determine if the number of acceptable errors per time - * period has been exceeded. It is not accurate as there are cases in which - * this scheme will not work, for example if there are periodic occurrences - * of errors that straddle updates to the start time. This scheme is - * sufficient for our usage. - * - * To use, embed a 'struct edc' somewhere, initialize it with - * edc_init() and when an error hits: - * - * if (do_something_fails_with_a_soft_error) { - * if (edc_inc(&my->edc, MAX_ERRORS, MAX_TIMEFRAME)) - * Ops, hard error, do something about it - * else - * Retry or ignore, depending on whatever - * } - */ -static inline int edc_inc(struct edc *edc, u16 max_err, u16 timeframe) -{ - unsigned long now; - - now = jiffies; - if (time_after(now, edc->timestart + timeframe)) { - edc->errorcount = 1; - edc->timestart = now; - } else if (++edc->errorcount > max_err) { - edc->errorcount = 0; - edc->timestart = now; - return 1; - } - return 0; -} - -/* Host-Device interface for USB */ -enum { - I2400M_USB_BOOT_RETRIES = 3, - I2400MU_MAX_NOTIFICATION_LEN = 256, - I2400MU_BLK_SIZE = 16, - I2400MU_PL_SIZE_MAX = 0x3EFF, - - /* Device IDs */ - USB_DEVICE_ID_I6050 = 0x0186, - USB_DEVICE_ID_I6050_2 = 0x0188, - USB_DEVICE_ID_I6150 = 0x07d6, - USB_DEVICE_ID_I6150_2 = 0x07d7, - USB_DEVICE_ID_I6150_3 = 0x07d9, - USB_DEVICE_ID_I6250 = 0x0187, -}; - - -/** - * struct i2400mu - descriptor for a USB connected i2400m - * - * @i2400m: bus-generic i2400m implementation; has to be first (see - * it's documentation in i2400m.h). - * - * @usb_dev: pointer to our USB device - * - * @usb_iface: pointer to our USB interface - * - * @urb_edc: error density counter; used to keep a density-on-time tab - * on how many soft (retryable or ignorable) errors we get. If we - * go over the threshold, we consider the bus transport is failing - * too much and reset. - * - * @notif_urb: URB for receiving notifications from the device. - * - * @tx_kthread: thread we use for data TX. We use a thread because in - * order to do deep power saving and put the device to sleep, we - * need to call usb_autopm_*() [blocking functions]. - * - * @tx_wq: waitqueue for the TX kthread to sleep when there is no data - * to be sent; when more data is available, it is woken up by - * i2400mu_bus_tx_kick(). - * - * @rx_kthread: thread we use for data RX. We use a thread because in - * order to do deep power saving and put the device to sleep, we - * need to call usb_autopm_*() [blocking functions]. - * - * @rx_wq: waitqueue for the RX kthread to sleep when there is no data - * to receive. When data is available, it is woken up by - * usb-notif.c:i2400mu_notification_grok(). - * - * @rx_pending_count: number of rx-data-ready notifications that were - * still not handled by the RX kthread. - * - * @rx_size: current RX buffer size that is being used. - * - * @rx_size_acc: accumulator of the sizes of the previous read - * transactions. - * - * @rx_size_cnt: number of read transactions accumulated in - * @rx_size_acc. - * - * @do_autopm: disable(0)/enable(>0) calling the - * usb_autopm_get/put_interface() barriers when executing - * commands. See doc in i2400mu_suspend() for more information. - * - * @rx_size_auto_shrink: if true, the rx_size is shrunk - * automatically based on the average size of the received - * transactions. This allows the receive code to allocate smaller - * chunks of memory and thus reduce pressure on the memory - * allocator by not wasting so much space. By default it is - * enabled. - * - * @debugfs_dentry: hookup for debugfs files. - * These have to be in a separate directory, a child of - * (wimax_dev->debugfs_dentry) so they can be removed when the - * module unloads, as we don't keep each dentry. - */ -struct i2400mu { - struct i2400m i2400m; /* FIRST! See doc */ - - struct usb_device *usb_dev; - struct usb_interface *usb_iface; - struct edc urb_edc; /* Error density counter */ - struct i2400m_endpoint_cfg endpoint_cfg; - - struct urb *notif_urb; - struct task_struct *tx_kthread; - wait_queue_head_t tx_wq; - - struct task_struct *rx_kthread; - wait_queue_head_t rx_wq; - atomic_t rx_pending_count; - size_t rx_size, rx_size_acc, rx_size_cnt; - atomic_t do_autopm; - u8 rx_size_auto_shrink; - - struct dentry *debugfs_dentry; - unsigned i6050:1; /* 1 if this is a 6050 based SKU */ -}; - - -static inline -void i2400mu_init(struct i2400mu *i2400mu) -{ - i2400m_init(&i2400mu->i2400m); - edc_init(&i2400mu->urb_edc); - init_waitqueue_head(&i2400mu->tx_wq); - atomic_set(&i2400mu->rx_pending_count, 0); - init_waitqueue_head(&i2400mu->rx_wq); - i2400mu->rx_size = PAGE_SIZE - sizeof(struct skb_shared_info); - atomic_set(&i2400mu->do_autopm, 1); - i2400mu->rx_size_auto_shrink = 1; -} - -int i2400mu_notification_setup(struct i2400mu *); -void i2400mu_notification_release(struct i2400mu *); - -int i2400mu_rx_setup(struct i2400mu *); -void i2400mu_rx_release(struct i2400mu *); -void i2400mu_rx_kick(struct i2400mu *); - -int i2400mu_tx_setup(struct i2400mu *); -void i2400mu_tx_release(struct i2400mu *); -void i2400mu_bus_tx_kick(struct i2400m *); - -ssize_t i2400mu_bus_bm_cmd_send(struct i2400m *, - const struct i2400m_bootrom_header *, size_t, - int); -ssize_t i2400mu_bus_bm_wait_for_ack(struct i2400m *, - struct i2400m_bootrom_header *, size_t); -#endif /* #ifndef __I2400M_USB_H__ */ diff --git a/drivers/net/wimax/i2400m/i2400m.h b/drivers/net/wimax/i2400m/i2400m.h deleted file mode 100644 index a3733a6d14f5..000000000000 --- a/drivers/net/wimax/i2400m/i2400m.h +++ /dev/null @@ -1,970 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * Declarations for bus-generic internal APIs - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Inaky Perez-Gonzalez - * Yanir Lubetkin - * - Initial implementation - * - * - * GENERAL DRIVER ARCHITECTURE - * - * The i2400m driver is split in the following two major parts: - * - * - bus specific driver - * - bus generic driver (this part) - * - * The bus specific driver sets up stuff specific to the bus the - * device is connected to (USB, PCI, tam-tam...non-authoritative - * nor binding list) which is basically the device-model management - * (probe/disconnect, etc), moving data from device to kernel and - * back, doing the power saving details and reseting the device. - * - * For details on each bus-specific driver, see it's include file, - * i2400m-BUSNAME.h - * - * The bus-generic functionality break up is: - * - * - Firmware upload: fw.c - takes care of uploading firmware to the - * device. bus-specific driver just needs to provides a way to - * execute boot-mode commands and to reset the device. - * - * - RX handling: rx.c - receives data from the bus-specific code and - * feeds it to the network or WiMAX stack or uses it to modify - * the driver state. bus-specific driver only has to receive - * frames and pass them to this module. - * - * - TX handling: tx.c - manages the TX FIFO queue and provides means - * for the bus-specific TX code to pull data from the FIFO - * queue. bus-specific code just pulls frames from this module - * to sends them to the device. - * - * - netdev glue: netdev.c - interface with Linux networking - * stack. Pass around data frames, and configure when the - * device is up and running or shutdown (through ifconfig up / - * down). Bus-generic only. - * - * - control ops: control.c - implements various commands for - * controlling the device. bus-generic only. - * - * - device model glue: driver.c - implements helpers for the - * device-model glue done by the bus-specific layer - * (setup/release the driver resources), turning the device on - * and off, handling the device reboots/resets and a few simple - * WiMAX stack ops. - * - * Code is also broken up in linux-glue / device-glue. - * - * Linux glue contains functions that deal mostly with gluing with the - * rest of the Linux kernel. - * - * Device-glue are functions that deal mostly with the way the device - * does things and talk the device's language. - * - * device-glue code is licensed BSD so other open source OSes can take - * it to implement their drivers. - * - * - * APIs AND HEADER FILES - * - * This bus generic code exports three APIs: - * - * - HDI (host-device interface) definitions common to all busses - * (include/linux/wimax/i2400m.h); these can be also used by user - * space code. - * - internal API for the bus-generic code - * - external API for the bus-specific drivers - * - * - * LIFE CYCLE: - * - * When the bus-specific driver probes, it allocates a network device - * with enough space for it's data structue, that must contain a - * &struct i2400m at the top. - * - * On probe, it needs to fill the i2400m members marked as [fill], as - * well as i2400m->wimax_dev.net_dev and call i2400m_setup(). The - * i2400m driver will only register with the WiMAX and network stacks; - * the only access done to the device is to read the MAC address so we - * can register a network device. - * - * The high-level call flow is: - * - * bus_probe() - * i2400m_setup() - * i2400m->bus_setup() - * boot rom initialization / read mac addr - * network / WiMAX stacks registration - * i2400m_dev_start() - * i2400m->bus_dev_start() - * i2400m_dev_initialize() - * - * The reverse applies for a disconnect() call: - * - * bus_disconnect() - * i2400m_release() - * i2400m_dev_stop() - * i2400m_dev_shutdown() - * i2400m->bus_dev_stop() - * network / WiMAX stack unregistration - * i2400m->bus_release() - * - * At this point, control and data communications are possible. - * - * While the device is up, it might reset. The bus-specific driver has - * to catch that situation and call i2400m_dev_reset_handle() to deal - * with it (reset the internal driver structures and go back to square - * one). - */ - -#ifndef __I2400M_H__ -#define __I2400M_H__ - -#include -#include -#include -#include -#include -#include -#include -#include - -enum { -/* netdev interface */ - /* - * Out of NWG spec (R1_v1.2.2), 3.3.3 ASN Bearer Plane MTU Size - * - * The MTU is 1400 or less - */ - I2400M_MAX_MTU = 1400, -}; - -/* Misc constants */ -enum { - /* Size of the Boot Mode Command buffer */ - I2400M_BM_CMD_BUF_SIZE = 16 * 1024, - I2400M_BM_ACK_BUF_SIZE = 256, -}; - -enum { - /* Maximum number of bus reset can be retried */ - I2400M_BUS_RESET_RETRIES = 3, -}; - -/** - * struct i2400m_poke_table - Hardware poke table for the Intel 2400m - * - * This structure will be used to create a device specific poke table - * to put the device in a consistent state at boot time. - * - * @address: The device address to poke - * - * @data: The data value to poke to the device address - * - */ -struct i2400m_poke_table{ - __le32 address; - __le32 data; -}; - -#define I2400M_FW_POKE(a, d) { \ - .address = cpu_to_le32(a), \ - .data = cpu_to_le32(d) \ -} - - -/** - * i2400m_reset_type - methods to reset a device - * - * @I2400M_RT_WARM: Reset without device disconnection, device handles - * are kept valid but state is back to power on, with firmware - * re-uploaded. - * @I2400M_RT_COLD: Tell the device to disconnect itself from the bus - * and reconnect. Renders all device handles invalid. - * @I2400M_RT_BUS: Tells the bus to reset the device; last measure - * used when both types above don't work. - */ -enum i2400m_reset_type { - I2400M_RT_WARM, /* first measure */ - I2400M_RT_COLD, /* second measure */ - I2400M_RT_BUS, /* call in artillery */ -}; - -struct i2400m_reset_ctx; -struct i2400m_roq; -struct i2400m_barker_db; - -/** - * struct i2400m - descriptor for an Intel 2400m - * - * Members marked with [fill] must be filled out/initialized before - * calling i2400m_setup(). - * - * Note the @bus_setup/@bus_release, @bus_dev_start/@bus_dev_release - * call pairs are very much doing almost the same, and depending on - * the underlying bus, some stuff has to be put in one or the - * other. The idea of setup/release is that they setup the minimal - * amount needed for loading firmware, where us dev_start/stop setup - * the rest needed to do full data/control traffic. - * - * @bus_tx_block_size: [fill] USB imposes a 16 block size, but other - * busses will differ. So we have a tx_blk_size variable that the - * bus layer sets to tell the engine how much of that we need. - * - * @bus_tx_room_min: [fill] Minimum room required while allocating - * TX queue's buffer space for message header. USB requires - * 16 bytes. Refer to bus specific driver code for details. - * - * @bus_pl_size_max: [fill] Maximum payload size. - * - * @bus_setup: [optional fill] Function called by the bus-generic code - * [i2400m_setup()] to setup the basic bus-specific communications - * to the the device needed to load firmware. See LIFE CYCLE above. - * - * NOTE: Doesn't need to upload the firmware, as that is taken - * care of by the bus-generic code. - * - * @bus_release: [optional fill] Function called by the bus-generic - * code [i2400m_release()] to shutdown the basic bus-specific - * communications to the the device needed to load firmware. See - * LIFE CYCLE above. - * - * This function does not need to reset the device, just tear down - * all the host resources created to handle communication with - * the device. - * - * @bus_dev_start: [optional fill] Function called by the bus-generic - * code [i2400m_dev_start()] to do things needed to start the - * device. See LIFE CYCLE above. - * - * NOTE: Doesn't need to upload the firmware, as that is taken - * care of by the bus-generic code. - * - * @bus_dev_stop: [optional fill] Function called by the bus-generic - * code [i2400m_dev_stop()] to do things needed for stopping the - * device. See LIFE CYCLE above. - * - * This function does not need to reset the device, just tear down - * all the host resources created to handle communication with - * the device. - * - * @bus_tx_kick: [fill] Function called by the bus-generic code to let - * the bus-specific code know that there is data available in the - * TX FIFO for transmission to the device. - * - * This function cannot sleep. - * - * @bus_reset: [fill] Function called by the bus-generic code to reset - * the device in in various ways. Doesn't need to wait for the - * reset to finish. - * - * If warm or cold reset fail, this function is expected to do a - * bus-specific reset (eg: USB reset) to get the device to a - * working state (even if it implies device disconecction). - * - * Note the warm reset is used by the firmware uploader to - * reinitialize the device. - * - * IMPORTANT: this is called very early in the device setup - * process, so it cannot rely on common infrastructure being laid - * out. - * - * IMPORTANT: don't call reset on RT_BUS with i2400m->init_mutex - * held, as the .pre/.post reset handlers will deadlock. - * - * @bus_bm_retries: [fill] How many times shall a firmware upload / - * device initialization be retried? Different models of the same - * device might need different values, hence it is set by the - * bus-specific driver. Note this value is used in two places, - * i2400m_fw_dnload() and __i2400m_dev_start(); they won't become - * multiplicative (__i2400m_dev_start() calling N times - * i2400m_fw_dnload() and this trying N times to download the - * firmware), as if __i2400m_dev_start() only retries if the - * firmware crashed while initializing the device (not in a - * general case). - * - * @bus_bm_cmd_send: [fill] Function called to send a boot-mode - * command. Flags are defined in 'enum i2400m_bm_cmd_flags'. This - * is synchronous and has to return 0 if ok or < 0 errno code in - * any error condition. - * - * @bus_bm_wait_for_ack: [fill] Function called to wait for a - * boot-mode notification (that can be a response to a previously - * issued command or an asynchronous one). Will read until all the - * indicated size is read or timeout. Reading more or less data - * than asked for is an error condition. Return 0 if ok, < 0 errno - * code on error. - * - * The caller to this function will check if the response is a - * barker that indicates the device going into reset mode. - * - * @bus_fw_names: [fill] a NULL-terminated array with the names of the - * firmware images to try loading. This is made a list so we can - * support backward compatibility of firmware releases (eg: if we - * can't find the default v1.4, we try v1.3). In general, the name - * should be i2400m-fw-X-VERSION.sbcf, where X is the bus name. - * The list is tried in order and the first one that loads is - * used. The fw loader will set i2400m->fw_name to point to the - * active firmware image. - * - * @bus_bm_mac_addr_impaired: [fill] Set to true if the device's MAC - * address provided in boot mode is kind of broken and needs to - * be re-read later on. - * - * @bus_bm_pokes_table: [fill/optional] A table of device addresses - * and values that will be poked at device init time to move the - * device to the correct state for the type of boot/firmware being - * used. This table MUST be terminated with (0x000000, - * 0x00000000) or bad things will happen. - * - * - * @wimax_dev: WiMAX generic device for linkage into the kernel WiMAX - * stack. Due to the way a net_device is allocated, we need to - * force this to be the first field so that we can get from - * netdev_priv() the right pointer. - * - * @updown: the device is up and ready for transmitting control and - * data packets. This implies @ready (communication infrastructure - * with the device is ready) and the device's firmware has been - * loaded and the device initialized. - * - * Write to it only inside a i2400m->init_mutex protected area - * followed with a wmb(); rmb() before accesing (unless locked - * inside i2400m->init_mutex). Read access can be loose like that - * [just using rmb()] because the paths that use this also do - * other error checks later on. - * - * @ready: Communication infrastructure with the device is ready, data - * frames can start to be passed around (this is lighter than - * using the WiMAX state for certain hot paths). - * - * Write to it only inside a i2400m->init_mutex protected area - * followed with a wmb(); rmb() before accesing (unless locked - * inside i2400m->init_mutex). Read access can be loose like that - * [just using rmb()] because the paths that use this also do - * other error checks later on. - * - * @rx_reorder: 1 if RX reordering is enabled; this can only be - * set at probe time. - * - * @state: device's state (as reported by it) - * - * @state_wq: waitqueue that is woken up whenever the state changes - * - * @tx_lock: spinlock to protect TX members - * - * @tx_buf: FIFO buffer for TX; we queue data here - * - * @tx_in: FIFO index for incoming data. Note this doesn't wrap around - * and it is always greater than @tx_out. - * - * @tx_out: FIFO index for outgoing data - * - * @tx_msg: current TX message that is active in the FIFO for - * appending payloads. - * - * @tx_sequence: current sequence number for TX messages from the - * device to the host. - * - * @tx_msg_size: size of the current message being transmitted by the - * bus-specific code. - * - * @tx_pl_num: total number of payloads sent - * - * @tx_pl_max: maximum number of payloads sent in a TX message - * - * @tx_pl_min: minimum number of payloads sent in a TX message - * - * @tx_num: number of TX messages sent - * - * @tx_size_acc: number of bytes in all TX messages sent - * (this is different to net_dev's statistics as it also counts - * control messages). - * - * @tx_size_min: smallest TX message sent. - * - * @tx_size_max: biggest TX message sent. - * - * @rx_lock: spinlock to protect RX members and rx_roq_refcount. - * - * @rx_pl_num: total number of payloads received - * - * @rx_pl_max: maximum number of payloads received in a RX message - * - * @rx_pl_min: minimum number of payloads received in a RX message - * - * @rx_num: number of RX messages received - * - * @rx_size_acc: number of bytes in all RX messages received - * (this is different to net_dev's statistics as it also counts - * control messages). - * - * @rx_size_min: smallest RX message received. - * - * @rx_size_max: buggest RX message received. - * - * @rx_roq: RX ReOrder queues. (fw >= v1.4) When packets are received - * out of order, the device will ask the driver to hold certain - * packets until the ones that are received out of order can be - * delivered. Then the driver can release them to the host. See - * drivers/net/i2400m/rx.c for details. - * - * @rx_roq_refcount: refcount rx_roq. This refcounts any access to - * rx_roq thus preventing rx_roq being destroyed when rx_roq - * is being accessed. rx_roq_refcount is protected by rx_lock. - * - * @rx_reports: reports received from the device that couldn't be - * processed because the driver wasn't still ready; when ready, - * they are pulled from here and chewed. - * - * @rx_reports_ws: Work struct used to kick a scan of the RX reports - * list and to process each. - * - * @src_mac_addr: MAC address used to make ethernet packets be coming - * from. This is generated at i2400m_setup() time and used during - * the life cycle of the instance. See i2400m_fake_eth_header(). - * - * @init_mutex: Mutex used for serializing the device bringup - * sequence; this way if the device reboots in the middle, we - * don't try to do a bringup again while we are tearing down the - * one that failed. - * - * Can't reuse @msg_mutex because from within the bringup sequence - * we need to send messages to the device and thus use @msg_mutex. - * - * @msg_mutex: mutex used to send control commands to the device (we - * only allow one at a time, per host-device interface design). - * - * @msg_completion: used to wait for an ack to a control command sent - * to the device. - * - * @ack_skb: used to store the actual ack to a control command if the - * reception of the command was successful. Otherwise, a ERR_PTR() - * errno code that indicates what failed with the ack reception. - * - * Only valid after @msg_completion is woken up. Only updateable - * if @msg_completion is armed. Only touched by - * i2400m_msg_to_dev(). - * - * Protected by @rx_lock. In theory the command execution flow is - * sequential, but in case the device sends an out-of-phase or - * very delayed response, we need to avoid it trampling current - * execution. - * - * @bm_cmd_buf: boot mode command buffer for composing firmware upload - * commands. - * - * USB can't r/w to stack, vmalloc, etc...as well, we end up - * having to alloc/free a lot to compose commands, so we use these - * for stagging and not having to realloc all the time. - * - * This assumes the code always runs serialized. Only one thread - * can call i2400m_bm_cmd() at the same time. - * - * @bm_ack_buf: boot mode acknoledge buffer for staging reception of - * responses to commands. - * - * See @bm_cmd_buf. - * - * @work_queue: work queue for processing device reports. This - * workqueue cannot be used for processing TX or RX to the device, - * as from it we'll process device reports, which might require - * further communication with the device. - * - * @debugfs_dentry: hookup for debugfs files. - * These have to be in a separate directory, a child of - * (wimax_dev->debugfs_dentry) so they can be removed when the - * module unloads, as we don't keep each dentry. - * - * @fw_name: name of the firmware image that is currently being used. - * - * @fw_version: version of the firmware interface, Major.minor, - * encoded in the high word and low word (major << 16 | minor). - * - * @fw_hdrs: NULL terminated array of pointers to the firmware - * headers. This is only available during firmware load time. - * - * @fw_cached: Used to cache firmware when the system goes to - * suspend/standby/hibernation (as on resume we can't read it). If - * NULL, no firmware was cached, read it. If ~0, you can't read - * any firmware files (the system still didn't come out of suspend - * and failed to cache one), so abort; otherwise, a valid cached - * firmware to be used. Access to this variable is protected by - * the spinlock i2400m->rx_lock. - * - * @barker: barker type that the device uses; this is initialized by - * i2400m_is_boot_barker() the first time it is called. Then it - * won't change during the life cycle of the device and every time - * a boot barker is received, it is just verified for it being the - * same. - * - * @pm_notifier: used to register for PM events - * - * @bus_reset_retries: counter for the number of bus resets attempted for - * this boot. It's not for tracking the number of bus resets during - * the whole driver life cycle (from insmod to rmmod) but for the - * number of dev_start() executed until dev_start() returns a success - * (ie: a good boot means a dev_stop() followed by a successful - * dev_start()). dev_reset_handler() increments this counter whenever - * it is triggering a bus reset. It checks this counter to decide if a - * subsequent bus reset should be retried. dev_reset_handler() retries - * the bus reset until dev_start() succeeds or the counter reaches - * I2400M_BUS_RESET_RETRIES. The counter is cleared to 0 in - * dev_reset_handle() when dev_start() returns a success, - * ie: a successul boot is completed. - * - * @alive: flag to denote if the device *should* be alive. This flag is - * everything like @updown (see doc for @updown) except reflecting - * the device state *we expect* rather than the actual state as denoted - * by @updown. It is set 1 whenever @updown is set 1 in dev_start(). - * Then the device is expected to be alive all the time - * (i2400m->alive remains 1) until the driver is removed. Therefore - * all the device reboot events detected can be still handled properly - * by either dev_reset_handle() or .pre_reset/.post_reset as long as - * the driver presents. It is set 0 along with @updown in dev_stop(). - * - * @error_recovery: flag to denote if we are ready to take an error recovery. - * 0 for ready to take an error recovery; 1 for not ready. It is - * initialized to 1 while probe() since we don't tend to take any error - * recovery during probe(). It is decremented by 1 whenever dev_start() - * succeeds to indicate we are ready to take error recovery from now on. - * It is checked every time we wanna schedule an error recovery. If an - * error recovery is already in place (error_recovery was set 1), we - * should not schedule another one until the last one is done. - */ -struct i2400m { - struct wimax_dev wimax_dev; /* FIRST! See doc */ - - unsigned updown:1; /* Network device is up or down */ - unsigned boot_mode:1; /* is the device in boot mode? */ - unsigned sboot:1; /* signed or unsigned fw boot */ - unsigned ready:1; /* Device comm infrastructure ready */ - unsigned rx_reorder:1; /* RX reorder is enabled */ - u8 trace_msg_from_user; /* echo rx msgs to 'trace' pipe */ - /* typed u8 so /sys/kernel/debug/u8 can tweak */ - enum i2400m_system_state state; - wait_queue_head_t state_wq; /* Woken up when on state updates */ - - size_t bus_tx_block_size; - size_t bus_tx_room_min; - size_t bus_pl_size_max; - unsigned bus_bm_retries; - - int (*bus_setup)(struct i2400m *); - int (*bus_dev_start)(struct i2400m *); - void (*bus_dev_stop)(struct i2400m *); - void (*bus_release)(struct i2400m *); - void (*bus_tx_kick)(struct i2400m *); - int (*bus_reset)(struct i2400m *, enum i2400m_reset_type); - ssize_t (*bus_bm_cmd_send)(struct i2400m *, - const struct i2400m_bootrom_header *, - size_t, int flags); - ssize_t (*bus_bm_wait_for_ack)(struct i2400m *, - struct i2400m_bootrom_header *, size_t); - const char **bus_fw_names; - unsigned bus_bm_mac_addr_impaired:1; - const struct i2400m_poke_table *bus_bm_pokes_table; - - spinlock_t tx_lock; /* protect TX state */ - void *tx_buf; - size_t tx_in, tx_out; - struct i2400m_msg_hdr *tx_msg; - size_t tx_sequence, tx_msg_size; - /* TX stats */ - unsigned tx_pl_num, tx_pl_max, tx_pl_min, - tx_num, tx_size_acc, tx_size_min, tx_size_max; - - /* RX stuff */ - /* protect RX state and rx_roq_refcount */ - spinlock_t rx_lock; - unsigned rx_pl_num, rx_pl_max, rx_pl_min, - rx_num, rx_size_acc, rx_size_min, rx_size_max; - struct i2400m_roq *rx_roq; /* access is refcounted */ - struct kref rx_roq_refcount; /* refcount access to rx_roq */ - u8 src_mac_addr[ETH_HLEN]; - struct list_head rx_reports; /* under rx_lock! */ - struct work_struct rx_report_ws; - - struct mutex msg_mutex; /* serialize command execution */ - struct completion msg_completion; - struct sk_buff *ack_skb; /* protected by rx_lock */ - - void *bm_ack_buf; /* for receiving acks over USB */ - void *bm_cmd_buf; /* for issuing commands over USB */ - - struct workqueue_struct *work_queue; - - struct mutex init_mutex; /* protect bringup seq */ - struct i2400m_reset_ctx *reset_ctx; /* protected by init_mutex */ - - struct work_struct wake_tx_ws; - struct sk_buff *wake_tx_skb; - - struct work_struct reset_ws; - const char *reset_reason; - - struct work_struct recovery_ws; - - struct dentry *debugfs_dentry; - const char *fw_name; /* name of the current firmware image */ - unsigned long fw_version; /* version of the firmware interface */ - const struct i2400m_bcf_hdr **fw_hdrs; - struct i2400m_fw *fw_cached; /* protected by rx_lock */ - struct i2400m_barker_db *barker; - - struct notifier_block pm_notifier; - - /* counting bus reset retries in this boot */ - atomic_t bus_reset_retries; - - /* if the device is expected to be alive */ - unsigned alive; - - /* 0 if we are ready for error recovery; 1 if not ready */ - atomic_t error_recovery; - -}; - - -/* - * Bus-generic internal APIs - * ------------------------- - */ - -static inline -struct i2400m *wimax_dev_to_i2400m(struct wimax_dev *wimax_dev) -{ - return container_of(wimax_dev, struct i2400m, wimax_dev); -} - -static inline -struct i2400m *net_dev_to_i2400m(struct net_device *net_dev) -{ - return wimax_dev_to_i2400m(netdev_priv(net_dev)); -} - -/* - * Boot mode support - */ - -/** - * i2400m_bm_cmd_flags - flags to i2400m_bm_cmd() - * - * @I2400M_BM_CMD_RAW: send the command block as-is, without doing any - * extra processing for adding CRC. - */ -enum i2400m_bm_cmd_flags { - I2400M_BM_CMD_RAW = 1 << 2, -}; - -/** - * i2400m_bri - Boot-ROM indicators - * - * Flags for i2400m_bootrom_init() and i2400m_dev_bootstrap() [which - * are passed from things like i2400m_setup()]. Can be combined with - * |. - * - * @I2400M_BRI_SOFT: The device rebooted already and a reboot - * barker received, proceed directly to ack the boot sequence. - * @I2400M_BRI_NO_REBOOT: Do not reboot the device and proceed - * directly to wait for a reboot barker from the device. - * @I2400M_BRI_MAC_REINIT: We need to reinitialize the boot - * rom after reading the MAC address. This is quite a dirty hack, - * if you ask me -- the device requires the bootrom to be - * initialized after reading the MAC address. - */ -enum i2400m_bri { - I2400M_BRI_SOFT = 1 << 1, - I2400M_BRI_NO_REBOOT = 1 << 2, - I2400M_BRI_MAC_REINIT = 1 << 3, -}; - -void i2400m_bm_cmd_prepare(struct i2400m_bootrom_header *); -int i2400m_dev_bootstrap(struct i2400m *, enum i2400m_bri); -int i2400m_read_mac_addr(struct i2400m *); -int i2400m_bootrom_init(struct i2400m *, enum i2400m_bri); -int i2400m_is_boot_barker(struct i2400m *, const void *, size_t); -static inline -int i2400m_is_d2h_barker(const void *buf) -{ - const __le32 *barker = buf; - return le32_to_cpu(*barker) == I2400M_D2H_MSG_BARKER; -} -void i2400m_unknown_barker(struct i2400m *, const void *, size_t); - -/* Make/grok boot-rom header commands */ - -static inline -__le32 i2400m_brh_command(enum i2400m_brh_opcode opcode, unsigned use_checksum, - unsigned direct_access) -{ - return cpu_to_le32( - I2400M_BRH_SIGNATURE - | (direct_access ? I2400M_BRH_DIRECT_ACCESS : 0) - | I2400M_BRH_RESPONSE_REQUIRED /* response always required */ - | (use_checksum ? I2400M_BRH_USE_CHECKSUM : 0) - | (opcode & I2400M_BRH_OPCODE_MASK)); -} - -static inline -void i2400m_brh_set_opcode(struct i2400m_bootrom_header *hdr, - enum i2400m_brh_opcode opcode) -{ - hdr->command = cpu_to_le32( - (le32_to_cpu(hdr->command) & ~I2400M_BRH_OPCODE_MASK) - | (opcode & I2400M_BRH_OPCODE_MASK)); -} - -static inline -unsigned i2400m_brh_get_opcode(const struct i2400m_bootrom_header *hdr) -{ - return le32_to_cpu(hdr->command) & I2400M_BRH_OPCODE_MASK; -} - -static inline -unsigned i2400m_brh_get_response(const struct i2400m_bootrom_header *hdr) -{ - return (le32_to_cpu(hdr->command) & I2400M_BRH_RESPONSE_MASK) - >> I2400M_BRH_RESPONSE_SHIFT; -} - -static inline -unsigned i2400m_brh_get_use_checksum(const struct i2400m_bootrom_header *hdr) -{ - return le32_to_cpu(hdr->command) & I2400M_BRH_USE_CHECKSUM; -} - -static inline -unsigned i2400m_brh_get_response_required( - const struct i2400m_bootrom_header *hdr) -{ - return le32_to_cpu(hdr->command) & I2400M_BRH_RESPONSE_REQUIRED; -} - -static inline -unsigned i2400m_brh_get_direct_access(const struct i2400m_bootrom_header *hdr) -{ - return le32_to_cpu(hdr->command) & I2400M_BRH_DIRECT_ACCESS; -} - -static inline -unsigned i2400m_brh_get_signature(const struct i2400m_bootrom_header *hdr) -{ - return (le32_to_cpu(hdr->command) & I2400M_BRH_SIGNATURE_MASK) - >> I2400M_BRH_SIGNATURE_SHIFT; -} - - -/* - * Driver / device setup and internal functions - */ -void i2400m_init(struct i2400m *); -int i2400m_reset(struct i2400m *, enum i2400m_reset_type); -void i2400m_netdev_setup(struct net_device *net_dev); -int i2400m_sysfs_setup(struct device_driver *); -void i2400m_sysfs_release(struct device_driver *); -int i2400m_tx_setup(struct i2400m *); -void i2400m_wake_tx_work(struct work_struct *); -void i2400m_tx_release(struct i2400m *); - -int i2400m_rx_setup(struct i2400m *); -void i2400m_rx_release(struct i2400m *); - -void i2400m_fw_cache(struct i2400m *); -void i2400m_fw_uncache(struct i2400m *); - -void i2400m_net_rx(struct i2400m *, struct sk_buff *, unsigned, const void *, - int); -void i2400m_net_erx(struct i2400m *, struct sk_buff *, enum i2400m_cs); -void i2400m_net_wake_stop(struct i2400m *); -enum i2400m_pt; -int i2400m_tx(struct i2400m *, const void *, size_t, enum i2400m_pt); - -#ifdef CONFIG_DEBUG_FS -void i2400m_debugfs_add(struct i2400m *); -void i2400m_debugfs_rm(struct i2400m *); -#else -static inline void i2400m_debugfs_add(struct i2400m *i2400m) {} -static inline void i2400m_debugfs_rm(struct i2400m *i2400m) {} -#endif - -/* Initialize/shutdown the device */ -int i2400m_dev_initialize(struct i2400m *); -void i2400m_dev_shutdown(struct i2400m *); - -extern struct attribute_group i2400m_dev_attr_group; - - -/* HDI message's payload description handling */ - -static inline -size_t i2400m_pld_size(const struct i2400m_pld *pld) -{ - return I2400M_PLD_SIZE_MASK & le32_to_cpu(pld->val); -} - -static inline -enum i2400m_pt i2400m_pld_type(const struct i2400m_pld *pld) -{ - return (I2400M_PLD_TYPE_MASK & le32_to_cpu(pld->val)) - >> I2400M_PLD_TYPE_SHIFT; -} - -static inline -void i2400m_pld_set(struct i2400m_pld *pld, size_t size, - enum i2400m_pt type) -{ - pld->val = cpu_to_le32( - ((type << I2400M_PLD_TYPE_SHIFT) & I2400M_PLD_TYPE_MASK) - | (size & I2400M_PLD_SIZE_MASK)); -} - - -/* - * API for the bus-specific drivers - * -------------------------------- - */ - -static inline -struct i2400m *i2400m_get(struct i2400m *i2400m) -{ - dev_hold(i2400m->wimax_dev.net_dev); - return i2400m; -} - -static inline -void i2400m_put(struct i2400m *i2400m) -{ - dev_put(i2400m->wimax_dev.net_dev); -} - -int i2400m_dev_reset_handle(struct i2400m *, const char *); -int i2400m_pre_reset(struct i2400m *); -int i2400m_post_reset(struct i2400m *); -void i2400m_error_recovery(struct i2400m *); - -/* - * _setup()/_release() are called by the probe/disconnect functions of - * the bus-specific drivers. - */ -int i2400m_setup(struct i2400m *, enum i2400m_bri bm_flags); -void i2400m_release(struct i2400m *); - -int i2400m_rx(struct i2400m *, struct sk_buff *); -struct i2400m_msg_hdr *i2400m_tx_msg_get(struct i2400m *, size_t *); -void i2400m_tx_msg_sent(struct i2400m *); - - -/* - * Utility functions - */ - -static inline -struct device *i2400m_dev(struct i2400m *i2400m) -{ - return i2400m->wimax_dev.net_dev->dev.parent; -} - -int i2400m_msg_check_status(const struct i2400m_l3l4_hdr *, char *, size_t); -int i2400m_msg_size_check(struct i2400m *, const struct i2400m_l3l4_hdr *, - size_t); -struct sk_buff *i2400m_msg_to_dev(struct i2400m *, const void *, size_t); -void i2400m_msg_to_dev_cancel_wait(struct i2400m *, int); -void i2400m_report_hook(struct i2400m *, const struct i2400m_l3l4_hdr *, - size_t); -void i2400m_report_hook_work(struct work_struct *); -int i2400m_cmd_enter_powersave(struct i2400m *); -int i2400m_cmd_exit_idle(struct i2400m *); -struct sk_buff *i2400m_get_device_info(struct i2400m *); -int i2400m_firmware_check(struct i2400m *); -int i2400m_set_idle_timeout(struct i2400m *, unsigned); - -static inline -struct usb_endpoint_descriptor *usb_get_epd(struct usb_interface *iface, int ep) -{ - return &iface->cur_altsetting->endpoint[ep].desc; -} - -int i2400m_op_rfkill_sw_toggle(struct wimax_dev *, enum wimax_rf_state); -void i2400m_report_tlv_rf_switches_status(struct i2400m *, - const struct i2400m_tlv_rf_switches_status *); - -/* - * Helpers for firmware backwards compatibility - * - * As we aim to support at least the firmware version that was - * released with the previous kernel/driver release, some code will be - * conditionally executed depending on the firmware version. On each - * release, the code to support fw releases past the last two ones - * will be purged. - * - * By making it depend on this macros, it is easier to keep it a tab - * on what has to go and what not. - */ -static inline -unsigned i2400m_le_v1_3(struct i2400m *i2400m) -{ - /* running fw is lower or v1.3 */ - return i2400m->fw_version <= 0x00090001; -} - -static inline -unsigned i2400m_ge_v1_4(struct i2400m *i2400m) -{ - /* running fw is higher or v1.4 */ - return i2400m->fw_version >= 0x00090002; -} - - -/* - * Do a millisecond-sleep for allowing wireshark to dump all the data - * packets. Used only for debugging. - */ -static inline -void __i2400m_msleep(unsigned ms) -{ -#if 1 -#else - msleep(ms); -#endif -} - - -/* module initialization helpers */ -int i2400m_barker_db_init(const char *); -void i2400m_barker_db_exit(void); - - - -#endif /* #ifndef __I2400M_H__ */ diff --git a/drivers/net/wimax/i2400m/netdev.c b/drivers/net/wimax/i2400m/netdev.c deleted file mode 100644 index a7fcbceb6e6b..000000000000 --- a/drivers/net/wimax/i2400m/netdev.c +++ /dev/null @@ -1,603 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Intel Wireless WiMAX Connection 2400m - * Glue with the networking stack - * - * Copyright (C) 2007 Intel Corporation - * Yanir Lubetkin - * Inaky Perez-Gonzalez - * - * This implements an ethernet device for the i2400m. - * - * We fake being an ethernet device to simplify the support from user - * space and from the other side. The world is (sadly) configured to - * take in only Ethernet devices... - * - * Because of this, when using firmwares <= v1.3, there is an - * copy-each-rxed-packet overhead on the RX path. Each IP packet has - * to be reallocated to add an ethernet header (as there is no space - * in what we get from the device). This is a known drawback and - * firmwares >= 1.4 add header space that can be used to insert the - * ethernet header without having to reallocate and copy. - * - * TX error handling is tricky; because we have to FIFO/queue the - * buffers for transmission (as the hardware likes it aggregated), we - * just give the skb to the TX subsystem and by the time it is - * transmitted, we have long forgotten about it. So we just don't care - * too much about it. - * - * Note that when the device is in idle mode with the basestation, we - * need to negotiate coming back up online. That involves negotiation - * and possible user space interaction. Thus, we defer to a workqueue - * to do all that. By default, we only queue a single packet and drop - * the rest, as potentially the time to go back from idle to normal is - * long. - * - * ROADMAP - * - * i2400m_open Called on ifconfig up - * i2400m_stop Called on ifconfig down - * - * i2400m_hard_start_xmit Called by the network stack to send a packet - * i2400m_net_wake_tx Wake up device from basestation-IDLE & TX - * i2400m_wake_tx_work - * i2400m_cmd_exit_idle - * i2400m_tx - * i2400m_net_tx TX a data frame - * i2400m_tx - * - * i2400m_change_mtu Called on ifconfig mtu XXX - * - * i2400m_tx_timeout Called when the device times out - * - * i2400m_net_rx Called by the RX code when a data frame is - * available (firmware <= 1.3) - * i2400m_net_erx Called by the RX code when a data frame is - * available (firmware >= 1.4). - * i2400m_netdev_setup Called to setup all the netdev stuff from - * alloc_netdev. - */ -#include -#include -#include -#include -#include -#include "i2400m.h" - - -#define D_SUBMODULE netdev -#include "debug-levels.h" - -enum { -/* netdev interface */ - /* 20 secs? yep, this is the maximum timeout that the device - * might take to get out of IDLE / negotiate it with the base - * station. We add 1sec for good measure. */ - I2400M_TX_TIMEOUT = 21 * HZ, - /* - * Experimentation has determined that, 20 to be a good value - * for minimizing the jitter in the throughput. - */ - I2400M_TX_QLEN = 20, -}; - - -static -int i2400m_open(struct net_device *net_dev) -{ - int result; - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(net_dev %p [i2400m %p])\n", net_dev, i2400m); - /* Make sure we wait until init is complete... */ - mutex_lock(&i2400m->init_mutex); - if (i2400m->updown) - result = 0; - else - result = -EBUSY; - mutex_unlock(&i2400m->init_mutex); - d_fnend(3, dev, "(net_dev %p [i2400m %p]) = %d\n", - net_dev, i2400m, result); - return result; -} - - -static -int i2400m_stop(struct net_device *net_dev) -{ - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(net_dev %p [i2400m %p])\n", net_dev, i2400m); - i2400m_net_wake_stop(i2400m); - d_fnend(3, dev, "(net_dev %p [i2400m %p]) = 0\n", net_dev, i2400m); - return 0; -} - - -/* - * Wake up the device and transmit a held SKB, then restart the net queue - * - * When the device goes into basestation-idle mode, we need to tell it - * to exit that mode; it will negotiate with the base station, user - * space may have to intervene to rehandshake crypto and then tell us - * when it is ready to transmit the packet we have "queued". Still we - * need to give it sometime after it reports being ok. - * - * On error, there is not much we can do. If the error was on TX, we - * still wake the queue up to see if the next packet will be luckier. - * - * If _cmd_exit_idle() fails...well, it could be many things; most - * commonly it is that something else took the device out of IDLE mode - * (for example, the base station). In that case we get an -EILSEQ and - * we are just going to ignore that one. If the device is back to - * connected, then fine -- if it is someother state, the packet will - * be dropped anyway. - */ -void i2400m_wake_tx_work(struct work_struct *ws) -{ - int result; - struct i2400m *i2400m = container_of(ws, struct i2400m, wake_tx_ws); - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *skb; - unsigned long flags; - - spin_lock_irqsave(&i2400m->tx_lock, flags); - skb = i2400m->wake_tx_skb; - i2400m->wake_tx_skb = NULL; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - - d_fnstart(3, dev, "(ws %p i2400m %p skb %p)\n", ws, i2400m, skb); - result = -EINVAL; - if (skb == NULL) { - dev_err(dev, "WAKE&TX: skb disappeared!\n"); - goto out_put; - } - /* If we have, somehow, lost the connection after this was - * queued, don't do anything; this might be the device got - * reset or just disconnected. */ - if (unlikely(!netif_carrier_ok(net_dev))) - goto out_kfree; - result = i2400m_cmd_exit_idle(i2400m); - if (result == -EILSEQ) - result = 0; - if (result < 0) { - dev_err(dev, "WAKE&TX: device didn't get out of idle: " - "%d - resetting\n", result); - i2400m_reset(i2400m, I2400M_RT_BUS); - goto error; - } - result = wait_event_timeout(i2400m->state_wq, - i2400m->state != I2400M_SS_IDLE, - net_dev->watchdog_timeo - HZ/2); - if (result == 0) - result = -ETIMEDOUT; - if (result < 0) { - dev_err(dev, "WAKE&TX: error waiting for device to exit IDLE: " - "%d - resetting\n", result); - i2400m_reset(i2400m, I2400M_RT_BUS); - goto error; - } - msleep(20); /* device still needs some time or it drops it */ - result = i2400m_tx(i2400m, skb->data, skb->len, I2400M_PT_DATA); -error: - netif_wake_queue(net_dev); -out_kfree: - kfree_skb(skb); /* refcount transferred by _hard_start_xmit() */ -out_put: - i2400m_put(i2400m); - d_fnend(3, dev, "(ws %p i2400m %p skb %p) = void [%d]\n", - ws, i2400m, skb, result); -} - - -/* - * Prepare the data payload TX header - * - * The i2400m expects a 4 byte header in front of a data packet. - * - * Because we pretend to be an ethernet device, this packet comes with - * an ethernet header. Pull it and push our header. - */ -static -void i2400m_tx_prep_header(struct sk_buff *skb) -{ - struct i2400m_pl_data_hdr *pl_hdr; - skb_pull(skb, ETH_HLEN); - pl_hdr = skb_push(skb, sizeof(*pl_hdr)); - pl_hdr->reserved = 0; -} - - - -/* - * Cleanup resources acquired during i2400m_net_wake_tx() - * - * This is called by __i2400m_dev_stop and means we have to make sure - * the workqueue is flushed from any pending work. - */ -void i2400m_net_wake_stop(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *wake_tx_skb; - unsigned long flags; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - /* - * See i2400m_hard_start_xmit(), references are taken there and - * here we release them if the packet was still pending. - */ - cancel_work_sync(&i2400m->wake_tx_ws); - - spin_lock_irqsave(&i2400m->tx_lock, flags); - wake_tx_skb = i2400m->wake_tx_skb; - i2400m->wake_tx_skb = NULL; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - - if (wake_tx_skb) { - i2400m_put(i2400m); - kfree_skb(wake_tx_skb); - } - - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} - - -/* - * TX an skb to an idle device - * - * When the device is in basestation-idle mode, we need to wake it up - * and then TX. So we queue a work_struct for doing so. - * - * We need to get an extra ref for the skb (so it is not dropped), as - * well as be careful not to queue more than one request (won't help - * at all). If more than one request comes or there are errors, we - * just drop the packets (see i2400m_hard_start_xmit()). - */ -static -int i2400m_net_wake_tx(struct i2400m *i2400m, struct net_device *net_dev, - struct sk_buff *skb) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - unsigned long flags; - - d_fnstart(3, dev, "(skb %p net_dev %p)\n", skb, net_dev); - if (net_ratelimit()) { - d_printf(3, dev, "WAKE&NETTX: " - "skb %p sending %d bytes to radio\n", - skb, skb->len); - d_dump(4, dev, skb->data, skb->len); - } - /* We hold a ref count for i2400m and skb, so when - * stopping() the device, we need to cancel that work - * and if pending, release those resources. */ - result = 0; - spin_lock_irqsave(&i2400m->tx_lock, flags); - if (!i2400m->wake_tx_skb) { - netif_stop_queue(net_dev); - i2400m_get(i2400m); - i2400m->wake_tx_skb = skb_get(skb); /* transfer ref count */ - i2400m_tx_prep_header(skb); - result = schedule_work(&i2400m->wake_tx_ws); - WARN_ON(result == 0); - } - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - if (result == 0) { - /* Yes, this happens even if we stopped the - * queue -- blame the queue disciplines that - * queue without looking -- I guess there is a reason - * for that. */ - if (net_ratelimit()) - d_printf(1, dev, "NETTX: device exiting idle, " - "dropping skb %p, queue running %d\n", - skb, netif_queue_stopped(net_dev)); - result = -EBUSY; - } - d_fnend(3, dev, "(skb %p net_dev %p) = %d\n", skb, net_dev, result); - return result; -} - - -/* - * Transmit a packet to the base station on behalf of the network stack. - * - * Returns: 0 if ok, < 0 errno code on error. - * - * We need to pull the ethernet header and add the hardware header, - * which is currently set to all zeroes and reserved. - */ -static -int i2400m_net_tx(struct i2400m *i2400m, struct net_device *net_dev, - struct sk_buff *skb) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p net_dev %p skb %p)\n", - i2400m, net_dev, skb); - /* FIXME: check eth hdr, only IPv4 is routed by the device as of now */ - netif_trans_update(net_dev); - i2400m_tx_prep_header(skb); - d_printf(3, dev, "NETTX: skb %p sending %d bytes to radio\n", - skb, skb->len); - d_dump(4, dev, skb->data, skb->len); - result = i2400m_tx(i2400m, skb->data, skb->len, I2400M_PT_DATA); - d_fnend(3, dev, "(i2400m %p net_dev %p skb %p) = %d\n", - i2400m, net_dev, skb, result); - return result; -} - - -/* - * Transmit a packet to the base station on behalf of the network stack - * - * - * Returns: NETDEV_TX_OK (always, even in case of error) - * - * In case of error, we just drop it. Reasons: - * - * - we add a hw header to each skb, and if the network stack - * retries, we have no way to know if that skb has it or not. - * - * - network protocols have their own drop-recovery mechanisms - * - * - there is not much else we can do - * - * If the device is idle, we need to wake it up; that is an operation - * that will sleep. See i2400m_net_wake_tx() for details. - */ -static -netdev_tx_t i2400m_hard_start_xmit(struct sk_buff *skb, - struct net_device *net_dev) -{ - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - struct device *dev = i2400m_dev(i2400m); - int result = -1; - - d_fnstart(3, dev, "(skb %p net_dev %p)\n", skb, net_dev); - - if (skb_cow_head(skb, 0)) - goto drop; - - if (i2400m->state == I2400M_SS_IDLE) - result = i2400m_net_wake_tx(i2400m, net_dev, skb); - else - result = i2400m_net_tx(i2400m, net_dev, skb); - if (result < 0) { -drop: - net_dev->stats.tx_dropped++; - } else { - net_dev->stats.tx_packets++; - net_dev->stats.tx_bytes += skb->len; - } - dev_kfree_skb(skb); - d_fnend(3, dev, "(skb %p net_dev %p) = %d\n", skb, net_dev, result); - return NETDEV_TX_OK; -} - - -static -void i2400m_tx_timeout(struct net_device *net_dev, unsigned int txqueue) -{ - /* - * We might want to kick the device - * - * There is not much we can do though, as the device requires - * that we send the data aggregated. By the time we receive - * this, there might be data pending to be sent or not... - */ - net_dev->stats.tx_errors++; -} - - -/* - * Create a fake ethernet header - * - * For emulating an ethernet device, every received IP header has to - * be prefixed with an ethernet header. Fake it with the given - * protocol. - */ -static -void i2400m_rx_fake_eth_header(struct net_device *net_dev, - void *_eth_hdr, __be16 protocol) -{ - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - struct ethhdr *eth_hdr = _eth_hdr; - - memcpy(eth_hdr->h_dest, net_dev->dev_addr, sizeof(eth_hdr->h_dest)); - memcpy(eth_hdr->h_source, i2400m->src_mac_addr, - sizeof(eth_hdr->h_source)); - eth_hdr->h_proto = protocol; -} - - -/* - * i2400m_net_rx - pass a network packet to the stack - * - * @i2400m: device instance - * @skb_rx: the skb where the buffer pointed to by @buf is - * @i: 1 if payload is the only one - * @buf: pointer to the buffer containing the data - * @len: buffer's length - * - * This is only used now for the v1.3 firmware. It will be deprecated - * in >= 2.6.31. - * - * Note that due to firmware limitations, we don't have space to add - * an ethernet header, so we need to copy each packet. Firmware - * versions >= v1.4 fix this [see i2400m_net_erx()]. - * - * We just clone the skb and set it up so that it's skb->data pointer - * points to "buf" and it's length. - * - * Note that if the payload is the last (or the only one) in a - * multi-payload message, we don't clone the SKB but just reuse it. - * - * This function is normally run from a thread context. However, we - * still use netif_rx() instead of netif_receive_skb() as was - * recommended in the mailing list. Reason is in some stress tests - * when sending/receiving a lot of data we seem to hit a softlock in - * the kernel's TCP implementation [aroudn tcp_delay_timer()]. Using - * netif_rx() took care of the issue. - * - * This is, of course, still open to do more research on why running - * with netif_receive_skb() hits this softlock. FIXME. - * - * FIXME: currently we don't do any efforts at distinguishing if what - * we got was an IPv4 or IPv6 header, to setup the protocol field - * correctly. - */ -void i2400m_net_rx(struct i2400m *i2400m, struct sk_buff *skb_rx, - unsigned i, const void *buf, int buf_len) -{ - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *skb; - - d_fnstart(2, dev, "(i2400m %p buf %p buf_len %d)\n", - i2400m, buf, buf_len); - if (i) { - skb = skb_get(skb_rx); - d_printf(2, dev, "RX: reusing first payload skb %p\n", skb); - skb_pull(skb, buf - (void *) skb->data); - skb_trim(skb, (void *) skb_end_pointer(skb) - buf); - } else { - /* Yes, this is bad -- a lot of overhead -- see - * comments at the top of the file */ - skb = __netdev_alloc_skb(net_dev, buf_len, GFP_KERNEL); - if (skb == NULL) { - dev_err(dev, "NETRX: no memory to realloc skb\n"); - net_dev->stats.rx_dropped++; - goto error_skb_realloc; - } - skb_put_data(skb, buf, buf_len); - } - i2400m_rx_fake_eth_header(i2400m->wimax_dev.net_dev, - skb->data - ETH_HLEN, - cpu_to_be16(ETH_P_IP)); - skb_set_mac_header(skb, -ETH_HLEN); - skb->dev = i2400m->wimax_dev.net_dev; - skb->protocol = htons(ETH_P_IP); - net_dev->stats.rx_packets++; - net_dev->stats.rx_bytes += buf_len; - d_printf(3, dev, "NETRX: receiving %d bytes to network stack\n", - buf_len); - d_dump(4, dev, buf, buf_len); - netif_rx_ni(skb); /* see notes in function header */ -error_skb_realloc: - d_fnend(2, dev, "(i2400m %p buf %p buf_len %d) = void\n", - i2400m, buf, buf_len); -} - - -/* - * i2400m_net_erx - pass a network packet to the stack (extended version) - * - * @i2400m: device descriptor - * @skb: the skb where the packet is - the skb should be set to point - * at the IP packet; this function will add ethernet headers if - * needed. - * @cs: packet type - * - * This is only used now for firmware >= v1.4. Note it is quite - * similar to i2400m_net_rx() (used only for v1.3 firmware). - * - * This function is normally run from a thread context. However, we - * still use netif_rx() instead of netif_receive_skb() as was - * recommended in the mailing list. Reason is in some stress tests - * when sending/receiving a lot of data we seem to hit a softlock in - * the kernel's TCP implementation [aroudn tcp_delay_timer()]. Using - * netif_rx() took care of the issue. - * - * This is, of course, still open to do more research on why running - * with netif_receive_skb() hits this softlock. FIXME. - */ -void i2400m_net_erx(struct i2400m *i2400m, struct sk_buff *skb, - enum i2400m_cs cs) -{ - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(2, dev, "(i2400m %p skb %p [%u] cs %d)\n", - i2400m, skb, skb->len, cs); - switch(cs) { - case I2400M_CS_IPV4_0: - case I2400M_CS_IPV4: - i2400m_rx_fake_eth_header(i2400m->wimax_dev.net_dev, - skb->data - ETH_HLEN, - cpu_to_be16(ETH_P_IP)); - skb_set_mac_header(skb, -ETH_HLEN); - skb->dev = i2400m->wimax_dev.net_dev; - skb->protocol = htons(ETH_P_IP); - net_dev->stats.rx_packets++; - net_dev->stats.rx_bytes += skb->len; - break; - default: - dev_err(dev, "ERX: BUG? CS type %u unsupported\n", cs); - goto error; - - } - d_printf(3, dev, "ERX: receiving %d bytes to the network stack\n", - skb->len); - d_dump(4, dev, skb->data, skb->len); - netif_rx_ni(skb); /* see notes in function header */ -error: - d_fnend(2, dev, "(i2400m %p skb %p [%u] cs %d) = void\n", - i2400m, skb, skb->len, cs); -} - -static const struct net_device_ops i2400m_netdev_ops = { - .ndo_open = i2400m_open, - .ndo_stop = i2400m_stop, - .ndo_start_xmit = i2400m_hard_start_xmit, - .ndo_tx_timeout = i2400m_tx_timeout, -}; - -static void i2400m_get_drvinfo(struct net_device *net_dev, - struct ethtool_drvinfo *info) -{ - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - - strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); - strlcpy(info->fw_version, i2400m->fw_name ? : "", - sizeof(info->fw_version)); - if (net_dev->dev.parent) - strlcpy(info->bus_info, dev_name(net_dev->dev.parent), - sizeof(info->bus_info)); -} - -static const struct ethtool_ops i2400m_ethtool_ops = { - .get_drvinfo = i2400m_get_drvinfo, - .get_link = ethtool_op_get_link, -}; - -/** - * i2400m_netdev_setup - Setup setup @net_dev's i2400m private data - * - * Called by alloc_netdev() - */ -void i2400m_netdev_setup(struct net_device *net_dev) -{ - d_fnstart(3, NULL, "(net_dev %p)\n", net_dev); - ether_setup(net_dev); - net_dev->mtu = I2400M_MAX_MTU; - net_dev->min_mtu = 0; - net_dev->max_mtu = I2400M_MAX_MTU; - net_dev->tx_queue_len = I2400M_TX_QLEN; - net_dev->features = - NETIF_F_VLAN_CHALLENGED - | NETIF_F_HIGHDMA; - net_dev->flags = - IFF_NOARP /* i2400m is apure IP device */ - & (~IFF_BROADCAST /* i2400m is P2P */ - & ~IFF_MULTICAST); - net_dev->watchdog_timeo = I2400M_TX_TIMEOUT; - net_dev->netdev_ops = &i2400m_netdev_ops; - net_dev->ethtool_ops = &i2400m_ethtool_ops; - d_fnend(3, NULL, "(net_dev %p) = void\n", net_dev); -} -EXPORT_SYMBOL_GPL(i2400m_netdev_setup); - diff --git a/drivers/net/wimax/i2400m/op-rfkill.c b/drivers/net/wimax/i2400m/op-rfkill.c deleted file mode 100644 index 5c79f052cad2..000000000000 --- a/drivers/net/wimax/i2400m/op-rfkill.c +++ /dev/null @@ -1,196 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Intel Wireless WiMAX Connection 2400m - * Implement backend for the WiMAX stack rfkill support - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - * - * The WiMAX kernel stack integrates into RF-Kill and keeps the - * switches's status. We just need to: - * - * - report changes in the HW RF Kill switch [with - * wimax_rfkill_{sw,hw}_report(), which happens when we detect those - * indications coming through hardware reports]. We also do it on - * initialization to let the stack know the initial HW state. - * - * - implement indications from the stack to change the SW RF Kill - * switch (coming from sysfs, the wimax stack or user space). - */ -#include "i2400m.h" -#include -#include - - - -#define D_SUBMODULE rfkill -#include "debug-levels.h" - -/* - * Return true if the i2400m radio is in the requested wimax_rf_state state - * - */ -static -int i2400m_radio_is(struct i2400m *i2400m, enum wimax_rf_state state) -{ - if (state == WIMAX_RF_OFF) - return i2400m->state == I2400M_SS_RF_OFF - || i2400m->state == I2400M_SS_RF_SHUTDOWN; - else if (state == WIMAX_RF_ON) - /* state == WIMAX_RF_ON */ - return i2400m->state != I2400M_SS_RF_OFF - && i2400m->state != I2400M_SS_RF_SHUTDOWN; - else { - BUG(); - return -EINVAL; /* shut gcc warnings on certain arches */ - } -} - - -/* - * WiMAX stack operation: implement SW RFKill toggling - * - * @wimax_dev: device descriptor - * @skb: skb where the message has been received; skb->data is - * expected to point to the message payload. - * @genl_info: passed by the generic netlink layer - * - * Generic Netlink will call this function when a message is sent from - * userspace to change the software RF-Kill switch status. - * - * This function will set the device's software RF-Kill switch state to - * match what is requested. - * - * NOTE: the i2400m has a strict state machine; we can only set the - * RF-Kill switch when it is on, the HW RF-Kill is on and the - * device is initialized. So we ignore errors steaming from not - * being in the right state (-EILSEQ). - */ -int i2400m_op_rfkill_sw_toggle(struct wimax_dev *wimax_dev, - enum wimax_rf_state state) -{ - int result; - struct i2400m *i2400m = wimax_dev_to_i2400m(wimax_dev); - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *ack_skb; - struct { - struct i2400m_l3l4_hdr hdr; - struct i2400m_tlv_rf_operation sw_rf; - } __packed *cmd; - char strerr[32]; - - d_fnstart(4, dev, "(wimax_dev %p state %d)\n", wimax_dev, state); - - result = -ENOMEM; - cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); - if (cmd == NULL) - goto error_alloc; - cmd->hdr.type = cpu_to_le16(I2400M_MT_CMD_RF_CONTROL); - cmd->hdr.length = sizeof(cmd->sw_rf); - cmd->hdr.version = cpu_to_le16(I2400M_L3L4_VERSION); - cmd->sw_rf.hdr.type = cpu_to_le16(I2400M_TLV_RF_OPERATION); - cmd->sw_rf.hdr.length = cpu_to_le16(sizeof(cmd->sw_rf.status)); - switch (state) { - case WIMAX_RF_OFF: /* RFKILL ON, radio OFF */ - cmd->sw_rf.status = cpu_to_le32(2); - break; - case WIMAX_RF_ON: /* RFKILL OFF, radio ON */ - cmd->sw_rf.status = cpu_to_le32(1); - break; - default: - BUG(); - } - - ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); - result = PTR_ERR(ack_skb); - if (IS_ERR(ack_skb)) { - dev_err(dev, "Failed to issue 'RF Control' command: %d\n", - result); - goto error_msg_to_dev; - } - result = i2400m_msg_check_status(wimax_msg_data(ack_skb), - strerr, sizeof(strerr)); - if (result < 0) { - dev_err(dev, "'RF Control' (0x%04x) command failed: %d - %s\n", - I2400M_MT_CMD_RF_CONTROL, result, strerr); - goto error_cmd; - } - - /* Now we wait for the state to change to RADIO_OFF or RADIO_ON */ - result = wait_event_timeout( - i2400m->state_wq, i2400m_radio_is(i2400m, state), - 5 * HZ); - if (result == 0) - result = -ETIMEDOUT; - if (result < 0) - dev_err(dev, "Error waiting for device to toggle RF state: " - "%d\n", result); - result = 0; -error_cmd: - kfree_skb(ack_skb); -error_msg_to_dev: -error_alloc: - d_fnend(4, dev, "(wimax_dev %p state %d) = %d\n", - wimax_dev, state, result); - kfree(cmd); - return result; -} - - -/* - * Inform the WiMAX stack of changes in the RF Kill switches reported - * by the device - * - * @i2400m: device descriptor - * @rfss: TLV for RF Switches status; already validated - * - * NOTE: the reports on RF switch status cannot be trusted - * or used until the device is in a state of RADIO_OFF - * or greater. - */ -void i2400m_report_tlv_rf_switches_status( - struct i2400m *i2400m, - const struct i2400m_tlv_rf_switches_status *rfss) -{ - struct device *dev = i2400m_dev(i2400m); - enum i2400m_rf_switch_status hw, sw; - enum wimax_st wimax_state; - - sw = le32_to_cpu(rfss->sw_rf_switch); - hw = le32_to_cpu(rfss->hw_rf_switch); - - d_fnstart(3, dev, "(i2400m %p rfss %p [hw %u sw %u])\n", - i2400m, rfss, hw, sw); - /* We only process rw switch evens when the device has been - * fully initialized */ - wimax_state = wimax_state_get(&i2400m->wimax_dev); - if (wimax_state < WIMAX_ST_RADIO_OFF) { - d_printf(3, dev, "ignoring RF switches report, state %u\n", - wimax_state); - goto out; - } - switch (sw) { - case I2400M_RF_SWITCH_ON: /* RF Kill disabled (radio on) */ - wimax_report_rfkill_sw(&i2400m->wimax_dev, WIMAX_RF_ON); - break; - case I2400M_RF_SWITCH_OFF: /* RF Kill enabled (radio off) */ - wimax_report_rfkill_sw(&i2400m->wimax_dev, WIMAX_RF_OFF); - break; - default: - dev_err(dev, "HW BUG? Unknown RF SW state 0x%x\n", sw); - } - - switch (hw) { - case I2400M_RF_SWITCH_ON: /* RF Kill disabled (radio on) */ - wimax_report_rfkill_hw(&i2400m->wimax_dev, WIMAX_RF_ON); - break; - case I2400M_RF_SWITCH_OFF: /* RF Kill enabled (radio off) */ - wimax_report_rfkill_hw(&i2400m->wimax_dev, WIMAX_RF_OFF); - break; - default: - dev_err(dev, "HW BUG? Unknown RF HW state 0x%x\n", hw); - } -out: - d_fnend(3, dev, "(i2400m %p rfss %p [hw %u sw %u]) = void\n", - i2400m, rfss, hw, sw); -} diff --git a/drivers/net/wimax/i2400m/rx.c b/drivers/net/wimax/i2400m/rx.c deleted file mode 100644 index c9fb619a9e01..000000000000 --- a/drivers/net/wimax/i2400m/rx.c +++ /dev/null @@ -1,1395 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * Handle incoming traffic and deliver it to the control or data planes - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * - Initial implementation - * Inaky Perez-Gonzalez - * - Use skb_clone(), break up processing in chunks - * - Split transport/device specific - * - Make buffer size dynamic to exert less memory pressure - * - RX reorder support - * - * This handles the RX path. - * - * We receive an RX message from the bus-specific driver, which - * contains one or more payloads that have potentially different - * destinataries (data or control paths). - * - * So we just take that payload from the transport specific code in - * the form of an skb, break it up in chunks (a cloned skb each in the - * case of network packets) and pass it to netdev or to the - * command/ack handler (and from there to the WiMAX stack). - * - * PROTOCOL FORMAT - * - * The format of the buffer is: - * - * HEADER (struct i2400m_msg_hdr) - * PAYLOAD DESCRIPTOR 0 (struct i2400m_pld) - * PAYLOAD DESCRIPTOR 1 - * ... - * PAYLOAD DESCRIPTOR N - * PAYLOAD 0 (raw bytes) - * PAYLOAD 1 - * ... - * PAYLOAD N - * - * See tx.c for a deeper description on alignment requirements and - * other fun facts of it. - * - * DATA PACKETS - * - * In firmwares <= v1.3, data packets have no header for RX, but they - * do for TX (currently unused). - * - * In firmware >= 1.4, RX packets have an extended header (16 - * bytes). This header conveys information for management of host - * reordering of packets (the device offloads storage of the packets - * for reordering to the host). Read below for more information. - * - * The header is used as dummy space to emulate an ethernet header and - * thus be able to act as an ethernet device without having to reallocate. - * - * DATA RX REORDERING - * - * Starting in firmware v1.4, the device can deliver packets for - * delivery with special reordering information; this allows it to - * more effectively do packet management when some frames were lost in - * the radio traffic. - * - * Thus, for RX packets that come out of order, the device gives the - * driver enough information to queue them properly and then at some - * point, the signal to deliver the whole (or part) of the queued - * packets to the networking stack. There are 16 such queues. - * - * This only happens when a packet comes in with the "need reorder" - * flag set in the RX header. When such bit is set, the following - * operations might be indicated: - * - * - reset queue: send all queued packets to the OS - * - * - queue: queue a packet - * - * - update ws: update the queue's window start and deliver queued - * packets that meet the criteria - * - * - queue & update ws: queue a packet, update the window start and - * deliver queued packets that meet the criteria - * - * (delivery criteria: the packet's [normalized] sequence number is - * lower than the new [normalized] window start). - * - * See the i2400m_roq_*() functions for details. - * - * ROADMAP - * - * i2400m_rx - * i2400m_rx_msg_hdr_check - * i2400m_rx_pl_descr_check - * i2400m_rx_payload - * i2400m_net_rx - * i2400m_rx_edata - * i2400m_net_erx - * i2400m_roq_reset - * i2400m_net_erx - * i2400m_roq_queue - * __i2400m_roq_queue - * i2400m_roq_update_ws - * __i2400m_roq_update_ws - * i2400m_net_erx - * i2400m_roq_queue_update_ws - * __i2400m_roq_queue - * __i2400m_roq_update_ws - * i2400m_net_erx - * i2400m_rx_ctl - * i2400m_msg_size_check - * i2400m_report_hook_work [in a workqueue] - * i2400m_report_hook - * wimax_msg_to_user - * i2400m_rx_ctl_ack - * wimax_msg_to_user_alloc - * i2400m_rx_trace - * i2400m_msg_size_check - * wimax_msg - */ -#include -#include -#include -#include -#include -#include -#include -#include "i2400m.h" - - -#define D_SUBMODULE rx -#include "debug-levels.h" - -static int i2400m_rx_reorder_disabled; /* 0 (rx reorder enabled) by default */ -module_param_named(rx_reorder_disabled, i2400m_rx_reorder_disabled, int, 0644); -MODULE_PARM_DESC(rx_reorder_disabled, - "If true, RX reordering will be disabled."); - -struct i2400m_report_hook_args { - struct sk_buff *skb_rx; - const struct i2400m_l3l4_hdr *l3l4_hdr; - size_t size; - struct list_head list_node; -}; - - -/* - * Execute i2400m_report_hook in a workqueue - * - * Goes over the list of queued reports in i2400m->rx_reports and - * processes them. - * - * NOTE: refcounts on i2400m are not needed because we flush the - * workqueue this runs on (i2400m->work_queue) before destroying - * i2400m. - */ -void i2400m_report_hook_work(struct work_struct *ws) -{ - struct i2400m *i2400m = container_of(ws, struct i2400m, rx_report_ws); - struct device *dev = i2400m_dev(i2400m); - struct i2400m_report_hook_args *args, *args_next; - LIST_HEAD(list); - unsigned long flags; - - while (1) { - spin_lock_irqsave(&i2400m->rx_lock, flags); - list_splice_init(&i2400m->rx_reports, &list); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - if (list_empty(&list)) - break; - else - d_printf(1, dev, "processing queued reports\n"); - list_for_each_entry_safe(args, args_next, &list, list_node) { - d_printf(2, dev, "processing queued report %p\n", args); - i2400m_report_hook(i2400m, args->l3l4_hdr, args->size); - kfree_skb(args->skb_rx); - list_del(&args->list_node); - kfree(args); - } - } -} - - -/* - * Flush the list of queued reports - */ -static -void i2400m_report_hook_flush(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - struct i2400m_report_hook_args *args, *args_next; - LIST_HEAD(list); - unsigned long flags; - - d_printf(1, dev, "flushing queued reports\n"); - spin_lock_irqsave(&i2400m->rx_lock, flags); - list_splice_init(&i2400m->rx_reports, &list); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - list_for_each_entry_safe(args, args_next, &list, list_node) { - d_printf(2, dev, "flushing queued report %p\n", args); - kfree_skb(args->skb_rx); - list_del(&args->list_node); - kfree(args); - } -} - - -/* - * Queue a report for later processing - * - * @i2400m: device descriptor - * @skb_rx: skb that contains the payload (for reference counting) - * @l3l4_hdr: pointer to the control - * @size: size of the message - */ -static -void i2400m_report_hook_queue(struct i2400m *i2400m, struct sk_buff *skb_rx, - const void *l3l4_hdr, size_t size) -{ - struct device *dev = i2400m_dev(i2400m); - unsigned long flags; - struct i2400m_report_hook_args *args; - - args = kzalloc(sizeof(*args), GFP_NOIO); - if (args) { - args->skb_rx = skb_get(skb_rx); - args->l3l4_hdr = l3l4_hdr; - args->size = size; - spin_lock_irqsave(&i2400m->rx_lock, flags); - list_add_tail(&args->list_node, &i2400m->rx_reports); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - d_printf(2, dev, "queued report %p\n", args); - rmb(); /* see i2400m->ready's documentation */ - if (likely(i2400m->ready)) /* only send if up */ - queue_work(i2400m->work_queue, &i2400m->rx_report_ws); - } else { - if (printk_ratelimit()) - dev_err(dev, "%s:%u: Can't allocate %zu B\n", - __func__, __LINE__, sizeof(*args)); - } -} - - -/* - * Process an ack to a command - * - * @i2400m: device descriptor - * @payload: pointer to message - * @size: size of the message - * - * Pass the acknodledgment (in an skb) to the thread that is waiting - * for it in i2400m->msg_completion. - * - * We need to coordinate properly with the thread waiting for the - * ack. Check if it is waiting or if it is gone. We loose the spinlock - * to avoid allocating on atomic contexts (yeah, could use GFP_ATOMIC, - * but this is not so speed critical). - */ -static -void i2400m_rx_ctl_ack(struct i2400m *i2400m, - const void *payload, size_t size) -{ - struct device *dev = i2400m_dev(i2400m); - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - unsigned long flags; - struct sk_buff *ack_skb; - - /* Anyone waiting for an answer? */ - spin_lock_irqsave(&i2400m->rx_lock, flags); - if (i2400m->ack_skb != ERR_PTR(-EINPROGRESS)) { - dev_err(dev, "Huh? reply to command with no waiters\n"); - goto error_no_waiter; - } - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - - ack_skb = wimax_msg_alloc(wimax_dev, NULL, payload, size, GFP_KERNEL); - - /* Check waiter didn't time out waiting for the answer... */ - spin_lock_irqsave(&i2400m->rx_lock, flags); - if (i2400m->ack_skb != ERR_PTR(-EINPROGRESS)) { - d_printf(1, dev, "Huh? waiter for command reply cancelled\n"); - goto error_waiter_cancelled; - } - if (IS_ERR(ack_skb)) - dev_err(dev, "CMD/GET/SET ack: cannot allocate SKB\n"); - i2400m->ack_skb = ack_skb; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - complete(&i2400m->msg_completion); - return; - -error_waiter_cancelled: - if (!IS_ERR(ack_skb)) - kfree_skb(ack_skb); -error_no_waiter: - spin_unlock_irqrestore(&i2400m->rx_lock, flags); -} - - -/* - * Receive and process a control payload - * - * @i2400m: device descriptor - * @skb_rx: skb that contains the payload (for reference counting) - * @payload: pointer to message - * @size: size of the message - * - * There are two types of control RX messages: reports (asynchronous, - * like your every day interrupts) and 'acks' (reponses to a command, - * get or set request). - * - * If it is a report, we run hooks on it (to extract information for - * things we need to do in the driver) and then pass it over to the - * WiMAX stack to send it to user space. - * - * NOTE: report processing is done in a workqueue specific to the - * generic driver, to avoid deadlocks in the system. - * - * If it is not a report, it is an ack to a previously executed - * command, set or get, so wake up whoever is waiting for it from - * i2400m_msg_to_dev(). i2400m_rx_ctl_ack() takes care of that. - * - * Note that the sizes we pass to other functions from here are the - * sizes of the _l3l4_hdr + payload, not full buffer sizes, as we have - * verified in _msg_size_check() that they are congruent. - * - * For reports: We can't clone the original skb where the data is - * because we need to send this up via netlink; netlink has to add - * headers and we can't overwrite what's preceding the payload...as - * it is another message. So we just dup them. - */ -static -void i2400m_rx_ctl(struct i2400m *i2400m, struct sk_buff *skb_rx, - const void *payload, size_t size) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_l3l4_hdr *l3l4_hdr = payload; - unsigned msg_type; - - result = i2400m_msg_size_check(i2400m, l3l4_hdr, size); - if (result < 0) { - dev_err(dev, "HW BUG? device sent a bad message: %d\n", - result); - goto error_check; - } - msg_type = le16_to_cpu(l3l4_hdr->type); - d_printf(1, dev, "%s 0x%04x: %zu bytes\n", - msg_type & I2400M_MT_REPORT_MASK ? "REPORT" : "CMD/SET/GET", - msg_type, size); - d_dump(2, dev, l3l4_hdr, size); - if (msg_type & I2400M_MT_REPORT_MASK) { - /* - * Process each report - * - * - has to be ran serialized as well - * - * - the handling might force the execution of - * commands. That might cause reentrancy issues with - * bus-specific subdrivers and workqueues, so the we - * run it in a separate workqueue. - * - * - when the driver is not yet ready to handle them, - * they are queued and at some point the queue is - * restarted [NOTE: we can't queue SKBs directly, as - * this might be a piece of a SKB, not the whole - * thing, and this is cheaper than cloning the - * SKB]. - * - * Note we don't do refcounting for the device - * structure; this is because before destroying - * 'i2400m', we make sure to flush the - * i2400m->work_queue, so there are no issues. - */ - i2400m_report_hook_queue(i2400m, skb_rx, l3l4_hdr, size); - if (unlikely(i2400m->trace_msg_from_user)) - wimax_msg(&i2400m->wimax_dev, "echo", - l3l4_hdr, size, GFP_KERNEL); - result = wimax_msg(&i2400m->wimax_dev, NULL, l3l4_hdr, size, - GFP_KERNEL); - if (result < 0) - dev_err(dev, "error sending report to userspace: %d\n", - result); - } else /* an ack to a CMD, GET or SET */ - i2400m_rx_ctl_ack(i2400m, payload, size); -error_check: - return; -} - - -/* - * Receive and send up a trace - * - * @i2400m: device descriptor - * @skb_rx: skb that contains the trace (for reference counting) - * @payload: pointer to trace message inside the skb - * @size: size of the message - * - * THe i2400m might produce trace information (diagnostics) and we - * send them through a different kernel-to-user pipe (to avoid - * clogging it). - * - * As in i2400m_rx_ctl(), we can't clone the original skb where the - * data is because we need to send this up via netlink; netlink has to - * add headers and we can't overwrite what's preceding the - * payload...as it is another message. So we just dup them. - */ -static -void i2400m_rx_trace(struct i2400m *i2400m, - const void *payload, size_t size) -{ - int result; - struct device *dev = i2400m_dev(i2400m); - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - const struct i2400m_l3l4_hdr *l3l4_hdr = payload; - unsigned msg_type; - - result = i2400m_msg_size_check(i2400m, l3l4_hdr, size); - if (result < 0) { - dev_err(dev, "HW BUG? device sent a bad trace message: %d\n", - result); - goto error_check; - } - msg_type = le16_to_cpu(l3l4_hdr->type); - d_printf(1, dev, "Trace %s 0x%04x: %zu bytes\n", - msg_type & I2400M_MT_REPORT_MASK ? "REPORT" : "CMD/SET/GET", - msg_type, size); - d_dump(2, dev, l3l4_hdr, size); - result = wimax_msg(wimax_dev, "trace", l3l4_hdr, size, GFP_KERNEL); - if (result < 0) - dev_err(dev, "error sending trace to userspace: %d\n", - result); -error_check: - return; -} - - -/* - * Reorder queue data stored on skb->cb while the skb is queued in the - * reorder queues. - */ -struct i2400m_roq_data { - unsigned sn; /* Serial number for the skb */ - enum i2400m_cs cs; /* packet type for the skb */ -}; - - -/* - * ReOrder Queue - * - * @ws: Window Start; sequence number where the current window start - * is for this queue - * @queue: the skb queue itself - * @log: circular ring buffer used to log information about the - * reorder process in this queue that can be displayed in case of - * error to help diagnose it. - * - * This is the head for a list of skbs. In the skb->cb member of the - * skb when queued here contains a 'struct i2400m_roq_data' were we - * store the sequence number (sn) and the cs (packet type) coming from - * the RX payload header from the device. - */ -struct i2400m_roq -{ - unsigned ws; - struct sk_buff_head queue; - struct i2400m_roq_log *log; -}; - - -static -void __i2400m_roq_init(struct i2400m_roq *roq) -{ - roq->ws = 0; - skb_queue_head_init(&roq->queue); -} - - -static -unsigned __i2400m_roq_index(struct i2400m *i2400m, struct i2400m_roq *roq) -{ - return ((unsigned long) roq - (unsigned long) i2400m->rx_roq) - / sizeof(*roq); -} - - -/* - * Normalize a sequence number based on the queue's window start - * - * nsn = (sn - ws) % 2048 - * - * Note that if @sn < @roq->ws, we still need a positive number; %'s - * sign is implementation specific, so we normalize it by adding 2048 - * to bring it to be positive. - */ -static -unsigned __i2400m_roq_nsn(struct i2400m_roq *roq, unsigned sn) -{ - int r; - r = ((int) sn - (int) roq->ws) % 2048; - if (r < 0) - r += 2048; - return r; -} - - -/* - * Circular buffer to keep the last N reorder operations - * - * In case something fails, dumb then to try to come up with what - * happened. - */ -enum { - I2400M_ROQ_LOG_LENGTH = 32, -}; - -struct i2400m_roq_log { - struct i2400m_roq_log_entry { - enum i2400m_ro_type type; - unsigned ws, count, sn, nsn, new_ws; - } entry[I2400M_ROQ_LOG_LENGTH]; - unsigned in, out; -}; - - -/* Print a log entry */ -static -void i2400m_roq_log_entry_print(struct i2400m *i2400m, unsigned index, - unsigned e_index, - struct i2400m_roq_log_entry *e) -{ - struct device *dev = i2400m_dev(i2400m); - - switch(e->type) { - case I2400M_RO_TYPE_RESET: - dev_err(dev, "q#%d reset ws %u cnt %u sn %u/%u" - " - new nws %u\n", - index, e->ws, e->count, e->sn, e->nsn, e->new_ws); - break; - case I2400M_RO_TYPE_PACKET: - dev_err(dev, "q#%d queue ws %u cnt %u sn %u/%u\n", - index, e->ws, e->count, e->sn, e->nsn); - break; - case I2400M_RO_TYPE_WS: - dev_err(dev, "q#%d update_ws ws %u cnt %u sn %u/%u" - " - new nws %u\n", - index, e->ws, e->count, e->sn, e->nsn, e->new_ws); - break; - case I2400M_RO_TYPE_PACKET_WS: - dev_err(dev, "q#%d queue_update_ws ws %u cnt %u sn %u/%u" - " - new nws %u\n", - index, e->ws, e->count, e->sn, e->nsn, e->new_ws); - break; - default: - dev_err(dev, "q#%d BUG? entry %u - unknown type %u\n", - index, e_index, e->type); - break; - } -} - - -static -void i2400m_roq_log_add(struct i2400m *i2400m, - struct i2400m_roq *roq, enum i2400m_ro_type type, - unsigned ws, unsigned count, unsigned sn, - unsigned nsn, unsigned new_ws) -{ - struct i2400m_roq_log_entry *e; - unsigned cnt_idx; - int index = __i2400m_roq_index(i2400m, roq); - - /* if we run out of space, we eat from the end */ - if (roq->log->in - roq->log->out == I2400M_ROQ_LOG_LENGTH) - roq->log->out++; - cnt_idx = roq->log->in++ % I2400M_ROQ_LOG_LENGTH; - e = &roq->log->entry[cnt_idx]; - - e->type = type; - e->ws = ws; - e->count = count; - e->sn = sn; - e->nsn = nsn; - e->new_ws = new_ws; - - if (d_test(1)) - i2400m_roq_log_entry_print(i2400m, index, cnt_idx, e); -} - - -/* Dump all the entries in the FIFO and reinitialize it */ -static -void i2400m_roq_log_dump(struct i2400m *i2400m, struct i2400m_roq *roq) -{ - unsigned cnt, cnt_idx; - struct i2400m_roq_log_entry *e; - int index = __i2400m_roq_index(i2400m, roq); - - BUG_ON(roq->log->out > roq->log->in); - for (cnt = roq->log->out; cnt < roq->log->in; cnt++) { - cnt_idx = cnt % I2400M_ROQ_LOG_LENGTH; - e = &roq->log->entry[cnt_idx]; - i2400m_roq_log_entry_print(i2400m, index, cnt_idx, e); - memset(e, 0, sizeof(*e)); - } - roq->log->in = roq->log->out = 0; -} - - -/* - * Backbone for the queuing of an skb (by normalized sequence number) - * - * @i2400m: device descriptor - * @roq: reorder queue where to add - * @skb: the skb to add - * @sn: the sequence number of the skb - * @nsn: the normalized sequence number of the skb (pre-computed by the - * caller from the @sn and @roq->ws). - * - * We try first a couple of quick cases: - * - * - the queue is empty - * - the skb would be appended to the queue - * - * These will be the most common operations. - * - * If these fail, then we have to do a sorted insertion in the queue, - * which is the slowest path. - * - * We don't have to acquire a reference count as we are going to own it. - */ -static -void __i2400m_roq_queue(struct i2400m *i2400m, struct i2400m_roq *roq, - struct sk_buff *skb, unsigned sn, unsigned nsn) -{ - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *skb_itr; - struct i2400m_roq_data *roq_data_itr, *roq_data; - unsigned nsn_itr; - - d_fnstart(4, dev, "(i2400m %p roq %p skb %p sn %u nsn %u)\n", - i2400m, roq, skb, sn, nsn); - - roq_data = (struct i2400m_roq_data *) &skb->cb; - BUILD_BUG_ON(sizeof(*roq_data) > sizeof(skb->cb)); - roq_data->sn = sn; - d_printf(3, dev, "ERX: roq %p [ws %u] nsn %d sn %u\n", - roq, roq->ws, nsn, roq_data->sn); - - /* Queues will be empty on not-so-bad environments, so try - * that first */ - if (skb_queue_empty(&roq->queue)) { - d_printf(2, dev, "ERX: roq %p - first one\n", roq); - __skb_queue_head(&roq->queue, skb); - goto out; - } - /* Now try append, as most of the operations will be that */ - skb_itr = skb_peek_tail(&roq->queue); - roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; - nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); - /* NSN bounds assumed correct (checked when it was queued) */ - if (nsn >= nsn_itr) { - d_printf(2, dev, "ERX: roq %p - appended after %p (nsn %d sn %u)\n", - roq, skb_itr, nsn_itr, roq_data_itr->sn); - __skb_queue_tail(&roq->queue, skb); - goto out; - } - /* None of the fast paths option worked. Iterate to find the - * right spot where to insert the packet; we know the queue is - * not empty, so we are not the first ones; we also know we - * are not going to be the last ones. The list is sorted, so - * we have to insert before the the first guy with an nsn_itr - * greater that our nsn. */ - skb_queue_walk(&roq->queue, skb_itr) { - roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; - nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); - /* NSN bounds assumed correct (checked when it was queued) */ - if (nsn_itr > nsn) { - d_printf(2, dev, "ERX: roq %p - queued before %p " - "(nsn %d sn %u)\n", roq, skb_itr, nsn_itr, - roq_data_itr->sn); - __skb_queue_before(&roq->queue, skb_itr, skb); - goto out; - } - } - /* If we get here, that is VERY bad -- print info to help - * diagnose and crash it */ - dev_err(dev, "SW BUG? failed to insert packet\n"); - dev_err(dev, "ERX: roq %p [ws %u] skb %p nsn %d sn %u\n", - roq, roq->ws, skb, nsn, roq_data->sn); - skb_queue_walk(&roq->queue, skb_itr) { - roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; - nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); - /* NSN bounds assumed correct (checked when it was queued) */ - dev_err(dev, "ERX: roq %p skb_itr %p nsn %d sn %u\n", - roq, skb_itr, nsn_itr, roq_data_itr->sn); - } - BUG(); -out: - d_fnend(4, dev, "(i2400m %p roq %p skb %p sn %u nsn %d) = void\n", - i2400m, roq, skb, sn, nsn); -} - - -/* - * Backbone for the update window start operation - * - * @i2400m: device descriptor - * @roq: Reorder queue - * @sn: New sequence number - * - * Updates the window start of a queue; when doing so, it must deliver - * to the networking stack all the queued skb's whose normalized - * sequence number is lower than the new normalized window start. - */ -static -unsigned __i2400m_roq_update_ws(struct i2400m *i2400m, struct i2400m_roq *roq, - unsigned sn) -{ - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *skb_itr, *tmp_itr; - struct i2400m_roq_data *roq_data_itr; - unsigned new_nws, nsn_itr; - - new_nws = __i2400m_roq_nsn(roq, sn); - /* - * For type 2(update_window_start) rx messages, there is no - * need to check if the normalized sequence number is greater 1023. - * Simply insert and deliver all packets to the host up to the - * window start. - */ - skb_queue_walk_safe(&roq->queue, skb_itr, tmp_itr) { - roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; - nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); - /* NSN bounds assumed correct (checked when it was queued) */ - if (nsn_itr < new_nws) { - d_printf(2, dev, "ERX: roq %p - release skb %p " - "(nsn %u/%u new nws %u)\n", - roq, skb_itr, nsn_itr, roq_data_itr->sn, - new_nws); - __skb_unlink(skb_itr, &roq->queue); - i2400m_net_erx(i2400m, skb_itr, roq_data_itr->cs); - } - else - break; /* rest of packets all nsn_itr > nws */ - } - roq->ws = sn; - return new_nws; -} - - -/* - * Reset a queue - * - * @i2400m: device descriptor - * @cin: Queue Index - * - * Deliver all the packets and reset the window-start to zero. Name is - * kind of misleading. - */ -static -void i2400m_roq_reset(struct i2400m *i2400m, struct i2400m_roq *roq) -{ - struct device *dev = i2400m_dev(i2400m); - struct sk_buff *skb_itr, *tmp_itr; - struct i2400m_roq_data *roq_data_itr; - - d_fnstart(2, dev, "(i2400m %p roq %p)\n", i2400m, roq); - i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_RESET, - roq->ws, skb_queue_len(&roq->queue), - ~0, ~0, 0); - skb_queue_walk_safe(&roq->queue, skb_itr, tmp_itr) { - roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; - d_printf(2, dev, "ERX: roq %p - release skb %p (sn %u)\n", - roq, skb_itr, roq_data_itr->sn); - __skb_unlink(skb_itr, &roq->queue); - i2400m_net_erx(i2400m, skb_itr, roq_data_itr->cs); - } - roq->ws = 0; - d_fnend(2, dev, "(i2400m %p roq %p) = void\n", i2400m, roq); -} - - -/* - * Queue a packet - * - * @i2400m: device descriptor - * @cin: Queue Index - * @skb: containing the packet data - * @fbn: First block number of the packet in @skb - * @lbn: Last block number of the packet in @skb - * - * The hardware is asking the driver to queue a packet for later - * delivery to the networking stack. - */ -static -void i2400m_roq_queue(struct i2400m *i2400m, struct i2400m_roq *roq, - struct sk_buff * skb, unsigned lbn) -{ - struct device *dev = i2400m_dev(i2400m); - unsigned nsn, len; - - d_fnstart(2, dev, "(i2400m %p roq %p skb %p lbn %u) = void\n", - i2400m, roq, skb, lbn); - len = skb_queue_len(&roq->queue); - nsn = __i2400m_roq_nsn(roq, lbn); - if (unlikely(nsn >= 1024)) { - dev_err(dev, "SW BUG? queue nsn %d (lbn %u ws %u)\n", - nsn, lbn, roq->ws); - i2400m_roq_log_dump(i2400m, roq); - i2400m_reset(i2400m, I2400M_RT_WARM); - } else { - __i2400m_roq_queue(i2400m, roq, skb, lbn, nsn); - i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_PACKET, - roq->ws, len, lbn, nsn, ~0); - } - d_fnend(2, dev, "(i2400m %p roq %p skb %p lbn %u) = void\n", - i2400m, roq, skb, lbn); -} - - -/* - * Update the window start in a reorder queue and deliver all skbs - * with a lower window start - * - * @i2400m: device descriptor - * @roq: Reorder queue - * @sn: New sequence number - */ -static -void i2400m_roq_update_ws(struct i2400m *i2400m, struct i2400m_roq *roq, - unsigned sn) -{ - struct device *dev = i2400m_dev(i2400m); - unsigned old_ws, nsn, len; - - d_fnstart(2, dev, "(i2400m %p roq %p sn %u)\n", i2400m, roq, sn); - old_ws = roq->ws; - len = skb_queue_len(&roq->queue); - nsn = __i2400m_roq_update_ws(i2400m, roq, sn); - i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_WS, - old_ws, len, sn, nsn, roq->ws); - d_fnstart(2, dev, "(i2400m %p roq %p sn %u) = void\n", i2400m, roq, sn); -} - - -/* - * Queue a packet and update the window start - * - * @i2400m: device descriptor - * @cin: Queue Index - * @skb: containing the packet data - * @fbn: First block number of the packet in @skb - * @sn: Last block number of the packet in @skb - * - * Note that unlike i2400m_roq_update_ws(), which sets the new window - * start to @sn, in here we'll set it to @sn + 1. - */ -static -void i2400m_roq_queue_update_ws(struct i2400m *i2400m, struct i2400m_roq *roq, - struct sk_buff * skb, unsigned sn) -{ - struct device *dev = i2400m_dev(i2400m); - unsigned nsn, old_ws, len; - - d_fnstart(2, dev, "(i2400m %p roq %p skb %p sn %u)\n", - i2400m, roq, skb, sn); - len = skb_queue_len(&roq->queue); - nsn = __i2400m_roq_nsn(roq, sn); - /* - * For type 3(queue_update_window_start) rx messages, there is no - * need to check if the normalized sequence number is greater 1023. - * Simply insert and deliver all packets to the host up to the - * window start. - */ - old_ws = roq->ws; - /* If the queue is empty, don't bother as we'd queue - * it and immediately unqueue it -- just deliver it. - */ - if (len == 0) { - struct i2400m_roq_data *roq_data; - roq_data = (struct i2400m_roq_data *) &skb->cb; - i2400m_net_erx(i2400m, skb, roq_data->cs); - } else - __i2400m_roq_queue(i2400m, roq, skb, sn, nsn); - - __i2400m_roq_update_ws(i2400m, roq, sn + 1); - i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_PACKET_WS, - old_ws, len, sn, nsn, roq->ws); - - d_fnend(2, dev, "(i2400m %p roq %p skb %p sn %u) = void\n", - i2400m, roq, skb, sn); -} - - -/* - * This routine destroys the memory allocated for rx_roq, when no - * other thread is accessing it. Access to rx_roq is refcounted by - * rx_roq_refcount, hence memory allocated must be destroyed when - * rx_roq_refcount becomes zero. This routine gets executed when - * rx_roq_refcount becomes zero. - */ -static void i2400m_rx_roq_destroy(struct kref *ref) -{ - unsigned itr; - struct i2400m *i2400m - = container_of(ref, struct i2400m, rx_roq_refcount); - for (itr = 0; itr < I2400M_RO_CIN + 1; itr++) - __skb_queue_purge(&i2400m->rx_roq[itr].queue); - kfree(i2400m->rx_roq[0].log); - kfree(i2400m->rx_roq); - i2400m->rx_roq = NULL; -} - -/* - * Receive and send up an extended data packet - * - * @i2400m: device descriptor - * @skb_rx: skb that contains the extended data packet - * @single_last: 1 if the payload is the only one or the last one of - * the skb. - * @payload: pointer to the packet's data inside the skb - * @size: size of the payload - * - * Starting in v1.4 of the i2400m's firmware, the device can send data - * packets to the host in an extended format that; this incudes a 16 - * byte header (struct i2400m_pl_edata_hdr). Using this header's space - * we can fake ethernet headers for ethernet device emulation without - * having to copy packets around. - * - * This function handles said path. - * - * - * Receive and send up an extended data packet that requires no reordering - * - * @i2400m: device descriptor - * @skb_rx: skb that contains the extended data packet - * @single_last: 1 if the payload is the only one or the last one of - * the skb. - * @payload: pointer to the packet's data (past the actual extended - * data payload header). - * @size: size of the payload - * - * Pass over to the networking stack a data packet that might have - * reordering requirements. - * - * This needs to the decide if the skb in which the packet is - * contained can be reused or if it needs to be cloned. Then it has to - * be trimmed in the edges so that the beginning is the space for eth - * header and then pass it to i2400m_net_erx() for the stack - * - * Assumes the caller has verified the sanity of the payload (size, - * etc) already. - */ -static -void i2400m_rx_edata(struct i2400m *i2400m, struct sk_buff *skb_rx, - unsigned single_last, const void *payload, size_t size) -{ - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_pl_edata_hdr *hdr = payload; - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - struct sk_buff *skb; - enum i2400m_cs cs; - u32 reorder; - unsigned ro_needed, ro_type, ro_cin, ro_sn; - struct i2400m_roq *roq; - struct i2400m_roq_data *roq_data; - unsigned long flags; - - BUILD_BUG_ON(ETH_HLEN > sizeof(*hdr)); - - d_fnstart(2, dev, "(i2400m %p skb_rx %p single %u payload %p " - "size %zu)\n", i2400m, skb_rx, single_last, payload, size); - if (size < sizeof(*hdr)) { - dev_err(dev, "ERX: HW BUG? message with short header (%zu " - "vs %zu bytes expected)\n", size, sizeof(*hdr)); - goto error; - } - - if (single_last) { - skb = skb_get(skb_rx); - d_printf(3, dev, "ERX: skb %p reusing\n", skb); - } else { - skb = skb_clone(skb_rx, GFP_KERNEL); - if (skb == NULL) { - dev_err(dev, "ERX: no memory to clone skb\n"); - net_dev->stats.rx_dropped++; - goto error_skb_clone; - } - d_printf(3, dev, "ERX: skb %p cloned from %p\n", skb, skb_rx); - } - /* now we have to pull and trim so that the skb points to the - * beginning of the IP packet; the netdev part will add the - * ethernet header as needed - we know there is enough space - * because we checked in i2400m_rx_edata(). */ - skb_pull(skb, payload + sizeof(*hdr) - (void *) skb->data); - skb_trim(skb, (void *) skb_end_pointer(skb) - payload - sizeof(*hdr)); - - reorder = le32_to_cpu(hdr->reorder); - ro_needed = reorder & I2400M_RO_NEEDED; - cs = hdr->cs; - if (ro_needed) { - ro_type = (reorder >> I2400M_RO_TYPE_SHIFT) & I2400M_RO_TYPE; - ro_cin = (reorder >> I2400M_RO_CIN_SHIFT) & I2400M_RO_CIN; - ro_sn = (reorder >> I2400M_RO_SN_SHIFT) & I2400M_RO_SN; - - spin_lock_irqsave(&i2400m->rx_lock, flags); - if (i2400m->rx_roq == NULL) { - kfree_skb(skb); /* rx_roq is already destroyed */ - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - goto error; - } - roq = &i2400m->rx_roq[ro_cin]; - kref_get(&i2400m->rx_roq_refcount); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - - roq_data = (struct i2400m_roq_data *) &skb->cb; - roq_data->sn = ro_sn; - roq_data->cs = cs; - d_printf(2, dev, "ERX: reorder needed: " - "type %u cin %u [ws %u] sn %u/%u len %zuB\n", - ro_type, ro_cin, roq->ws, ro_sn, - __i2400m_roq_nsn(roq, ro_sn), size); - d_dump(2, dev, payload, size); - switch(ro_type) { - case I2400M_RO_TYPE_RESET: - i2400m_roq_reset(i2400m, roq); - kfree_skb(skb); /* no data here */ - break; - case I2400M_RO_TYPE_PACKET: - i2400m_roq_queue(i2400m, roq, skb, ro_sn); - break; - case I2400M_RO_TYPE_WS: - i2400m_roq_update_ws(i2400m, roq, ro_sn); - kfree_skb(skb); /* no data here */ - break; - case I2400M_RO_TYPE_PACKET_WS: - i2400m_roq_queue_update_ws(i2400m, roq, skb, ro_sn); - break; - default: - dev_err(dev, "HW BUG? unknown reorder type %u\n", ro_type); - } - - spin_lock_irqsave(&i2400m->rx_lock, flags); - kref_put(&i2400m->rx_roq_refcount, i2400m_rx_roq_destroy); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - } - else - i2400m_net_erx(i2400m, skb, cs); -error_skb_clone: -error: - d_fnend(2, dev, "(i2400m %p skb_rx %p single %u payload %p " - "size %zu) = void\n", i2400m, skb_rx, single_last, payload, size); -} - - -/* - * Act on a received payload - * - * @i2400m: device instance - * @skb_rx: skb where the transaction was received - * @single_last: 1 this is the only payload or the last one (so the - * skb can be reused instead of cloned). - * @pld: payload descriptor - * @payload: payload data - * - * Upon reception of a payload, look at its guts in the payload - * descriptor and decide what to do with it. If it is a single payload - * skb or if the last skb is a data packet, the skb will be referenced - * and modified (so it doesn't have to be cloned). - */ -static -void i2400m_rx_payload(struct i2400m *i2400m, struct sk_buff *skb_rx, - unsigned single_last, const struct i2400m_pld *pld, - const void *payload) -{ - struct device *dev = i2400m_dev(i2400m); - size_t pl_size = i2400m_pld_size(pld); - enum i2400m_pt pl_type = i2400m_pld_type(pld); - - d_printf(7, dev, "RX: received payload type %u, %zu bytes\n", - pl_type, pl_size); - d_dump(8, dev, payload, pl_size); - - switch (pl_type) { - case I2400M_PT_DATA: - d_printf(3, dev, "RX: data payload %zu bytes\n", pl_size); - i2400m_net_rx(i2400m, skb_rx, single_last, payload, pl_size); - break; - case I2400M_PT_CTRL: - i2400m_rx_ctl(i2400m, skb_rx, payload, pl_size); - break; - case I2400M_PT_TRACE: - i2400m_rx_trace(i2400m, payload, pl_size); - break; - case I2400M_PT_EDATA: - d_printf(3, dev, "ERX: data payload %zu bytes\n", pl_size); - i2400m_rx_edata(i2400m, skb_rx, single_last, payload, pl_size); - break; - default: /* Anything else shouldn't come to the host */ - if (printk_ratelimit()) - dev_err(dev, "RX: HW BUG? unexpected payload type %u\n", - pl_type); - } -} - - -/* - * Check a received transaction's message header - * - * @i2400m: device descriptor - * @msg_hdr: message header - * @buf_size: size of the received buffer - * - * Check that the declarations done by a RX buffer message header are - * sane and consistent with the amount of data that was received. - */ -static -int i2400m_rx_msg_hdr_check(struct i2400m *i2400m, - const struct i2400m_msg_hdr *msg_hdr, - size_t buf_size) -{ - int result = -EIO; - struct device *dev = i2400m_dev(i2400m); - if (buf_size < sizeof(*msg_hdr)) { - dev_err(dev, "RX: HW BUG? message with short header (%zu " - "vs %zu bytes expected)\n", buf_size, sizeof(*msg_hdr)); - goto error; - } - if (msg_hdr->barker != cpu_to_le32(I2400M_D2H_MSG_BARKER)) { - dev_err(dev, "RX: HW BUG? message received with unknown " - "barker 0x%08x (buf_size %zu bytes)\n", - le32_to_cpu(msg_hdr->barker), buf_size); - goto error; - } - if (msg_hdr->num_pls == 0) { - dev_err(dev, "RX: HW BUG? zero payload packets in message\n"); - goto error; - } - if (le16_to_cpu(msg_hdr->num_pls) > I2400M_MAX_PLS_IN_MSG) { - dev_err(dev, "RX: HW BUG? message contains more payload " - "than maximum; ignoring.\n"); - goto error; - } - result = 0; -error: - return result; -} - - -/* - * Check a payload descriptor against the received data - * - * @i2400m: device descriptor - * @pld: payload descriptor - * @pl_itr: offset (in bytes) in the received buffer the payload is - * located - * @buf_size: size of the received buffer - * - * Given a payload descriptor (part of a RX buffer), check it is sane - * and that the data it declares fits in the buffer. - */ -static -int i2400m_rx_pl_descr_check(struct i2400m *i2400m, - const struct i2400m_pld *pld, - size_t pl_itr, size_t buf_size) -{ - int result = -EIO; - struct device *dev = i2400m_dev(i2400m); - size_t pl_size = i2400m_pld_size(pld); - enum i2400m_pt pl_type = i2400m_pld_type(pld); - - if (pl_size > i2400m->bus_pl_size_max) { - dev_err(dev, "RX: HW BUG? payload @%zu: size %zu is " - "bigger than maximum %zu; ignoring message\n", - pl_itr, pl_size, i2400m->bus_pl_size_max); - goto error; - } - if (pl_itr + pl_size > buf_size) { /* enough? */ - dev_err(dev, "RX: HW BUG? payload @%zu: size %zu " - "goes beyond the received buffer " - "size (%zu bytes); ignoring message\n", - pl_itr, pl_size, buf_size); - goto error; - } - if (pl_type >= I2400M_PT_ILLEGAL) { - dev_err(dev, "RX: HW BUG? illegal payload type %u; " - "ignoring message\n", pl_type); - goto error; - } - result = 0; -error: - return result; -} - - -/** - * i2400m_rx - Receive a buffer of data from the device - * - * @i2400m: device descriptor - * @skb: skbuff where the data has been received - * - * Parse in a buffer of data that contains an RX message sent from the - * device. See the file header for the format. Run all checks on the - * buffer header, then run over each payload's descriptors, verify - * their consistency and act on each payload's contents. If - * everything is successful, update the device's statistics. - * - * Note: You need to set the skb to contain only the length of the - * received buffer; for that, use skb_trim(skb, RECEIVED_SIZE). - * - * Returns: - * - * 0 if ok, < 0 errno on error - * - * If ok, this function owns now the skb and the caller DOESN'T have - * to run kfree_skb() on it. However, on error, the caller still owns - * the skb and it is responsible for releasing it. - */ -int i2400m_rx(struct i2400m *i2400m, struct sk_buff *skb) -{ - int i, result; - struct device *dev = i2400m_dev(i2400m); - const struct i2400m_msg_hdr *msg_hdr; - size_t pl_itr, pl_size; - unsigned long flags; - unsigned num_pls, single_last, skb_len; - - skb_len = skb->len; - d_fnstart(4, dev, "(i2400m %p skb %p [size %u])\n", - i2400m, skb, skb_len); - msg_hdr = (void *) skb->data; - result = i2400m_rx_msg_hdr_check(i2400m, msg_hdr, skb_len); - if (result < 0) - goto error_msg_hdr_check; - result = -EIO; - num_pls = le16_to_cpu(msg_hdr->num_pls); - /* Check payload descriptor(s) */ - pl_itr = struct_size(msg_hdr, pld, num_pls); - pl_itr = ALIGN(pl_itr, I2400M_PL_ALIGN); - if (pl_itr > skb_len) { /* got all the payload descriptors? */ - dev_err(dev, "RX: HW BUG? message too short (%u bytes) for " - "%u payload descriptors (%zu each, total %zu)\n", - skb_len, num_pls, sizeof(msg_hdr->pld[0]), pl_itr); - goto error_pl_descr_short; - } - /* Walk each payload payload--check we really got it */ - for (i = 0; i < num_pls; i++) { - /* work around old gcc warnings */ - pl_size = i2400m_pld_size(&msg_hdr->pld[i]); - result = i2400m_rx_pl_descr_check(i2400m, &msg_hdr->pld[i], - pl_itr, skb_len); - if (result < 0) - goto error_pl_descr_check; - single_last = num_pls == 1 || i == num_pls - 1; - i2400m_rx_payload(i2400m, skb, single_last, &msg_hdr->pld[i], - skb->data + pl_itr); - pl_itr += ALIGN(pl_size, I2400M_PL_ALIGN); - cond_resched(); /* Don't monopolize */ - } - kfree_skb(skb); - /* Update device statistics */ - spin_lock_irqsave(&i2400m->rx_lock, flags); - i2400m->rx_pl_num += i; - if (i > i2400m->rx_pl_max) - i2400m->rx_pl_max = i; - if (i < i2400m->rx_pl_min) - i2400m->rx_pl_min = i; - i2400m->rx_num++; - i2400m->rx_size_acc += skb_len; - if (skb_len < i2400m->rx_size_min) - i2400m->rx_size_min = skb_len; - if (skb_len > i2400m->rx_size_max) - i2400m->rx_size_max = skb_len; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); -error_pl_descr_check: -error_pl_descr_short: -error_msg_hdr_check: - d_fnend(4, dev, "(i2400m %p skb %p [size %u]) = %d\n", - i2400m, skb, skb_len, result); - return result; -} -EXPORT_SYMBOL_GPL(i2400m_rx); - - -void i2400m_unknown_barker(struct i2400m *i2400m, - const void *buf, size_t size) -{ - struct device *dev = i2400m_dev(i2400m); - char prefix[64]; - const __le32 *barker = buf; - dev_err(dev, "RX: HW BUG? unknown barker %08x, " - "dropping %zu bytes\n", le32_to_cpu(*barker), size); - snprintf(prefix, sizeof(prefix), "%s %s: ", - dev_driver_string(dev), dev_name(dev)); - if (size > 64) { - print_hex_dump(KERN_ERR, prefix, DUMP_PREFIX_OFFSET, - 8, 4, buf, 64, 0); - printk(KERN_ERR "%s... (only first 64 bytes " - "dumped)\n", prefix); - } else - print_hex_dump(KERN_ERR, prefix, DUMP_PREFIX_OFFSET, - 8, 4, buf, size, 0); -} -EXPORT_SYMBOL(i2400m_unknown_barker); - - -/* - * Initialize the RX queue and infrastructure - * - * This sets up all the RX reordering infrastructures, which will not - * be used if reordering is not enabled or if the firmware does not - * support it. The device is told to do reordering in - * i2400m_dev_initialize(), where it also looks at the value of the - * i2400m->rx_reorder switch before taking a decission. - * - * Note we allocate the roq queues in one chunk and the actual logging - * support for it (logging) in another one and then we setup the - * pointers from the first to the last. - */ -int i2400m_rx_setup(struct i2400m *i2400m) -{ - int result = 0; - - i2400m->rx_reorder = i2400m_rx_reorder_disabled? 0 : 1; - if (i2400m->rx_reorder) { - unsigned itr; - struct i2400m_roq_log *rd; - - result = -ENOMEM; - - i2400m->rx_roq = kcalloc(I2400M_RO_CIN + 1, - sizeof(i2400m->rx_roq[0]), GFP_KERNEL); - if (i2400m->rx_roq == NULL) - goto error_roq_alloc; - - rd = kcalloc(I2400M_RO_CIN + 1, sizeof(*i2400m->rx_roq[0].log), - GFP_KERNEL); - if (rd == NULL) { - result = -ENOMEM; - goto error_roq_log_alloc; - } - - for(itr = 0; itr < I2400M_RO_CIN + 1; itr++) { - __i2400m_roq_init(&i2400m->rx_roq[itr]); - i2400m->rx_roq[itr].log = &rd[itr]; - } - kref_init(&i2400m->rx_roq_refcount); - } - return 0; - -error_roq_log_alloc: - kfree(i2400m->rx_roq); -error_roq_alloc: - return result; -} - - -/* Tear down the RX queue and infrastructure */ -void i2400m_rx_release(struct i2400m *i2400m) -{ - unsigned long flags; - - if (i2400m->rx_reorder) { - spin_lock_irqsave(&i2400m->rx_lock, flags); - kref_put(&i2400m->rx_roq_refcount, i2400m_rx_roq_destroy); - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - } - /* at this point, nothing can be received... */ - i2400m_report_hook_flush(i2400m); -} diff --git a/drivers/net/wimax/i2400m/sysfs.c b/drivers/net/wimax/i2400m/sysfs.c deleted file mode 100644 index 895ee265909b..000000000000 --- a/drivers/net/wimax/i2400m/sysfs.c +++ /dev/null @@ -1,65 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Intel Wireless WiMAX Connection 2400m - * Sysfs interfaces to show driver and device information - * - * Copyright (C) 2007 Intel Corporation - * Inaky Perez-Gonzalez - */ - -#include -#include -#include -#include -#include "i2400m.h" - - -#define D_SUBMODULE sysfs -#include "debug-levels.h" - - -/* - * Set the idle timeout (msecs) - * - * FIXME: eventually this should be a common WiMAX stack method, but - * would like to wait to see how other devices manage it. - */ -static -ssize_t i2400m_idle_timeout_store(struct device *dev, - struct device_attribute *attr, - const char *buf, size_t size) -{ - ssize_t result; - struct i2400m *i2400m = net_dev_to_i2400m(to_net_dev(dev)); - unsigned val; - - result = -EINVAL; - if (sscanf(buf, "%u\n", &val) != 1) - goto error_no_unsigned; - if (val != 0 && (val < 100 || val > 300000 || val % 100 != 0)) { - dev_err(dev, "idle_timeout: %u: invalid msecs specification; " - "valid values are 0, 100-300000 in 100 increments\n", - val); - goto error_bad_value; - } - result = i2400m_set_idle_timeout(i2400m, val); - if (result >= 0) - result = size; -error_no_unsigned: -error_bad_value: - return result; -} - -static -DEVICE_ATTR_WO(i2400m_idle_timeout); - -static -struct attribute *i2400m_dev_attrs[] = { - &dev_attr_i2400m_idle_timeout.attr, - NULL, -}; - -struct attribute_group i2400m_dev_attr_group = { - .name = NULL, /* we want them in the same directory */ - .attrs = i2400m_dev_attrs, -}; diff --git a/drivers/net/wimax/i2400m/tx.c b/drivers/net/wimax/i2400m/tx.c deleted file mode 100644 index 1255302e251e..000000000000 --- a/drivers/net/wimax/i2400m/tx.c +++ /dev/null @@ -1,1011 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * Generic (non-bus specific) TX handling - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * - Initial implementation - * - * Intel Corporation - * Inaky Perez-Gonzalez - * - Rewritten to use a single FIFO to lower the memory allocation - * pressure and optimize cache hits when copying to the queue, as - * well as splitting out bus-specific code. - * - * - * Implements data transmission to the device; this is done through a - * software FIFO, as data/control frames can be coalesced (while the - * device is reading the previous tx transaction, others accumulate). - * - * A FIFO is used because at the end it is resource-cheaper that trying - * to implement scatter/gather over USB. As well, most traffic is going - * to be download (vs upload). - * - * The format for sending/receiving data to/from the i2400m is - * described in detail in rx.c:PROTOCOL FORMAT. In here we implement - * the transmission of that. This is split between a bus-independent - * part that just prepares everything and a bus-specific part that - * does the actual transmission over the bus to the device (in the - * bus-specific driver). - * - * - * The general format of a device-host transaction is MSG-HDR, PLD1, - * PLD2...PLDN, PL1, PL2,...PLN, PADDING. - * - * Because we need the send payload descriptors and then payloads and - * because it is kind of expensive to do scatterlists in USB (one URB - * per node), it becomes cheaper to append all the data to a FIFO - * (copying to a FIFO potentially in cache is cheaper). - * - * Then the bus-specific code takes the parts of that FIFO that are - * written and passes them to the device. - * - * So the concepts to keep in mind there are: - * - * We use a FIFO to queue the data in a linear buffer. We first append - * a MSG-HDR, space for I2400M_TX_PLD_MAX payload descriptors and then - * go appending payloads until we run out of space or of payload - * descriptors. Then we append padding to make the whole transaction a - * multiple of i2400m->bus_tx_block_size (as defined by the bus layer). - * - * - A TX message: a combination of a message header, payload - * descriptors and payloads. - * - * Open: it is marked as active (i2400m->tx_msg is valid) and we - * can keep adding payloads to it. - * - * Closed: we are not appending more payloads to this TX message - * (exahusted space in the queue, too many payloads or - * whichever). We have appended padding so the whole message - * length is aligned to i2400m->bus_tx_block_size (as set by the - * bus/transport layer). - * - * - Most of the time we keep a TX message open to which we append - * payloads. - * - * - If we are going to append and there is no more space (we are at - * the end of the FIFO), we close the message, mark the rest of the - * FIFO space unusable (skip_tail), create a new message at the - * beginning of the FIFO (if there is space) and append the message - * there. - * - * This is because we need to give linear TX messages to the bus - * engine. So we don't write a message to the remaining FIFO space - * until the tail and continue at the head of it. - * - * - We overload one of the fields in the message header to use it as - * 'size' of the TX message, so we can iterate over them. It also - * contains a flag that indicates if we have to skip it or not. - * When we send the buffer, we update that to its real on-the-wire - * value. - * - * - The MSG-HDR PLD1...PLD2 stuff has to be a size multiple of 16. - * - * It follows that if MSG-HDR says we have N messages, the whole - * header + descriptors is 16 + 4*N; for those to be a multiple of - * 16, it follows that N can be 4, 8, 12, ... (32, 48, 64, 80... - * bytes). - * - * So if we have only 1 payload, we have to submit a header that in - * all truth has space for 4. - * - * The implication is that we reserve space for 12 (64 bytes); but - * if we fill up only (eg) 2, our header becomes 32 bytes only. So - * the TX engine has to shift those 32 bytes of msg header and 2 - * payloads and padding so that right after it the payloads start - * and the TX engine has to know about that. - * - * It is cheaper to move the header up than the whole payloads down. - * - * We do this in i2400m_tx_close(). See 'i2400m_msg_hdr->offset'. - * - * - Each payload has to be size-padded to 16 bytes; before appending - * it, we just do it. - * - * - The whole message has to be padded to i2400m->bus_tx_block_size; - * we do this at close time. Thus, when reserving space for the - * payload, we always make sure there is also free space for this - * padding that sooner or later will happen. - * - * When we append a message, we tell the bus specific code to kick in - * TXs. It will TX (in parallel) until the buffer is exhausted--hence - * the lockin we do. The TX code will only send a TX message at the - * time (which remember, might contain more than one payload). Of - * course, when the bus-specific driver attempts to TX a message that - * is still open, it gets closed first. - * - * Gee, this is messy; well a picture. In the example below we have a - * partially full FIFO, with a closed message ready to be delivered - * (with a moved message header to make sure it is size-aligned to - * 16), TAIL room that was unusable (and thus is marked with a message - * header that says 'skip this') and at the head of the buffer, an - * incomplete message with a couple of payloads. - * - * N ___________________________________________________ - * | | - * | TAIL room | - * | | - * | msg_hdr to skip (size |= 0x80000) | - * |---------------------------------------------------|------- - * | | /|\ - * | | | - * | TX message padding | | - * | | | - * | | | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -| | - * | | | - * | payload 1 | | - * | | N * tx_block_size - * | | | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -| | - * | | | - * | payload 1 | | - * | | | - * | | | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -|- -|- - - - - * | padding 3 /|\ | | /|\ - * | padding 2 | | | | - * | pld 1 32 bytes (2 * 16) | | | - * | pld 0 | | | | - * | moved msg_hdr \|/ | \|/ | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -|- - - | - * | | _PLD_SIZE - * | unused | | - * | | | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -| | - * | msg_hdr (size X) [this message is closed] | \|/ - * |===================================================|========== <=== OUT - * | | - * | | - * | | - * | Free rooom | - * | | - * | | - * | | - * | | - * | | - * | | - * | | - * | | - * | | - * |===================================================|========== <=== IN - * | | - * | | - * | | - * | | - * | payload 1 | - * | | - * | | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -| - * | | - * | payload 0 | - * | | - * | | - * |- - - - - - - - - - - - - - - - - - - - - - - - - -| - * | pld 11 /|\ | - * | ... | | - * | pld 1 64 bytes (2 * 16) | - * | pld 0 | | - * | msg_hdr (size X) \|/ [message is open] | - * 0 --------------------------------------------------- - * - * - * ROADMAP - * - * i2400m_tx_setup() Called by i2400m_setup - * i2400m_tx_release() Called by i2400m_release() - * - * i2400m_tx() Called to send data or control frames - * i2400m_tx_fifo_push() Allocates append-space in the FIFO - * i2400m_tx_new() Opens a new message in the FIFO - * i2400m_tx_fits() Checks if a new payload fits in the message - * i2400m_tx_close() Closes an open message in the FIFO - * i2400m_tx_skip_tail() Marks unusable FIFO tail space - * i2400m->bus_tx_kick() - * - * Now i2400m->bus_tx_kick() is the the bus-specific driver backend - * implementation; that would do: - * - * i2400m->bus_tx_kick() - * i2400m_tx_msg_get() Gets first message ready to go - * ...sends it... - * i2400m_tx_msg_sent() Ack the message is sent; repeat from - * _tx_msg_get() until it returns NULL - * (FIFO empty). - */ -#include -#include -#include -#include "i2400m.h" - - -#define D_SUBMODULE tx -#include "debug-levels.h" - -enum { - /** - * TX Buffer size - * - * Doc says maximum transaction is 16KiB. If we had 16KiB en - * route and 16KiB being queued, it boils down to needing - * 32KiB. - * 32KiB is insufficient for 1400 MTU, hence increasing - * tx buffer size to 64KiB. - */ - I2400M_TX_BUF_SIZE = 65536, - /** - * Message header and payload descriptors have to be 16 - * aligned (16 + 4 * N = 16 * M). If we take that average sent - * packets are MTU size (~1400-~1500) it follows that we could - * fit at most 10-11 payloads in one transaction. To meet the - * alignment requirement, that means we need to leave space - * for 12 (64 bytes). To simplify, we leave space for that. If - * at the end there are less, we pad up to the nearest - * multiple of 16. - */ - /* - * According to Intel Wimax i3200, i5x50 and i6x50 specification - * documents, the maximum number of payloads per message can be - * up to 60. Increasing the number of payloads to 60 per message - * helps to accommodate smaller payloads in a single transaction. - */ - I2400M_TX_PLD_MAX = 60, - I2400M_TX_PLD_SIZE = sizeof(struct i2400m_msg_hdr) - + I2400M_TX_PLD_MAX * sizeof(struct i2400m_pld), - I2400M_TX_SKIP = 0x80000000, - /* - * According to Intel Wimax i3200, i5x50 and i6x50 specification - * documents, the maximum size of each message can be up to 16KiB. - */ - I2400M_TX_MSG_SIZE = 16384, -}; - -#define TAIL_FULL ((void *)~(unsigned long)NULL) - -/* - * Calculate how much tail room is available - * - * Note the trick here. This path is ONLY caleed for Case A (see - * i2400m_tx_fifo_push() below), where we have: - * - * Case A - * N ___________ - * | tail room | - * | | - * |<- IN ->| - * | | - * | data | - * | | - * |<- OUT ->| - * | | - * | head room | - * 0 ----------- - * - * When calculating the tail_room, tx_in might get to be zero if - * i2400m->tx_in is right at the end of the buffer (really full - * buffer) if there is no head room. In this case, tail_room would be - * I2400M_TX_BUF_SIZE, although it is actually zero. Hence the final - * mod (%) operation. However, when doing this kind of optimization, - * i2400m->tx_in being zero would fail, so we treat is an a special - * case. - */ -static inline -size_t __i2400m_tx_tail_room(struct i2400m *i2400m) -{ - size_t tail_room; - size_t tx_in; - - if (unlikely(i2400m->tx_in == 0)) - return I2400M_TX_BUF_SIZE; - tx_in = i2400m->tx_in % I2400M_TX_BUF_SIZE; - tail_room = I2400M_TX_BUF_SIZE - tx_in; - tail_room %= I2400M_TX_BUF_SIZE; - return tail_room; -} - - -/* - * Allocate @size bytes in the TX fifo, return a pointer to it - * - * @i2400m: device descriptor - * @size: size of the buffer we need to allocate - * @padding: ensure that there is at least this many bytes of free - * contiguous space in the fifo. This is needed because later on - * we might need to add padding. - * @try_head: specify either to allocate head room or tail room space - * in the TX FIFO. This boolean is required to avoids a system hang - * due to an infinite loop caused by i2400m_tx_fifo_push(). - * The caller must always try to allocate tail room space first by - * calling this routine with try_head = 0. In case if there - * is not enough tail room space but there is enough head room space, - * (i2400m_tx_fifo_push() returns TAIL_FULL) try to allocate head - * room space, by calling this routine again with try_head = 1. - * - * Returns: - * - * Pointer to the allocated space. NULL if there is no - * space. TAIL_FULL if there is no space at the tail but there is at - * the head (Case B below). - * - * These are the two basic cases we need to keep an eye for -- it is - * much better explained in linux/kernel/kfifo.c, but this code - * basically does the same. No rocket science here. - * - * Case A Case B - * N ___________ ___________ - * | tail room | | data | - * | | | | - * |<- IN ->| |<- OUT ->| - * | | | | - * | data | | room | - * | | | | - * |<- OUT ->| |<- IN ->| - * | | | | - * | head room | | data | - * 0 ----------- ----------- - * - * We allocate only *contiguous* space. - * - * We can allocate only from 'room'. In Case B, it is simple; in case - * A, we only try from the tail room; if it is not enough, we just - * fail and return TAIL_FULL and let the caller figure out if we wants to - * skip the tail room and try to allocate from the head. - * - * There is a corner case, wherein i2400m_tx_new() can get into - * an infinite loop calling i2400m_tx_fifo_push(). - * In certain situations, tx_in would have reached on the top of TX FIFO - * and i2400m_tx_tail_room() returns 0, as described below: - * - * N ___________ tail room is zero - * |<- IN ->| - * | | - * | | - * | | - * | data | - * |<- OUT ->| - * | | - * | | - * | head room | - * 0 ----------- - * During such a time, where tail room is zero in the TX FIFO and if there - * is a request to add a payload to TX FIFO, which calls: - * i2400m_tx() - * ->calls i2400m_tx_close() - * ->calls i2400m_tx_skip_tail() - * goto try_new; - * ->calls i2400m_tx_new() - * |----> [try_head:] - * infinite loop | ->calls i2400m_tx_fifo_push() - * | if (tail_room < needed) - * | if (head_room => needed) - * | return TAIL_FULL; - * |<---- goto try_head; - * - * i2400m_tx() calls i2400m_tx_close() to close the message, since there - * is no tail room to accommodate the payload and calls - * i2400m_tx_skip_tail() to skip the tail space. Now i2400m_tx() calls - * i2400m_tx_new() to allocate space for new message header calling - * i2400m_tx_fifo_push() that returns TAIL_FULL, since there is no tail space - * to accommodate the message header, but there is enough head space. - * The i2400m_tx_new() keeps re-retrying by calling i2400m_tx_fifo_push() - * ending up in a loop causing system freeze. - * - * This corner case is avoided by using a try_head boolean, - * as an argument to i2400m_tx_fifo_push(). - * - * Note: - * - * Assumes i2400m->tx_lock is taken, and we use that as a barrier - * - * The indexes keep increasing and we reset them to zero when we - * pop data off the queue - */ -static -void *i2400m_tx_fifo_push(struct i2400m *i2400m, size_t size, - size_t padding, bool try_head) -{ - struct device *dev = i2400m_dev(i2400m); - size_t room, tail_room, needed_size; - void *ptr; - - needed_size = size + padding; - room = I2400M_TX_BUF_SIZE - (i2400m->tx_in - i2400m->tx_out); - if (room < needed_size) { /* this takes care of Case B */ - d_printf(2, dev, "fifo push %zu/%zu: no space\n", - size, padding); - return NULL; - } - /* Is there space at the tail? */ - tail_room = __i2400m_tx_tail_room(i2400m); - if (!try_head && tail_room < needed_size) { - /* - * If the tail room space is not enough to push the message - * in the TX FIFO, then there are two possibilities: - * 1. There is enough head room space to accommodate - * this message in the TX FIFO. - * 2. There is not enough space in the head room and - * in tail room of the TX FIFO to accommodate the message. - * In the case (1), return TAIL_FULL so that the caller - * can figure out, if the caller wants to push the message - * into the head room space. - * In the case (2), return NULL, indicating that the TX FIFO - * cannot accommodate the message. - */ - if (room - tail_room >= needed_size) { - d_printf(2, dev, "fifo push %zu/%zu: tail full\n", - size, padding); - return TAIL_FULL; /* There might be head space */ - } else { - d_printf(2, dev, "fifo push %zu/%zu: no head space\n", - size, padding); - return NULL; /* There is no space */ - } - } - ptr = i2400m->tx_buf + i2400m->tx_in % I2400M_TX_BUF_SIZE; - d_printf(2, dev, "fifo push %zu/%zu: at @%zu\n", size, padding, - i2400m->tx_in % I2400M_TX_BUF_SIZE); - i2400m->tx_in += size; - return ptr; -} - - -/* - * Mark the tail of the FIFO buffer as 'to-skip' - * - * We should never hit the BUG_ON() because all the sizes we push to - * the FIFO are padded to be a multiple of 16 -- the size of *msg - * (I2400M_PL_PAD for the payloads, I2400M_TX_PLD_SIZE for the - * header). - * - * Tail room can get to be zero if a message was opened when there was - * space only for a header. _tx_close() will mark it as to-skip (as it - * will have no payloads) and there will be no more space to flush, so - * nothing has to be done here. This is probably cheaper than ensuring - * in _tx_new() that there is some space for payloads...as we could - * always possibly hit the same problem if the payload wouldn't fit. - * - * Note: - * - * Assumes i2400m->tx_lock is taken, and we use that as a barrier - * - * This path is only taken for Case A FIFO situations [see - * i2400m_tx_fifo_push()] - */ -static -void i2400m_tx_skip_tail(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - size_t tx_in = i2400m->tx_in % I2400M_TX_BUF_SIZE; - size_t tail_room = __i2400m_tx_tail_room(i2400m); - struct i2400m_msg_hdr *msg = i2400m->tx_buf + tx_in; - if (unlikely(tail_room == 0)) - return; - BUG_ON(tail_room < sizeof(*msg)); - msg->size = tail_room | I2400M_TX_SKIP; - d_printf(2, dev, "skip tail: skipping %zu bytes @%zu\n", - tail_room, tx_in); - i2400m->tx_in += tail_room; -} - - -/* - * Check if a skb will fit in the TX queue's current active TX - * message (if there are still descriptors left unused). - * - * Returns: - * 0 if the message won't fit, 1 if it will. - * - * Note: - * - * Assumes a TX message is active (i2400m->tx_msg). - * - * Assumes i2400m->tx_lock is taken, and we use that as a barrier - */ -static -unsigned i2400m_tx_fits(struct i2400m *i2400m) -{ - struct i2400m_msg_hdr *msg_hdr = i2400m->tx_msg; - return le16_to_cpu(msg_hdr->num_pls) < I2400M_TX_PLD_MAX; - -} - - -/* - * Start a new TX message header in the queue. - * - * Reserve memory from the base FIFO engine and then just initialize - * the message header. - * - * We allocate the biggest TX message header we might need (one that'd - * fit I2400M_TX_PLD_MAX payloads) -- when it is closed it will be - * 'ironed it out' and the unneeded parts removed. - * - * NOTE: - * - * Assumes that the previous message is CLOSED (eg: either - * there was none or 'i2400m_tx_close()' was called on it). - * - * Assumes i2400m->tx_lock is taken, and we use that as a barrier - */ -static -void i2400m_tx_new(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - struct i2400m_msg_hdr *tx_msg; - bool try_head = false; - BUG_ON(i2400m->tx_msg != NULL); - /* - * In certain situations, TX queue might have enough space to - * accommodate the new message header I2400M_TX_PLD_SIZE, but - * might not have enough space to accommodate the payloads. - * Adding bus_tx_room_min padding while allocating a new TX message - * increases the possibilities of including at least one payload of the - * size <= bus_tx_room_min. - */ -try_head: - tx_msg = i2400m_tx_fifo_push(i2400m, I2400M_TX_PLD_SIZE, - i2400m->bus_tx_room_min, try_head); - if (tx_msg == NULL) - goto out; - else if (tx_msg == TAIL_FULL) { - i2400m_tx_skip_tail(i2400m); - d_printf(2, dev, "new TX message: tail full, trying head\n"); - try_head = true; - goto try_head; - } - memset(tx_msg, 0, I2400M_TX_PLD_SIZE); - tx_msg->size = I2400M_TX_PLD_SIZE; -out: - i2400m->tx_msg = tx_msg; - d_printf(2, dev, "new TX message: %p @%zu\n", - tx_msg, (void *) tx_msg - i2400m->tx_buf); -} - - -/* - * Finalize the current TX message header - * - * Sets the message header to be at the proper location depending on - * how many descriptors we have (check documentation at the file's - * header for more info on that). - * - * Appends padding bytes to make sure the whole TX message (counting - * from the 'relocated' message header) is aligned to - * tx_block_size. We assume the _append() code has left enough space - * in the FIFO for that. If there are no payloads, just pass, as it - * won't be transferred. - * - * The amount of padding bytes depends on how many payloads are in the - * TX message, as the "msg header and payload descriptors" will be - * shifted up in the buffer. - */ -static -void i2400m_tx_close(struct i2400m *i2400m) -{ - struct device *dev = i2400m_dev(i2400m); - struct i2400m_msg_hdr *tx_msg = i2400m->tx_msg; - struct i2400m_msg_hdr *tx_msg_moved; - size_t aligned_size, padding, hdr_size; - void *pad_buf; - unsigned num_pls; - - if (tx_msg->size & I2400M_TX_SKIP) /* a skipper? nothing to do */ - goto out; - num_pls = le16_to_cpu(tx_msg->num_pls); - /* We can get this situation when a new message was started - * and there was no space to add payloads before hitting the - tail (and taking padding into consideration). */ - if (num_pls == 0) { - tx_msg->size |= I2400M_TX_SKIP; - goto out; - } - /* Relocate the message header - * - * Find the current header size, align it to 16 and if we need - * to move it so the tail is next to the payloads, move it and - * set the offset. - * - * If it moved, this header is good only for transmission; the - * original one (it is kept if we moved) is still used to - * figure out where the next TX message starts (and where the - * offset to the moved header is). - */ - hdr_size = struct_size(tx_msg, pld, le16_to_cpu(tx_msg->num_pls)); - hdr_size = ALIGN(hdr_size, I2400M_PL_ALIGN); - tx_msg->offset = I2400M_TX_PLD_SIZE - hdr_size; - tx_msg_moved = (void *) tx_msg + tx_msg->offset; - memmove(tx_msg_moved, tx_msg, hdr_size); - tx_msg_moved->size -= tx_msg->offset; - /* - * Now figure out how much we have to add to the (moved!) - * message so the size is a multiple of i2400m->bus_tx_block_size. - */ - aligned_size = ALIGN(tx_msg_moved->size, i2400m->bus_tx_block_size); - padding = aligned_size - tx_msg_moved->size; - if (padding > 0) { - pad_buf = i2400m_tx_fifo_push(i2400m, padding, 0, 0); - if (WARN_ON(pad_buf == NULL || pad_buf == TAIL_FULL)) { - /* This should not happen -- append should verify - * there is always space left at least to append - * tx_block_size */ - dev_err(dev, - "SW BUG! Possible data leakage from memory the " - "device should not read for padding - " - "size %lu aligned_size %zu tx_buf %p in " - "%zu out %zu\n", - (unsigned long) tx_msg_moved->size, - aligned_size, i2400m->tx_buf, i2400m->tx_in, - i2400m->tx_out); - } else - memset(pad_buf, 0xad, padding); - } - tx_msg_moved->padding = cpu_to_le16(padding); - tx_msg_moved->size += padding; - if (tx_msg != tx_msg_moved) - tx_msg->size += padding; -out: - i2400m->tx_msg = NULL; -} - - -/** - * i2400m_tx - send the data in a buffer to the device - * - * @buf: pointer to the buffer to transmit - * - * @buf_len: buffer size - * - * @pl_type: type of the payload we are sending. - * - * Returns: - * 0 if ok, < 0 errno code on error (-ENOSPC, if there is no more - * room for the message in the queue). - * - * Appends the buffer to the TX FIFO and notifies the bus-specific - * part of the driver that there is new data ready to transmit. - * Once this function returns, the buffer has been copied, so it can - * be reused. - * - * The steps followed to append are explained in detail in the file - * header. - * - * Whenever we write to a message, we increase msg->size, so it - * reflects exactly how big the message is. This is needed so that if - * we concatenate two messages before they can be sent, the code that - * sends the messages can find the boundaries (and it will replace the - * size with the real barker before sending). - * - * Note: - * - * Cold and warm reset payloads need to be sent as a single - * payload, so we handle that. - */ -int i2400m_tx(struct i2400m *i2400m, const void *buf, size_t buf_len, - enum i2400m_pt pl_type) -{ - int result = -ENOSPC; - struct device *dev = i2400m_dev(i2400m); - unsigned long flags; - size_t padded_len; - void *ptr; - bool try_head = false; - unsigned is_singleton = pl_type == I2400M_PT_RESET_WARM - || pl_type == I2400M_PT_RESET_COLD; - - d_fnstart(3, dev, "(i2400m %p skb %p [%zu bytes] pt %u)\n", - i2400m, buf, buf_len, pl_type); - padded_len = ALIGN(buf_len, I2400M_PL_ALIGN); - d_printf(5, dev, "padded_len %zd buf_len %zd\n", padded_len, buf_len); - /* If there is no current TX message, create one; if the - * current one is out of payload slots or we have a singleton, - * close it and start a new one */ - spin_lock_irqsave(&i2400m->tx_lock, flags); - /* If tx_buf is NULL, device is shutdown */ - if (i2400m->tx_buf == NULL) { - result = -ESHUTDOWN; - goto error_tx_new; - } -try_new: - if (unlikely(i2400m->tx_msg == NULL)) - i2400m_tx_new(i2400m); - else if (unlikely(!i2400m_tx_fits(i2400m) - || (is_singleton && i2400m->tx_msg->num_pls != 0))) { - d_printf(2, dev, "closing TX message (fits %u singleton " - "%u num_pls %u)\n", i2400m_tx_fits(i2400m), - is_singleton, i2400m->tx_msg->num_pls); - i2400m_tx_close(i2400m); - i2400m_tx_new(i2400m); - } - if (i2400m->tx_msg == NULL) - goto error_tx_new; - /* - * Check if this skb will fit in the TX queue's current active - * TX message. The total message size must not exceed the maximum - * size of each message I2400M_TX_MSG_SIZE. If it exceeds, - * close the current message and push this skb into the new message. - */ - if (i2400m->tx_msg->size + padded_len > I2400M_TX_MSG_SIZE) { - d_printf(2, dev, "TX: message too big, going new\n"); - i2400m_tx_close(i2400m); - i2400m_tx_new(i2400m); - } - if (i2400m->tx_msg == NULL) - goto error_tx_new; - /* So we have a current message header; now append space for - * the message -- if there is not enough, try the head */ - ptr = i2400m_tx_fifo_push(i2400m, padded_len, - i2400m->bus_tx_block_size, try_head); - if (ptr == TAIL_FULL) { /* Tail is full, try head */ - d_printf(2, dev, "pl append: tail full\n"); - i2400m_tx_close(i2400m); - i2400m_tx_skip_tail(i2400m); - try_head = true; - goto try_new; - } else if (ptr == NULL) { /* All full */ - result = -ENOSPC; - d_printf(2, dev, "pl append: all full\n"); - } else { /* Got space, copy it, set padding */ - struct i2400m_msg_hdr *tx_msg = i2400m->tx_msg; - unsigned num_pls = le16_to_cpu(tx_msg->num_pls); - memcpy(ptr, buf, buf_len); - memset(ptr + buf_len, 0xad, padded_len - buf_len); - i2400m_pld_set(&tx_msg->pld[num_pls], buf_len, pl_type); - d_printf(3, dev, "pld 0x%08x (type 0x%1x len 0x%04zx\n", - le32_to_cpu(tx_msg->pld[num_pls].val), - pl_type, buf_len); - tx_msg->num_pls = le16_to_cpu(num_pls+1); - tx_msg->size += padded_len; - d_printf(2, dev, "TX: appended %zu b (up to %u b) pl #%u\n", - padded_len, tx_msg->size, num_pls+1); - d_printf(2, dev, - "TX: appended hdr @%zu %zu b pl #%u @%zu %zu/%zu b\n", - (void *)tx_msg - i2400m->tx_buf, (size_t)tx_msg->size, - num_pls+1, ptr - i2400m->tx_buf, buf_len, padded_len); - result = 0; - if (is_singleton) - i2400m_tx_close(i2400m); - } -error_tx_new: - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - /* kick in most cases, except when the TX subsys is down, as - * it might free space */ - if (likely(result != -ESHUTDOWN)) - i2400m->bus_tx_kick(i2400m); - d_fnend(3, dev, "(i2400m %p skb %p [%zu bytes] pt %u) = %d\n", - i2400m, buf, buf_len, pl_type, result); - return result; -} -EXPORT_SYMBOL_GPL(i2400m_tx); - - -/** - * i2400m_tx_msg_get - Get the first TX message in the FIFO to start sending it - * - * @i2400m: device descriptors - * @bus_size: where to place the size of the TX message - * - * Called by the bus-specific driver to get the first TX message at - * the FIF that is ready for transmission. - * - * It sets the state in @i2400m to indicate the bus-specific driver is - * transferring that message (i2400m->tx_msg_size). - * - * Once the transfer is completed, call i2400m_tx_msg_sent(). - * - * Notes: - * - * The size of the TX message to be transmitted might be smaller than - * that of the TX message in the FIFO (in case the header was - * shorter). Hence, we copy it in @bus_size, for the bus layer to - * use. We keep the message's size in i2400m->tx_msg_size so that - * when the bus later is done transferring we know how much to - * advance the fifo. - * - * We collect statistics here as all the data is available and we - * assume it is going to work [see i2400m_tx_msg_sent()]. - */ -struct i2400m_msg_hdr *i2400m_tx_msg_get(struct i2400m *i2400m, - size_t *bus_size) -{ - struct device *dev = i2400m_dev(i2400m); - struct i2400m_msg_hdr *tx_msg, *tx_msg_moved; - unsigned long flags, pls; - - d_fnstart(3, dev, "(i2400m %p bus_size %p)\n", i2400m, bus_size); - spin_lock_irqsave(&i2400m->tx_lock, flags); - tx_msg_moved = NULL; - if (i2400m->tx_buf == NULL) - goto out_unlock; -skip: - tx_msg_moved = NULL; - if (i2400m->tx_in == i2400m->tx_out) { /* Empty FIFO? */ - i2400m->tx_in = 0; - i2400m->tx_out = 0; - d_printf(2, dev, "TX: FIFO empty: resetting\n"); - goto out_unlock; - } - tx_msg = i2400m->tx_buf + i2400m->tx_out % I2400M_TX_BUF_SIZE; - if (tx_msg->size & I2400M_TX_SKIP) { /* skip? */ - d_printf(2, dev, "TX: skip: msg @%zu (%zu b)\n", - i2400m->tx_out % I2400M_TX_BUF_SIZE, - (size_t) tx_msg->size & ~I2400M_TX_SKIP); - i2400m->tx_out += tx_msg->size & ~I2400M_TX_SKIP; - goto skip; - } - - if (tx_msg->num_pls == 0) { /* No payloads? */ - if (tx_msg == i2400m->tx_msg) { /* open, we are done */ - d_printf(2, dev, - "TX: FIFO empty: open msg w/o payloads @%zu\n", - (void *) tx_msg - i2400m->tx_buf); - tx_msg = NULL; - goto out_unlock; - } else { /* closed, skip it */ - d_printf(2, dev, - "TX: skip msg w/o payloads @%zu (%zu b)\n", - (void *) tx_msg - i2400m->tx_buf, - (size_t) tx_msg->size); - i2400m->tx_out += tx_msg->size & ~I2400M_TX_SKIP; - goto skip; - } - } - if (tx_msg == i2400m->tx_msg) /* open msg? */ - i2400m_tx_close(i2400m); - - /* Now we have a valid TX message (with payloads) to TX */ - tx_msg_moved = (void *) tx_msg + tx_msg->offset; - i2400m->tx_msg_size = tx_msg->size; - *bus_size = tx_msg_moved->size; - d_printf(2, dev, "TX: pid %d msg hdr at @%zu offset +@%zu " - "size %zu bus_size %zu\n", - current->pid, (void *) tx_msg - i2400m->tx_buf, - (size_t) tx_msg->offset, (size_t) tx_msg->size, - (size_t) tx_msg_moved->size); - tx_msg_moved->barker = le32_to_cpu(I2400M_H2D_PREVIEW_BARKER); - tx_msg_moved->sequence = le32_to_cpu(i2400m->tx_sequence++); - - pls = le32_to_cpu(tx_msg_moved->num_pls); - i2400m->tx_pl_num += pls; /* Update stats */ - if (pls > i2400m->tx_pl_max) - i2400m->tx_pl_max = pls; - if (pls < i2400m->tx_pl_min) - i2400m->tx_pl_min = pls; - i2400m->tx_num++; - i2400m->tx_size_acc += *bus_size; - if (*bus_size < i2400m->tx_size_min) - i2400m->tx_size_min = *bus_size; - if (*bus_size > i2400m->tx_size_max) - i2400m->tx_size_max = *bus_size; -out_unlock: - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - d_fnstart(3, dev, "(i2400m %p bus_size %p [%zu]) = %p\n", - i2400m, bus_size, *bus_size, tx_msg_moved); - return tx_msg_moved; -} -EXPORT_SYMBOL_GPL(i2400m_tx_msg_get); - - -/** - * i2400m_tx_msg_sent - indicate the transmission of a TX message - * - * @i2400m: device descriptor - * - * Called by the bus-specific driver when a message has been sent; - * this pops it from the FIFO; and as there is space, start the queue - * in case it was stopped. - * - * Should be called even if the message send failed and we are - * dropping this TX message. - */ -void i2400m_tx_msg_sent(struct i2400m *i2400m) -{ - unsigned n; - unsigned long flags; - struct device *dev = i2400m_dev(i2400m); - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - spin_lock_irqsave(&i2400m->tx_lock, flags); - if (i2400m->tx_buf == NULL) - goto out_unlock; - i2400m->tx_out += i2400m->tx_msg_size; - d_printf(2, dev, "TX: sent %zu b\n", (size_t) i2400m->tx_msg_size); - i2400m->tx_msg_size = 0; - BUG_ON(i2400m->tx_out > i2400m->tx_in); - /* level them FIFO markers off */ - n = i2400m->tx_out / I2400M_TX_BUF_SIZE; - i2400m->tx_out %= I2400M_TX_BUF_SIZE; - i2400m->tx_in -= n * I2400M_TX_BUF_SIZE; -out_unlock: - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} -EXPORT_SYMBOL_GPL(i2400m_tx_msg_sent); - - -/** - * i2400m_tx_setup - Initialize the TX queue and infrastructure - * - * Make sure we reset the TX sequence to zero, as when this function - * is called, the firmware has been just restarted. Same rational - * for tx_in, tx_out, tx_msg_size and tx_msg. We reset them since - * the memory for TX queue is reallocated. - */ -int i2400m_tx_setup(struct i2400m *i2400m) -{ - int result = 0; - void *tx_buf; - unsigned long flags; - - /* Do this here only once -- can't do on - * i2400m_hard_start_xmit() as we'll cause race conditions if - * the WS was scheduled on another CPU */ - INIT_WORK(&i2400m->wake_tx_ws, i2400m_wake_tx_work); - - tx_buf = kmalloc(I2400M_TX_BUF_SIZE, GFP_ATOMIC); - if (tx_buf == NULL) { - result = -ENOMEM; - goto error_kmalloc; - } - - /* - * Fail the build if we can't fit at least two maximum size messages - * on the TX FIFO [one being delivered while one is constructed]. - */ - BUILD_BUG_ON(2 * I2400M_TX_MSG_SIZE > I2400M_TX_BUF_SIZE); - spin_lock_irqsave(&i2400m->tx_lock, flags); - i2400m->tx_sequence = 0; - i2400m->tx_in = 0; - i2400m->tx_out = 0; - i2400m->tx_msg_size = 0; - i2400m->tx_msg = NULL; - i2400m->tx_buf = tx_buf; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - /* Huh? the bus layer has to define this... */ - BUG_ON(i2400m->bus_tx_block_size == 0); -error_kmalloc: - return result; - -} - - -/** - * i2400m_tx_release - Tear down the TX queue and infrastructure - */ -void i2400m_tx_release(struct i2400m *i2400m) -{ - unsigned long flags; - spin_lock_irqsave(&i2400m->tx_lock, flags); - kfree(i2400m->tx_buf); - i2400m->tx_buf = NULL; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); -} diff --git a/drivers/net/wimax/i2400m/usb-debug-levels.h b/drivers/net/wimax/i2400m/usb-debug-levels.h deleted file mode 100644 index b6f7335de765..000000000000 --- a/drivers/net/wimax/i2400m/usb-debug-levels.h +++ /dev/null @@ -1,28 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Intel Wireless WiMAX Connection 2400m - * Debug levels control file for the i2400m-usb module - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - */ -#ifndef __debug_levels__h__ -#define __debug_levels__h__ - -/* Maximum compile and run time debug level for all submodules */ -#define D_MODULENAME i2400m_usb -#define D_MASTER CONFIG_WIMAX_I2400M_DEBUG_LEVEL - -#include - -/* List of all the enabled modules */ -enum d_module { - D_SUBMODULE_DECLARE(usb), - D_SUBMODULE_DECLARE(fw), - D_SUBMODULE_DECLARE(notif), - D_SUBMODULE_DECLARE(rx), - D_SUBMODULE_DECLARE(tx), -}; - - -#endif /* #ifndef __debug_levels__h__ */ diff --git a/drivers/net/wimax/i2400m/usb-fw.c b/drivers/net/wimax/i2400m/usb-fw.c deleted file mode 100644 index 27ab233650d5..000000000000 --- a/drivers/net/wimax/i2400m/usb-fw.c +++ /dev/null @@ -1,365 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * Firmware uploader's USB specifics - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * Inaky Perez-Gonzalez - * - Initial implementation - * - * Inaky Perez-Gonzalez - * - bus generic/specific split - * - * THE PROCEDURE - * - * See fw.c for the generic description of this procedure. - * - * This file implements only the USB specifics. It boils down to how - * to send a command and waiting for an acknowledgement from the - * device. - * - * This code (and process) is single threaded. It assumes it is the - * only thread poking around (guaranteed by fw.c). - * - * COMMAND EXECUTION - * - * A write URB is posted with the buffer to the bulk output endpoint. - * - * ACK RECEPTION - * - * We just post a URB to the notification endpoint and wait for - * data. We repeat until we get all the data we expect (as indicated - * by the call from the bus generic code). - * - * The data is not read from the bulk in endpoint for boot mode. - * - * ROADMAP - * - * i2400mu_bus_bm_cmd_send - * i2400m_bm_cmd_prepare... - * i2400mu_tx_bulk_out - * - * i2400mu_bus_bm_wait_for_ack - * i2400m_notif_submit - */ -#include -#include -#include "i2400m-usb.h" - - -#define D_SUBMODULE fw -#include "usb-debug-levels.h" - - -/* - * Synchronous write to the device - * - * Takes care of updating EDC counts and thus, handle device errors. - */ -static -ssize_t i2400mu_tx_bulk_out(struct i2400mu *i2400mu, void *buf, size_t buf_size) -{ - int result; - struct device *dev = &i2400mu->usb_iface->dev; - int len; - struct usb_endpoint_descriptor *epd; - int pipe, do_autopm = 1; - - result = usb_autopm_get_interface(i2400mu->usb_iface); - if (result < 0) { - dev_err(dev, "BM-CMD: can't get autopm: %d\n", result); - do_autopm = 0; - } - epd = usb_get_epd(i2400mu->usb_iface, i2400mu->endpoint_cfg.bulk_out); - pipe = usb_sndbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); -retry: - result = usb_bulk_msg(i2400mu->usb_dev, pipe, buf, buf_size, &len, 200); - switch (result) { - case 0: - if (len != buf_size) { - dev_err(dev, "BM-CMD: short write (%u B vs %zu " - "expected)\n", len, buf_size); - result = -EIO; - break; - } - result = len; - break; - case -EPIPE: - /* - * Stall -- maybe the device is choking with our - * requests. Clear it and give it some time. If they - * happen to often, it might be another symptom, so we - * reset. - * - * No error handling for usb_clear_halt(0; if it - * works, the retry works; if it fails, this switch - * does the error handling for us. - */ - if (edc_inc(&i2400mu->urb_edc, - 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "BM-CMD: too many stalls in " - "URB; resetting device\n"); - usb_queue_reset_device(i2400mu->usb_iface); - } else { - usb_clear_halt(i2400mu->usb_dev, pipe); - msleep(10); /* give the device some time */ - goto retry; - } - fallthrough; - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* just ignore it */ - case -ESHUTDOWN: /* and exit */ - case -ECONNRESET: - result = -ESHUTDOWN; - break; - case -ETIMEDOUT: /* bah... */ - break; - default: /* any other? */ - if (edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "BM-CMD: maximum errors in " - "URB exceeded; resetting device\n"); - usb_queue_reset_device(i2400mu->usb_iface); - result = -ENODEV; - break; - } - dev_err(dev, "BM-CMD: URB error %d, retrying\n", - result); - goto retry; - } - if (do_autopm) - usb_autopm_put_interface(i2400mu->usb_iface); - return result; -} - - -/* - * Send a boot-mode command over the bulk-out pipe - * - * Command can be a raw command, which requires no preparation (and - * which might not even be following the command format). Checks that - * the right amount of data was transferred. - * - * To satisfy USB requirements (no onstack, vmalloc or in data segment - * buffers), we copy the command to i2400m->bm_cmd_buf and send it from - * there. - * - * @flags: pass thru from i2400m_bm_cmd() - * @return: cmd_size if ok, < 0 errno code on error. - */ -ssize_t i2400mu_bus_bm_cmd_send(struct i2400m *i2400m, - const struct i2400m_bootrom_header *_cmd, - size_t cmd_size, int flags) -{ - ssize_t result; - struct device *dev = i2400m_dev(i2400m); - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - int opcode = _cmd == NULL ? -1 : i2400m_brh_get_opcode(_cmd); - struct i2400m_bootrom_header *cmd; - size_t cmd_size_a = ALIGN(cmd_size, 16); /* USB restriction */ - - d_fnstart(8, dev, "(i2400m %p cmd %p size %zu)\n", - i2400m, _cmd, cmd_size); - result = -E2BIG; - if (cmd_size > I2400M_BM_CMD_BUF_SIZE) - goto error_too_big; - if (_cmd != i2400m->bm_cmd_buf) - memmove(i2400m->bm_cmd_buf, _cmd, cmd_size); - cmd = i2400m->bm_cmd_buf; - if (cmd_size_a > cmd_size) /* Zero pad space */ - memset(i2400m->bm_cmd_buf + cmd_size, 0, cmd_size_a - cmd_size); - if ((flags & I2400M_BM_CMD_RAW) == 0) { - if (WARN_ON(i2400m_brh_get_response_required(cmd) == 0)) - dev_warn(dev, "SW BUG: response_required == 0\n"); - i2400m_bm_cmd_prepare(cmd); - } - result = i2400mu_tx_bulk_out(i2400mu, i2400m->bm_cmd_buf, cmd_size); - if (result < 0) { - dev_err(dev, "boot-mode cmd %d: cannot send: %zd\n", - opcode, result); - goto error_cmd_send; - } - if (result != cmd_size) { /* all was transferred? */ - dev_err(dev, "boot-mode cmd %d: incomplete transfer " - "(%zd vs %zu submitted)\n", opcode, result, cmd_size); - result = -EIO; - goto error_cmd_size; - } -error_cmd_size: -error_cmd_send: -error_too_big: - d_fnend(8, dev, "(i2400m %p cmd %p size %zu) = %zd\n", - i2400m, _cmd, cmd_size, result); - return result; -} - - -static -void __i2400mu_bm_notif_cb(struct urb *urb) -{ - complete(urb->context); -} - - -/* - * submit a read to the notification endpoint - * - * @i2400m: device descriptor - * @urb: urb to use - * @completion: completion variable to complete when done - * - * Data is always read to i2400m->bm_ack_buf - */ -static -int i2400mu_notif_submit(struct i2400mu *i2400mu, struct urb *urb, - struct completion *completion) -{ - struct i2400m *i2400m = &i2400mu->i2400m; - struct usb_endpoint_descriptor *epd; - int pipe; - - epd = usb_get_epd(i2400mu->usb_iface, - i2400mu->endpoint_cfg.notification); - pipe = usb_rcvintpipe(i2400mu->usb_dev, epd->bEndpointAddress); - usb_fill_int_urb(urb, i2400mu->usb_dev, pipe, - i2400m->bm_ack_buf, I2400M_BM_ACK_BUF_SIZE, - __i2400mu_bm_notif_cb, completion, - epd->bInterval); - return usb_submit_urb(urb, GFP_KERNEL); -} - - -/* - * Read an ack from the notification endpoint - * - * @i2400m: - * @_ack: pointer to where to store the read data - * @ack_size: how many bytes we should read - * - * Returns: < 0 errno code on error; otherwise, amount of received bytes. - * - * Submits a notification read, appends the read data to the given ack - * buffer and then repeats (until @ack_size bytes have been - * received). - */ -ssize_t i2400mu_bus_bm_wait_for_ack(struct i2400m *i2400m, - struct i2400m_bootrom_header *_ack, - size_t ack_size) -{ - ssize_t result = -ENOMEM; - struct device *dev = i2400m_dev(i2400m); - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - struct urb notif_urb; - void *ack = _ack; - size_t offset, len; - long val; - int do_autopm = 1; - DECLARE_COMPLETION_ONSTACK(notif_completion); - - d_fnstart(8, dev, "(i2400m %p ack %p size %zu)\n", - i2400m, ack, ack_size); - BUG_ON(_ack == i2400m->bm_ack_buf); - result = usb_autopm_get_interface(i2400mu->usb_iface); - if (result < 0) { - dev_err(dev, "BM-ACK: can't get autopm: %d\n", (int) result); - do_autopm = 0; - } - usb_init_urb(¬if_urb); /* ready notifications */ - usb_get_urb(¬if_urb); - offset = 0; - while (offset < ack_size) { - init_completion(¬if_completion); - result = i2400mu_notif_submit(i2400mu, ¬if_urb, - ¬if_completion); - if (result < 0) - goto error_notif_urb_submit; - val = wait_for_completion_interruptible_timeout( - ¬if_completion, HZ); - if (val == 0) { - result = -ETIMEDOUT; - usb_kill_urb(¬if_urb); /* Timedout */ - goto error_notif_wait; - } - if (val == -ERESTARTSYS) { - result = -EINTR; /* Interrupted */ - usb_kill_urb(¬if_urb); - goto error_notif_wait; - } - result = notif_urb.status; /* How was the ack? */ - switch (result) { - case 0: - break; - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* just ignore it */ - case -ESHUTDOWN: /* and exit */ - case -ECONNRESET: - result = -ESHUTDOWN; - goto error_dev_gone; - default: /* any other? */ - usb_kill_urb(¬if_urb); /* Timedout */ - if (edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) - goto error_exceeded; - dev_err(dev, "BM-ACK: URB error %d, " - "retrying\n", notif_urb.status); - continue; /* retry */ - } - if (notif_urb.actual_length == 0) { - d_printf(6, dev, "ZLP received, retrying\n"); - continue; - } - /* Got data, append it to the buffer */ - len = min(ack_size - offset, (size_t) notif_urb.actual_length); - memcpy(ack + offset, i2400m->bm_ack_buf, len); - offset += len; - } - result = offset; -error_notif_urb_submit: -error_notif_wait: -error_dev_gone: -out: - if (do_autopm) - usb_autopm_put_interface(i2400mu->usb_iface); - d_fnend(8, dev, "(i2400m %p ack %p size %zu) = %ld\n", - i2400m, ack, ack_size, (long) result); - usb_put_urb(¬if_urb); - return result; - -error_exceeded: - dev_err(dev, "bm: maximum errors in notification URB exceeded; " - "resetting device\n"); - usb_queue_reset_device(i2400mu->usb_iface); - goto out; -} diff --git a/drivers/net/wimax/i2400m/usb-notif.c b/drivers/net/wimax/i2400m/usb-notif.c deleted file mode 100644 index 5d429f816125..000000000000 --- a/drivers/net/wimax/i2400m/usb-notif.c +++ /dev/null @@ -1,258 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m over USB - * Notification handling - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * Inaky Perez-Gonzalez - * - Initial implementation - * - * - * The notification endpoint is active when the device is not in boot - * mode; in here we just read and get notifications; based on those, - * we act to either reinitialize the device after a reboot or to - * submit a RX request. - * - * ROADMAP - * - * i2400mu_usb_notification_setup() - * - * i2400mu_usb_notification_release() - * - * i2400mu_usb_notification_cb() Called when a URB is ready - * i2400mu_notif_grok() - * i2400m_is_boot_barker() - * i2400m_dev_reset_handle() - * i2400mu_rx_kick() - */ -#include -#include -#include "i2400m-usb.h" - - -#define D_SUBMODULE notif -#include "usb-debug-levels.h" - - -static const -__le32 i2400m_ZERO_BARKER[4] = { 0, 0, 0, 0 }; - - -/* - * Process a received notification - * - * In normal operation mode, we can only receive two types of payloads - * on the notification endpoint: - * - * - a reboot barker, we do a bootstrap (the device has reseted). - * - * - a block of zeroes: there is pending data in the IN endpoint - */ -static -int i2400mu_notification_grok(struct i2400mu *i2400mu, const void *buf, - size_t buf_len) -{ - int ret; - struct device *dev = &i2400mu->usb_iface->dev; - struct i2400m *i2400m = &i2400mu->i2400m; - - d_fnstart(4, dev, "(i2400m %p buf %p buf_len %zu)\n", - i2400mu, buf, buf_len); - ret = -EIO; - if (buf_len < sizeof(i2400m_ZERO_BARKER)) - /* Not a bug, just ignore */ - goto error_bad_size; - ret = 0; - if (!memcmp(i2400m_ZERO_BARKER, buf, sizeof(i2400m_ZERO_BARKER))) { - i2400mu_rx_kick(i2400mu); - goto out; - } - ret = i2400m_is_boot_barker(i2400m, buf, buf_len); - if (unlikely(ret >= 0)) - ret = i2400m_dev_reset_handle(i2400m, "device rebooted"); - else /* Unknown or unexpected data in the notif message */ - i2400m_unknown_barker(i2400m, buf, buf_len); -error_bad_size: -out: - d_fnend(4, dev, "(i2400m %p buf %p buf_len %zu) = %d\n", - i2400mu, buf, buf_len, ret); - return ret; -} - - -/* - * URB callback for the notification endpoint - * - * @urb: the urb received from the notification endpoint - * - * This function will just process the USB side of the transaction, - * checking everything is fine, pass the processing to - * i2400m_notification_grok() and resubmit the URB. - */ -static -void i2400mu_notification_cb(struct urb *urb) -{ - int ret; - struct i2400mu *i2400mu = urb->context; - struct device *dev = &i2400mu->usb_iface->dev; - - d_fnstart(4, dev, "(urb %p status %d actual_length %d)\n", - urb, urb->status, urb->actual_length); - ret = urb->status; - switch (ret) { - case 0: - ret = i2400mu_notification_grok(i2400mu, urb->transfer_buffer, - urb->actual_length); - if (ret == -EIO && edc_inc(&i2400mu->urb_edc, EDC_MAX_ERRORS, - EDC_ERROR_TIMEFRAME)) - goto error_exceeded; - if (ret == -ENOMEM) /* uff...power cycle? shutdown? */ - goto error_exceeded; - break; - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* ditto */ - case -ESHUTDOWN: /* URB killed */ - case -ECONNRESET: /* disconnection */ - goto out; /* Notify around */ - default: /* Some error? */ - if (edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) - goto error_exceeded; - dev_err(dev, "notification: URB error %d, retrying\n", - urb->status); - } - usb_mark_last_busy(i2400mu->usb_dev); - ret = usb_submit_urb(i2400mu->notif_urb, GFP_ATOMIC); - switch (ret) { - case 0: - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* ditto */ - case -ESHUTDOWN: /* URB killed */ - case -ECONNRESET: /* disconnection */ - break; /* just ignore */ - default: /* Some error? */ - dev_err(dev, "notification: cannot submit URB: %d\n", ret); - goto error_submit; - } - d_fnend(4, dev, "(urb %p status %d actual_length %d) = void\n", - urb, urb->status, urb->actual_length); - return; - -error_exceeded: - dev_err(dev, "maximum errors in notification URB exceeded; " - "resetting device\n"); -error_submit: - usb_queue_reset_device(i2400mu->usb_iface); -out: - d_fnend(4, dev, "(urb %p status %d actual_length %d) = void\n", - urb, urb->status, urb->actual_length); -} - - -/* - * setup the notification endpoint - * - * @i2400m: device descriptor - * - * This procedure prepares the notification urb and handler for receiving - * unsolicited barkers from the device. - */ -int i2400mu_notification_setup(struct i2400mu *i2400mu) -{ - struct device *dev = &i2400mu->usb_iface->dev; - int usb_pipe, ret = 0; - struct usb_endpoint_descriptor *epd; - char *buf; - - d_fnstart(4, dev, "(i2400m %p)\n", i2400mu); - buf = kmalloc(I2400MU_MAX_NOTIFICATION_LEN, GFP_KERNEL | GFP_DMA); - if (buf == NULL) { - ret = -ENOMEM; - goto error_buf_alloc; - } - - i2400mu->notif_urb = usb_alloc_urb(0, GFP_KERNEL); - if (!i2400mu->notif_urb) { - ret = -ENOMEM; - goto error_alloc_urb; - } - epd = usb_get_epd(i2400mu->usb_iface, - i2400mu->endpoint_cfg.notification); - usb_pipe = usb_rcvintpipe(i2400mu->usb_dev, epd->bEndpointAddress); - usb_fill_int_urb(i2400mu->notif_urb, i2400mu->usb_dev, usb_pipe, - buf, I2400MU_MAX_NOTIFICATION_LEN, - i2400mu_notification_cb, i2400mu, epd->bInterval); - ret = usb_submit_urb(i2400mu->notif_urb, GFP_KERNEL); - if (ret != 0) { - dev_err(dev, "notification: cannot submit URB: %d\n", ret); - goto error_submit; - } - d_fnend(4, dev, "(i2400m %p) = %d\n", i2400mu, ret); - return ret; - -error_submit: - usb_free_urb(i2400mu->notif_urb); -error_alloc_urb: - kfree(buf); -error_buf_alloc: - d_fnend(4, dev, "(i2400m %p) = %d\n", i2400mu, ret); - return ret; -} - - -/* - * Tear down of the notification mechanism - * - * @i2400m: device descriptor - * - * Kill the interrupt endpoint urb, free any allocated resources. - * - * We need to check if we have done it before as for example, - * _suspend() call this; if after a suspend() we get a _disconnect() - * (as the case is when hibernating), nothing bad happens. - */ -void i2400mu_notification_release(struct i2400mu *i2400mu) -{ - struct device *dev = &i2400mu->usb_iface->dev; - - d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); - if (i2400mu->notif_urb != NULL) { - usb_kill_urb(i2400mu->notif_urb); - kfree(i2400mu->notif_urb->transfer_buffer); - usb_free_urb(i2400mu->notif_urb); - i2400mu->notif_urb = NULL; - } - d_fnend(4, dev, "(i2400mu %p)\n", i2400mu); -} diff --git a/drivers/net/wimax/i2400m/usb-rx.c b/drivers/net/wimax/i2400m/usb-rx.c deleted file mode 100644 index 5b64bda7d9e7..000000000000 --- a/drivers/net/wimax/i2400m/usb-rx.c +++ /dev/null @@ -1,462 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * USB RX handling - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * - Initial implementation - * Inaky Perez-Gonzalez - * - Use skb_clone(), break up processing in chunks - * - Split transport/device specific - * - Make buffer size dynamic to exert less memory pressure - * - * - * This handles the RX path on USB. - * - * When a notification is received that says 'there is RX data ready', - * we call i2400mu_rx_kick(); that wakes up the RX kthread, which - * reads a buffer from USB and passes it to i2400m_rx() in the generic - * handling code. The RX buffer has an specific format that is - * described in rx.c. - * - * We use a kernel thread in a loop because: - * - * - we want to be able to call the USB power management get/put - * functions (blocking) before each transaction. - * - * - We might get a lot of notifications and we don't want to submit - * a zillion reads; by serializing, we are throttling. - * - * - RX data processing can get heavy enough so that it is not - * appropriate for doing it in the USB callback; thus we run it in a - * process context. - * - * We provide a read buffer of an arbitrary size (short of a page); if - * the callback reports -EOVERFLOW, it means it was too small, so we - * just double the size and retry (being careful to append, as - * sometimes the device provided some data). Every now and then we - * check if the average packet size is smaller than the current packet - * size and if so, we halve it. At the end, the size of the - * preallocated buffer should be following the average received - * transaction size, adapting dynamically to it. - * - * ROADMAP - * - * i2400mu_rx_kick() Called from notif.c when we get a - * 'data ready' notification - * i2400mu_rxd() Kernel RX daemon - * i2400mu_rx() Receive USB data - * i2400m_rx() Send data to generic i2400m RX handling - * - * i2400mu_rx_setup() called from i2400mu_bus_dev_start() - * - * i2400mu_rx_release() called from i2400mu_bus_dev_stop() - */ -#include -#include -#include -#include "i2400m-usb.h" - - -#define D_SUBMODULE rx -#include "usb-debug-levels.h" - -/* - * Dynamic RX size - * - * We can't let the rx_size be a multiple of 512 bytes (the RX - * endpoint's max packet size). On some USB host controllers (we - * haven't been able to fully characterize which), if the device is - * about to send (for example) X bytes and we only post a buffer to - * receive n*512, it will fail to mark that as babble (so that - * i2400mu_rx() [case -EOVERFLOW] can resize the buffer and get the - * rest). - * - * So on growing or shrinking, if it is a multiple of the - * maxpacketsize, we remove some (instead of incresing some, so in a - * buddy allocator we try to waste less space). - * - * Note we also need a hook for this on i2400mu_rx() -- when we do the - * first read, we are sure we won't hit this spot because - * i240mm->rx_size has been set properly. However, if we have to - * double because of -EOVERFLOW, when we launch the read to get the - * rest of the data, we *have* to make sure that also is not a - * multiple of the max_pkt_size. - */ - -static -size_t i2400mu_rx_size_grow(struct i2400mu *i2400mu) -{ - struct device *dev = &i2400mu->usb_iface->dev; - size_t rx_size; - const size_t max_pkt_size = 512; - - rx_size = 2 * i2400mu->rx_size; - if (rx_size % max_pkt_size == 0) { - rx_size -= 8; - d_printf(1, dev, - "RX: expected size grew to %zu [adjusted -8] " - "from %zu\n", - rx_size, i2400mu->rx_size); - } else - d_printf(1, dev, - "RX: expected size grew to %zu from %zu\n", - rx_size, i2400mu->rx_size); - return rx_size; -} - - -static -void i2400mu_rx_size_maybe_shrink(struct i2400mu *i2400mu) -{ - const size_t max_pkt_size = 512; - struct device *dev = &i2400mu->usb_iface->dev; - - if (unlikely(i2400mu->rx_size_cnt >= 100 - && i2400mu->rx_size_auto_shrink)) { - size_t avg_rx_size = - i2400mu->rx_size_acc / i2400mu->rx_size_cnt; - size_t new_rx_size = i2400mu->rx_size / 2; - if (avg_rx_size < new_rx_size) { - if (new_rx_size % max_pkt_size == 0) { - new_rx_size -= 8; - d_printf(1, dev, - "RX: expected size shrank to %zu " - "[adjusted -8] from %zu\n", - new_rx_size, i2400mu->rx_size); - } else - d_printf(1, dev, - "RX: expected size shrank to %zu " - "from %zu\n", - new_rx_size, i2400mu->rx_size); - i2400mu->rx_size = new_rx_size; - i2400mu->rx_size_cnt = 0; - i2400mu->rx_size_acc = i2400mu->rx_size; - } - } -} - -/* - * Receive a message with payloads from the USB bus into an skb - * - * @i2400mu: USB device descriptor - * @rx_skb: skb where to place the received message - * - * Deals with all the USB-specifics of receiving, dynamically - * increasing the buffer size if so needed. Returns the payload in the - * skb, ready to process. On a zero-length packet, we retry. - * - * On soft USB errors, we retry (until they become too frequent and - * then are promoted to hard); on hard USB errors, we reset the - * device. On other errors (skb realloacation, we just drop it and - * hope for the next invocation to solve it). - * - * Returns: pointer to the skb if ok, ERR_PTR on error. - * NOTE: this function might realloc the skb (if it is too small), - * so always update with the one returned. - * ERR_PTR() is < 0 on error. - * Will return NULL if it cannot reallocate -- this can be - * considered a transient retryable error. - */ -static -struct sk_buff *i2400mu_rx(struct i2400mu *i2400mu, struct sk_buff *rx_skb) -{ - int result = 0; - struct device *dev = &i2400mu->usb_iface->dev; - int usb_pipe, read_size, rx_size, do_autopm; - struct usb_endpoint_descriptor *epd; - const size_t max_pkt_size = 512; - - d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); - do_autopm = atomic_read(&i2400mu->do_autopm); - result = do_autopm ? - usb_autopm_get_interface(i2400mu->usb_iface) : 0; - if (result < 0) { - dev_err(dev, "RX: can't get autopm: %d\n", result); - do_autopm = 0; - } - epd = usb_get_epd(i2400mu->usb_iface, i2400mu->endpoint_cfg.bulk_in); - usb_pipe = usb_rcvbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); -retry: - rx_size = skb_end_pointer(rx_skb) - rx_skb->data - rx_skb->len; - if (unlikely(rx_size % max_pkt_size == 0)) { - rx_size -= 8; - d_printf(1, dev, "RX: rx_size adapted to %d [-8]\n", rx_size); - } - result = usb_bulk_msg( - i2400mu->usb_dev, usb_pipe, rx_skb->data + rx_skb->len, - rx_size, &read_size, 200); - usb_mark_last_busy(i2400mu->usb_dev); - switch (result) { - case 0: - if (read_size == 0) - goto retry; /* ZLP, just resubmit */ - skb_put(rx_skb, read_size); - break; - case -EPIPE: - /* - * Stall -- maybe the device is choking with our - * requests. Clear it and give it some time. If they - * happen to often, it might be another symptom, so we - * reset. - * - * No error handling for usb_clear_halt(0; if it - * works, the retry works; if it fails, this switch - * does the error handling for us. - */ - if (edc_inc(&i2400mu->urb_edc, - 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "BM-CMD: too many stalls in " - "URB; resetting device\n"); - goto do_reset; - } - usb_clear_halt(i2400mu->usb_dev, usb_pipe); - msleep(10); /* give the device some time */ - goto retry; - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* just ignore it */ - case -ESHUTDOWN: - case -ECONNRESET: - break; - case -EOVERFLOW: { /* too small, reallocate */ - struct sk_buff *new_skb; - rx_size = i2400mu_rx_size_grow(i2400mu); - if (rx_size <= (1 << 16)) /* cap it */ - i2400mu->rx_size = rx_size; - else if (printk_ratelimit()) { - dev_err(dev, "BUG? rx_size up to %d\n", rx_size); - result = -EINVAL; - goto out; - } - skb_put(rx_skb, read_size); - new_skb = skb_copy_expand(rx_skb, 0, rx_size - rx_skb->len, - GFP_KERNEL); - if (new_skb == NULL) { - kfree_skb(rx_skb); - rx_skb = NULL; - goto out; /* drop it...*/ - } - kfree_skb(rx_skb); - rx_skb = new_skb; - i2400mu->rx_size_cnt = 0; - i2400mu->rx_size_acc = i2400mu->rx_size; - d_printf(1, dev, "RX: size changed to %d, received %d, " - "copied %d, capacity %ld\n", - rx_size, read_size, rx_skb->len, - (long) skb_end_offset(new_skb)); - goto retry; - } - /* In most cases, it happens due to the hardware scheduling a - * read when there was no data - unfortunately, we have no way - * to tell this timeout from a USB timeout. So we just ignore - * it. */ - case -ETIMEDOUT: - dev_err(dev, "RX: timeout: %d\n", result); - result = 0; - break; - default: /* Any error */ - if (edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) - goto error_reset; - dev_err(dev, "RX: error receiving URB: %d, retrying\n", result); - goto retry; - } -out: - if (do_autopm) - usb_autopm_put_interface(i2400mu->usb_iface); - d_fnend(4, dev, "(i2400mu %p) = %p\n", i2400mu, rx_skb); - return rx_skb; - -error_reset: - dev_err(dev, "RX: maximum errors in URB exceeded; " - "resetting device\n"); -do_reset: - usb_queue_reset_device(i2400mu->usb_iface); - rx_skb = ERR_PTR(result); - goto out; -} - - -/* - * Kernel thread for USB reception of data - * - * This thread waits for a kick; once kicked, it will allocate an skb - * and receive a single message to it from USB (using - * i2400mu_rx()). Once received, it is passed to the generic i2400m RX - * code for processing. - * - * When done processing, it runs some dirty statistics to verify if - * the last 100 messages received were smaller than half of the - * current RX buffer size. In that case, the RX buffer size is - * halved. This will helps lowering the pressure on the memory - * allocator. - * - * Hard errors force the thread to exit. - */ -static -int i2400mu_rxd(void *_i2400mu) -{ - int result = 0; - struct i2400mu *i2400mu = _i2400mu; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = &i2400mu->usb_iface->dev; - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - size_t pending; - int rx_size; - struct sk_buff *rx_skb; - unsigned long flags; - - d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); - spin_lock_irqsave(&i2400m->rx_lock, flags); - BUG_ON(i2400mu->rx_kthread != NULL); - i2400mu->rx_kthread = current; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - while (1) { - d_printf(2, dev, "RX: waiting for messages\n"); - pending = 0; - wait_event_interruptible( - i2400mu->rx_wq, - (kthread_should_stop() /* check this first! */ - || (pending = atomic_read(&i2400mu->rx_pending_count))) - ); - if (kthread_should_stop()) - break; - if (pending == 0) - continue; - rx_size = i2400mu->rx_size; - d_printf(2, dev, "RX: reading up to %d bytes\n", rx_size); - rx_skb = __netdev_alloc_skb(net_dev, rx_size, GFP_KERNEL); - if (rx_skb == NULL) { - dev_err(dev, "RX: can't allocate skb [%d bytes]\n", - rx_size); - msleep(50); /* give it some time? */ - continue; - } - - /* Receive the message with the payloads */ - rx_skb = i2400mu_rx(i2400mu, rx_skb); - result = PTR_ERR(rx_skb); - if (IS_ERR(rx_skb)) - goto out; - atomic_dec(&i2400mu->rx_pending_count); - if (rx_skb == NULL || rx_skb->len == 0) { - /* some "ignorable" condition */ - kfree_skb(rx_skb); - continue; - } - - /* Deliver the message to the generic i2400m code */ - i2400mu->rx_size_cnt++; - i2400mu->rx_size_acc += rx_skb->len; - result = i2400m_rx(i2400m, rx_skb); - if (result == -EIO - && edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - goto error_reset; - } - - /* Maybe adjust RX buffer size */ - i2400mu_rx_size_maybe_shrink(i2400mu); - } - result = 0; -out: - spin_lock_irqsave(&i2400m->rx_lock, flags); - i2400mu->rx_kthread = NULL; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - d_fnend(4, dev, "(i2400mu %p) = %d\n", i2400mu, result); - return result; - -error_reset: - dev_err(dev, "RX: maximum errors in received buffer exceeded; " - "resetting device\n"); - usb_queue_reset_device(i2400mu->usb_iface); - goto out; -} - - -/* - * Start reading from the device - * - * @i2400m: device instance - * - * Notify the RX thread that there is data pending. - */ -void i2400mu_rx_kick(struct i2400mu *i2400mu) -{ - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = &i2400mu->usb_iface->dev; - - d_fnstart(3, dev, "(i2400mu %p)\n", i2400m); - atomic_inc(&i2400mu->rx_pending_count); - wake_up_all(&i2400mu->rx_wq); - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} - - -int i2400mu_rx_setup(struct i2400mu *i2400mu) -{ - int result = 0; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = &i2400mu->usb_iface->dev; - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - struct task_struct *kthread; - - kthread = kthread_run(i2400mu_rxd, i2400mu, "%s-rx", - wimax_dev->name); - /* the kthread function sets i2400mu->rx_thread */ - if (IS_ERR(kthread)) { - result = PTR_ERR(kthread); - dev_err(dev, "RX: cannot start thread: %d\n", result); - } - return result; -} - - -void i2400mu_rx_release(struct i2400mu *i2400mu) -{ - unsigned long flags; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = i2400m_dev(i2400m); - struct task_struct *kthread; - - spin_lock_irqsave(&i2400m->rx_lock, flags); - kthread = i2400mu->rx_kthread; - i2400mu->rx_kthread = NULL; - spin_unlock_irqrestore(&i2400m->rx_lock, flags); - if (kthread) - kthread_stop(kthread); - else - d_printf(1, dev, "RX: kthread had already exited\n"); -} - diff --git a/drivers/net/wimax/i2400m/usb-tx.c b/drivers/net/wimax/i2400m/usb-tx.c deleted file mode 100644 index 3ba9d70cca1b..000000000000 --- a/drivers/net/wimax/i2400m/usb-tx.c +++ /dev/null @@ -1,273 +0,0 @@ -/* - * Intel Wireless WiMAX Connection 2400m - * USB specific TX handling - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Yanir Lubetkin - * - Initial implementation - * Inaky Perez-Gonzalez - * - Split transport/device specific - * - * - * Takes the TX messages in the i2400m's driver TX FIFO and sends them - * to the device until there are no more. - * - * If we fail sending the message, we just drop it. There isn't much - * we can do at this point. We could also retry, but the USB stack has - * already retried and still failed, so there is not much of a - * point. As well, most of the traffic is network, which has recovery - * methods for dropped packets. - * - * For sending we just obtain a FIFO buffer to send, send it to the - * USB bulk out, tell the TX FIFO code we have sent it; query for - * another one, etc... until done. - * - * We use a thread so we can call usb_autopm_enable() and - * usb_autopm_disable() for each transaction; this way when the device - * goes idle, it will suspend. It also has less overhead than a - * dedicated workqueue, as it is being used for a single task. - * - * ROADMAP - * - * i2400mu_tx_setup() - * i2400mu_tx_release() - * - * i2400mu_bus_tx_kick() - Called by the tx.c code when there - * is new data in the FIFO. - * i2400mu_txd() - * i2400m_tx_msg_get() - * i2400m_tx_msg_sent() - */ -#include "i2400m-usb.h" - - -#define D_SUBMODULE tx -#include "usb-debug-levels.h" - - -/* - * Get the next TX message in the TX FIFO and send it to the device - * - * Note that any iteration consumes a message to be sent, no matter if - * it succeeds or fails (we have no real way to retry or complain). - * - * Return: 0 if ok, < 0 errno code on hard error. - */ -static -int i2400mu_tx(struct i2400mu *i2400mu, struct i2400m_msg_hdr *tx_msg, - size_t tx_msg_size) -{ - int result = 0; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = &i2400mu->usb_iface->dev; - int usb_pipe, sent_size, do_autopm; - struct usb_endpoint_descriptor *epd; - - d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); - do_autopm = atomic_read(&i2400mu->do_autopm); - result = do_autopm ? - usb_autopm_get_interface(i2400mu->usb_iface) : 0; - if (result < 0) { - dev_err(dev, "TX: can't get autopm: %d\n", result); - do_autopm = 0; - } - epd = usb_get_epd(i2400mu->usb_iface, i2400mu->endpoint_cfg.bulk_out); - usb_pipe = usb_sndbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); -retry: - result = usb_bulk_msg(i2400mu->usb_dev, usb_pipe, - tx_msg, tx_msg_size, &sent_size, 200); - usb_mark_last_busy(i2400mu->usb_dev); - switch (result) { - case 0: - if (sent_size != tx_msg_size) { /* Too short? drop it */ - dev_err(dev, "TX: short write (%d B vs %zu " - "expected)\n", sent_size, tx_msg_size); - result = -EIO; - } - break; - case -EPIPE: - /* - * Stall -- maybe the device is choking with our - * requests. Clear it and give it some time. If they - * happen to often, it might be another symptom, so we - * reset. - * - * No error handling for usb_clear_halt(0; if it - * works, the retry works; if it fails, this switch - * does the error handling for us. - */ - if (edc_inc(&i2400mu->urb_edc, - 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "BM-CMD: too many stalls in " - "URB; resetting device\n"); - usb_queue_reset_device(i2400mu->usb_iface); - } else { - usb_clear_halt(i2400mu->usb_dev, usb_pipe); - msleep(10); /* give the device some time */ - goto retry; - } - fallthrough; - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* just ignore it */ - case -ESHUTDOWN: /* and exit */ - case -ECONNRESET: - result = -ESHUTDOWN; - break; - default: /* Some error? */ - if (edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "TX: maximum errors in URB " - "exceeded; resetting device\n"); - usb_queue_reset_device(i2400mu->usb_iface); - } else { - dev_err(dev, "TX: cannot send URB; retrying. " - "tx_msg @%zu %zu B [%d sent]: %d\n", - (void *) tx_msg - i2400m->tx_buf, - tx_msg_size, sent_size, result); - goto retry; - } - } - if (do_autopm) - usb_autopm_put_interface(i2400mu->usb_iface); - d_fnend(4, dev, "(i2400mu %p) = result\n", i2400mu); - return result; -} - - -/* - * Get the next TX message in the TX FIFO and send it to the device - * - * Note we exit the loop if i2400mu_tx() fails; that function only - * fails on hard error (failing to tx a buffer not being one of them, - * see its doc). - * - * Return: 0 - */ -static -int i2400mu_txd(void *_i2400mu) -{ - struct i2400mu *i2400mu = _i2400mu; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = &i2400mu->usb_iface->dev; - struct i2400m_msg_hdr *tx_msg; - size_t tx_msg_size; - unsigned long flags; - - d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); - - spin_lock_irqsave(&i2400m->tx_lock, flags); - BUG_ON(i2400mu->tx_kthread != NULL); - i2400mu->tx_kthread = current; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - - while (1) { - d_printf(2, dev, "TX: waiting for messages\n"); - tx_msg = NULL; - wait_event_interruptible( - i2400mu->tx_wq, - (kthread_should_stop() /* check this first! */ - || (tx_msg = i2400m_tx_msg_get(i2400m, &tx_msg_size))) - ); - if (kthread_should_stop()) - break; - WARN_ON(tx_msg == NULL); /* should not happen...*/ - d_printf(2, dev, "TX: submitting %zu bytes\n", tx_msg_size); - d_dump(5, dev, tx_msg, tx_msg_size); - /* Yeah, we ignore errors ... not much we can do */ - i2400mu_tx(i2400mu, tx_msg, tx_msg_size); - i2400m_tx_msg_sent(i2400m); /* ack it, advance the FIFO */ - } - - spin_lock_irqsave(&i2400m->tx_lock, flags); - i2400mu->tx_kthread = NULL; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - - d_fnend(4, dev, "(i2400mu %p)\n", i2400mu); - return 0; -} - - -/* - * i2400m TX engine notifies us that there is data in the FIFO ready - * for TX - * - * If there is a URB in flight, don't do anything; when it finishes, - * it will see there is data in the FIFO and send it. Else, just - * submit a write. - */ -void i2400mu_bus_tx_kick(struct i2400m *i2400m) -{ - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - struct device *dev = &i2400mu->usb_iface->dev; - - d_fnstart(3, dev, "(i2400m %p) = void\n", i2400m); - wake_up_all(&i2400mu->tx_wq); - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} - - -int i2400mu_tx_setup(struct i2400mu *i2400mu) -{ - int result = 0; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = &i2400mu->usb_iface->dev; - struct wimax_dev *wimax_dev = &i2400m->wimax_dev; - struct task_struct *kthread; - - kthread = kthread_run(i2400mu_txd, i2400mu, "%s-tx", - wimax_dev->name); - /* the kthread function sets i2400mu->tx_thread */ - if (IS_ERR(kthread)) { - result = PTR_ERR(kthread); - dev_err(dev, "TX: cannot start thread: %d\n", result); - } - return result; -} - -void i2400mu_tx_release(struct i2400mu *i2400mu) -{ - unsigned long flags; - struct i2400m *i2400m = &i2400mu->i2400m; - struct device *dev = i2400m_dev(i2400m); - struct task_struct *kthread; - - spin_lock_irqsave(&i2400m->tx_lock, flags); - kthread = i2400mu->tx_kthread; - i2400mu->tx_kthread = NULL; - spin_unlock_irqrestore(&i2400m->tx_lock, flags); - if (kthread) - kthread_stop(kthread); - else - d_printf(1, dev, "TX: kthread had already exited\n"); -} diff --git a/drivers/net/wimax/i2400m/usb.c b/drivers/net/wimax/i2400m/usb.c deleted file mode 100644 index b684e97ac976..000000000000 --- a/drivers/net/wimax/i2400m/usb.c +++ /dev/null @@ -1,764 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Intel Wireless WiMAX Connection 2400m - * Linux driver model glue for USB device, reset & fw upload - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - * Yanir Lubetkin - * - * See i2400m-usb.h for a general description of this driver. - * - * This file implements driver model glue, and hook ups for the - * generic driver to implement the bus-specific functions (device - * communication setup/tear down, firmware upload and resetting). - * - * ROADMAP - * - * i2400mu_probe() - * alloc_netdev()... - * i2400mu_netdev_setup() - * i2400mu_init() - * i2400m_netdev_setup() - * i2400m_setup()... - * - * i2400mu_disconnect - * i2400m_release() - * free_netdev() - * - * i2400mu_suspend() - * i2400m_cmd_enter_powersave() - * i2400mu_notification_release() - * - * i2400mu_resume() - * i2400mu_notification_setup() - * - * i2400mu_bus_dev_start() Called by i2400m_dev_start() [who is - * i2400mu_tx_setup() called by i2400m_setup()] - * i2400mu_rx_setup() - * i2400mu_notification_setup() - * - * i2400mu_bus_dev_stop() Called by i2400m_dev_stop() [who is - * i2400mu_notification_release() called by i2400m_release()] - * i2400mu_rx_release() - * i2400mu_tx_release() - * - * i2400mu_bus_reset() Called by i2400m_reset - * __i2400mu_reset() - * __i2400mu_send_barker() - * usb_reset_device() - */ -#include "i2400m-usb.h" -#include -#include -#include -#include - - -#define D_SUBMODULE usb -#include "usb-debug-levels.h" - -static char i2400mu_debug_params[128]; -module_param_string(debug, i2400mu_debug_params, sizeof(i2400mu_debug_params), - 0644); -MODULE_PARM_DESC(debug, - "String of space-separated NAME:VALUE pairs, where NAMEs " - "are the different debug submodules and VALUE are the " - "initial debug value to set."); - -/* Our firmware file name */ -static const char *i2400mu_bus_fw_names_5x50[] = { -#define I2400MU_FW_FILE_NAME_v1_5 "i2400m-fw-usb-1.5.sbcf" - I2400MU_FW_FILE_NAME_v1_5, -#define I2400MU_FW_FILE_NAME_v1_4 "i2400m-fw-usb-1.4.sbcf" - I2400MU_FW_FILE_NAME_v1_4, - NULL, -}; - - -static const char *i2400mu_bus_fw_names_6050[] = { -#define I6050U_FW_FILE_NAME_v1_5 "i6050-fw-usb-1.5.sbcf" - I6050U_FW_FILE_NAME_v1_5, - NULL, -}; - - -static -int i2400mu_bus_dev_start(struct i2400m *i2400m) -{ - int result; - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - struct device *dev = &i2400mu->usb_iface->dev; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - result = i2400mu_tx_setup(i2400mu); - if (result < 0) - goto error_usb_tx_setup; - result = i2400mu_rx_setup(i2400mu); - if (result < 0) - goto error_usb_rx_setup; - result = i2400mu_notification_setup(i2400mu); - if (result < 0) - goto error_notif_setup; - d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); - return result; - -error_notif_setup: - i2400mu_rx_release(i2400mu); -error_usb_rx_setup: - i2400mu_tx_release(i2400mu); -error_usb_tx_setup: - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); - return result; -} - - -static -void i2400mu_bus_dev_stop(struct i2400m *i2400m) -{ - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - struct device *dev = &i2400mu->usb_iface->dev; - - d_fnstart(3, dev, "(i2400m %p)\n", i2400m); - i2400mu_notification_release(i2400mu); - i2400mu_rx_release(i2400mu); - i2400mu_tx_release(i2400mu); - d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); -} - - -/* - * Sends a barker buffer to the device - * - * This helper will allocate a kmalloced buffer and use it to transmit - * (then free it). Reason for this is that other arches cannot use - * stack/vmalloc/text areas for DMA transfers. - * - * Error recovery here is simpler: anything is considered a hard error - * and will move the reset code to use a last-resort bus-based reset. - */ -static -int __i2400mu_send_barker(struct i2400mu *i2400mu, - const __le32 *barker, - size_t barker_size, - unsigned endpoint) -{ - struct usb_endpoint_descriptor *epd = NULL; - int pipe, actual_len, ret; - struct device *dev = &i2400mu->usb_iface->dev; - void *buffer; - int do_autopm = 1; - - ret = usb_autopm_get_interface(i2400mu->usb_iface); - if (ret < 0) { - dev_err(dev, "RESET: can't get autopm: %d\n", ret); - do_autopm = 0; - } - ret = -ENOMEM; - buffer = kmalloc(barker_size, GFP_KERNEL); - if (buffer == NULL) - goto error_kzalloc; - epd = usb_get_epd(i2400mu->usb_iface, endpoint); - pipe = usb_sndbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); - memcpy(buffer, barker, barker_size); -retry: - ret = usb_bulk_msg(i2400mu->usb_dev, pipe, buffer, barker_size, - &actual_len, 200); - switch (ret) { - case 0: - if (actual_len != barker_size) { /* Too short? drop it */ - dev_err(dev, "E: %s: short write (%d B vs %zu " - "expected)\n", - __func__, actual_len, barker_size); - ret = -EIO; - } - break; - case -EPIPE: - /* - * Stall -- maybe the device is choking with our - * requests. Clear it and give it some time. If they - * happen to often, it might be another symptom, so we - * reset. - * - * No error handling for usb_clear_halt(0; if it - * works, the retry works; if it fails, this switch - * does the error handling for us. - */ - if (edc_inc(&i2400mu->urb_edc, - 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "E: %s: too many stalls in " - "URB; resetting device\n", __func__); - usb_queue_reset_device(i2400mu->usb_iface); - /* fallthrough */ - } else { - usb_clear_halt(i2400mu->usb_dev, pipe); - msleep(10); /* give the device some time */ - goto retry; - } - fallthrough; - case -EINVAL: /* while removing driver */ - case -ENODEV: /* dev disconnect ... */ - case -ENOENT: /* just ignore it */ - case -ESHUTDOWN: /* and exit */ - case -ECONNRESET: - ret = -ESHUTDOWN; - break; - default: /* Some error? */ - if (edc_inc(&i2400mu->urb_edc, - EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { - dev_err(dev, "E: %s: maximum errors in URB " - "exceeded; resetting device\n", - __func__); - usb_queue_reset_device(i2400mu->usb_iface); - } else { - dev_warn(dev, "W: %s: cannot send URB: %d\n", - __func__, ret); - goto retry; - } - } - kfree(buffer); -error_kzalloc: - if (do_autopm) - usb_autopm_put_interface(i2400mu->usb_iface); - return ret; -} - - -/* - * Reset a device at different levels (warm, cold or bus) - * - * @i2400m: device descriptor - * @reset_type: soft, warm or bus reset (I2400M_RT_WARM/SOFT/BUS) - * - * Warm and cold resets get a USB reset if they fail. - * - * Warm reset: - * - * The device will be fully reset internally, but won't be - * disconnected from the USB bus (so no reenumeration will - * happen). Firmware upload will be necessary. - * - * The device will send a reboot barker in the notification endpoint - * that will trigger the driver to reinitialize the state - * automatically from notif.c:i2400m_notification_grok() into - * i2400m_dev_bootstrap_delayed(). - * - * Cold and bus (USB) reset: - * - * The device will be fully reset internally, disconnected from the - * USB bus an a reenumeration will happen. Firmware upload will be - * necessary. Thus, we don't do any locking or struct - * reinitialization, as we are going to be fully disconnected and - * reenumerated. - * - * Note we need to return -ENODEV if a warm reset was requested and we - * had to resort to a bus reset. See i2400m_op_reset(), wimax_reset() - * and wimax_dev->op_reset. - * - * WARNING: no driver state saved/fixed - */ -static -int i2400mu_bus_reset(struct i2400m *i2400m, enum i2400m_reset_type rt) -{ - int result; - struct i2400mu *i2400mu = - container_of(i2400m, struct i2400mu, i2400m); - struct device *dev = i2400m_dev(i2400m); - static const __le32 i2400m_WARM_BOOT_BARKER[4] = { - cpu_to_le32(I2400M_WARM_RESET_BARKER), - cpu_to_le32(I2400M_WARM_RESET_BARKER), - cpu_to_le32(I2400M_WARM_RESET_BARKER), - cpu_to_le32(I2400M_WARM_RESET_BARKER), - }; - static const __le32 i2400m_COLD_BOOT_BARKER[4] = { - cpu_to_le32(I2400M_COLD_RESET_BARKER), - cpu_to_le32(I2400M_COLD_RESET_BARKER), - cpu_to_le32(I2400M_COLD_RESET_BARKER), - cpu_to_le32(I2400M_COLD_RESET_BARKER), - }; - - d_fnstart(3, dev, "(i2400m %p rt %u)\n", i2400m, rt); - if (rt == I2400M_RT_WARM) - result = __i2400mu_send_barker( - i2400mu, i2400m_WARM_BOOT_BARKER, - sizeof(i2400m_WARM_BOOT_BARKER), - i2400mu->endpoint_cfg.bulk_out); - else if (rt == I2400M_RT_COLD) - result = __i2400mu_send_barker( - i2400mu, i2400m_COLD_BOOT_BARKER, - sizeof(i2400m_COLD_BOOT_BARKER), - i2400mu->endpoint_cfg.reset_cold); - else if (rt == I2400M_RT_BUS) { - result = usb_reset_device(i2400mu->usb_dev); - switch (result) { - case 0: - case -EINVAL: /* device is gone */ - case -ENODEV: - case -ENOENT: - case -ESHUTDOWN: - result = 0; - break; /* We assume the device is disconnected */ - default: - dev_err(dev, "USB reset failed (%d), giving up!\n", - result); - } - } else { - result = -EINVAL; /* shut gcc up in certain arches */ - BUG(); - } - if (result < 0 - && result != -EINVAL /* device is gone */ - && rt != I2400M_RT_BUS) { - /* - * Things failed -- resort to lower level reset, that - * we queue in another context; the reason for this is - * that the pre and post reset functionality requires - * the i2400m->init_mutex; RT_WARM and RT_COLD can - * come from areas where i2400m->init_mutex is taken. - */ - dev_err(dev, "%s reset failed (%d); trying USB reset\n", - rt == I2400M_RT_WARM ? "warm" : "cold", result); - usb_queue_reset_device(i2400mu->usb_iface); - result = -ENODEV; - } - d_fnend(3, dev, "(i2400m %p rt %u) = %d\n", i2400m, rt, result); - return result; -} - -static void i2400mu_get_drvinfo(struct net_device *net_dev, - struct ethtool_drvinfo *info) -{ - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - struct usb_device *udev = i2400mu->usb_dev; - - strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); - strlcpy(info->fw_version, i2400m->fw_name ? : "", - sizeof(info->fw_version)); - usb_make_path(udev, info->bus_info, sizeof(info->bus_info)); -} - -static const struct ethtool_ops i2400mu_ethtool_ops = { - .get_drvinfo = i2400mu_get_drvinfo, - .get_link = ethtool_op_get_link, -}; - -static -void i2400mu_netdev_setup(struct net_device *net_dev) -{ - struct i2400m *i2400m = net_dev_to_i2400m(net_dev); - struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); - i2400mu_init(i2400mu); - i2400m_netdev_setup(net_dev); - net_dev->ethtool_ops = &i2400mu_ethtool_ops; -} - - -/* - * Debug levels control; see debug.h - */ -struct d_level D_LEVEL[] = { - D_SUBMODULE_DEFINE(usb), - D_SUBMODULE_DEFINE(fw), - D_SUBMODULE_DEFINE(notif), - D_SUBMODULE_DEFINE(rx), - D_SUBMODULE_DEFINE(tx), -}; -size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); - -static -void i2400mu_debugfs_add(struct i2400mu *i2400mu) -{ - struct dentry *dentry = i2400mu->i2400m.wimax_dev.debugfs_dentry; - - dentry = debugfs_create_dir("i2400m-usb", dentry); - i2400mu->debugfs_dentry = dentry; - - d_level_register_debugfs("dl_", usb, dentry); - d_level_register_debugfs("dl_", fw, dentry); - d_level_register_debugfs("dl_", notif, dentry); - d_level_register_debugfs("dl_", rx, dentry); - d_level_register_debugfs("dl_", tx, dentry); - - /* Don't touch these if you don't know what you are doing */ - debugfs_create_u8("rx_size_auto_shrink", 0600, dentry, - &i2400mu->rx_size_auto_shrink); - - debugfs_create_size_t("rx_size", 0600, dentry, &i2400mu->rx_size); -} - - -static struct device_type i2400mu_type = { - .name = "wimax", -}; - -/* - * Probe a i2400m interface and register it - * - * @iface: USB interface to link to - * @id: USB class/subclass/protocol id - * @returns: 0 if ok, < 0 errno code on error. - * - * Alloc a net device, initialize the bus-specific details and then - * calls the bus-generic initialization routine. That will register - * the wimax and netdev devices, upload the firmware [using - * _bus_bm_*()], call _bus_dev_start() to finalize the setup of the - * communication with the device and then will start to talk to it to - * finnish setting it up. - */ -static -int i2400mu_probe(struct usb_interface *iface, - const struct usb_device_id *id) -{ - int result; - struct net_device *net_dev; - struct device *dev = &iface->dev; - struct i2400m *i2400m; - struct i2400mu *i2400mu; - struct usb_device *usb_dev = interface_to_usbdev(iface); - - if (iface->cur_altsetting->desc.bNumEndpoints < 4) - return -ENODEV; - - if (usb_dev->speed != USB_SPEED_HIGH) - dev_err(dev, "device not connected as high speed\n"); - - /* Allocate instance [calls i2400m_netdev_setup() on it]. */ - result = -ENOMEM; - net_dev = alloc_netdev(sizeof(*i2400mu), "wmx%d", NET_NAME_UNKNOWN, - i2400mu_netdev_setup); - if (net_dev == NULL) { - dev_err(dev, "no memory for network device instance\n"); - goto error_alloc_netdev; - } - SET_NETDEV_DEV(net_dev, dev); - SET_NETDEV_DEVTYPE(net_dev, &i2400mu_type); - i2400m = net_dev_to_i2400m(net_dev); - i2400mu = container_of(i2400m, struct i2400mu, i2400m); - i2400m->wimax_dev.net_dev = net_dev; - i2400mu->usb_dev = usb_get_dev(usb_dev); - i2400mu->usb_iface = iface; - usb_set_intfdata(iface, i2400mu); - - i2400m->bus_tx_block_size = I2400MU_BLK_SIZE; - /* - * Room required in the Tx queue for USB message to accommodate - * a smallest payload while allocating header space is 16 bytes. - * Adding this room for the new tx message increases the - * possibilities of including any payload with size <= 16 bytes. - */ - i2400m->bus_tx_room_min = I2400MU_BLK_SIZE; - i2400m->bus_pl_size_max = I2400MU_PL_SIZE_MAX; - i2400m->bus_setup = NULL; - i2400m->bus_dev_start = i2400mu_bus_dev_start; - i2400m->bus_dev_stop = i2400mu_bus_dev_stop; - i2400m->bus_release = NULL; - i2400m->bus_tx_kick = i2400mu_bus_tx_kick; - i2400m->bus_reset = i2400mu_bus_reset; - i2400m->bus_bm_retries = I2400M_USB_BOOT_RETRIES; - i2400m->bus_bm_cmd_send = i2400mu_bus_bm_cmd_send; - i2400m->bus_bm_wait_for_ack = i2400mu_bus_bm_wait_for_ack; - i2400m->bus_bm_mac_addr_impaired = 0; - - switch (id->idProduct) { - case USB_DEVICE_ID_I6050: - case USB_DEVICE_ID_I6050_2: - case USB_DEVICE_ID_I6150: - case USB_DEVICE_ID_I6150_2: - case USB_DEVICE_ID_I6150_3: - case USB_DEVICE_ID_I6250: - i2400mu->i6050 = 1; - break; - default: - break; - } - - if (i2400mu->i6050) { - i2400m->bus_fw_names = i2400mu_bus_fw_names_6050; - i2400mu->endpoint_cfg.bulk_out = 0; - i2400mu->endpoint_cfg.notification = 3; - i2400mu->endpoint_cfg.reset_cold = 2; - i2400mu->endpoint_cfg.bulk_in = 1; - } else { - i2400m->bus_fw_names = i2400mu_bus_fw_names_5x50; - i2400mu->endpoint_cfg.bulk_out = 0; - i2400mu->endpoint_cfg.notification = 1; - i2400mu->endpoint_cfg.reset_cold = 2; - i2400mu->endpoint_cfg.bulk_in = 3; - } -#ifdef CONFIG_PM - iface->needs_remote_wakeup = 1; /* autosuspend (15s delay) */ - device_init_wakeup(dev, 1); - pm_runtime_set_autosuspend_delay(&usb_dev->dev, 15000); - usb_enable_autosuspend(usb_dev); -#endif - - result = i2400m_setup(i2400m, I2400M_BRI_MAC_REINIT); - if (result < 0) { - dev_err(dev, "cannot setup device: %d\n", result); - goto error_setup; - } - i2400mu_debugfs_add(i2400mu); - return 0; - -error_setup: - usb_set_intfdata(iface, NULL); - usb_put_dev(i2400mu->usb_dev); - free_netdev(net_dev); -error_alloc_netdev: - return result; -} - - -/* - * Disconnect a i2400m from the system. - * - * i2400m_stop() has been called before, so al the rx and tx contexts - * have been taken down already. Make sure the queue is stopped, - * unregister netdev and i2400m, free and kill. - */ -static -void i2400mu_disconnect(struct usb_interface *iface) -{ - struct i2400mu *i2400mu = usb_get_intfdata(iface); - struct i2400m *i2400m = &i2400mu->i2400m; - struct net_device *net_dev = i2400m->wimax_dev.net_dev; - struct device *dev = &iface->dev; - - d_fnstart(3, dev, "(iface %p i2400m %p)\n", iface, i2400m); - - debugfs_remove_recursive(i2400mu->debugfs_dentry); - i2400m_release(i2400m); - usb_set_intfdata(iface, NULL); - usb_put_dev(i2400mu->usb_dev); - free_netdev(net_dev); - d_fnend(3, dev, "(iface %p i2400m %p) = void\n", iface, i2400m); -} - - -/* - * Get the device ready for USB port or system standby and hibernation - * - * USB port and system standby are handled the same. - * - * When the system hibernates, the USB device is powered down and then - * up, so we don't really have to do much here, as it will be seen as - * a reconnect. Still for simplicity we consider this case the same as - * suspend, so that the device has a chance to do notify the base - * station (if connected). - * - * So at the end, the three cases require common handling. - * - * If at the time of this call the device's firmware is not loaded, - * nothing has to be done. Note we can be "loose" about not reading - * i2400m->updown under i2400m->init_mutex. If it happens to change - * inmediately, other parts of the call flow will fail and effectively - * catch it. - * - * If the firmware is loaded, we need to: - * - * - tell the device to go into host interface power save mode, wait - * for it to ack - * - * This is quite more interesting than it is; we need to execute a - * command, but this time, we don't want the code in usb-{tx,rx}.c - * to call the usb_autopm_get/put_interface() barriers as it'd - * deadlock, so we need to decrement i2400mu->do_autopm, that acts - * as a poor man's semaphore. Ugly, but it works. - * - * As well, the device might refuse going to sleep for whichever - * reason. In this case we just fail. For system suspend/hibernate, - * we *can't* fail. We check PMSG_IS_AUTO to see if the - * suspend call comes from the USB stack or from the system and act - * in consequence. - * - * - stop the notification endpoint polling - */ -static -int i2400mu_suspend(struct usb_interface *iface, pm_message_t pm_msg) -{ - int result = 0; - struct device *dev = &iface->dev; - struct i2400mu *i2400mu = usb_get_intfdata(iface); - unsigned is_autosuspend = 0; - struct i2400m *i2400m = &i2400mu->i2400m; - -#ifdef CONFIG_PM - if (PMSG_IS_AUTO(pm_msg)) - is_autosuspend = 1; -#endif - - d_fnstart(3, dev, "(iface %p pm_msg %u)\n", iface, pm_msg.event); - rmb(); /* see i2400m->updown's documentation */ - if (i2400m->updown == 0) - goto no_firmware; - if (i2400m->state == I2400M_SS_DATA_PATH_CONNECTED && is_autosuspend) { - /* ugh -- the device is connected and this suspend - * request is an autosuspend one (not a system standby - * / hibernate). - * - * The only way the device can go to standby is if the - * link with the base station is in IDLE mode; that - * were the case, we'd be in status - * I2400M_SS_CONNECTED_IDLE. But we are not. - * - * If we *tell* him to go power save now, it'll reset - * as a precautionary measure, so if this is an - * autosuspend thing, say no and it'll come back - * later, when the link is IDLE - */ - result = -EBADF; - d_printf(1, dev, "fw up, link up, not-idle, autosuspend: " - "not entering powersave\n"); - goto error_not_now; - } - d_printf(1, dev, "fw up: entering powersave\n"); - atomic_dec(&i2400mu->do_autopm); - result = i2400m_cmd_enter_powersave(i2400m); - atomic_inc(&i2400mu->do_autopm); - if (result < 0 && !is_autosuspend) { - /* System suspend, can't fail */ - dev_err(dev, "failed to suspend, will reset on resume\n"); - result = 0; - } - if (result < 0) - goto error_enter_powersave; - i2400mu_notification_release(i2400mu); - d_printf(1, dev, "powersave requested\n"); -error_enter_powersave: -error_not_now: -no_firmware: - d_fnend(3, dev, "(iface %p pm_msg %u) = %d\n", - iface, pm_msg.event, result); - return result; -} - - -static -int i2400mu_resume(struct usb_interface *iface) -{ - int ret = 0; - struct device *dev = &iface->dev; - struct i2400mu *i2400mu = usb_get_intfdata(iface); - struct i2400m *i2400m = &i2400mu->i2400m; - - d_fnstart(3, dev, "(iface %p)\n", iface); - rmb(); /* see i2400m->updown's documentation */ - if (i2400m->updown == 0) { - d_printf(1, dev, "fw was down, no resume needed\n"); - goto out; - } - d_printf(1, dev, "fw was up, resuming\n"); - i2400mu_notification_setup(i2400mu); - /* USB has flow control, so we don't need to give it time to - * come back; otherwise, we'd use something like a get-state - * command... */ -out: - d_fnend(3, dev, "(iface %p) = %d\n", iface, ret); - return ret; -} - - -static -int i2400mu_reset_resume(struct usb_interface *iface) -{ - int result; - struct device *dev = &iface->dev; - struct i2400mu *i2400mu = usb_get_intfdata(iface); - struct i2400m *i2400m = &i2400mu->i2400m; - - d_fnstart(3, dev, "(iface %p)\n", iface); - result = i2400m_dev_reset_handle(i2400m, "device reset on resume"); - d_fnend(3, dev, "(iface %p) = %d\n", iface, result); - return result < 0 ? result : 0; -} - - -/* - * Another driver or user space is triggering a reset on the device - * which contains the interface passed as an argument. Cease IO and - * save any device state you need to restore. - * - * If you need to allocate memory here, use GFP_NOIO or GFP_ATOMIC, if - * you are in atomic context. - */ -static -int i2400mu_pre_reset(struct usb_interface *iface) -{ - struct i2400mu *i2400mu = usb_get_intfdata(iface); - return i2400m_pre_reset(&i2400mu->i2400m); -} - - -/* - * The reset has completed. Restore any saved device state and begin - * using the device again. - * - * If you need to allocate memory here, use GFP_NOIO or GFP_ATOMIC, if - * you are in atomic context. - */ -static -int i2400mu_post_reset(struct usb_interface *iface) -{ - struct i2400mu *i2400mu = usb_get_intfdata(iface); - return i2400m_post_reset(&i2400mu->i2400m); -} - - -static -struct usb_device_id i2400mu_id_table[] = { - { USB_DEVICE(0x8086, USB_DEVICE_ID_I6050) }, - { USB_DEVICE(0x8086, USB_DEVICE_ID_I6050_2) }, - { USB_DEVICE(0x8087, USB_DEVICE_ID_I6150) }, - { USB_DEVICE(0x8087, USB_DEVICE_ID_I6150_2) }, - { USB_DEVICE(0x8087, USB_DEVICE_ID_I6150_3) }, - { USB_DEVICE(0x8086, USB_DEVICE_ID_I6250) }, - { USB_DEVICE(0x8086, 0x0181) }, - { USB_DEVICE(0x8086, 0x1403) }, - { USB_DEVICE(0x8086, 0x1405) }, - { USB_DEVICE(0x8086, 0x0180) }, - { USB_DEVICE(0x8086, 0x0182) }, - { USB_DEVICE(0x8086, 0x1406) }, - { USB_DEVICE(0x8086, 0x1403) }, - { }, -}; -MODULE_DEVICE_TABLE(usb, i2400mu_id_table); - - -static -struct usb_driver i2400mu_driver = { - .name = KBUILD_MODNAME, - .suspend = i2400mu_suspend, - .resume = i2400mu_resume, - .reset_resume = i2400mu_reset_resume, - .probe = i2400mu_probe, - .disconnect = i2400mu_disconnect, - .pre_reset = i2400mu_pre_reset, - .post_reset = i2400mu_post_reset, - .id_table = i2400mu_id_table, - .supports_autosuspend = 1, -}; - -static -int __init i2400mu_driver_init(void) -{ - d_parse_params(D_LEVEL, D_LEVEL_SIZE, i2400mu_debug_params, - "i2400m_usb.debug"); - return usb_register(&i2400mu_driver); -} -module_init(i2400mu_driver_init); - - -static -void __exit i2400mu_driver_exit(void) -{ - usb_deregister(&i2400mu_driver); -} -module_exit(i2400mu_driver_exit); - -MODULE_AUTHOR("Intel Corporation "); -MODULE_DESCRIPTION("Driver for USB based Intel Wireless WiMAX Connection 2400M " - "(5x50 & 6050)"); -MODULE_LICENSE("GPL"); -MODULE_FIRMWARE(I2400MU_FW_FILE_NAME_v1_5); -MODULE_FIRMWARE(I6050U_FW_FILE_NAME_v1_5); diff --git a/drivers/staging/Kconfig b/drivers/staging/Kconfig index 2d0310448eba..443ca3f3cdf0 100644 --- a/drivers/staging/Kconfig +++ b/drivers/staging/Kconfig @@ -114,6 +114,8 @@ source "drivers/staging/kpc2000/Kconfig" source "drivers/staging/qlge/Kconfig" +source "drivers/staging/wimax/Kconfig" + source "drivers/staging/wfx/Kconfig" source "drivers/staging/hikey9xx/Kconfig" diff --git a/drivers/staging/Makefile b/drivers/staging/Makefile index 757a892ab5b9..dc45128ef525 100644 --- a/drivers/staging/Makefile +++ b/drivers/staging/Makefile @@ -47,5 +47,6 @@ obj-$(CONFIG_XIL_AXIS_FIFO) += axis-fifo/ obj-$(CONFIG_FIELDBUS_DEV) += fieldbus/ obj-$(CONFIG_KPC2000) += kpc2000/ obj-$(CONFIG_QLGE) += qlge/ +obj-$(CONFIG_WIMAX) += wimax/ obj-$(CONFIG_WFX) += wfx/ obj-y += hikey9xx/ diff --git a/drivers/staging/wimax/Documentation/i2400m.rst b/drivers/staging/wimax/Documentation/i2400m.rst new file mode 100644 index 000000000000..194388c0c351 --- /dev/null +++ b/drivers/staging/wimax/Documentation/i2400m.rst @@ -0,0 +1,283 @@ +.. include:: + +==================================================== +Driver for the Intel Wireless Wimax Connection 2400m +==================================================== + +:Copyright: |copy| 2008 Intel Corporation < linux-wimax@intel.com > + + This provides a driver for the Intel Wireless WiMAX Connection 2400m + and a basic Linux kernel WiMAX stack. + +1. Requirements +=============== + + * Linux installation with Linux kernel 2.6.22 or newer (if building + from a separate tree) + * Intel i2400m Echo Peak or Baxter Peak; this includes the Intel + Wireless WiMAX/WiFi Link 5x50 series. + * build tools: + + + Linux kernel development package for the target kernel; to + build against your currently running kernel, you need to have + the kernel development package corresponding to the running + image installed (usually if your kernel is named + linux-VERSION, the development package is called + linux-dev-VERSION or linux-headers-VERSION). + + GNU C Compiler, make + +2. Compilation and installation +=============================== + +2.1. Compilation of the drivers included in the kernel +------------------------------------------------------ + + Configure the kernel; to enable the WiMAX drivers select Drivers > + Networking Drivers > WiMAX device support. Enable all of them as + modules (easier). + + If USB or SDIO are not enabled in the kernel configuration, the options + to build the i2400m USB or SDIO drivers will not show. Enable said + subsystems and go back to the WiMAX menu to enable the drivers. + + Compile and install your kernel as usual. + +2.2. Compilation of the drivers distributed as an standalone module +------------------------------------------------------------------- + + To compile:: + + $ cd source/directory + $ make + + Once built you can load and unload using the provided load.sh script; + load.sh will load the modules, load.sh u will unload them. + + To install in the default kernel directories (and enable auto loading + when the device is plugged):: + + $ make install + $ depmod -a + + If your kernel development files are located in a non standard + directory or if you want to build for a kernel that is not the + currently running one, set KDIR to the right location:: + + $ make KDIR=/path/to/kernel/dev/tree + + For more information, please contact linux-wimax@intel.com. + +3. Installing the firmware +-------------------------- + + The firmware can be obtained from http://linuxwimax.org or might have + been supplied with your hardware. + + It has to be installed in the target system:: + + $ cp FIRMWAREFILE.sbcf /lib/firmware/i2400m-fw-BUSTYPE-1.3.sbcf + + * NOTE: if your firmware came in an .rpm or .deb file, just install + it as normal, with the rpm (rpm -i FIRMWARE.rpm) or dpkg + (dpkg -i FIRMWARE.deb) commands. No further action is needed. + * BUSTYPE will be usb or sdio, depending on the hardware you have. + Each hardware type comes with its own firmware and will not work + with other types. + +4. Design +========= + + This package contains two major parts: a WiMAX kernel stack and a + driver for the Intel i2400m. + + The WiMAX stack is designed to provide for common WiMAX control + services to current and future WiMAX devices from any vendor; please + see README.wimax for details. + + The i2400m kernel driver is broken up in two main parts: the bus + generic driver and the bus-specific drivers. The bus generic driver + forms the drivercore and contain no knowledge of the actual method we + use to connect to the device. The bus specific drivers are just the + glue to connect the bus-generic driver and the device. Currently only + USB and SDIO are supported. See drivers/net/wimax/i2400m/i2400m.h for + more information. + + The bus generic driver is logically broken up in two parts: OS-glue and + hardware-glue. The OS-glue interfaces with Linux. The hardware-glue + interfaces with the device on using an interface provided by the + bus-specific driver. The reason for this breakup is to be able to + easily reuse the hardware-glue to write drivers for other OSes; note + the hardware glue part is written as a native Linux driver; no + abstraction layers are used, so to port to another OS, the Linux kernel + API calls should be replaced with the target OS's. + +5. Usage +======== + + To load the driver, follow the instructions in the install section; + once the driver is loaded, plug in the device (unless it is permanently + plugged in). The driver will enumerate the device, upload the firmware + and output messages in the kernel log (dmesg, /var/log/messages or + /var/log/kern.log) such as:: + + ... + i2400m_usb 5-4:1.0: firmware interface version 8.0.0 + i2400m_usb 5-4:1.0: WiMAX interface wmx0 (00:1d:e1:01:94:2c) ready + + At this point the device is ready to work. + + Current versions require the Intel WiMAX Network Service in userspace + to make things work. See the network service's README for instructions + on how to scan, connect and disconnect. + +5.1. Module parameters +---------------------- + + Module parameters can be set at kernel or module load time or by + echoing values:: + + $ echo VALUE > /sys/module/MODULENAME/parameters/PARAMETERNAME + + To make changes permanent, for example, for the i2400m module, you can + also create a file named /etc/modprobe.d/i2400m containing:: + + options i2400m idle_mode_disabled=1 + + To find which parameters are supported by a module, run:: + + $ modinfo path/to/module.ko + + During kernel bootup (if the driver is linked in the kernel), specify + the following to the kernel command line:: + + i2400m.PARAMETER=VALUE + +5.1.1. i2400m: idle_mode_disabled +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + The i2400m module supports a parameter to disable idle mode. This + parameter, once set, will take effect only when the device is + reinitialized by the driver (eg: following a reset or a reconnect). + +5.2. Debug operations: debugfs entries +-------------------------------------- + + The driver will register debugfs entries that allow the user to tweak + debug settings. There are three main container directories where + entries are placed, which correspond to the three blocks a i2400m WiMAX + driver has: + + * /sys/kernel/debug/wimax:DEVNAME/ for the generic WiMAX stack + controls + * /sys/kernel/debug/wimax:DEVNAME/i2400m for the i2400m generic + driver controls + * /sys/kernel/debug/wimax:DEVNAME/i2400m-usb (or -sdio) for the + bus-specific i2400m-usb or i2400m-sdio controls). + + Of course, if debugfs is mounted in a directory other than + /sys/kernel/debug, those paths will change. + +5.2.1. Increasing debug output +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + The files named *dl_* indicate knobs for controlling the debug output + of different submodules:: + + # find /sys/kernel/debug/wimax\:wmx0 -name \*dl_\* + /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_tx + /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_rx + /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_notif + /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_fw + /sys/kernel/debug/wimax:wmx0/i2400m-usb/dl_usb + /sys/kernel/debug/wimax:wmx0/i2400m/dl_tx + /sys/kernel/debug/wimax:wmx0/i2400m/dl_rx + /sys/kernel/debug/wimax:wmx0/i2400m/dl_rfkill + /sys/kernel/debug/wimax:wmx0/i2400m/dl_netdev + /sys/kernel/debug/wimax:wmx0/i2400m/dl_fw + /sys/kernel/debug/wimax:wmx0/i2400m/dl_debugfs + /sys/kernel/debug/wimax:wmx0/i2400m/dl_driver + /sys/kernel/debug/wimax:wmx0/i2400m/dl_control + /sys/kernel/debug/wimax:wmx0/wimax_dl_stack + /sys/kernel/debug/wimax:wmx0/wimax_dl_op_rfkill + /sys/kernel/debug/wimax:wmx0/wimax_dl_op_reset + /sys/kernel/debug/wimax:wmx0/wimax_dl_op_msg + /sys/kernel/debug/wimax:wmx0/wimax_dl_id_table + /sys/kernel/debug/wimax:wmx0/wimax_dl_debugfs + + By reading the file you can obtain the current value of said debug + level; by writing to it, you can set it. + + To increase the debug level of, for example, the i2400m's generic TX + engine, just write:: + + $ echo 3 > /sys/kernel/debug/wimax:wmx0/i2400m/dl_tx + + Increasing numbers yield increasing debug information; for details of + what is printed and the available levels, check the source. The code + uses 0 for disabled and increasing values until 8. + +5.2.2. RX and TX statistics +^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + The i2400m/rx_stats and i2400m/tx_stats provide statistics about the + data reception/delivery from the device:: + + $ cat /sys/kernel/debug/wimax:wmx0/i2400m/rx_stats + 45 1 3 34 3104 48 480 + + The numbers reported are: + + * packets/RX-buffer: total, min, max + * RX-buffers: total RX buffers received, accumulated RX buffer size + in bytes, min size received, max size received + + Thus, to find the average buffer size received, divide accumulated + RX-buffer / total RX-buffers. + + To clear the statistics back to 0, write anything to the rx_stats file:: + + $ echo 1 > /sys/kernel/debug/wimax:wmx0/i2400m_rx_stats + + Likewise for TX. + + Note the packets this debug file refers to are not network packet, but + packets in the sense of the device-specific protocol for communication + to the host. See drivers/net/wimax/i2400m/tx.c. + +5.2.3. Tracing messages received from user space +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + To echo messages received from user space into the trace pipe that the + i2400m driver creates, set the debug file i2400m/trace_msg_from_user to + 1:: + + $ echo 1 > /sys/kernel/debug/wimax:wmx0/i2400m/trace_msg_from_user + +5.2.4. Performing a device reset +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + By writing a 0, a 1 or a 2 to the file + /sys/kernel/debug/wimax:wmx0/reset, the driver performs a warm (without + disconnecting from the bus), cold (disconnecting from the bus) or bus + (bus specific) reset on the device. + +5.2.5. Asking the device to enter power saving mode +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + By writing any value to the /sys/kernel/debug/wimax:wmx0 file, the + device will attempt to enter power saving mode. + +6. Troubleshooting +================== + +6.1. Driver complains about ``i2400m-fw-usb-1.2.sbcf: request failed`` +---------------------------------------------------------------------- + + If upon connecting the device, the following is output in the kernel + log:: + + i2400m_usb 5-4:1.0: fw i2400m-fw-usb-1.3.sbcf: request failed: -2 + + This means that the driver cannot locate the firmware file named + /lib/firmware/i2400m-fw-usb-1.2.sbcf. Check that the file is present in + the right location. diff --git a/drivers/staging/wimax/Documentation/index.rst b/drivers/staging/wimax/Documentation/index.rst new file mode 100644 index 000000000000..fdf7c1f99ff5 --- /dev/null +++ b/drivers/staging/wimax/Documentation/index.rst @@ -0,0 +1,19 @@ +.. SPDX-License-Identifier: GPL-2.0 + +=============== +WiMAX subsystem +=============== + +.. toctree:: + :maxdepth: 2 + + wimax + + i2400m + +.. only:: subproject and html + + Indices + ======= + + * :ref:`genindex` diff --git a/drivers/staging/wimax/Documentation/wimax.rst b/drivers/staging/wimax/Documentation/wimax.rst new file mode 100644 index 000000000000..817ee8ba2732 --- /dev/null +++ b/drivers/staging/wimax/Documentation/wimax.rst @@ -0,0 +1,89 @@ +.. include:: + +======================== +Linux kernel WiMAX stack +======================== + +:Copyright: |copy| 2008 Intel Corporation < linux-wimax@intel.com > + + This provides a basic Linux kernel WiMAX stack to provide a common + control API for WiMAX devices, usable from kernel and user space. + +1. Design +========= + + The WiMAX stack is designed to provide for common WiMAX control + services to current and future WiMAX devices from any vendor. + + Because currently there is only one and we don't know what would be the + common services, the APIs it currently provides are very minimal. + However, it is done in such a way that it is easily extensible to + accommodate future requirements. + + The stack works by embedding a struct wimax_dev in your device's + control structures. This provides a set of callbacks that the WiMAX + stack will call in order to implement control operations requested by + the user. As well, the stack provides API functions that the driver + calls to notify about changes of state in the device. + + The stack exports the API calls needed to control the device to user + space using generic netlink as a marshalling mechanism. You can access + them using your own code or use the wrappers provided for your + convenience in libwimax (in the wimax-tools package). + + For detailed information on the stack, please see + include/linux/wimax.h. + +2. Usage +======== + + For usage in a driver (registration, API, etc) please refer to the + instructions in the header file include/linux/wimax.h. + + When a device is registered with the WiMAX stack, a set of debugfs + files will appear in /sys/kernel/debug/wimax:wmxX can tweak for + control. + +2.1. Obtaining debug information: debugfs entries +------------------------------------------------- + + The WiMAX stack is compiled, by default, with debug messages that can + be used to diagnose issues. By default, said messages are disabled. + + The drivers will register debugfs entries that allow the user to tweak + debug settings. + + Each driver, when registering with the stack, will cause a debugfs + directory named wimax:DEVICENAME to be created; optionally, it might + create more subentries below it. + +2.1.1. Increasing debug output +------------------------------ + + The files named *dl_* indicate knobs for controlling the debug output + of different submodules of the WiMAX stack:: + + # find /sys/kernel/debug/wimax\:wmx0 -name \*dl_\* + /sys/kernel/debug/wimax:wmx0/wimax_dl_stack + /sys/kernel/debug/wimax:wmx0/wimax_dl_op_rfkill + /sys/kernel/debug/wimax:wmx0/wimax_dl_op_reset + /sys/kernel/debug/wimax:wmx0/wimax_dl_op_msg + /sys/kernel/debug/wimax:wmx0/wimax_dl_id_table + /sys/kernel/debug/wimax:wmx0/wimax_dl_debugfs + /sys/kernel/debug/wimax:wmx0/.... # other driver specific files + + NOTE: + Of course, if debugfs is mounted in a directory other than + /sys/kernel/debug, those paths will change. + + By reading the file you can obtain the current value of said debug + level; by writing to it, you can set it. + + To increase the debug level of, for example, the id-table submodule, + just write: + + $ echo 3 > /sys/kernel/debug/wimax:wmx0/wimax_dl_id_table + + Increasing numbers yield increasing debug information; for details of + what is printed and the available levels, check the source. The code + uses 0 for disabled and increasing values until 8. diff --git a/drivers/staging/wimax/Kconfig b/drivers/staging/wimax/Kconfig new file mode 100644 index 000000000000..ded8b70b25ee --- /dev/null +++ b/drivers/staging/wimax/Kconfig @@ -0,0 +1,46 @@ +# SPDX-License-Identifier: GPL-2.0-only +# +# WiMAX LAN device configuration +# + +menuconfig WIMAX + tristate "WiMAX Wireless Broadband support" + depends on RFKILL || !RFKILL + help + + Select to configure support for devices that provide + wireless broadband connectivity using the WiMAX protocol + (IEEE 802.16). + + Please note that most of these devices require signing up + for a service plan with a provider. + + The different WiMAX drivers can be enabled in the menu entry + + Device Drivers > Network device support > WiMAX Wireless + Broadband devices + + If unsure, it is safe to select M (module). + +if WIMAX + +config WIMAX_DEBUG_LEVEL + int "WiMAX debug level" + depends on WIMAX + default 8 + help + + Select the maximum debug verbosity level to be compiled into + the WiMAX stack code. + + By default, debug messages are disabled at runtime and can + be selectively enabled for different parts of the code using + the sysfs debug-levels file. + + If set at zero, this will compile out all the debug code. + + It is recommended that it is left at 8. + +source "drivers/staging/wimax/i2400m/Kconfig" + +endif diff --git a/drivers/staging/wimax/Makefile b/drivers/staging/wimax/Makefile new file mode 100644 index 000000000000..0e3f988656aa --- /dev/null +++ b/drivers/staging/wimax/Makefile @@ -0,0 +1,15 @@ +# SPDX-License-Identifier: GPL-2.0 + +obj-$(CONFIG_WIMAX) += wimax.o + +wimax-y := \ + id-table.o \ + op-msg.o \ + op-reset.o \ + op-rfkill.o \ + op-state-get.o \ + stack.o + +wimax-$(CONFIG_DEBUG_FS) += debugfs.o + +obj-$(CONFIG_WIMAX_I2400M) += i2400m/ diff --git a/drivers/staging/wimax/TODO b/drivers/staging/wimax/TODO new file mode 100644 index 000000000000..26e4cb9e9599 --- /dev/null +++ b/drivers/staging/wimax/TODO @@ -0,0 +1,18 @@ +There are no known users of this driver as of October 2020, and it will +be removed unless someone turns out to still need it in future releases. + +According to https://en.wikipedia.org/wiki/List_of_WiMAX_networks, there +have been many public wimax networks, but it appears that many of these +have migrated to LTE or discontinued their service altogether. As most +PCs and phones lack WiMAX hardware support, the remaining networks tend +to use standalone routers. These almost certainly run Linux, but not a +modern kernel or the mainline wimax driver stack. + +NetworkManager appears to have dropped userspace support in 2015 +https://bugzilla.gnome.org/show_bug.cgi?id=747846, the www.linuxwimax.org +site had already shut down earlier. + +WiMax is apparently still being deployed on airport campus networks +("AeroMACS"), but in a frequency band that was not supported by the old +Intel 2400m (used in Sandy Bridge laptops and earlier), which is the +only driver using the kernel's wimax stack. diff --git a/drivers/staging/wimax/debug-levels.h b/drivers/staging/wimax/debug-levels.h new file mode 100644 index 000000000000..b854802d1d00 --- /dev/null +++ b/drivers/staging/wimax/debug-levels.h @@ -0,0 +1,29 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Linux WiMAX Stack + * Debug levels control file for the wimax module + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + */ +#ifndef __debug_levels__h__ +#define __debug_levels__h__ + +/* Maximum compile and run time debug level for all submodules */ +#define D_MODULENAME wimax +#define D_MASTER CONFIG_WIMAX_DEBUG_LEVEL + +#include "linux-wimax-debug.h" + +/* List of all the enabled modules */ +enum d_module { + D_SUBMODULE_DECLARE(debugfs), + D_SUBMODULE_DECLARE(id_table), + D_SUBMODULE_DECLARE(op_msg), + D_SUBMODULE_DECLARE(op_reset), + D_SUBMODULE_DECLARE(op_rfkill), + D_SUBMODULE_DECLARE(op_state_get), + D_SUBMODULE_DECLARE(stack), +}; + +#endif /* #ifndef __debug_levels__h__ */ diff --git a/drivers/staging/wimax/debugfs.c b/drivers/staging/wimax/debugfs.c new file mode 100644 index 000000000000..e11bff61ffcf --- /dev/null +++ b/drivers/staging/wimax/debugfs.c @@ -0,0 +1,38 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * Debugfs support + * + * Copyright (C) 2005-2006 Intel Corporation + * Inaky Perez-Gonzalez + */ +#include +#include "linux-wimax.h" +#include "wimax-internal.h" + +#define D_SUBMODULE debugfs +#include "debug-levels.h" + +void wimax_debugfs_add(struct wimax_dev *wimax_dev) +{ + struct net_device *net_dev = wimax_dev->net_dev; + struct dentry *dentry; + char buf[128]; + + snprintf(buf, sizeof(buf), "wimax:%s", net_dev->name); + dentry = debugfs_create_dir(buf, NULL); + wimax_dev->debugfs_dentry = dentry; + + d_level_register_debugfs("wimax_dl_", debugfs, dentry); + d_level_register_debugfs("wimax_dl_", id_table, dentry); + d_level_register_debugfs("wimax_dl_", op_msg, dentry); + d_level_register_debugfs("wimax_dl_", op_reset, dentry); + d_level_register_debugfs("wimax_dl_", op_rfkill, dentry); + d_level_register_debugfs("wimax_dl_", op_state_get, dentry); + d_level_register_debugfs("wimax_dl_", stack, dentry); +} + +void wimax_debugfs_rm(struct wimax_dev *wimax_dev) +{ + debugfs_remove_recursive(wimax_dev->debugfs_dentry); +} diff --git a/drivers/staging/wimax/i2400m/Kconfig b/drivers/staging/wimax/i2400m/Kconfig new file mode 100644 index 000000000000..843b905a26a3 --- /dev/null +++ b/drivers/staging/wimax/i2400m/Kconfig @@ -0,0 +1,37 @@ +# SPDX-License-Identifier: GPL-2.0-only + +config WIMAX_I2400M + tristate + depends on WIMAX + select FW_LOADER + +comment "Enable USB support to see WiMAX USB drivers" + depends on USB = n + +config WIMAX_I2400M_USB + tristate "Intel Wireless WiMAX Connection 2400 over USB (including 5x50)" + depends on WIMAX && USB + select WIMAX_I2400M + help + Select if you have a device based on the Intel WiMAX + Connection 2400 over USB (like any of the Intel Wireless + WiMAX/WiFi Link 5x50 series). + + If unsure, it is safe to select M (module). + +config WIMAX_I2400M_DEBUG_LEVEL + int "WiMAX i2400m debug level" + depends on WIMAX_I2400M + default 8 + help + + Select the maximum debug verbosity level to be compiled into + the WiMAX i2400m driver code. + + By default, this is disabled at runtime and can be + selectively enabled at runtime for different parts of the + code using the sysfs debug-levels file. + + If set at zero, this will compile out all the debug code. + + It is recommended that it is left at 8. diff --git a/drivers/staging/wimax/i2400m/Makefile b/drivers/staging/wimax/i2400m/Makefile new file mode 100644 index 000000000000..b1db1eff0648 --- /dev/null +++ b/drivers/staging/wimax/i2400m/Makefile @@ -0,0 +1,23 @@ +# SPDX-License-Identifier: GPL-2.0 + +obj-$(CONFIG_WIMAX_I2400M) += i2400m.o +obj-$(CONFIG_WIMAX_I2400M_USB) += i2400m-usb.o + +i2400m-y := \ + control.o \ + driver.o \ + fw.o \ + op-rfkill.o \ + sysfs.o \ + netdev.o \ + tx.o \ + rx.o + +i2400m-$(CONFIG_DEBUG_FS) += debugfs.o + +i2400m-usb-y := \ + usb-fw.o \ + usb-notif.o \ + usb-tx.o \ + usb-rx.o \ + usb.o diff --git a/drivers/staging/wimax/i2400m/control.c b/drivers/staging/wimax/i2400m/control.c new file mode 100644 index 000000000000..fe885aa56cf3 --- /dev/null +++ b/drivers/staging/wimax/i2400m/control.c @@ -0,0 +1,1434 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * Miscellaneous control functions for managing the device + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Inaky Perez-Gonzalez + * - Initial implementation + * + * This is a collection of functions used to control the device (plus + * a few helpers). + * + * There are utilities for handling TLV buffers, hooks on the device's + * reports to act on device changes of state [i2400m_report_hook()], + * on acks to commands [i2400m_msg_ack_hook()], a helper for sending + * commands to the device and blocking until a reply arrives + * [i2400m_msg_to_dev()], a few high level commands for manipulating + * the device state, powersving mode and configuration plus the + * routines to setup the device once communication is stablished with + * it [i2400m_dev_initialize()]. + * + * ROADMAP + * + * i2400m_dev_initialize() Called by i2400m_dev_start() + * i2400m_set_init_config() + * i2400m_cmd_get_state() + * i2400m_dev_shutdown() Called by i2400m_dev_stop() + * i2400m_reset() + * + * i2400m_{cmd,get,set}_*() + * i2400m_msg_to_dev() + * i2400m_msg_check_status() + * + * i2400m_report_hook() Called on reception of an event + * i2400m_report_state_hook() + * i2400m_tlv_buffer_walk() + * i2400m_tlv_match() + * i2400m_report_tlv_system_state() + * i2400m_report_tlv_rf_switches_status() + * i2400m_report_tlv_media_status() + * i2400m_cmd_enter_powersave() + * + * i2400m_msg_ack_hook() Called on reception of a reply to a + * command, get or set + */ + +#include +#include "i2400m.h" +#include +#include +#include "linux-wimax-i2400m.h" +#include +#include + + +#define D_SUBMODULE control +#include "debug-levels.h" + +static int i2400m_idle_mode_disabled;/* 0 (idle mode enabled) by default */ +module_param_named(idle_mode_disabled, i2400m_idle_mode_disabled, int, 0644); +MODULE_PARM_DESC(idle_mode_disabled, + "If true, the device will not enable idle mode negotiation " + "with the base station (when connected) to save power."); + +/* 0 (power saving enabled) by default */ +static int i2400m_power_save_disabled; +module_param_named(power_save_disabled, i2400m_power_save_disabled, int, 0644); +MODULE_PARM_DESC(power_save_disabled, + "If true, the driver will not tell the device to enter " + "power saving mode when it reports it is ready for it. " + "False by default (so the device is told to do power " + "saving)."); + +static int i2400m_passive_mode; /* 0 (passive mode disabled) by default */ +module_param_named(passive_mode, i2400m_passive_mode, int, 0644); +MODULE_PARM_DESC(passive_mode, + "If true, the driver will not do any device setup " + "and leave it up to user space, who must be properly " + "setup."); + + +/* + * Return if a TLV is of a give type and size + * + * @tlv_hdr: pointer to the TLV + * @tlv_type: type of the TLV we are looking for + * @tlv_size: expected size of the TLV we are looking for (if -1, + * don't check the size). This includes the header + * Returns: 0 if the TLV matches + * < 0 if it doesn't match at all + * > 0 total TLV + payload size, if the type matches, but not + * the size + */ +static +ssize_t i2400m_tlv_match(const struct i2400m_tlv_hdr *tlv, + enum i2400m_tlv tlv_type, ssize_t tlv_size) +{ + if (le16_to_cpu(tlv->type) != tlv_type) /* Not our type? skip */ + return -1; + if (tlv_size != -1 + && le16_to_cpu(tlv->length) + sizeof(*tlv) != tlv_size) { + size_t size = le16_to_cpu(tlv->length) + sizeof(*tlv); + printk(KERN_WARNING "W: tlv type 0x%x mismatched because of " + "size (got %zu vs %zd expected)\n", + tlv_type, size, tlv_size); + return size; + } + return 0; +} + + +/* + * Given a buffer of TLVs, iterate over them + * + * @i2400m: device instance + * @tlv_buf: pointer to the beginning of the TLV buffer + * @buf_size: buffer size in bytes + * @tlv_pos: seek position; this is assumed to be a pointer returned + * by i2400m_tlv_buffer_walk() [and thus, validated]. The + * TLV returned will be the one following this one. + * + * Usage: + * + * tlv_itr = NULL; + * while (tlv_itr = i2400m_tlv_buffer_walk(i2400m, buf, size, tlv_itr)) { + * ... + * // Do stuff with tlv_itr, DON'T MODIFY IT + * ... + * } + */ +static +const struct i2400m_tlv_hdr *i2400m_tlv_buffer_walk( + struct i2400m *i2400m, + const void *tlv_buf, size_t buf_size, + const struct i2400m_tlv_hdr *tlv_pos) +{ + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_tlv_hdr *tlv_top = tlv_buf + buf_size; + size_t offset, length, avail_size; + unsigned type; + + if (tlv_pos == NULL) /* Take the first one? */ + tlv_pos = tlv_buf; + else /* Nope, the next one */ + tlv_pos = (void *) tlv_pos + + le16_to_cpu(tlv_pos->length) + sizeof(*tlv_pos); + if (tlv_pos == tlv_top) { /* buffer done */ + tlv_pos = NULL; + goto error_beyond_end; + } + if (tlv_pos > tlv_top) { + tlv_pos = NULL; + WARN_ON(1); + goto error_beyond_end; + } + offset = (void *) tlv_pos - (void *) tlv_buf; + avail_size = buf_size - offset; + if (avail_size < sizeof(*tlv_pos)) { + dev_err(dev, "HW BUG? tlv_buf %p [%zu bytes], tlv @%zu: " + "short header\n", tlv_buf, buf_size, offset); + goto error_short_header; + } + type = le16_to_cpu(tlv_pos->type); + length = le16_to_cpu(tlv_pos->length); + if (avail_size < sizeof(*tlv_pos) + length) { + dev_err(dev, "HW BUG? tlv_buf %p [%zu bytes], " + "tlv type 0x%04x @%zu: " + "short data (%zu bytes vs %zu needed)\n", + tlv_buf, buf_size, type, offset, avail_size, + sizeof(*tlv_pos) + length); + goto error_short_header; + } +error_short_header: +error_beyond_end: + return tlv_pos; +} + + +/* + * Find a TLV in a buffer of sequential TLVs + * + * @i2400m: device descriptor + * @tlv_hdr: pointer to the first TLV in the sequence + * @size: size of the buffer in bytes; all TLVs are assumed to fit + * fully in the buffer (otherwise we'll complain). + * @tlv_type: type of the TLV we are looking for + * @tlv_size: expected size of the TLV we are looking for (if -1, + * don't check the size). This includes the header + * + * Returns: NULL if the TLV is not found, otherwise a pointer to + * it. If the sizes don't match, an error is printed and NULL + * returned. + */ +static +const struct i2400m_tlv_hdr *i2400m_tlv_find( + struct i2400m *i2400m, + const struct i2400m_tlv_hdr *tlv_hdr, size_t size, + enum i2400m_tlv tlv_type, ssize_t tlv_size) +{ + ssize_t match; + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_tlv_hdr *tlv = NULL; + while ((tlv = i2400m_tlv_buffer_walk(i2400m, tlv_hdr, size, tlv))) { + match = i2400m_tlv_match(tlv, tlv_type, tlv_size); + if (match == 0) /* found it :) */ + break; + if (match > 0) + dev_warn(dev, "TLV type 0x%04x found with size " + "mismatch (%zu vs %zd needed)\n", + tlv_type, match, tlv_size); + } + return tlv; +} + + +static const struct +{ + char *msg; + int errno; +} ms_to_errno[I2400M_MS_MAX] = { + [I2400M_MS_DONE_OK] = { "", 0 }, + [I2400M_MS_DONE_IN_PROGRESS] = { "", 0 }, + [I2400M_MS_INVALID_OP] = { "invalid opcode", -ENOSYS }, + [I2400M_MS_BAD_STATE] = { "invalid state", -EILSEQ }, + [I2400M_MS_ILLEGAL_VALUE] = { "illegal value", -EINVAL }, + [I2400M_MS_MISSING_PARAMS] = { "missing parameters", -ENOMSG }, + [I2400M_MS_VERSION_ERROR] = { "bad version", -EIO }, + [I2400M_MS_ACCESSIBILITY_ERROR] = { "accesibility error", -EIO }, + [I2400M_MS_BUSY] = { "busy", -EBUSY }, + [I2400M_MS_CORRUPTED_TLV] = { "corrupted TLV", -EILSEQ }, + [I2400M_MS_UNINITIALIZED] = { "uninitialized", -EILSEQ }, + [I2400M_MS_UNKNOWN_ERROR] = { "unknown error", -EIO }, + [I2400M_MS_PRODUCTION_ERROR] = { "production error", -EIO }, + [I2400M_MS_NO_RF] = { "no RF", -EIO }, + [I2400M_MS_NOT_READY_FOR_POWERSAVE] = + { "not ready for powersave", -EACCES }, + [I2400M_MS_THERMAL_CRITICAL] = { "thermal critical", -EL3HLT }, +}; + + +/* + * i2400m_msg_check_status - translate a message's status code + * + * @i2400m: device descriptor + * @l3l4_hdr: message header + * @strbuf: buffer to place a formatted error message (unless NULL). + * @strbuf_size: max amount of available space; larger messages will + * be truncated. + * + * Returns: errno code corresponding to the status code in @l3l4_hdr + * and a message in @strbuf describing the error. + */ +int i2400m_msg_check_status(const struct i2400m_l3l4_hdr *l3l4_hdr, + char *strbuf, size_t strbuf_size) +{ + int result; + enum i2400m_ms status = le16_to_cpu(l3l4_hdr->status); + const char *str; + + if (status == 0) + return 0; + if (status >= ARRAY_SIZE(ms_to_errno)) { + str = "unknown status code"; + result = -EBADR; + } else { + str = ms_to_errno[status].msg; + result = ms_to_errno[status].errno; + } + if (strbuf) + snprintf(strbuf, strbuf_size, "%s (%d)", str, status); + return result; +} + + +/* + * Act on a TLV System State reported by the device + * + * @i2400m: device descriptor + * @ss: validated System State TLV + */ +static +void i2400m_report_tlv_system_state(struct i2400m *i2400m, + const struct i2400m_tlv_system_state *ss) +{ + struct device *dev = i2400m_dev(i2400m); + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + enum i2400m_system_state i2400m_state = le32_to_cpu(ss->state); + + d_fnstart(3, dev, "(i2400m %p ss %p [%u])\n", i2400m, ss, i2400m_state); + + if (i2400m->state != i2400m_state) { + i2400m->state = i2400m_state; + wake_up_all(&i2400m->state_wq); + } + switch (i2400m_state) { + case I2400M_SS_UNINITIALIZED: + case I2400M_SS_INIT: + case I2400M_SS_CONFIG: + case I2400M_SS_PRODUCTION: + wimax_state_change(wimax_dev, WIMAX_ST_UNINITIALIZED); + break; + + case I2400M_SS_RF_OFF: + case I2400M_SS_RF_SHUTDOWN: + wimax_state_change(wimax_dev, WIMAX_ST_RADIO_OFF); + break; + + case I2400M_SS_READY: + case I2400M_SS_STANDBY: + case I2400M_SS_SLEEPACTIVE: + wimax_state_change(wimax_dev, WIMAX_ST_READY); + break; + + case I2400M_SS_CONNECTING: + case I2400M_SS_WIMAX_CONNECTED: + wimax_state_change(wimax_dev, WIMAX_ST_READY); + break; + + case I2400M_SS_SCAN: + case I2400M_SS_OUT_OF_ZONE: + wimax_state_change(wimax_dev, WIMAX_ST_SCANNING); + break; + + case I2400M_SS_IDLE: + d_printf(1, dev, "entering BS-negotiated idle mode\n"); + fallthrough; + case I2400M_SS_DISCONNECTING: + case I2400M_SS_DATA_PATH_CONNECTED: + wimax_state_change(wimax_dev, WIMAX_ST_CONNECTED); + break; + + default: + /* Huh? just in case, shut it down */ + dev_err(dev, "HW BUG? unknown state %u: shutting down\n", + i2400m_state); + i2400m_reset(i2400m, I2400M_RT_WARM); + break; + } + d_fnend(3, dev, "(i2400m %p ss %p [%u]) = void\n", + i2400m, ss, i2400m_state); +} + + +/* + * Parse and act on a TLV Media Status sent by the device + * + * @i2400m: device descriptor + * @ms: validated Media Status TLV + * + * This will set the carrier up on down based on the device's link + * report. This is done asides of what the WiMAX stack does based on + * the device's state as sometimes we need to do a link-renew (the BS + * wants us to renew a DHCP lease, for example). + * + * In fact, doc says that every time we get a link-up, we should do a + * DHCP negotiation... + */ +static +void i2400m_report_tlv_media_status(struct i2400m *i2400m, + const struct i2400m_tlv_media_status *ms) +{ + struct device *dev = i2400m_dev(i2400m); + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + struct net_device *net_dev = wimax_dev->net_dev; + enum i2400m_media_status status = le32_to_cpu(ms->media_status); + + d_fnstart(3, dev, "(i2400m %p ms %p [%u])\n", i2400m, ms, status); + + switch (status) { + case I2400M_MEDIA_STATUS_LINK_UP: + netif_carrier_on(net_dev); + break; + case I2400M_MEDIA_STATUS_LINK_DOWN: + netif_carrier_off(net_dev); + break; + /* + * This is the network telling us we need to retrain the DHCP + * lease -- so far, we are trusting the WiMAX Network Service + * in user space to pick this up and poke the DHCP client. + */ + case I2400M_MEDIA_STATUS_LINK_RENEW: + netif_carrier_on(net_dev); + break; + default: + dev_err(dev, "HW BUG? unknown media status %u\n", + status); + } + d_fnend(3, dev, "(i2400m %p ms %p [%u]) = void\n", + i2400m, ms, status); +} + + +/* + * Process a TLV from a 'state report' + * + * @i2400m: device descriptor + * @tlv: pointer to the TLV header; it has been already validated for + * consistent size. + * @tag: for error messages + * + * Act on the TLVs from a 'state report'. + */ +static +void i2400m_report_state_parse_tlv(struct i2400m *i2400m, + const struct i2400m_tlv_hdr *tlv, + const char *tag) +{ + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_tlv_media_status *ms; + const struct i2400m_tlv_system_state *ss; + const struct i2400m_tlv_rf_switches_status *rfss; + + if (0 == i2400m_tlv_match(tlv, I2400M_TLV_SYSTEM_STATE, sizeof(*ss))) { + ss = container_of(tlv, typeof(*ss), hdr); + d_printf(2, dev, "%s: system state TLV " + "found (0x%04x), state 0x%08x\n", + tag, I2400M_TLV_SYSTEM_STATE, + le32_to_cpu(ss->state)); + i2400m_report_tlv_system_state(i2400m, ss); + } + if (0 == i2400m_tlv_match(tlv, I2400M_TLV_RF_STATUS, sizeof(*rfss))) { + rfss = container_of(tlv, typeof(*rfss), hdr); + d_printf(2, dev, "%s: RF status TLV " + "found (0x%04x), sw 0x%02x hw 0x%02x\n", + tag, I2400M_TLV_RF_STATUS, + le32_to_cpu(rfss->sw_rf_switch), + le32_to_cpu(rfss->hw_rf_switch)); + i2400m_report_tlv_rf_switches_status(i2400m, rfss); + } + if (0 == i2400m_tlv_match(tlv, I2400M_TLV_MEDIA_STATUS, sizeof(*ms))) { + ms = container_of(tlv, typeof(*ms), hdr); + d_printf(2, dev, "%s: Media Status TLV: %u\n", + tag, le32_to_cpu(ms->media_status)); + i2400m_report_tlv_media_status(i2400m, ms); + } +} + + +/* + * Parse a 'state report' and extract information + * + * @i2400m: device descriptor + * @l3l4_hdr: pointer to message; it has been already validated for + * consistent size. + * @size: size of the message (header + payload). The header length + * declaration is assumed to be congruent with @size (as in + * sizeof(*l3l4_hdr) + l3l4_hdr->length == size) + * + * Walk over the TLVs in a report state and act on them. + */ +static +void i2400m_report_state_hook(struct i2400m *i2400m, + const struct i2400m_l3l4_hdr *l3l4_hdr, + size_t size, const char *tag) +{ + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_tlv_hdr *tlv; + size_t tlv_size = le16_to_cpu(l3l4_hdr->length); + + d_fnstart(4, dev, "(i2400m %p, l3l4_hdr %p, size %zu, %s)\n", + i2400m, l3l4_hdr, size, tag); + tlv = NULL; + + while ((tlv = i2400m_tlv_buffer_walk(i2400m, &l3l4_hdr->pl, + tlv_size, tlv))) + i2400m_report_state_parse_tlv(i2400m, tlv, tag); + d_fnend(4, dev, "(i2400m %p, l3l4_hdr %p, size %zu, %s) = void\n", + i2400m, l3l4_hdr, size, tag); +} + + +/* + * i2400m_report_hook - (maybe) act on a report + * + * @i2400m: device descriptor + * @l3l4_hdr: pointer to message; it has been already validated for + * consistent size. + * @size: size of the message (header + payload). The header length + * declaration is assumed to be congruent with @size (as in + * sizeof(*l3l4_hdr) + l3l4_hdr->length == size) + * + * Extract information we might need (like carrien on/off) from a + * device report. + */ +void i2400m_report_hook(struct i2400m *i2400m, + const struct i2400m_l3l4_hdr *l3l4_hdr, size_t size) +{ + struct device *dev = i2400m_dev(i2400m); + unsigned msg_type; + + d_fnstart(3, dev, "(i2400m %p l3l4_hdr %p size %zu)\n", + i2400m, l3l4_hdr, size); + /* Chew on the message, we might need some information from + * here */ + msg_type = le16_to_cpu(l3l4_hdr->type); + switch (msg_type) { + case I2400M_MT_REPORT_STATE: /* carrier detection... */ + i2400m_report_state_hook(i2400m, + l3l4_hdr, size, "REPORT STATE"); + break; + /* If the device is ready for power save, then ask it to do + * it. */ + case I2400M_MT_REPORT_POWERSAVE_READY: /* zzzzz */ + if (l3l4_hdr->status == cpu_to_le16(I2400M_MS_DONE_OK)) { + if (i2400m_power_save_disabled) + d_printf(1, dev, "ready for powersave, " + "not requesting (disabled by module " + "parameter)\n"); + else { + d_printf(1, dev, "ready for powersave, " + "requesting\n"); + i2400m_cmd_enter_powersave(i2400m); + } + } + break; + } + d_fnend(3, dev, "(i2400m %p l3l4_hdr %p size %zu) = void\n", + i2400m, l3l4_hdr, size); +} + + +/* + * i2400m_msg_ack_hook - process cmd/set/get ack for internal status + * + * @i2400m: device descriptor + * @l3l4_hdr: pointer to message; it has been already validated for + * consistent size. + * @size: size of the message + * + * Extract information we might need from acks to commands and act on + * it. This is akin to i2400m_report_hook(). Note most of this + * processing should be done in the function that calls the + * command. This is here for some cases where it can't happen... + */ +static void i2400m_msg_ack_hook(struct i2400m *i2400m, + const struct i2400m_l3l4_hdr *l3l4_hdr, + size_t size) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + unsigned int ack_type; + char strerr[32]; + + /* Chew on the message, we might need some information from + * here */ + ack_type = le16_to_cpu(l3l4_hdr->type); + switch (ack_type) { + case I2400M_MT_CMD_ENTER_POWERSAVE: + /* This is just left here for the sake of example, as + * the processing is done somewhere else. */ + if (0) { + result = i2400m_msg_check_status( + l3l4_hdr, strerr, sizeof(strerr)); + if (result >= 0) + d_printf(1, dev, "ready for power save: %zd\n", + size); + } + break; + } +} + + +/* + * i2400m_msg_size_check() - verify message size and header are congruent + * + * It is ok if the total message size is larger than the expected + * size, as there can be padding. + */ +int i2400m_msg_size_check(struct i2400m *i2400m, + const struct i2400m_l3l4_hdr *l3l4_hdr, + size_t msg_size) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + size_t expected_size; + d_fnstart(4, dev, "(i2400m %p l3l4_hdr %p msg_size %zu)\n", + i2400m, l3l4_hdr, msg_size); + if (msg_size < sizeof(*l3l4_hdr)) { + dev_err(dev, "bad size for message header " + "(expected at least %zu, got %zu)\n", + (size_t) sizeof(*l3l4_hdr), msg_size); + result = -EIO; + goto error_hdr_size; + } + expected_size = le16_to_cpu(l3l4_hdr->length) + sizeof(*l3l4_hdr); + if (msg_size < expected_size) { + dev_err(dev, "bad size for message code 0x%04x (expected %zu, " + "got %zu)\n", le16_to_cpu(l3l4_hdr->type), + expected_size, msg_size); + result = -EIO; + } else + result = 0; +error_hdr_size: + d_fnend(4, dev, + "(i2400m %p l3l4_hdr %p msg_size %zu) = %d\n", + i2400m, l3l4_hdr, msg_size, result); + return result; +} + + + +/* + * Cancel a wait for a command ACK + * + * @i2400m: device descriptor + * @code: [negative] errno code to cancel with (don't use + * -EINPROGRESS) + * + * If there is an ack already filled out, free it. + */ +void i2400m_msg_to_dev_cancel_wait(struct i2400m *i2400m, int code) +{ + struct sk_buff *ack_skb; + unsigned long flags; + + spin_lock_irqsave(&i2400m->rx_lock, flags); + ack_skb = i2400m->ack_skb; + if (ack_skb && !IS_ERR(ack_skb)) + kfree_skb(ack_skb); + i2400m->ack_skb = ERR_PTR(code); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); +} + + +/** + * i2400m_msg_to_dev - Send a control message to the device and get a response + * + * @i2400m: device descriptor + * + * @buf: pointer to the buffer containing the message to be sent; it + * has to start with a &struct i2400M_l3l4_hdr and then + * followed by the payload. Once this function returns, the + * buffer can be reused. + * + * @buf_len: buffer size + * + * Returns: + * + * Pointer to skb containing the ack message. You need to check the + * pointer with IS_ERR(), as it might be an error code. Error codes + * could happen because: + * + * - the message wasn't formatted correctly + * - couldn't send the message + * - failed waiting for a response + * - the ack message wasn't formatted correctly + * + * The returned skb has been allocated with wimax_msg_to_user_alloc(), + * it contains the response in a netlink attribute and is ready to be + * passed up to user space with wimax_msg_to_user_send(). To access + * the payload and its length, use wimax_msg_{data,len}() on the skb. + * + * The skb has to be freed with kfree_skb() once done. + * + * Description: + * + * This function delivers a message/command to the device and waits + * for an ack to be received. The format is described in + * linux/wimax/i2400m.h. In summary, a command/get/set is followed by an + * ack. + * + * This function will not check the ack status, that's left up to the + * caller. Once done with the ack skb, it has to be kfree_skb()ed. + * + * The i2400m handles only one message at the same time, thus we need + * the mutex to exclude other players. + * + * We write the message and then wait for an answer to come back. The + * RX path intercepts control messages and handles them in + * i2400m_rx_ctl(). Reports (notifications) are (maybe) processed + * locally and then forwarded (as needed) to user space on the WiMAX + * stack message pipe. Acks are saved and passed back to us through an + * skb in i2400m->ack_skb which is ready to be given to generic + * netlink if need be. + */ +struct sk_buff *i2400m_msg_to_dev(struct i2400m *i2400m, + const void *buf, size_t buf_len) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_l3l4_hdr *msg_l3l4_hdr; + struct sk_buff *ack_skb; + const struct i2400m_l3l4_hdr *ack_l3l4_hdr; + size_t ack_len; + int ack_timeout; + unsigned msg_type; + unsigned long flags; + + d_fnstart(3, dev, "(i2400m %p buf %p len %zu)\n", + i2400m, buf, buf_len); + + rmb(); /* Make sure we see what i2400m_dev_reset_handle() */ + if (i2400m->boot_mode) + return ERR_PTR(-EL3RST); + + msg_l3l4_hdr = buf; + /* Check msg & payload consistency */ + result = i2400m_msg_size_check(i2400m, msg_l3l4_hdr, buf_len); + if (result < 0) + goto error_bad_msg; + msg_type = le16_to_cpu(msg_l3l4_hdr->type); + d_printf(1, dev, "CMD/GET/SET 0x%04x %zu bytes\n", + msg_type, buf_len); + d_dump(2, dev, buf, buf_len); + + /* Setup the completion, ack_skb ("we are waiting") and send + * the message to the device */ + mutex_lock(&i2400m->msg_mutex); + spin_lock_irqsave(&i2400m->rx_lock, flags); + i2400m->ack_skb = ERR_PTR(-EINPROGRESS); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + init_completion(&i2400m->msg_completion); + result = i2400m_tx(i2400m, buf, buf_len, I2400M_PT_CTRL); + if (result < 0) { + dev_err(dev, "can't send message 0x%04x: %d\n", + le16_to_cpu(msg_l3l4_hdr->type), result); + goto error_tx; + } + + /* Some commands take longer to execute because of crypto ops, + * so we give them some more leeway on timeout */ + switch (msg_type) { + case I2400M_MT_GET_TLS_OPERATION_RESULT: + case I2400M_MT_CMD_SEND_EAP_RESPONSE: + ack_timeout = 5 * HZ; + break; + default: + ack_timeout = HZ; + } + + if (unlikely(i2400m->trace_msg_from_user)) + wimax_msg(&i2400m->wimax_dev, "echo", buf, buf_len, GFP_KERNEL); + /* The RX path in rx.c will put any response for this message + * in i2400m->ack_skb and wake us up. If we cancel the wait, + * we need to change the value of i2400m->ack_skb to something + * not -EINPROGRESS so RX knows there is no one waiting. */ + result = wait_for_completion_interruptible_timeout( + &i2400m->msg_completion, ack_timeout); + if (result == 0) { + dev_err(dev, "timeout waiting for reply to message 0x%04x\n", + msg_type); + result = -ETIMEDOUT; + i2400m_msg_to_dev_cancel_wait(i2400m, result); + goto error_wait_for_completion; + } else if (result < 0) { + dev_err(dev, "error waiting for reply to message 0x%04x: %d\n", + msg_type, result); + i2400m_msg_to_dev_cancel_wait(i2400m, result); + goto error_wait_for_completion; + } + + /* Pull out the ack data from i2400m->ack_skb -- see if it is + * an error and act accordingly */ + spin_lock_irqsave(&i2400m->rx_lock, flags); + ack_skb = i2400m->ack_skb; + if (IS_ERR(ack_skb)) + result = PTR_ERR(ack_skb); + else + result = 0; + i2400m->ack_skb = NULL; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + if (result < 0) + goto error_ack_status; + ack_l3l4_hdr = wimax_msg_data_len(ack_skb, &ack_len); + + /* Check the ack and deliver it if it is ok */ + if (unlikely(i2400m->trace_msg_from_user)) + wimax_msg(&i2400m->wimax_dev, "echo", + ack_l3l4_hdr, ack_len, GFP_KERNEL); + result = i2400m_msg_size_check(i2400m, ack_l3l4_hdr, ack_len); + if (result < 0) { + dev_err(dev, "HW BUG? reply to message 0x%04x: %d\n", + msg_type, result); + goto error_bad_ack_len; + } + if (msg_type != le16_to_cpu(ack_l3l4_hdr->type)) { + dev_err(dev, "HW BUG? bad reply 0x%04x to message 0x%04x\n", + le16_to_cpu(ack_l3l4_hdr->type), msg_type); + result = -EIO; + goto error_bad_ack_type; + } + i2400m_msg_ack_hook(i2400m, ack_l3l4_hdr, ack_len); + mutex_unlock(&i2400m->msg_mutex); + d_fnend(3, dev, "(i2400m %p buf %p len %zu) = %p\n", + i2400m, buf, buf_len, ack_skb); + return ack_skb; + +error_bad_ack_type: +error_bad_ack_len: + kfree_skb(ack_skb); +error_ack_status: +error_wait_for_completion: +error_tx: + mutex_unlock(&i2400m->msg_mutex); +error_bad_msg: + d_fnend(3, dev, "(i2400m %p buf %p len %zu) = %d\n", + i2400m, buf, buf_len, result); + return ERR_PTR(result); +} + + +/* + * Definitions for the Enter Power Save command + * + * The Enter Power Save command requests the device to go into power + * saving mode. The device will ack or nak the command depending on it + * being ready for it. If it acks, we tell the USB subsystem to + * + * As well, the device might request to go into power saving mode by + * sending a report (REPORT_POWERSAVE_READY), in which case, we issue + * this command. The hookups in the RX coder allow + */ +enum { + I2400M_WAKEUP_ENABLED = 0x01, + I2400M_WAKEUP_DISABLED = 0x02, + I2400M_TLV_TYPE_WAKEUP_MODE = 144, +}; + +struct i2400m_cmd_enter_power_save { + struct i2400m_l3l4_hdr hdr; + struct i2400m_tlv_hdr tlv; + __le32 val; +} __packed; + + +/* + * Request entering power save + * + * This command is (mainly) executed when the device indicates that it + * is ready to go into powersave mode via a REPORT_POWERSAVE_READY. + */ +int i2400m_cmd_enter_powersave(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct i2400m_cmd_enter_power_save *cmd; + char strerr[32]; + + result = -ENOMEM; + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->hdr.type = cpu_to_le16(I2400M_MT_CMD_ENTER_POWERSAVE); + cmd->hdr.length = cpu_to_le16(sizeof(*cmd) - sizeof(cmd->hdr)); + cmd->hdr.version = cpu_to_le16(I2400M_L3L4_VERSION); + cmd->tlv.type = cpu_to_le16(I2400M_TLV_TYPE_WAKEUP_MODE); + cmd->tlv.length = cpu_to_le16(sizeof(cmd->val)); + cmd->val = cpu_to_le32(I2400M_WAKEUP_ENABLED); + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + result = PTR_ERR(ack_skb); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'Enter power save' command: %d\n", + result); + goto error_msg_to_dev; + } + result = i2400m_msg_check_status(wimax_msg_data(ack_skb), + strerr, sizeof(strerr)); + if (result == -EACCES) + d_printf(1, dev, "Cannot enter power save mode\n"); + else if (result < 0) + dev_err(dev, "'Enter power save' (0x%04x) command failed: " + "%d - %s\n", I2400M_MT_CMD_ENTER_POWERSAVE, + result, strerr); + else + d_printf(1, dev, "device ready to power save\n"); + kfree_skb(ack_skb); +error_msg_to_dev: + kfree(cmd); +error_alloc: + return result; +} +EXPORT_SYMBOL_GPL(i2400m_cmd_enter_powersave); + + +/* + * Definitions for getting device information + */ +enum { + I2400M_TLV_DETAILED_DEVICE_INFO = 140 +}; + +/** + * i2400m_get_device_info - Query the device for detailed device information + * + * @i2400m: device descriptor + * + * Returns: an skb whose skb->data points to a 'struct + * i2400m_tlv_detailed_device_info'. When done, kfree_skb() it. The + * skb is *guaranteed* to contain the whole TLV data structure. + * + * On error, IS_ERR(skb) is true and ERR_PTR(skb) is the error + * code. + */ +struct sk_buff *i2400m_get_device_info(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct i2400m_l3l4_hdr *cmd; + const struct i2400m_l3l4_hdr *ack; + size_t ack_len; + const struct i2400m_tlv_hdr *tlv; + const struct i2400m_tlv_detailed_device_info *ddi; + char strerr[32]; + + ack_skb = ERR_PTR(-ENOMEM); + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->type = cpu_to_le16(I2400M_MT_GET_DEVICE_INFO); + cmd->length = 0; + cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'get device info' command: %ld\n", + PTR_ERR(ack_skb)); + goto error_msg_to_dev; + } + ack = wimax_msg_data_len(ack_skb, &ack_len); + result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); + if (result < 0) { + dev_err(dev, "'get device info' (0x%04x) command failed: " + "%d - %s\n", I2400M_MT_GET_DEVICE_INFO, result, + strerr); + goto error_cmd_failed; + } + tlv = i2400m_tlv_find(i2400m, ack->pl, ack_len - sizeof(*ack), + I2400M_TLV_DETAILED_DEVICE_INFO, sizeof(*ddi)); + if (tlv == NULL) { + dev_err(dev, "GET DEVICE INFO: " + "detailed device info TLV not found (0x%04x)\n", + I2400M_TLV_DETAILED_DEVICE_INFO); + result = -EIO; + goto error_no_tlv; + } + skb_pull(ack_skb, (void *) tlv - (void *) ack_skb->data); +error_msg_to_dev: + kfree(cmd); +error_alloc: + return ack_skb; + +error_no_tlv: +error_cmd_failed: + kfree_skb(ack_skb); + kfree(cmd); + return ERR_PTR(result); +} + + +/* Firmware interface versions we support */ +enum { + I2400M_HDIv_MAJOR = 9, + I2400M_HDIv_MINOR = 1, + I2400M_HDIv_MINOR_2 = 2, +}; + + +/** + * i2400m_firmware_check - check firmware versions are compatible with + * the driver + * + * @i2400m: device descriptor + * + * Returns: 0 if ok, < 0 errno code an error and a message in the + * kernel log. + * + * Long function, but quite simple; first chunk launches the command + * and double checks the reply for the right TLV. Then we process the + * TLV (where the meat is). + * + * Once we process the TLV that gives us the firmware's interface + * version, we encode it and save it in i2400m->fw_version for future + * reference. + */ +int i2400m_firmware_check(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct i2400m_l3l4_hdr *cmd; + const struct i2400m_l3l4_hdr *ack; + size_t ack_len; + const struct i2400m_tlv_hdr *tlv; + const struct i2400m_tlv_l4_message_versions *l4mv; + char strerr[32]; + unsigned major, minor, branch; + + result = -ENOMEM; + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->type = cpu_to_le16(I2400M_MT_GET_LM_VERSION); + cmd->length = 0; + cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + if (IS_ERR(ack_skb)) { + result = PTR_ERR(ack_skb); + dev_err(dev, "Failed to issue 'get lm version' command: %-d\n", + result); + goto error_msg_to_dev; + } + ack = wimax_msg_data_len(ack_skb, &ack_len); + result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); + if (result < 0) { + dev_err(dev, "'get lm version' (0x%04x) command failed: " + "%d - %s\n", I2400M_MT_GET_LM_VERSION, result, + strerr); + goto error_cmd_failed; + } + tlv = i2400m_tlv_find(i2400m, ack->pl, ack_len - sizeof(*ack), + I2400M_TLV_L4_MESSAGE_VERSIONS, sizeof(*l4mv)); + if (tlv == NULL) { + dev_err(dev, "get lm version: TLV not found (0x%04x)\n", + I2400M_TLV_L4_MESSAGE_VERSIONS); + result = -EIO; + goto error_no_tlv; + } + l4mv = container_of(tlv, typeof(*l4mv), hdr); + major = le16_to_cpu(l4mv->major); + minor = le16_to_cpu(l4mv->minor); + branch = le16_to_cpu(l4mv->branch); + result = -EINVAL; + if (major != I2400M_HDIv_MAJOR) { + dev_err(dev, "unsupported major fw version " + "%u.%u.%u\n", major, minor, branch); + goto error_bad_major; + } + result = 0; + if (minor > I2400M_HDIv_MINOR_2 || minor < I2400M_HDIv_MINOR) + dev_warn(dev, "untested minor fw version %u.%u.%u\n", + major, minor, branch); + /* Yes, we ignore the branch -- we don't have to track it */ + i2400m->fw_version = major << 16 | minor; + dev_info(dev, "firmware interface version %u.%u.%u\n", + major, minor, branch); +error_bad_major: +error_no_tlv: +error_cmd_failed: + kfree_skb(ack_skb); +error_msg_to_dev: + kfree(cmd); +error_alloc: + return result; +} + + +/* + * Send an DoExitIdle command to the device to ask it to go out of + * basestation-idle mode. + * + * @i2400m: device descriptor + * + * This starts a renegotiation with the basestation that might involve + * another crypto handshake with user space. + * + * Returns: 0 if ok, < 0 errno code on error. + */ +int i2400m_cmd_exit_idle(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct i2400m_l3l4_hdr *cmd; + char strerr[32]; + + result = -ENOMEM; + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->type = cpu_to_le16(I2400M_MT_CMD_EXIT_IDLE); + cmd->length = 0; + cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + result = PTR_ERR(ack_skb); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'exit idle' command: %d\n", + result); + goto error_msg_to_dev; + } + result = i2400m_msg_check_status(wimax_msg_data(ack_skb), + strerr, sizeof(strerr)); + kfree_skb(ack_skb); +error_msg_to_dev: + kfree(cmd); +error_alloc: + return result; + +} + + +/* + * Query the device for its state, update the WiMAX stack's idea of it + * + * @i2400m: device descriptor + * + * Returns: 0 if ok, < 0 errno code on error. + * + * Executes a 'Get State' command and parses the returned + * TLVs. + * + * Because this is almost identical to a 'Report State', we use + * i2400m_report_state_hook() to parse the answer. This will set the + * carrier state, as well as the RF Kill switches state. + */ +static int i2400m_cmd_get_state(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct i2400m_l3l4_hdr *cmd; + const struct i2400m_l3l4_hdr *ack; + size_t ack_len; + char strerr[32]; + + result = -ENOMEM; + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->type = cpu_to_le16(I2400M_MT_GET_STATE); + cmd->length = 0; + cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'get state' command: %ld\n", + PTR_ERR(ack_skb)); + result = PTR_ERR(ack_skb); + goto error_msg_to_dev; + } + ack = wimax_msg_data_len(ack_skb, &ack_len); + result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); + if (result < 0) { + dev_err(dev, "'get state' (0x%04x) command failed: " + "%d - %s\n", I2400M_MT_GET_STATE, result, strerr); + goto error_cmd_failed; + } + i2400m_report_state_hook(i2400m, ack, ack_len - sizeof(*ack), + "GET STATE"); + result = 0; + kfree_skb(ack_skb); +error_cmd_failed: +error_msg_to_dev: + kfree(cmd); +error_alloc: + return result; +} + +/** + * Set basic configuration settings + * + * @i2400m: device descriptor + * @args: array of pointers to the TLV headers to send for + * configuration (each followed by its payload). + * TLV headers and payloads must be properly initialized, with the + * right endianess (LE). + * @arg_size: number of pointers in the @args array + */ +static int i2400m_set_init_config(struct i2400m *i2400m, + const struct i2400m_tlv_hdr **arg, + size_t args) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct i2400m_l3l4_hdr *cmd; + char strerr[32]; + unsigned argc, argsize, tlv_size; + const struct i2400m_tlv_hdr *tlv_hdr; + void *buf, *itr; + + d_fnstart(3, dev, "(i2400m %p arg %p args %zu)\n", i2400m, arg, args); + result = 0; + if (args == 0) + goto none; + /* Compute the size of all the TLVs, so we can alloc a + * contiguous command block to copy them. */ + argsize = 0; + for (argc = 0; argc < args; argc++) { + tlv_hdr = arg[argc]; + argsize += sizeof(*tlv_hdr) + le16_to_cpu(tlv_hdr->length); + } + WARN_ON(argc >= 9); /* As per hw spec */ + + /* Alloc the space for the command and TLVs*/ + result = -ENOMEM; + buf = kzalloc(sizeof(*cmd) + argsize, GFP_KERNEL); + if (buf == NULL) + goto error_alloc; + cmd = buf; + cmd->type = cpu_to_le16(I2400M_MT_SET_INIT_CONFIG); + cmd->length = cpu_to_le16(argsize); + cmd->version = cpu_to_le16(I2400M_L3L4_VERSION); + + /* Copy the TLVs */ + itr = buf + sizeof(*cmd); + for (argc = 0; argc < args; argc++) { + tlv_hdr = arg[argc]; + tlv_size = sizeof(*tlv_hdr) + le16_to_cpu(tlv_hdr->length); + memcpy(itr, tlv_hdr, tlv_size); + itr += tlv_size; + } + + /* Send the message! */ + ack_skb = i2400m_msg_to_dev(i2400m, buf, sizeof(*cmd) + argsize); + result = PTR_ERR(ack_skb); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'init config' command: %d\n", + result); + + goto error_msg_to_dev; + } + result = i2400m_msg_check_status(wimax_msg_data(ack_skb), + strerr, sizeof(strerr)); + if (result < 0) + dev_err(dev, "'init config' (0x%04x) command failed: %d - %s\n", + I2400M_MT_SET_INIT_CONFIG, result, strerr); + kfree_skb(ack_skb); +error_msg_to_dev: + kfree(buf); +error_alloc: +none: + d_fnend(3, dev, "(i2400m %p arg %p args %zu) = %d\n", + i2400m, arg, args, result); + return result; + +} + +/** + * i2400m_set_idle_timeout - Set the device's idle mode timeout + * + * @i2400m: i2400m device descriptor + * + * @msecs: milliseconds for the timeout to enter idle mode. Between + * 100 to 300000 (5m); 0 to disable. In increments of 100. + * + * After this @msecs of the link being idle (no data being sent or + * received), the device will negotiate with the basestation entering + * idle mode for saving power. The connection is maintained, but + * getting out of it (done in tx.c) will require some negotiation, + * possible crypto re-handshake and a possible DHCP re-lease. + * + * Only available if fw_version >= 0x00090002. + * + * Returns: 0 if ok, < 0 errno code on error. + */ +int i2400m_set_idle_timeout(struct i2400m *i2400m, unsigned msecs) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct { + struct i2400m_l3l4_hdr hdr; + struct i2400m_tlv_config_idle_timeout cit; + } *cmd; + const struct i2400m_l3l4_hdr *ack; + size_t ack_len; + char strerr[32]; + + result = -ENOSYS; + if (i2400m_le_v1_3(i2400m)) + goto error_alloc; + result = -ENOMEM; + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->hdr.type = cpu_to_le16(I2400M_MT_GET_STATE); + cmd->hdr.length = cpu_to_le16(sizeof(*cmd) - sizeof(cmd->hdr)); + cmd->hdr.version = cpu_to_le16(I2400M_L3L4_VERSION); + + cmd->cit.hdr.type = + cpu_to_le16(I2400M_TLV_CONFIG_IDLE_TIMEOUT); + cmd->cit.hdr.length = cpu_to_le16(sizeof(cmd->cit.timeout)); + cmd->cit.timeout = cpu_to_le32(msecs); + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'set idle timeout' command: " + "%ld\n", PTR_ERR(ack_skb)); + result = PTR_ERR(ack_skb); + goto error_msg_to_dev; + } + ack = wimax_msg_data_len(ack_skb, &ack_len); + result = i2400m_msg_check_status(ack, strerr, sizeof(strerr)); + if (result < 0) { + dev_err(dev, "'set idle timeout' (0x%04x) command failed: " + "%d - %s\n", I2400M_MT_GET_STATE, result, strerr); + goto error_cmd_failed; + } + result = 0; + kfree_skb(ack_skb); +error_cmd_failed: +error_msg_to_dev: + kfree(cmd); +error_alloc: + return result; +} + + +/** + * i2400m_dev_initialize - Initialize the device once communications are ready + * + * @i2400m: device descriptor + * + * Returns: 0 if ok, < 0 errno code on error. + * + * Configures the device to work the way we like it. + * + * At the point of this call, the device is registered with the WiMAX + * and netdev stacks, firmware is uploaded and we can talk to the + * device normally. + */ +int i2400m_dev_initialize(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct i2400m_tlv_config_idle_parameters idle_params; + struct i2400m_tlv_config_idle_timeout idle_timeout; + struct i2400m_tlv_config_d2h_data_format df; + struct i2400m_tlv_config_dl_host_reorder dlhr; + const struct i2400m_tlv_hdr *args[9]; + unsigned argc = 0; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + if (i2400m_passive_mode) + goto out_passive; + /* Disable idle mode? (enabled by default) */ + if (i2400m_idle_mode_disabled) { + if (i2400m_le_v1_3(i2400m)) { + idle_params.hdr.type = + cpu_to_le16(I2400M_TLV_CONFIG_IDLE_PARAMETERS); + idle_params.hdr.length = cpu_to_le16( + sizeof(idle_params) - sizeof(idle_params.hdr)); + idle_params.idle_timeout = 0; + idle_params.idle_paging_interval = 0; + args[argc++] = &idle_params.hdr; + } else { + idle_timeout.hdr.type = + cpu_to_le16(I2400M_TLV_CONFIG_IDLE_TIMEOUT); + idle_timeout.hdr.length = cpu_to_le16( + sizeof(idle_timeout) - sizeof(idle_timeout.hdr)); + idle_timeout.timeout = 0; + args[argc++] = &idle_timeout.hdr; + } + } + if (i2400m_ge_v1_4(i2400m)) { + /* Enable extended RX data format? */ + df.hdr.type = + cpu_to_le16(I2400M_TLV_CONFIG_D2H_DATA_FORMAT); + df.hdr.length = cpu_to_le16( + sizeof(df) - sizeof(df.hdr)); + df.format = 1; + args[argc++] = &df.hdr; + + /* Enable RX data reordering? + * (switch flipped in rx.c:i2400m_rx_setup() after fw upload) */ + if (i2400m->rx_reorder) { + dlhr.hdr.type = + cpu_to_le16(I2400M_TLV_CONFIG_DL_HOST_REORDER); + dlhr.hdr.length = cpu_to_le16( + sizeof(dlhr) - sizeof(dlhr.hdr)); + dlhr.reorder = 1; + args[argc++] = &dlhr.hdr; + } + } + result = i2400m_set_init_config(i2400m, args, argc); + if (result < 0) + goto error; +out_passive: + /* + * Update state: Here it just calls a get state; parsing the + * result (System State TLV and RF Status TLV [done in the rx + * path hooks]) will set the hardware and software RF-Kill + * status. + */ + result = i2400m_cmd_get_state(i2400m); +error: + if (result < 0) + dev_err(dev, "failed to initialize the device: %d\n", result); + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; +} + + +/** + * i2400m_dev_shutdown - Shutdown a running device + * + * @i2400m: device descriptor + * + * Release resources acquired during the running of the device; in + * theory, should also tell the device to go to sleep, switch off the + * radio, all that, but at this point, in most cases (driver + * disconnection, reset handling) we can't even talk to the device. + */ +void i2400m_dev_shutdown(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} diff --git a/drivers/staging/wimax/i2400m/debug-levels.h b/drivers/staging/wimax/i2400m/debug-levels.h new file mode 100644 index 000000000000..a317e9fbb734 --- /dev/null +++ b/drivers/staging/wimax/i2400m/debug-levels.h @@ -0,0 +1,32 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Intel Wireless WiMAX Connection 2400m + * Debug levels control file for the i2400m module + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + */ +#ifndef __debug_levels__h__ +#define __debug_levels__h__ + +/* Maximum compile and run time debug level for all submodules */ +#define D_MODULENAME i2400m +#define D_MASTER CONFIG_WIMAX_I2400M_DEBUG_LEVEL + +#include "../linux-wimax-debug.h" + +/* List of all the enabled modules */ +enum d_module { + D_SUBMODULE_DECLARE(control), + D_SUBMODULE_DECLARE(driver), + D_SUBMODULE_DECLARE(debugfs), + D_SUBMODULE_DECLARE(fw), + D_SUBMODULE_DECLARE(netdev), + D_SUBMODULE_DECLARE(rfkill), + D_SUBMODULE_DECLARE(rx), + D_SUBMODULE_DECLARE(sysfs), + D_SUBMODULE_DECLARE(tx), +}; + + +#endif /* #ifndef __debug_levels__h__ */ diff --git a/drivers/staging/wimax/i2400m/debugfs.c b/drivers/staging/wimax/i2400m/debugfs.c new file mode 100644 index 000000000000..1c640b41ea4c --- /dev/null +++ b/drivers/staging/wimax/i2400m/debugfs.c @@ -0,0 +1,253 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Wireless WiMAX Connection 2400m + * Debugfs interfaces to manipulate driver and device information + * + * Copyright (C) 2007 Intel Corporation + * Inaky Perez-Gonzalez + */ + +#include +#include +#include +#include +#include +#include +#include "i2400m.h" + + +#define D_SUBMODULE debugfs +#include "debug-levels.h" + +static +int debugfs_netdev_queue_stopped_get(void *data, u64 *val) +{ + struct i2400m *i2400m = data; + *val = netif_queue_stopped(i2400m->wimax_dev.net_dev); + return 0; +} +DEFINE_DEBUGFS_ATTRIBUTE(fops_netdev_queue_stopped, + debugfs_netdev_queue_stopped_get, + NULL, "%llu\n"); + +/* + * We don't allow partial reads of this file, as then the reader would + * get weirdly confused data as it is updated. + * + * So or you read it all or nothing; if you try to read with an offset + * != 0, we consider you are done reading. + */ +static +ssize_t i2400m_rx_stats_read(struct file *filp, char __user *buffer, + size_t count, loff_t *ppos) +{ + struct i2400m *i2400m = filp->private_data; + char buf[128]; + unsigned long flags; + + if (*ppos != 0) + return 0; + if (count < sizeof(buf)) + return -ENOSPC; + spin_lock_irqsave(&i2400m->rx_lock, flags); + snprintf(buf, sizeof(buf), "%u %u %u %u %u %u %u\n", + i2400m->rx_pl_num, i2400m->rx_pl_min, + i2400m->rx_pl_max, i2400m->rx_num, + i2400m->rx_size_acc, + i2400m->rx_size_min, i2400m->rx_size_max); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + return simple_read_from_buffer(buffer, count, ppos, buf, strlen(buf)); +} + + +/* Any write clears the stats */ +static +ssize_t i2400m_rx_stats_write(struct file *filp, const char __user *buffer, + size_t count, loff_t *ppos) +{ + struct i2400m *i2400m = filp->private_data; + unsigned long flags; + + spin_lock_irqsave(&i2400m->rx_lock, flags); + i2400m->rx_pl_num = 0; + i2400m->rx_pl_max = 0; + i2400m->rx_pl_min = UINT_MAX; + i2400m->rx_num = 0; + i2400m->rx_size_acc = 0; + i2400m->rx_size_min = UINT_MAX; + i2400m->rx_size_max = 0; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + return count; +} + +static +const struct file_operations i2400m_rx_stats_fops = { + .owner = THIS_MODULE, + .open = simple_open, + .read = i2400m_rx_stats_read, + .write = i2400m_rx_stats_write, + .llseek = default_llseek, +}; + + +/* See i2400m_rx_stats_read() */ +static +ssize_t i2400m_tx_stats_read(struct file *filp, char __user *buffer, + size_t count, loff_t *ppos) +{ + struct i2400m *i2400m = filp->private_data; + char buf[128]; + unsigned long flags; + + if (*ppos != 0) + return 0; + if (count < sizeof(buf)) + return -ENOSPC; + spin_lock_irqsave(&i2400m->tx_lock, flags); + snprintf(buf, sizeof(buf), "%u %u %u %u %u %u %u\n", + i2400m->tx_pl_num, i2400m->tx_pl_min, + i2400m->tx_pl_max, i2400m->tx_num, + i2400m->tx_size_acc, + i2400m->tx_size_min, i2400m->tx_size_max); + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + return simple_read_from_buffer(buffer, count, ppos, buf, strlen(buf)); +} + +/* Any write clears the stats */ +static +ssize_t i2400m_tx_stats_write(struct file *filp, const char __user *buffer, + size_t count, loff_t *ppos) +{ + struct i2400m *i2400m = filp->private_data; + unsigned long flags; + + spin_lock_irqsave(&i2400m->tx_lock, flags); + i2400m->tx_pl_num = 0; + i2400m->tx_pl_max = 0; + i2400m->tx_pl_min = UINT_MAX; + i2400m->tx_num = 0; + i2400m->tx_size_acc = 0; + i2400m->tx_size_min = UINT_MAX; + i2400m->tx_size_max = 0; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + return count; +} + +static +const struct file_operations i2400m_tx_stats_fops = { + .owner = THIS_MODULE, + .open = simple_open, + .read = i2400m_tx_stats_read, + .write = i2400m_tx_stats_write, + .llseek = default_llseek, +}; + + +/* Write 1 to ask the device to go into suspend */ +static +int debugfs_i2400m_suspend_set(void *data, u64 val) +{ + int result; + struct i2400m *i2400m = data; + result = i2400m_cmd_enter_powersave(i2400m); + if (result >= 0) + result = 0; + return result; +} +DEFINE_DEBUGFS_ATTRIBUTE(fops_i2400m_suspend, + NULL, debugfs_i2400m_suspend_set, + "%llu\n"); + +/* + * Reset the device + * + * Write 0 to ask the device to soft reset, 1 to cold reset, 2 to bus + * reset (as defined by enum i2400m_reset_type). + */ +static +int debugfs_i2400m_reset_set(void *data, u64 val) +{ + int result; + struct i2400m *i2400m = data; + enum i2400m_reset_type rt = val; + switch(rt) { + case I2400M_RT_WARM: + case I2400M_RT_COLD: + case I2400M_RT_BUS: + result = i2400m_reset(i2400m, rt); + if (result >= 0) + result = 0; + break; + default: + result = -EINVAL; + } + return result; +} +DEFINE_DEBUGFS_ATTRIBUTE(fops_i2400m_reset, + NULL, debugfs_i2400m_reset_set, + "%llu\n"); + +void i2400m_debugfs_add(struct i2400m *i2400m) +{ + struct dentry *dentry = i2400m->wimax_dev.debugfs_dentry; + + dentry = debugfs_create_dir("i2400m", dentry); + i2400m->debugfs_dentry = dentry; + + d_level_register_debugfs("dl_", control, dentry); + d_level_register_debugfs("dl_", driver, dentry); + d_level_register_debugfs("dl_", debugfs, dentry); + d_level_register_debugfs("dl_", fw, dentry); + d_level_register_debugfs("dl_", netdev, dentry); + d_level_register_debugfs("dl_", rfkill, dentry); + d_level_register_debugfs("dl_", rx, dentry); + d_level_register_debugfs("dl_", tx, dentry); + + debugfs_create_size_t("tx_in", 0400, dentry, &i2400m->tx_in); + debugfs_create_size_t("tx_out", 0400, dentry, &i2400m->tx_out); + debugfs_create_u32("state", 0600, dentry, &i2400m->state); + + /* + * Trace received messages from user space + * + * In order to tap the bidirectional message stream in the + * 'msg' pipe, user space can read from the 'msg' pipe; + * however, due to limitations in libnl, we can't know what + * the different applications are sending down to the kernel. + * + * So we have this hack where the driver will echo any message + * received on the msg pipe from user space [through a call to + * wimax_dev->op_msg_from_user() into + * i2400m_op_msg_from_user()] into the 'trace' pipe that this + * driver creates. + * + * So then, reading from both the 'trace' and 'msg' pipes in + * user space will provide a full dump of the traffic. + * + * Write 1 to activate, 0 to clear. + * + * It is not really very atomic, but it is also not too + * critical. + */ + debugfs_create_u8("trace_msg_from_user", 0600, dentry, + &i2400m->trace_msg_from_user); + + debugfs_create_file("netdev_queue_stopped", 0400, dentry, i2400m, + &fops_netdev_queue_stopped); + + debugfs_create_file("rx_stats", 0600, dentry, i2400m, + &i2400m_rx_stats_fops); + + debugfs_create_file("tx_stats", 0600, dentry, i2400m, + &i2400m_tx_stats_fops); + + debugfs_create_file("suspend", 0200, dentry, i2400m, + &fops_i2400m_suspend); + + debugfs_create_file("reset", 0200, dentry, i2400m, &fops_i2400m_reset); +} + +void i2400m_debugfs_rm(struct i2400m *i2400m) +{ + debugfs_remove_recursive(i2400m->debugfs_dentry); +} diff --git a/drivers/staging/wimax/i2400m/driver.c b/drivers/staging/wimax/i2400m/driver.c new file mode 100644 index 000000000000..dc8939ff78c0 --- /dev/null +++ b/drivers/staging/wimax/i2400m/driver.c @@ -0,0 +1,1002 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Wireless WiMAX Connection 2400m + * Generic probe/disconnect, reset and message passing + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + * + * See i2400m.h for driver documentation. This contains helpers for + * the driver model glue [_setup()/_release()], handling device resets + * [_dev_reset_handle()], and the backends for the WiMAX stack ops + * reset [_op_reset()] and message from user [_op_msg_from_user()]. + * + * ROADMAP: + * + * i2400m_op_msg_from_user() + * i2400m_msg_to_dev() + * wimax_msg_to_user_send() + * + * i2400m_op_reset() + * i240m->bus_reset() + * + * i2400m_dev_reset_handle() + * __i2400m_dev_reset_handle() + * __i2400m_dev_stop() + * __i2400m_dev_start() + * + * i2400m_setup() + * i2400m->bus_setup() + * i2400m_bootrom_init() + * register_netdev() + * wimax_dev_add() + * i2400m_dev_start() + * __i2400m_dev_start() + * i2400m_dev_bootstrap() + * i2400m_tx_setup() + * i2400m->bus_dev_start() + * i2400m_firmware_check() + * i2400m_check_mac_addr() + * + * i2400m_release() + * i2400m_dev_stop() + * __i2400m_dev_stop() + * i2400m_dev_shutdown() + * i2400m->bus_dev_stop() + * i2400m_tx_release() + * i2400m->bus_release() + * wimax_dev_rm() + * unregister_netdev() + */ +#include "i2400m.h" +#include +#include "linux-wimax-i2400m.h" +#include +#include +#include +#include + +#define D_SUBMODULE driver +#include "debug-levels.h" + + +static char i2400m_debug_params[128]; +module_param_string(debug, i2400m_debug_params, sizeof(i2400m_debug_params), + 0644); +MODULE_PARM_DESC(debug, + "String of space-separated NAME:VALUE pairs, where NAMEs " + "are the different debug submodules and VALUE are the " + "initial debug value to set."); + +static char i2400m_barkers_params[128]; +module_param_string(barkers, i2400m_barkers_params, + sizeof(i2400m_barkers_params), 0644); +MODULE_PARM_DESC(barkers, + "String of comma-separated 32-bit values; each is " + "recognized as the value the device sends as a reboot " + "signal; values are appended to a list--setting one value " + "as zero cleans the existing list and starts a new one."); + +/* + * WiMAX stack operation: relay a message from user space + * + * @wimax_dev: device descriptor + * @pipe_name: named pipe the message is for + * @msg_buf: pointer to the message bytes + * @msg_len: length of the buffer + * @genl_info: passed by the generic netlink layer + * + * The WiMAX stack will call this function when a message was received + * from user space. + * + * For the i2400m, this is an L3L4 message, as specified in + * include/linux/wimax/i2400m.h, and thus prefixed with a 'struct + * i2400m_l3l4_hdr'. Driver (and device) expect the messages to be + * coded in Little Endian. + * + * This function just verifies that the header declaration and the + * payload are consistent and then deals with it, either forwarding it + * to the device or procesing it locally. + * + * In the i2400m, messages are basically commands that will carry an + * ack, so we use i2400m_msg_to_dev() and then deliver the ack back to + * user space. The rx.c code might intercept the response and use it + * to update the driver's state, but then it will pass it on so it can + * be relayed back to user space. + * + * Note that asynchronous events from the device are processed and + * sent to user space in rx.c. + */ +static +int i2400m_op_msg_from_user(struct wimax_dev *wimax_dev, + const char *pipe_name, + const void *msg_buf, size_t msg_len, + const struct genl_info *genl_info) +{ + int result; + struct i2400m *i2400m = wimax_dev_to_i2400m(wimax_dev); + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + + d_fnstart(4, dev, "(wimax_dev %p [i2400m %p] msg_buf %p " + "msg_len %zu genl_info %p)\n", wimax_dev, i2400m, + msg_buf, msg_len, genl_info); + ack_skb = i2400m_msg_to_dev(i2400m, msg_buf, msg_len); + result = PTR_ERR(ack_skb); + if (IS_ERR(ack_skb)) + goto error_msg_to_dev; + result = wimax_msg_send(&i2400m->wimax_dev, ack_skb); +error_msg_to_dev: + d_fnend(4, dev, "(wimax_dev %p [i2400m %p] msg_buf %p msg_len %zu " + "genl_info %p) = %d\n", wimax_dev, i2400m, msg_buf, msg_len, + genl_info, result); + return result; +} + + +/* + * Context to wait for a reset to finalize + */ +struct i2400m_reset_ctx { + struct completion completion; + int result; +}; + + +/* + * WiMAX stack operation: reset a device + * + * @wimax_dev: device descriptor + * + * See the documentation for wimax_reset() and wimax_dev->op_reset for + * the requirements of this function. The WiMAX stack guarantees + * serialization on calls to this function. + * + * Do a warm reset on the device; if it fails, resort to a cold reset + * and return -ENODEV. On successful warm reset, we need to block + * until it is complete. + * + * The bus-driver implementation of reset takes care of falling back + * to cold reset if warm fails. + */ +static +int i2400m_op_reset(struct wimax_dev *wimax_dev) +{ + int result; + struct i2400m *i2400m = wimax_dev_to_i2400m(wimax_dev); + struct device *dev = i2400m_dev(i2400m); + struct i2400m_reset_ctx ctx = { + .completion = COMPLETION_INITIALIZER_ONSTACK(ctx.completion), + .result = 0, + }; + + d_fnstart(4, dev, "(wimax_dev %p)\n", wimax_dev); + mutex_lock(&i2400m->init_mutex); + i2400m->reset_ctx = &ctx; + mutex_unlock(&i2400m->init_mutex); + result = i2400m_reset(i2400m, I2400M_RT_WARM); + if (result < 0) + goto out; + result = wait_for_completion_timeout(&ctx.completion, 4*HZ); + if (result == 0) + result = -ETIMEDOUT; + else if (result > 0) + result = ctx.result; + /* if result < 0, pass it on */ + mutex_lock(&i2400m->init_mutex); + i2400m->reset_ctx = NULL; + mutex_unlock(&i2400m->init_mutex); +out: + d_fnend(4, dev, "(wimax_dev %p) = %d\n", wimax_dev, result); + return result; +} + + +/* + * Check the MAC address we got from boot mode is ok + * + * @i2400m: device descriptor + * + * Returns: 0 if ok, < 0 errno code on error. + */ +static +int i2400m_check_mac_addr(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *skb; + const struct i2400m_tlv_detailed_device_info *ddi; + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + skb = i2400m_get_device_info(i2400m); + if (IS_ERR(skb)) { + result = PTR_ERR(skb); + dev_err(dev, "Cannot verify MAC address, error reading: %d\n", + result); + goto error; + } + /* Extract MAC address */ + ddi = (void *) skb->data; + BUILD_BUG_ON(ETH_ALEN != sizeof(ddi->mac_address)); + d_printf(2, dev, "GET DEVICE INFO: mac addr %pM\n", + ddi->mac_address); + if (!memcmp(net_dev->perm_addr, ddi->mac_address, + sizeof(ddi->mac_address))) + goto ok; + dev_warn(dev, "warning: device reports a different MAC address " + "to that of boot mode's\n"); + dev_warn(dev, "device reports %pM\n", ddi->mac_address); + dev_warn(dev, "boot mode reported %pM\n", net_dev->perm_addr); + if (is_zero_ether_addr(ddi->mac_address)) + dev_err(dev, "device reports an invalid MAC address, " + "not updating\n"); + else { + dev_warn(dev, "updating MAC address\n"); + net_dev->addr_len = ETH_ALEN; + memcpy(net_dev->perm_addr, ddi->mac_address, ETH_ALEN); + memcpy(net_dev->dev_addr, ddi->mac_address, ETH_ALEN); + } +ok: + result = 0; + kfree_skb(skb); +error: + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; +} + + +/** + * __i2400m_dev_start - Bring up driver communication with the device + * + * @i2400m: device descriptor + * @flags: boot mode flags + * + * Returns: 0 if ok, < 0 errno code on error. + * + * Uploads firmware and brings up all the resources needed to be able + * to communicate with the device. + * + * The workqueue has to be setup early, at least before RX handling + * (it's only real user for now) so it can process reports as they + * arrive. We also want to destroy it if we retry, to make sure it is + * flushed...easier like this. + * + * TX needs to be setup before the bus-specific code (otherwise on + * shutdown, the bus-tx code could try to access it). + */ +static +int __i2400m_dev_start(struct i2400m *i2400m, enum i2400m_bri flags) +{ + int result; + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + struct net_device *net_dev = wimax_dev->net_dev; + struct device *dev = i2400m_dev(i2400m); + int times = i2400m->bus_bm_retries; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); +retry: + result = i2400m_dev_bootstrap(i2400m, flags); + if (result < 0) { + dev_err(dev, "cannot bootstrap device: %d\n", result); + goto error_bootstrap; + } + result = i2400m_tx_setup(i2400m); + if (result < 0) + goto error_tx_setup; + result = i2400m_rx_setup(i2400m); + if (result < 0) + goto error_rx_setup; + i2400m->work_queue = create_singlethread_workqueue(wimax_dev->name); + if (i2400m->work_queue == NULL) { + result = -ENOMEM; + dev_err(dev, "cannot create workqueue\n"); + goto error_create_workqueue; + } + if (i2400m->bus_dev_start) { + result = i2400m->bus_dev_start(i2400m); + if (result < 0) + goto error_bus_dev_start; + } + i2400m->ready = 1; + wmb(); /* see i2400m->ready's documentation */ + /* process pending reports from the device */ + queue_work(i2400m->work_queue, &i2400m->rx_report_ws); + result = i2400m_firmware_check(i2400m); /* fw versions ok? */ + if (result < 0) + goto error_fw_check; + /* At this point is ok to send commands to the device */ + result = i2400m_check_mac_addr(i2400m); + if (result < 0) + goto error_check_mac_addr; + result = i2400m_dev_initialize(i2400m); + if (result < 0) + goto error_dev_initialize; + + /* We don't want any additional unwanted error recovery triggered + * from any other context so if anything went wrong before we come + * here, let's keep i2400m->error_recovery untouched and leave it to + * dev_reset_handle(). See dev_reset_handle(). */ + + atomic_dec(&i2400m->error_recovery); + /* Every thing works so far, ok, now we are ready to + * take error recovery if it's required. */ + + /* At this point, reports will come for the device and set it + * to the right state if it is different than UNINITIALIZED */ + d_fnend(3, dev, "(net_dev %p [i2400m %p]) = %d\n", + net_dev, i2400m, result); + return result; + +error_dev_initialize: +error_check_mac_addr: +error_fw_check: + i2400m->ready = 0; + wmb(); /* see i2400m->ready's documentation */ + flush_workqueue(i2400m->work_queue); + if (i2400m->bus_dev_stop) + i2400m->bus_dev_stop(i2400m); +error_bus_dev_start: + destroy_workqueue(i2400m->work_queue); +error_create_workqueue: + i2400m_rx_release(i2400m); +error_rx_setup: + i2400m_tx_release(i2400m); +error_tx_setup: +error_bootstrap: + if (result == -EL3RST && times-- > 0) { + flags = I2400M_BRI_SOFT|I2400M_BRI_MAC_REINIT; + goto retry; + } + d_fnend(3, dev, "(net_dev %p [i2400m %p]) = %d\n", + net_dev, i2400m, result); + return result; +} + + +static +int i2400m_dev_start(struct i2400m *i2400m, enum i2400m_bri bm_flags) +{ + int result = 0; + mutex_lock(&i2400m->init_mutex); /* Well, start the device */ + if (i2400m->updown == 0) { + result = __i2400m_dev_start(i2400m, bm_flags); + if (result >= 0) { + i2400m->updown = 1; + i2400m->alive = 1; + wmb();/* see i2400m->updown and i2400m->alive's doc */ + } + } + mutex_unlock(&i2400m->init_mutex); + return result; +} + + +/** + * i2400m_dev_stop - Tear down driver communication with the device + * + * @i2400m: device descriptor + * + * Returns: 0 if ok, < 0 errno code on error. + * + * Releases all the resources allocated to communicate with the + * device. Note we cannot destroy the workqueue earlier as until RX is + * fully destroyed, it could still try to schedule jobs. + */ +static +void __i2400m_dev_stop(struct i2400m *i2400m) +{ + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + wimax_state_change(wimax_dev, __WIMAX_ST_QUIESCING); + i2400m_msg_to_dev_cancel_wait(i2400m, -EL3RST); + complete(&i2400m->msg_completion); + i2400m_net_wake_stop(i2400m); + i2400m_dev_shutdown(i2400m); + /* + * Make sure no report hooks are running *before* we stop the + * communication infrastructure with the device. + */ + i2400m->ready = 0; /* nobody can queue work anymore */ + wmb(); /* see i2400m->ready's documentation */ + flush_workqueue(i2400m->work_queue); + + if (i2400m->bus_dev_stop) + i2400m->bus_dev_stop(i2400m); + destroy_workqueue(i2400m->work_queue); + i2400m_rx_release(i2400m); + i2400m_tx_release(i2400m); + wimax_state_change(wimax_dev, WIMAX_ST_DOWN); + d_fnend(3, dev, "(i2400m %p) = 0\n", i2400m); +} + + +/* + * Watch out -- we only need to stop if there is a need for it. The + * device could have reset itself and failed to come up again (see + * _i2400m_dev_reset_handle()). + */ +static +void i2400m_dev_stop(struct i2400m *i2400m) +{ + mutex_lock(&i2400m->init_mutex); + if (i2400m->updown) { + __i2400m_dev_stop(i2400m); + i2400m->updown = 0; + i2400m->alive = 0; + wmb(); /* see i2400m->updown and i2400m->alive's doc */ + } + mutex_unlock(&i2400m->init_mutex); +} + + +/* + * Listen to PM events to cache the firmware before suspend/hibernation + * + * When the device comes out of suspend, it might go into reset and + * firmware has to be uploaded again. At resume, most of the times, we + * can't load firmware images from disk, so we need to cache it. + * + * i2400m_fw_cache() will allocate a kobject and attach the firmware + * to it; that way we don't have to worry too much about the fw loader + * hitting a race condition. + * + * Note: modus operandi stolen from the Orinoco driver; thx. + */ +static +int i2400m_pm_notifier(struct notifier_block *notifier, + unsigned long pm_event, + void *unused) +{ + struct i2400m *i2400m = + container_of(notifier, struct i2400m, pm_notifier); + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p pm_event %lx)\n", i2400m, pm_event); + switch (pm_event) { + case PM_HIBERNATION_PREPARE: + case PM_SUSPEND_PREPARE: + i2400m_fw_cache(i2400m); + break; + case PM_POST_RESTORE: + /* Restore from hibernation failed. We need to clean + * up in exactly the same way, so fall through. */ + case PM_POST_HIBERNATION: + case PM_POST_SUSPEND: + i2400m_fw_uncache(i2400m); + break; + + case PM_RESTORE_PREPARE: + default: + break; + } + d_fnend(3, dev, "(i2400m %p pm_event %lx) = void\n", i2400m, pm_event); + return NOTIFY_DONE; +} + + +/* + * pre-reset is called before a device is going on reset + * + * This has to be followed by a call to i2400m_post_reset(), otherwise + * bad things might happen. + */ +int i2400m_pre_reset(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + d_printf(1, dev, "pre-reset shut down\n"); + + mutex_lock(&i2400m->init_mutex); + if (i2400m->updown) { + netif_tx_disable(i2400m->wimax_dev.net_dev); + __i2400m_dev_stop(i2400m); + /* down't set updown to zero -- this way + * post_reset can restore properly */ + } + mutex_unlock(&i2400m->init_mutex); + if (i2400m->bus_release) + i2400m->bus_release(i2400m); + d_fnend(3, dev, "(i2400m %p) = 0\n", i2400m); + return 0; +} +EXPORT_SYMBOL_GPL(i2400m_pre_reset); + + +/* + * Restore device state after a reset + * + * Do the work needed after a device reset to bring it up to the same + * state as it was before the reset. + * + * NOTE: this requires i2400m->init_mutex taken + */ +int i2400m_post_reset(struct i2400m *i2400m) +{ + int result = 0; + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + d_printf(1, dev, "post-reset start\n"); + if (i2400m->bus_setup) { + result = i2400m->bus_setup(i2400m); + if (result < 0) { + dev_err(dev, "bus-specific setup failed: %d\n", + result); + goto error_bus_setup; + } + } + mutex_lock(&i2400m->init_mutex); + if (i2400m->updown) { + result = __i2400m_dev_start( + i2400m, I2400M_BRI_SOFT | I2400M_BRI_MAC_REINIT); + if (result < 0) + goto error_dev_start; + } + mutex_unlock(&i2400m->init_mutex); + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; + +error_dev_start: + if (i2400m->bus_release) + i2400m->bus_release(i2400m); + /* even if the device was up, it could not be recovered, so we + * mark it as down. */ + i2400m->updown = 0; + wmb(); /* see i2400m->updown's documentation */ + mutex_unlock(&i2400m->init_mutex); +error_bus_setup: + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; +} +EXPORT_SYMBOL_GPL(i2400m_post_reset); + + +/* + * The device has rebooted; fix up the device and the driver + * + * Tear down the driver communication with the device, reload the + * firmware and reinitialize the communication with the device. + * + * If someone calls a reset when the device's firmware is down, in + * theory we won't see it because we are not listening. However, just + * in case, leave the code to handle it. + * + * If there is a reset context, use it; this means someone is waiting + * for us to tell him when the reset operation is complete and the + * device is ready to rock again. + * + * NOTE: if we are in the process of bringing up or down the + * communication with the device [running i2400m_dev_start() or + * _stop()], don't do anything, let it fail and handle it. + * + * This function is ran always in a thread context + * + * This function gets passed, as payload to i2400m_work() a 'const + * char *' ptr with a "reason" why the reset happened (for messages). + */ +static +void __i2400m_dev_reset_handle(struct work_struct *ws) +{ + struct i2400m *i2400m = container_of(ws, struct i2400m, reset_ws); + const char *reason = i2400m->reset_reason; + struct device *dev = i2400m_dev(i2400m); + struct i2400m_reset_ctx *ctx = i2400m->reset_ctx; + int result; + + d_fnstart(3, dev, "(ws %p i2400m %p reason %s)\n", ws, i2400m, reason); + + i2400m->boot_mode = 1; + wmb(); /* Make sure i2400m_msg_to_dev() sees boot_mode */ + + result = 0; + if (mutex_trylock(&i2400m->init_mutex) == 0) { + /* We are still in i2400m_dev_start() [let it fail] or + * i2400m_dev_stop() [we are shutting down anyway, so + * ignore it] or we are resetting somewhere else. */ + dev_err(dev, "device rebooted somewhere else?\n"); + i2400m_msg_to_dev_cancel_wait(i2400m, -EL3RST); + complete(&i2400m->msg_completion); + goto out; + } + + dev_err(dev, "%s: reinitializing driver\n", reason); + rmb(); + if (i2400m->updown) { + __i2400m_dev_stop(i2400m); + i2400m->updown = 0; + wmb(); /* see i2400m->updown's documentation */ + } + + if (i2400m->alive) { + result = __i2400m_dev_start(i2400m, + I2400M_BRI_SOFT | I2400M_BRI_MAC_REINIT); + if (result < 0) { + dev_err(dev, "%s: cannot start the device: %d\n", + reason, result); + result = -EUCLEAN; + if (atomic_read(&i2400m->bus_reset_retries) + >= I2400M_BUS_RESET_RETRIES) { + result = -ENODEV; + dev_err(dev, "tried too many times to " + "reset the device, giving up\n"); + } + } + } + + if (i2400m->reset_ctx) { + ctx->result = result; + complete(&ctx->completion); + } + mutex_unlock(&i2400m->init_mutex); + if (result == -EUCLEAN) { + /* + * We come here because the reset during operational mode + * wasn't successfully done and need to proceed to a bus + * reset. For the dev_reset_handle() to be able to handle + * the reset event later properly, we restore boot_mode back + * to the state before previous reset. ie: just like we are + * issuing the bus reset for the first time + */ + i2400m->boot_mode = 0; + wmb(); + + atomic_inc(&i2400m->bus_reset_retries); + /* ops, need to clean up [w/ init_mutex not held] */ + result = i2400m_reset(i2400m, I2400M_RT_BUS); + if (result >= 0) + result = -ENODEV; + } else { + rmb(); + if (i2400m->alive) { + /* great, we expect the device state up and + * dev_start() actually brings the device state up */ + i2400m->updown = 1; + wmb(); + atomic_set(&i2400m->bus_reset_retries, 0); + } + } +out: + d_fnend(3, dev, "(ws %p i2400m %p reason %s) = void\n", + ws, i2400m, reason); +} + + +/** + * i2400m_dev_reset_handle - Handle a device's reset in a thread context + * + * Schedule a device reset handling out on a thread context, so it + * is safe to call from atomic context. We can't use the i2400m's + * queue as we are going to destroy it and reinitialize it as part of + * the driver bringup/bringup process. + * + * See __i2400m_dev_reset_handle() for details; that takes care of + * reinitializing the driver to handle the reset, calling into the + * bus-specific functions ops as needed. + */ +int i2400m_dev_reset_handle(struct i2400m *i2400m, const char *reason) +{ + i2400m->reset_reason = reason; + return schedule_work(&i2400m->reset_ws); +} +EXPORT_SYMBOL_GPL(i2400m_dev_reset_handle); + + + /* + * The actual work of error recovery. + * + * The current implementation of error recovery is to trigger a bus reset. + */ +static +void __i2400m_error_recovery(struct work_struct *ws) +{ + struct i2400m *i2400m = container_of(ws, struct i2400m, recovery_ws); + + i2400m_reset(i2400m, I2400M_RT_BUS); +} + +/* + * Schedule a work struct for error recovery. + * + * The intention of error recovery is to bring back the device to some + * known state whenever TX sees -110 (-ETIMEOUT) on copying the data to + * the device. The TX failure could mean a device bus stuck, so the current + * error recovery implementation is to trigger a bus reset to the device + * and hopefully it can bring back the device. + * + * The actual work of error recovery has to be in a thread context because + * it is kicked off in the TX thread (i2400ms->tx_workqueue) which is to be + * destroyed by the error recovery mechanism (currently a bus reset). + * + * Also, there may be already a queue of TX works that all hit + * the -ETIMEOUT error condition because the device is stuck already. + * Since bus reset is used as the error recovery mechanism and we don't + * want consecutive bus resets simply because the multiple TX works + * in the queue all hit the same device erratum, the flag "error_recovery" + * is introduced for preventing unwanted consecutive bus resets. + * + * Error recovery shall only be invoked again if previous one was completed. + * The flag error_recovery is set when error recovery mechanism is scheduled, + * and is checked when we need to schedule another error recovery. If it is + * in place already, then we shouldn't schedule another one. + */ +void i2400m_error_recovery(struct i2400m *i2400m) +{ + if (atomic_add_return(1, &i2400m->error_recovery) == 1) + schedule_work(&i2400m->recovery_ws); + else + atomic_dec(&i2400m->error_recovery); +} +EXPORT_SYMBOL_GPL(i2400m_error_recovery); + +/* + * Alloc the command and ack buffers for boot mode + * + * Get the buffers needed to deal with boot mode messages. + */ +static +int i2400m_bm_buf_alloc(struct i2400m *i2400m) +{ + i2400m->bm_cmd_buf = kzalloc(I2400M_BM_CMD_BUF_SIZE, GFP_KERNEL); + if (i2400m->bm_cmd_buf == NULL) + goto error_bm_cmd_kzalloc; + i2400m->bm_ack_buf = kzalloc(I2400M_BM_ACK_BUF_SIZE, GFP_KERNEL); + if (i2400m->bm_ack_buf == NULL) + goto error_bm_ack_buf_kzalloc; + return 0; + +error_bm_ack_buf_kzalloc: + kfree(i2400m->bm_cmd_buf); +error_bm_cmd_kzalloc: + return -ENOMEM; +} + + +/* + * Free boot mode command and ack buffers. + */ +static +void i2400m_bm_buf_free(struct i2400m *i2400m) +{ + kfree(i2400m->bm_ack_buf); + kfree(i2400m->bm_cmd_buf); +} + + +/** + * i2400m_init - Initialize a 'struct i2400m' from all zeroes + * + * This is a bus-generic API call. + */ +void i2400m_init(struct i2400m *i2400m) +{ + wimax_dev_init(&i2400m->wimax_dev); + + i2400m->boot_mode = 1; + i2400m->rx_reorder = 1; + init_waitqueue_head(&i2400m->state_wq); + + spin_lock_init(&i2400m->tx_lock); + i2400m->tx_pl_min = UINT_MAX; + i2400m->tx_size_min = UINT_MAX; + + spin_lock_init(&i2400m->rx_lock); + i2400m->rx_pl_min = UINT_MAX; + i2400m->rx_size_min = UINT_MAX; + INIT_LIST_HEAD(&i2400m->rx_reports); + INIT_WORK(&i2400m->rx_report_ws, i2400m_report_hook_work); + + mutex_init(&i2400m->msg_mutex); + init_completion(&i2400m->msg_completion); + + mutex_init(&i2400m->init_mutex); + /* wake_tx_ws is initialized in i2400m_tx_setup() */ + + INIT_WORK(&i2400m->reset_ws, __i2400m_dev_reset_handle); + INIT_WORK(&i2400m->recovery_ws, __i2400m_error_recovery); + + atomic_set(&i2400m->bus_reset_retries, 0); + + i2400m->alive = 0; + + /* initialize error_recovery to 1 for denoting we + * are not yet ready to take any error recovery */ + atomic_set(&i2400m->error_recovery, 1); +} +EXPORT_SYMBOL_GPL(i2400m_init); + + +int i2400m_reset(struct i2400m *i2400m, enum i2400m_reset_type rt) +{ + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + + /* + * Make sure we stop TXs and down the carrier before + * resetting; this is needed to avoid things like + * i2400m_wake_tx() scheduling stuff in parallel. + */ + if (net_dev->reg_state == NETREG_REGISTERED) { + netif_tx_disable(net_dev); + netif_carrier_off(net_dev); + } + return i2400m->bus_reset(i2400m, rt); +} +EXPORT_SYMBOL_GPL(i2400m_reset); + + +/** + * i2400m_setup - bus-generic setup function for the i2400m device + * + * @i2400m: device descriptor (bus-specific parts have been initialized) + * + * Returns: 0 if ok, < 0 errno code on error. + * + * Sets up basic device comunication infrastructure, boots the ROM to + * read the MAC address, registers with the WiMAX and network stacks + * and then brings up the device. + */ +int i2400m_setup(struct i2400m *i2400m, enum i2400m_bri bm_flags) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + + snprintf(wimax_dev->name, sizeof(wimax_dev->name), + "i2400m-%s:%s", dev->bus->name, dev_name(dev)); + + result = i2400m_bm_buf_alloc(i2400m); + if (result < 0) { + dev_err(dev, "cannot allocate bootmode scratch buffers\n"); + goto error_bm_buf_alloc; + } + + if (i2400m->bus_setup) { + result = i2400m->bus_setup(i2400m); + if (result < 0) { + dev_err(dev, "bus-specific setup failed: %d\n", + result); + goto error_bus_setup; + } + } + + result = i2400m_bootrom_init(i2400m, bm_flags); + if (result < 0) { + dev_err(dev, "read mac addr: bootrom init " + "failed: %d\n", result); + goto error_bootrom_init; + } + result = i2400m_read_mac_addr(i2400m); + if (result < 0) + goto error_read_mac_addr; + eth_random_addr(i2400m->src_mac_addr); + + i2400m->pm_notifier.notifier_call = i2400m_pm_notifier; + register_pm_notifier(&i2400m->pm_notifier); + + result = register_netdev(net_dev); /* Okey dokey, bring it up */ + if (result < 0) { + dev_err(dev, "cannot register i2400m network device: %d\n", + result); + goto error_register_netdev; + } + netif_carrier_off(net_dev); + + i2400m->wimax_dev.op_msg_from_user = i2400m_op_msg_from_user; + i2400m->wimax_dev.op_rfkill_sw_toggle = i2400m_op_rfkill_sw_toggle; + i2400m->wimax_dev.op_reset = i2400m_op_reset; + + result = wimax_dev_add(&i2400m->wimax_dev, net_dev); + if (result < 0) + goto error_wimax_dev_add; + + /* Now setup all that requires a registered net and wimax device. */ + result = sysfs_create_group(&net_dev->dev.kobj, &i2400m_dev_attr_group); + if (result < 0) { + dev_err(dev, "cannot setup i2400m's sysfs: %d\n", result); + goto error_sysfs_setup; + } + + i2400m_debugfs_add(i2400m); + + result = i2400m_dev_start(i2400m, bm_flags); + if (result < 0) + goto error_dev_start; + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; + +error_dev_start: + i2400m_debugfs_rm(i2400m); + sysfs_remove_group(&i2400m->wimax_dev.net_dev->dev.kobj, + &i2400m_dev_attr_group); +error_sysfs_setup: + wimax_dev_rm(&i2400m->wimax_dev); +error_wimax_dev_add: + unregister_netdev(net_dev); +error_register_netdev: + unregister_pm_notifier(&i2400m->pm_notifier); +error_read_mac_addr: +error_bootrom_init: + if (i2400m->bus_release) + i2400m->bus_release(i2400m); +error_bus_setup: + i2400m_bm_buf_free(i2400m); +error_bm_buf_alloc: + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; +} +EXPORT_SYMBOL_GPL(i2400m_setup); + + +/** + * i2400m_release - release the bus-generic driver resources + * + * Sends a disconnect message and undoes any setup done by i2400m_setup() + */ +void i2400m_release(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + netif_stop_queue(i2400m->wimax_dev.net_dev); + + i2400m_dev_stop(i2400m); + + cancel_work_sync(&i2400m->reset_ws); + cancel_work_sync(&i2400m->recovery_ws); + + i2400m_debugfs_rm(i2400m); + sysfs_remove_group(&i2400m->wimax_dev.net_dev->dev.kobj, + &i2400m_dev_attr_group); + wimax_dev_rm(&i2400m->wimax_dev); + unregister_netdev(i2400m->wimax_dev.net_dev); + unregister_pm_notifier(&i2400m->pm_notifier); + if (i2400m->bus_release) + i2400m->bus_release(i2400m); + i2400m_bm_buf_free(i2400m); + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} +EXPORT_SYMBOL_GPL(i2400m_release); + + +/* + * Debug levels control; see debug.h + */ +struct d_level D_LEVEL[] = { + D_SUBMODULE_DEFINE(control), + D_SUBMODULE_DEFINE(driver), + D_SUBMODULE_DEFINE(debugfs), + D_SUBMODULE_DEFINE(fw), + D_SUBMODULE_DEFINE(netdev), + D_SUBMODULE_DEFINE(rfkill), + D_SUBMODULE_DEFINE(rx), + D_SUBMODULE_DEFINE(sysfs), + D_SUBMODULE_DEFINE(tx), +}; +size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); + + +static +int __init i2400m_driver_init(void) +{ + d_parse_params(D_LEVEL, D_LEVEL_SIZE, i2400m_debug_params, + "i2400m.debug"); + return i2400m_barker_db_init(i2400m_barkers_params); +} +module_init(i2400m_driver_init); + +static +void __exit i2400m_driver_exit(void) +{ + i2400m_barker_db_exit(); +} +module_exit(i2400m_driver_exit); + +MODULE_AUTHOR("Intel Corporation "); +MODULE_DESCRIPTION("Intel 2400M WiMAX networking bus-generic driver"); +MODULE_LICENSE("GPL"); diff --git a/drivers/staging/wimax/i2400m/fw.c b/drivers/staging/wimax/i2400m/fw.c new file mode 100644 index 000000000000..6c9a41bff2e0 --- /dev/null +++ b/drivers/staging/wimax/i2400m/fw.c @@ -0,0 +1,1653 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * Firmware uploader + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * Inaky Perez-Gonzalez + * - Initial implementation + * + * + * THE PROCEDURE + * + * The 2400m and derived devices work in two modes: boot-mode or + * normal mode. In boot mode we can execute only a handful of commands + * targeted at uploading the firmware and launching it. + * + * The 2400m enters boot mode when it is first connected to the + * system, when it crashes and when you ask it to reboot. There are + * two submodes of the boot mode: signed and non-signed. Signed takes + * firmwares signed with a certain private key, non-signed takes any + * firmware. Normal hardware takes only signed firmware. + * + * On boot mode, in USB, we write to the device using the bulk out + * endpoint and read from it in the notification endpoint. + * + * Upon entrance to boot mode, the device sends (preceded with a few + * zero length packets (ZLPs) on the notification endpoint in USB) a + * reboot barker (4 le32 words with the same value). We ack it by + * sending the same barker to the device. The device acks with a + * reboot ack barker (4 le32 words with value I2400M_ACK_BARKER) and + * then is fully booted. At this point we can upload the firmware. + * + * Note that different iterations of the device and EEPROM + * configurations will send different [re]boot barkers; these are + * collected in i2400m_barker_db along with the firmware + * characteristics they require. + * + * This process is accomplished by the i2400m_bootrom_init() + * function. All the device interaction happens through the + * i2400m_bm_cmd() [boot mode command]. Special return values will + * indicate if the device did reset during the process. + * + * After this, we read the MAC address and then (if needed) + * reinitialize the device. We need to read it ahead of time because + * in the future, we might not upload the firmware until userspace + * 'ifconfig up's the device. + * + * We can then upload the firmware file. The file is composed of a BCF + * header (basic data, keys and signatures) and a list of write + * commands and payloads. Optionally more BCF headers might follow the + * main payload. We first upload the header [i2400m_dnload_init()] and + * then pass the commands and payloads verbatim to the i2400m_bm_cmd() + * function [i2400m_dnload_bcf()]. Then we tell the device to jump to + * the new firmware [i2400m_dnload_finalize()]. + * + * Once firmware is uploaded, we are good to go :) + * + * When we don't know in which mode we are, we first try by sending a + * warm reset request that will take us to boot-mode. If we time out + * waiting for a reboot barker, that means maybe we are already in + * boot mode, so we send a reboot barker. + * + * COMMAND EXECUTION + * + * This code (and process) is single threaded; for executing commands, + * we post a URB to the notification endpoint, post the command, wait + * for data on the notification buffer. We don't need to worry about + * others as we know we are the only ones in there. + * + * BACKEND IMPLEMENTATION + * + * This code is bus-generic; the bus-specific driver provides back end + * implementations to send a boot mode command to the device and to + * read an acknolwedgement from it (or an asynchronous notification) + * from it. + * + * FIRMWARE LOADING + * + * Note that in some cases, we can't just load a firmware file (for + * example, when resuming). For that, we might cache the firmware + * file. Thus, when doing the bootstrap, if there is a cache firmware + * file, it is used; if not, loading from disk is attempted. + * + * ROADMAP + * + * i2400m_barker_db_init Called by i2400m_driver_init() + * i2400m_barker_db_add + * + * i2400m_barker_db_exit Called by i2400m_driver_exit() + * + * i2400m_dev_bootstrap Called by __i2400m_dev_start() + * request_firmware + * i2400m_fw_bootstrap + * i2400m_fw_check + * i2400m_fw_hdr_check + * i2400m_fw_dnload + * release_firmware + * + * i2400m_fw_dnload + * i2400m_bootrom_init + * i2400m_bm_cmd + * i2400m_reset + * i2400m_dnload_init + * i2400m_dnload_init_signed + * i2400m_dnload_init_nonsigned + * i2400m_download_chunk + * i2400m_bm_cmd + * i2400m_dnload_bcf + * i2400m_bm_cmd + * i2400m_dnload_finalize + * i2400m_bm_cmd + * + * i2400m_bm_cmd + * i2400m->bus_bm_cmd_send() + * i2400m->bus_bm_wait_for_ack + * __i2400m_bm_ack_verify + * i2400m_is_boot_barker + * + * i2400m_bm_cmd_prepare Used by bus-drivers to prep + * commands before sending + * + * i2400m_pm_notifier Called on Power Management events + * i2400m_fw_cache + * i2400m_fw_uncache + */ +#include +#include +#include +#include +#include +#include "i2400m.h" + + +#define D_SUBMODULE fw +#include "debug-levels.h" + + +static const __le32 i2400m_ACK_BARKER[4] = { + cpu_to_le32(I2400M_ACK_BARKER), + cpu_to_le32(I2400M_ACK_BARKER), + cpu_to_le32(I2400M_ACK_BARKER), + cpu_to_le32(I2400M_ACK_BARKER) +}; + + +/** + * Prepare a boot-mode command for delivery + * + * @cmd: pointer to bootrom header to prepare + * + * Computes checksum if so needed. After calling this function, DO NOT + * modify the command or header as the checksum won't work anymore. + * + * We do it from here because some times we cannot do it in the + * original context the command was sent (it is a const), so when we + * copy it to our staging buffer, we add the checksum there. + */ +void i2400m_bm_cmd_prepare(struct i2400m_bootrom_header *cmd) +{ + if (i2400m_brh_get_use_checksum(cmd)) { + int i; + u32 checksum = 0; + const u32 *checksum_ptr = (void *) cmd->payload; + for (i = 0; i < cmd->data_size / 4; i++) + checksum += cpu_to_le32(*checksum_ptr++); + checksum += cmd->command + cmd->target_addr + cmd->data_size; + cmd->block_checksum = cpu_to_le32(checksum); + } +} +EXPORT_SYMBOL_GPL(i2400m_bm_cmd_prepare); + + +/* + * Database of known barkers. + * + * A barker is what the device sends indicating he is ready to be + * bootloaded. Different versions of the device will send different + * barkers. Depending on the barker, it might mean the device wants + * some kind of firmware or the other. + */ +static struct i2400m_barker_db { + __le32 data[4]; +} *i2400m_barker_db; +static size_t i2400m_barker_db_used, i2400m_barker_db_size; + + +static +int i2400m_zrealloc_2x(void **ptr, size_t *_count, size_t el_size, + gfp_t gfp_flags) +{ + size_t old_count = *_count, + new_count = old_count ? 2 * old_count : 2, + old_size = el_size * old_count, + new_size = el_size * new_count; + void *nptr = krealloc(*ptr, new_size, gfp_flags); + if (nptr) { + /* zero the other half or the whole thing if old_count + * was zero */ + if (old_size == 0) + memset(nptr, 0, new_size); + else + memset(nptr + old_size, 0, old_size); + *_count = new_count; + *ptr = nptr; + return 0; + } else + return -ENOMEM; +} + + +/* + * Add a barker to the database + * + * This cannot used outside of this module and only at at module_init + * time. This is to avoid the need to do locking. + */ +static +int i2400m_barker_db_add(u32 barker_id) +{ + int result; + + struct i2400m_barker_db *barker; + if (i2400m_barker_db_used >= i2400m_barker_db_size) { + result = i2400m_zrealloc_2x( + (void **) &i2400m_barker_db, &i2400m_barker_db_size, + sizeof(i2400m_barker_db[0]), GFP_KERNEL); + if (result < 0) + return result; + } + barker = i2400m_barker_db + i2400m_barker_db_used++; + barker->data[0] = le32_to_cpu(barker_id); + barker->data[1] = le32_to_cpu(barker_id); + barker->data[2] = le32_to_cpu(barker_id); + barker->data[3] = le32_to_cpu(barker_id); + return 0; +} + + +void i2400m_barker_db_exit(void) +{ + kfree(i2400m_barker_db); + i2400m_barker_db = NULL; + i2400m_barker_db_size = 0; + i2400m_barker_db_used = 0; +} + + +/* + * Helper function to add all the known stable barkers to the barker + * database. + */ +static +int i2400m_barker_db_known_barkers(void) +{ + int result; + + result = i2400m_barker_db_add(I2400M_NBOOT_BARKER); + if (result < 0) + goto error_add; + result = i2400m_barker_db_add(I2400M_SBOOT_BARKER); + if (result < 0) + goto error_add; + result = i2400m_barker_db_add(I2400M_SBOOT_BARKER_6050); + if (result < 0) + goto error_add; +error_add: + return result; +} + + +/* + * Initialize the barker database + * + * This can only be used from the module_init function for this + * module; this is to avoid the need to do locking. + * + * @options: command line argument with extra barkers to + * recognize. This is a comma-separated list of 32-bit hex + * numbers. They are appended to the existing list. Setting 0 + * cleans the existing list and starts a new one. + */ +int i2400m_barker_db_init(const char *_options) +{ + int result; + char *options = NULL, *options_orig, *token; + + i2400m_barker_db = NULL; + i2400m_barker_db_size = 0; + i2400m_barker_db_used = 0; + + result = i2400m_barker_db_known_barkers(); + if (result < 0) + goto error_add; + /* parse command line options from i2400m.barkers */ + if (_options != NULL) { + unsigned barker; + + options_orig = kstrdup(_options, GFP_KERNEL); + if (options_orig == NULL) { + result = -ENOMEM; + goto error_parse; + } + options = options_orig; + + while ((token = strsep(&options, ",")) != NULL) { + if (*token == '\0') /* eat joint commas */ + continue; + if (sscanf(token, "%x", &barker) != 1 + || barker > 0xffffffff) { + printk(KERN_ERR "%s: can't recognize " + "i2400m.barkers value '%s' as " + "a 32-bit number\n", + __func__, token); + result = -EINVAL; + goto error_parse; + } + if (barker == 0) { + /* clean list and start new */ + i2400m_barker_db_exit(); + continue; + } + result = i2400m_barker_db_add(barker); + if (result < 0) + goto error_parse_add; + } + kfree(options_orig); + } + return 0; + +error_parse_add: +error_parse: + kfree(options_orig); +error_add: + kfree(i2400m_barker_db); + return result; +} + + +/* + * Recognize a boot barker + * + * @buf: buffer where the boot barker. + * @buf_size: size of the buffer (has to be 16 bytes). It is passed + * here so the function can check it for the caller. + * + * Note that as a side effect, upon identifying the obtained boot + * barker, this function will set i2400m->barker to point to the right + * barker database entry. Subsequent calls to the function will result + * in verifying that the same type of boot barker is returned when the + * device [re]boots (as long as the same device instance is used). + * + * Return: 0 if @buf matches a known boot barker. -ENOENT if the + * buffer in @buf doesn't match any boot barker in the database or + * -EILSEQ if the buffer doesn't have the right size. + */ +int i2400m_is_boot_barker(struct i2400m *i2400m, + const void *buf, size_t buf_size) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct i2400m_barker_db *barker; + int i; + + result = -ENOENT; + if (buf_size != sizeof(i2400m_barker_db[i].data)) + return result; + + /* Short circuit if we have already discovered the barker + * associated with the device. */ + if (i2400m->barker && + !memcmp(buf, i2400m->barker, sizeof(i2400m->barker->data))) + return 0; + + for (i = 0; i < i2400m_barker_db_used; i++) { + barker = &i2400m_barker_db[i]; + BUILD_BUG_ON(sizeof(barker->data) != 16); + if (memcmp(buf, barker->data, sizeof(barker->data))) + continue; + + if (i2400m->barker == NULL) { + i2400m->barker = barker; + d_printf(1, dev, "boot barker set to #%u/%08x\n", + i, le32_to_cpu(barker->data[0])); + if (barker->data[0] == le32_to_cpu(I2400M_NBOOT_BARKER)) + i2400m->sboot = 0; + else + i2400m->sboot = 1; + } else if (i2400m->barker != barker) { + dev_err(dev, "HW inconsistency: device " + "reports a different boot barker " + "than set (from %08x to %08x)\n", + le32_to_cpu(i2400m->barker->data[0]), + le32_to_cpu(barker->data[0])); + result = -EIO; + } else + d_printf(2, dev, "boot barker confirmed #%u/%08x\n", + i, le32_to_cpu(barker->data[0])); + result = 0; + break; + } + return result; +} +EXPORT_SYMBOL_GPL(i2400m_is_boot_barker); + + +/* + * Verify the ack data received + * + * Given a reply to a boot mode command, chew it and verify everything + * is ok. + * + * @opcode: opcode which generated this ack. For error messages. + * @ack: pointer to ack data we received + * @ack_size: size of that data buffer + * @flags: I2400M_BM_CMD_* flags we called the command with. + * + * Way too long function -- maybe it should be further split + */ +static +ssize_t __i2400m_bm_ack_verify(struct i2400m *i2400m, int opcode, + struct i2400m_bootrom_header *ack, + size_t ack_size, int flags) +{ + ssize_t result = -ENOMEM; + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(8, dev, "(i2400m %p opcode %d ack %p size %zu)\n", + i2400m, opcode, ack, ack_size); + if (ack_size < sizeof(*ack)) { + result = -EIO; + dev_err(dev, "boot-mode cmd %d: HW BUG? notification didn't " + "return enough data (%zu bytes vs %zu expected)\n", + opcode, ack_size, sizeof(*ack)); + goto error_ack_short; + } + result = i2400m_is_boot_barker(i2400m, ack, ack_size); + if (result >= 0) { + result = -ERESTARTSYS; + d_printf(6, dev, "boot-mode cmd %d: HW boot barker\n", opcode); + goto error_reboot; + } + if (ack_size == sizeof(i2400m_ACK_BARKER) + && memcmp(ack, i2400m_ACK_BARKER, sizeof(*ack)) == 0) { + result = -EISCONN; + d_printf(3, dev, "boot-mode cmd %d: HW reboot ack barker\n", + opcode); + goto error_reboot_ack; + } + result = 0; + if (flags & I2400M_BM_CMD_RAW) + goto out_raw; + ack->data_size = le32_to_cpu(ack->data_size); + ack->target_addr = le32_to_cpu(ack->target_addr); + ack->block_checksum = le32_to_cpu(ack->block_checksum); + d_printf(5, dev, "boot-mode cmd %d: notification for opcode %u " + "response %u csum %u rr %u da %u\n", + opcode, i2400m_brh_get_opcode(ack), + i2400m_brh_get_response(ack), + i2400m_brh_get_use_checksum(ack), + i2400m_brh_get_response_required(ack), + i2400m_brh_get_direct_access(ack)); + result = -EIO; + if (i2400m_brh_get_signature(ack) != 0xcbbc) { + dev_err(dev, "boot-mode cmd %d: HW BUG? wrong signature " + "0x%04x\n", opcode, i2400m_brh_get_signature(ack)); + goto error_ack_signature; + } + if (opcode != -1 && opcode != i2400m_brh_get_opcode(ack)) { + dev_err(dev, "boot-mode cmd %d: HW BUG? " + "received response for opcode %u, expected %u\n", + opcode, i2400m_brh_get_opcode(ack), opcode); + goto error_ack_opcode; + } + if (i2400m_brh_get_response(ack) != 0) { /* failed? */ + dev_err(dev, "boot-mode cmd %d: error; hw response %u\n", + opcode, i2400m_brh_get_response(ack)); + goto error_ack_failed; + } + if (ack_size < ack->data_size + sizeof(*ack)) { + dev_err(dev, "boot-mode cmd %d: SW BUG " + "driver provided only %zu bytes for %zu bytes " + "of data\n", opcode, ack_size, + (size_t) le32_to_cpu(ack->data_size) + sizeof(*ack)); + goto error_ack_short_buffer; + } + result = ack_size; + /* Don't you love this stack of empty targets? Well, I don't + * either, but it helps track exactly who comes in here and + * why :) */ +error_ack_short_buffer: +error_ack_failed: +error_ack_opcode: +error_ack_signature: +out_raw: +error_reboot_ack: +error_reboot: +error_ack_short: + d_fnend(8, dev, "(i2400m %p opcode %d ack %p size %zu) = %d\n", + i2400m, opcode, ack, ack_size, (int) result); + return result; +} + + +/** + * i2400m_bm_cmd - Execute a boot mode command + * + * @cmd: buffer containing the command data (pointing at the header). + * This data can be ANYWHERE (for USB, we will copy it to an + * specific buffer). Make sure everything is in proper little + * endian. + * + * A raw buffer can be also sent, just cast it and set flags to + * I2400M_BM_CMD_RAW. + * + * This function will generate a checksum for you if the + * checksum bit in the command is set (unless I2400M_BM_CMD_RAW + * is set). + * + * You can use the i2400m->bm_cmd_buf to stage your commands and + * send them. + * + * If NULL, no command is sent (we just wait for an ack). + * + * @cmd_size: size of the command. Will be auto padded to the + * bus-specific drivers padding requirements. + * + * @ack: buffer where to place the acknowledgement. If it is a regular + * command response, all fields will be returned with the right, + * native endianess. + * + * You *cannot* use i2400m->bm_ack_buf for this buffer. + * + * @ack_size: size of @ack, 16 aligned; you need to provide at least + * sizeof(*ack) bytes and then enough to contain the return data + * from the command + * + * @flags: see I2400M_BM_CMD_* above. + * + * @returns: bytes received by the notification; if < 0, an errno code + * denoting an error or: + * + * -ERESTARTSYS The device has rebooted + * + * Executes a boot-mode command and waits for a response, doing basic + * validation on it; if a zero length response is received, it retries + * waiting for a response until a non-zero one is received (timing out + * after %I2400M_BOOT_RETRIES retries). + */ +static +ssize_t i2400m_bm_cmd(struct i2400m *i2400m, + const struct i2400m_bootrom_header *cmd, size_t cmd_size, + struct i2400m_bootrom_header *ack, size_t ack_size, + int flags) +{ + ssize_t result = -ENOMEM, rx_bytes; + struct device *dev = i2400m_dev(i2400m); + int opcode = cmd == NULL ? -1 : i2400m_brh_get_opcode(cmd); + + d_fnstart(6, dev, "(i2400m %p cmd %p size %zu ack %p size %zu)\n", + i2400m, cmd, cmd_size, ack, ack_size); + BUG_ON(ack_size < sizeof(*ack)); + BUG_ON(i2400m->boot_mode == 0); + + if (cmd != NULL) { /* send the command */ + result = i2400m->bus_bm_cmd_send(i2400m, cmd, cmd_size, flags); + if (result < 0) + goto error_cmd_send; + if ((flags & I2400M_BM_CMD_RAW) == 0) + d_printf(5, dev, + "boot-mode cmd %d csum %u rr %u da %u: " + "addr 0x%04x size %u block csum 0x%04x\n", + opcode, i2400m_brh_get_use_checksum(cmd), + i2400m_brh_get_response_required(cmd), + i2400m_brh_get_direct_access(cmd), + cmd->target_addr, cmd->data_size, + cmd->block_checksum); + } + result = i2400m->bus_bm_wait_for_ack(i2400m, ack, ack_size); + if (result < 0) { + dev_err(dev, "boot-mode cmd %d: error waiting for an ack: %d\n", + opcode, (int) result); /* bah, %zd doesn't work */ + goto error_wait_for_ack; + } + rx_bytes = result; + /* verify the ack and read more if necessary [result is the + * final amount of bytes we get in the ack] */ + result = __i2400m_bm_ack_verify(i2400m, opcode, ack, ack_size, flags); + if (result < 0) + goto error_bad_ack; + /* Don't you love this stack of empty targets? Well, I don't + * either, but it helps track exactly who comes in here and + * why :) */ + result = rx_bytes; +error_bad_ack: +error_wait_for_ack: +error_cmd_send: + d_fnend(6, dev, "(i2400m %p cmd %p size %zu ack %p size %zu) = %d\n", + i2400m, cmd, cmd_size, ack, ack_size, (int) result); + return result; +} + + +/** + * i2400m_download_chunk - write a single chunk of data to the device's memory + * + * @i2400m: device descriptor + * @buf: the buffer to write + * @buf_len: length of the buffer to write + * @addr: address in the device memory space + * @direct: bootrom write mode + * @do_csum: should a checksum validation be performed + */ +static int i2400m_download_chunk(struct i2400m *i2400m, const void *chunk, + size_t __chunk_len, unsigned long addr, + unsigned int direct, unsigned int do_csum) +{ + int ret; + size_t chunk_len = ALIGN(__chunk_len, I2400M_PL_ALIGN); + struct device *dev = i2400m_dev(i2400m); + struct { + struct i2400m_bootrom_header cmd; + u8 cmd_payload[]; + } __packed *buf; + struct i2400m_bootrom_header ack; + + d_fnstart(5, dev, "(i2400m %p chunk %p __chunk_len %zu addr 0x%08lx " + "direct %u do_csum %u)\n", i2400m, chunk, __chunk_len, + addr, direct, do_csum); + buf = i2400m->bm_cmd_buf; + memcpy(buf->cmd_payload, chunk, __chunk_len); + memset(buf->cmd_payload + __chunk_len, 0xad, chunk_len - __chunk_len); + + buf->cmd.command = i2400m_brh_command(I2400M_BRH_WRITE, + __chunk_len & 0x3 ? 0 : do_csum, + __chunk_len & 0xf ? 0 : direct); + buf->cmd.target_addr = cpu_to_le32(addr); + buf->cmd.data_size = cpu_to_le32(__chunk_len); + ret = i2400m_bm_cmd(i2400m, &buf->cmd, sizeof(buf->cmd) + chunk_len, + &ack, sizeof(ack), 0); + if (ret >= 0) + ret = 0; + d_fnend(5, dev, "(i2400m %p chunk %p __chunk_len %zu addr 0x%08lx " + "direct %u do_csum %u) = %d\n", i2400m, chunk, __chunk_len, + addr, direct, do_csum, ret); + return ret; +} + + +/* + * Download a BCF file's sections to the device + * + * @i2400m: device descriptor + * @bcf: pointer to firmware data (first header followed by the + * payloads). Assumed verified and consistent. + * @bcf_len: length (in bytes) of the @bcf buffer. + * + * Returns: < 0 errno code on error or the offset to the jump instruction. + * + * Given a BCF file, downloads each section (a command and a payload) + * to the device's address space. Actually, it just executes each + * command i the BCF file. + * + * The section size has to be aligned to 4 bytes AND the padding has + * to be taken from the firmware file, as the signature takes it into + * account. + */ +static +ssize_t i2400m_dnload_bcf(struct i2400m *i2400m, + const struct i2400m_bcf_hdr *bcf, size_t bcf_len) +{ + ssize_t ret; + struct device *dev = i2400m_dev(i2400m); + size_t offset, /* iterator offset */ + data_size, /* Size of the data payload */ + section_size, /* Size of the whole section (cmd + payload) */ + section = 1; + const struct i2400m_bootrom_header *bh; + struct i2400m_bootrom_header ack; + + d_fnstart(3, dev, "(i2400m %p bcf %p bcf_len %zu)\n", + i2400m, bcf, bcf_len); + /* Iterate over the command blocks in the BCF file that start + * after the header */ + offset = le32_to_cpu(bcf->header_len) * sizeof(u32); + while (1) { /* start sending the file */ + bh = (void *) bcf + offset; + data_size = le32_to_cpu(bh->data_size); + section_size = ALIGN(sizeof(*bh) + data_size, 4); + d_printf(7, dev, + "downloading section #%zu (@%zu %zu B) to 0x%08x\n", + section, offset, sizeof(*bh) + data_size, + le32_to_cpu(bh->target_addr)); + /* + * We look for JUMP cmd from the bootmode header, + * either I2400M_BRH_SIGNED_JUMP for secure boot + * or I2400M_BRH_JUMP for unsecure boot, the last chunk + * should be the bootmode header with JUMP cmd. + */ + if (i2400m_brh_get_opcode(bh) == I2400M_BRH_SIGNED_JUMP || + i2400m_brh_get_opcode(bh) == I2400M_BRH_JUMP) { + d_printf(5, dev, "jump found @%zu\n", offset); + break; + } + if (offset + section_size > bcf_len) { + dev_err(dev, "fw %s: bad section #%zu, " + "end (@%zu) beyond EOF (@%zu)\n", + i2400m->fw_name, section, + offset + section_size, bcf_len); + ret = -EINVAL; + goto error_section_beyond_eof; + } + __i2400m_msleep(20); + ret = i2400m_bm_cmd(i2400m, bh, section_size, + &ack, sizeof(ack), I2400M_BM_CMD_RAW); + if (ret < 0) { + dev_err(dev, "fw %s: section #%zu (@%zu %zu B) " + "failed %d\n", i2400m->fw_name, section, + offset, sizeof(*bh) + data_size, (int) ret); + goto error_send; + } + offset += section_size; + section++; + } + ret = offset; +error_section_beyond_eof: +error_send: + d_fnend(3, dev, "(i2400m %p bcf %p bcf_len %zu) = %d\n", + i2400m, bcf, bcf_len, (int) ret); + return ret; +} + + +/* + * Indicate if the device emitted a reboot barker that indicates + * "signed boot" + */ +static +unsigned i2400m_boot_is_signed(struct i2400m *i2400m) +{ + return likely(i2400m->sboot); +} + + +/* + * Do the final steps of uploading firmware + * + * @bcf_hdr: BCF header we are actually using + * @bcf: pointer to the firmware image (which matches the first header + * that is followed by the actual payloads). + * @offset: [byte] offset into @bcf for the command we need to send. + * + * Depending on the boot mode (signed vs non-signed), different + * actions need to be taken. + */ +static +int i2400m_dnload_finalize(struct i2400m *i2400m, + const struct i2400m_bcf_hdr *bcf_hdr, + const struct i2400m_bcf_hdr *bcf, size_t offset) +{ + int ret = 0; + struct device *dev = i2400m_dev(i2400m); + struct i2400m_bootrom_header *cmd, ack; + struct { + struct i2400m_bootrom_header cmd; + u8 cmd_pl[0]; + } __packed *cmd_buf; + size_t signature_block_offset, signature_block_size; + + d_fnstart(3, dev, "offset %zu\n", offset); + cmd = (void *) bcf + offset; + if (i2400m_boot_is_signed(i2400m) == 0) { + struct i2400m_bootrom_header jump_ack; + d_printf(1, dev, "unsecure boot, jumping to 0x%08x\n", + le32_to_cpu(cmd->target_addr)); + cmd_buf = i2400m->bm_cmd_buf; + memcpy(&cmd_buf->cmd, cmd, sizeof(*cmd)); + cmd = &cmd_buf->cmd; + /* now cmd points to the actual bootrom_header in cmd_buf */ + i2400m_brh_set_opcode(cmd, I2400M_BRH_JUMP); + cmd->data_size = 0; + ret = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), + &jump_ack, sizeof(jump_ack), 0); + } else { + d_printf(1, dev, "secure boot, jumping to 0x%08x\n", + le32_to_cpu(cmd->target_addr)); + cmd_buf = i2400m->bm_cmd_buf; + memcpy(&cmd_buf->cmd, cmd, sizeof(*cmd)); + signature_block_offset = + sizeof(*bcf_hdr) + + le32_to_cpu(bcf_hdr->key_size) * sizeof(u32) + + le32_to_cpu(bcf_hdr->exponent_size) * sizeof(u32); + signature_block_size = + le32_to_cpu(bcf_hdr->modulus_size) * sizeof(u32); + memcpy(cmd_buf->cmd_pl, + (void *) bcf_hdr + signature_block_offset, + signature_block_size); + ret = i2400m_bm_cmd(i2400m, &cmd_buf->cmd, + sizeof(cmd_buf->cmd) + signature_block_size, + &ack, sizeof(ack), I2400M_BM_CMD_RAW); + } + d_fnend(3, dev, "returning %d\n", ret); + return ret; +} + + +/** + * i2400m_bootrom_init - Reboots a powered device into boot mode + * + * @i2400m: device descriptor + * @flags: + * I2400M_BRI_SOFT: a reboot barker has been seen + * already, so don't wait for it. + * + * I2400M_BRI_NO_REBOOT: Don't send a reboot command, but wait + * for a reboot barker notification. This is a one shot; if + * the state machine needs to send a reboot command it will. + * + * Returns: + * + * < 0 errno code on error, 0 if ok. + * + * Description: + * + * Tries hard enough to put the device in boot-mode. There are two + * main phases to this: + * + * a. (1) send a reboot command and (2) get a reboot barker + * + * b. (1) echo/ack the reboot sending the reboot barker back and (2) + * getting an ack barker in return + * + * We want to skip (a) in some cases [soft]. The state machine is + * horrible, but it is basically: on each phase, send what has to be + * sent (if any), wait for the answer and act on the answer. We might + * have to backtrack and retry, so we keep a max tries counter for + * that. + * + * It sucks because we don't know ahead of time which is going to be + * the reboot barker (the device might send different ones depending + * on its EEPROM config) and once the device reboots and waits for the + * echo/ack reboot barker being sent back, it doesn't understand + * anything else. So we can be left at the point where we don't know + * what to send to it -- cold reset and bus reset seem to have little + * effect. So the function iterates (in this case) through all the + * known barkers and tries them all until an ACK is + * received. Otherwise, it gives up. + * + * If we get a timeout after sending a warm reset, we do it again. + */ +int i2400m_bootrom_init(struct i2400m *i2400m, enum i2400m_bri flags) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct i2400m_bootrom_header *cmd; + struct i2400m_bootrom_header ack; + int count = i2400m->bus_bm_retries; + int ack_timeout_cnt = 1; + unsigned i; + + BUILD_BUG_ON(sizeof(*cmd) != sizeof(i2400m_barker_db[0].data)); + BUILD_BUG_ON(sizeof(ack) != sizeof(i2400m_ACK_BARKER)); + + d_fnstart(4, dev, "(i2400m %p flags 0x%08x)\n", i2400m, flags); + result = -ENOMEM; + cmd = i2400m->bm_cmd_buf; + if (flags & I2400M_BRI_SOFT) + goto do_reboot_ack; +do_reboot: + ack_timeout_cnt = 1; + if (--count < 0) + goto error_timeout; + d_printf(4, dev, "device reboot: reboot command [%d # left]\n", + count); + if ((flags & I2400M_BRI_NO_REBOOT) == 0) + i2400m_reset(i2400m, I2400M_RT_WARM); + result = i2400m_bm_cmd(i2400m, NULL, 0, &ack, sizeof(ack), + I2400M_BM_CMD_RAW); + flags &= ~I2400M_BRI_NO_REBOOT; + switch (result) { + case -ERESTARTSYS: + /* + * at this point, i2400m_bm_cmd(), through + * __i2400m_bm_ack_process(), has updated + * i2400m->barker and we are good to go. + */ + d_printf(4, dev, "device reboot: got reboot barker\n"); + break; + case -EISCONN: /* we don't know how it got here...but we follow it */ + d_printf(4, dev, "device reboot: got ack barker - whatever\n"); + goto do_reboot; + case -ETIMEDOUT: + /* + * Device has timed out, we might be in boot mode + * already and expecting an ack; if we don't know what + * the barker is, we just send them all. Cold reset + * and bus reset don't work. Beats me. + */ + if (i2400m->barker != NULL) { + dev_err(dev, "device boot: reboot barker timed out, " + "trying (set) %08x echo/ack\n", + le32_to_cpu(i2400m->barker->data[0])); + goto do_reboot_ack; + } + for (i = 0; i < i2400m_barker_db_used; i++) { + struct i2400m_barker_db *barker = &i2400m_barker_db[i]; + memcpy(cmd, barker->data, sizeof(barker->data)); + result = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), + &ack, sizeof(ack), + I2400M_BM_CMD_RAW); + if (result == -EISCONN) { + dev_warn(dev, "device boot: got ack barker " + "after sending echo/ack barker " + "#%d/%08x; rebooting j.i.c.\n", + i, le32_to_cpu(barker->data[0])); + flags &= ~I2400M_BRI_NO_REBOOT; + goto do_reboot; + } + } + dev_err(dev, "device boot: tried all the echo/acks, could " + "not get device to respond; giving up"); + result = -ESHUTDOWN; + case -EPROTO: + case -ESHUTDOWN: /* dev is gone */ + case -EINTR: /* user cancelled */ + goto error_dev_gone; + default: + dev_err(dev, "device reboot: error %d while waiting " + "for reboot barker - rebooting\n", result); + d_dump(1, dev, &ack, result); + goto do_reboot; + } + /* At this point we ack back with 4 REBOOT barkers and expect + * 4 ACK barkers. This is ugly, as we send a raw command -- + * hence the cast. _bm_cmd() will catch the reboot ack + * notification and report it as -EISCONN. */ +do_reboot_ack: + d_printf(4, dev, "device reboot ack: sending ack [%d # left]\n", count); + memcpy(cmd, i2400m->barker->data, sizeof(i2400m->barker->data)); + result = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), + &ack, sizeof(ack), I2400M_BM_CMD_RAW); + switch (result) { + case -ERESTARTSYS: + d_printf(4, dev, "reboot ack: got reboot barker - retrying\n"); + if (--count < 0) + goto error_timeout; + goto do_reboot_ack; + case -EISCONN: + d_printf(4, dev, "reboot ack: got ack barker - good\n"); + break; + case -ETIMEDOUT: /* no response, maybe it is the other type? */ + if (ack_timeout_cnt-- < 0) { + d_printf(4, dev, "reboot ack timedout: retrying\n"); + goto do_reboot_ack; + } else { + dev_err(dev, "reboot ack timedout too long: " + "trying reboot\n"); + goto do_reboot; + } + break; + case -EPROTO: + case -ESHUTDOWN: /* dev is gone */ + goto error_dev_gone; + default: + dev_err(dev, "device reboot ack: error %d while waiting for " + "reboot ack barker - rebooting\n", result); + goto do_reboot; + } + d_printf(2, dev, "device reboot ack: got ack barker - boot done\n"); + result = 0; +exit_timeout: +error_dev_gone: + d_fnend(4, dev, "(i2400m %p flags 0x%08x) = %d\n", + i2400m, flags, result); + return result; + +error_timeout: + dev_err(dev, "Timed out waiting for reboot ack\n"); + result = -ETIMEDOUT; + goto exit_timeout; +} + + +/* + * Read the MAC addr + * + * The position this function reads is fixed in device memory and + * always available, even without firmware. + * + * Note we specify we want to read only six bytes, but provide space + * for 16, as we always get it rounded up. + */ +int i2400m_read_mac_addr(struct i2400m *i2400m) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + struct i2400m_bootrom_header *cmd; + struct { + struct i2400m_bootrom_header ack; + u8 ack_pl[16]; + } __packed ack_buf; + + d_fnstart(5, dev, "(i2400m %p)\n", i2400m); + cmd = i2400m->bm_cmd_buf; + cmd->command = i2400m_brh_command(I2400M_BRH_READ, 0, 1); + cmd->target_addr = cpu_to_le32(0x00203fe8); + cmd->data_size = cpu_to_le32(6); + result = i2400m_bm_cmd(i2400m, cmd, sizeof(*cmd), + &ack_buf.ack, sizeof(ack_buf), 0); + if (result < 0) { + dev_err(dev, "BM: read mac addr failed: %d\n", result); + goto error_read_mac; + } + d_printf(2, dev, "mac addr is %pM\n", ack_buf.ack_pl); + if (i2400m->bus_bm_mac_addr_impaired == 1) { + ack_buf.ack_pl[0] = 0x00; + ack_buf.ack_pl[1] = 0x16; + ack_buf.ack_pl[2] = 0xd3; + get_random_bytes(&ack_buf.ack_pl[3], 3); + dev_err(dev, "BM is MAC addr impaired, faking MAC addr to " + "mac addr is %pM\n", ack_buf.ack_pl); + result = 0; + } + net_dev->addr_len = ETH_ALEN; + memcpy(net_dev->dev_addr, ack_buf.ack_pl, ETH_ALEN); +error_read_mac: + d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; +} + + +/* + * Initialize a non signed boot + * + * This implies sending some magic values to the device's memory. Note + * we convert the values to little endian in the same array + * declaration. + */ +static +int i2400m_dnload_init_nonsigned(struct i2400m *i2400m) +{ + unsigned i = 0; + int ret = 0; + struct device *dev = i2400m_dev(i2400m); + d_fnstart(5, dev, "(i2400m %p)\n", i2400m); + if (i2400m->bus_bm_pokes_table) { + while (i2400m->bus_bm_pokes_table[i].address) { + ret = i2400m_download_chunk( + i2400m, + &i2400m->bus_bm_pokes_table[i].data, + sizeof(i2400m->bus_bm_pokes_table[i].data), + i2400m->bus_bm_pokes_table[i].address, 1, 1); + if (ret < 0) + break; + i++; + } + } + d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, ret); + return ret; +} + + +/* + * Initialize the signed boot process + * + * @i2400m: device descriptor + * + * @bcf_hdr: pointer to the firmware header; assumes it is fully in + * memory (it has gone through basic validation). + * + * Returns: 0 if ok, < 0 errno code on error, -ERESTARTSYS if the hw + * rebooted. + * + * This writes the firmware BCF header to the device using the + * HASH_PAYLOAD_ONLY command. + */ +static +int i2400m_dnload_init_signed(struct i2400m *i2400m, + const struct i2400m_bcf_hdr *bcf_hdr) +{ + int ret; + struct device *dev = i2400m_dev(i2400m); + struct { + struct i2400m_bootrom_header cmd; + struct i2400m_bcf_hdr cmd_pl; + } __packed *cmd_buf; + struct i2400m_bootrom_header ack; + + d_fnstart(5, dev, "(i2400m %p bcf_hdr %p)\n", i2400m, bcf_hdr); + cmd_buf = i2400m->bm_cmd_buf; + cmd_buf->cmd.command = + i2400m_brh_command(I2400M_BRH_HASH_PAYLOAD_ONLY, 0, 0); + cmd_buf->cmd.target_addr = 0; + cmd_buf->cmd.data_size = cpu_to_le32(sizeof(cmd_buf->cmd_pl)); + memcpy(&cmd_buf->cmd_pl, bcf_hdr, sizeof(*bcf_hdr)); + ret = i2400m_bm_cmd(i2400m, &cmd_buf->cmd, sizeof(*cmd_buf), + &ack, sizeof(ack), 0); + if (ret >= 0) + ret = 0; + d_fnend(5, dev, "(i2400m %p bcf_hdr %p) = %d\n", i2400m, bcf_hdr, ret); + return ret; +} + + +/* + * Initialize the firmware download at the device size + * + * Multiplex to the one that matters based on the device's mode + * (signed or non-signed). + */ +static +int i2400m_dnload_init(struct i2400m *i2400m, + const struct i2400m_bcf_hdr *bcf_hdr) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + + if (i2400m_boot_is_signed(i2400m)) { + d_printf(1, dev, "signed boot\n"); + result = i2400m_dnload_init_signed(i2400m, bcf_hdr); + if (result == -ERESTARTSYS) + return result; + if (result < 0) + dev_err(dev, "firmware %s: signed boot download " + "initialization failed: %d\n", + i2400m->fw_name, result); + } else { + /* non-signed boot process without pokes */ + d_printf(1, dev, "non-signed boot\n"); + result = i2400m_dnload_init_nonsigned(i2400m); + if (result == -ERESTARTSYS) + return result; + if (result < 0) + dev_err(dev, "firmware %s: non-signed download " + "initialization failed: %d\n", + i2400m->fw_name, result); + } + return result; +} + + +/* + * Run consistency tests on the firmware file and load up headers + * + * Check for the firmware being made for the i2400m device, + * etc...These checks are mostly informative, as the device will make + * them too; but the driver's response is more informative on what + * went wrong. + * + * This will also look at all the headers present on the firmware + * file, and update i2400m->fw_bcf_hdr to point to them. + */ +static +int i2400m_fw_hdr_check(struct i2400m *i2400m, + const struct i2400m_bcf_hdr *bcf_hdr, + size_t index, size_t offset) +{ + struct device *dev = i2400m_dev(i2400m); + + unsigned module_type, header_len, major_version, minor_version, + module_id, module_vendor, date, size; + + module_type = le32_to_cpu(bcf_hdr->module_type); + header_len = sizeof(u32) * le32_to_cpu(bcf_hdr->header_len); + major_version = (le32_to_cpu(bcf_hdr->header_version) & 0xffff0000) + >> 16; + minor_version = le32_to_cpu(bcf_hdr->header_version) & 0x0000ffff; + module_id = le32_to_cpu(bcf_hdr->module_id); + module_vendor = le32_to_cpu(bcf_hdr->module_vendor); + date = le32_to_cpu(bcf_hdr->date); + size = sizeof(u32) * le32_to_cpu(bcf_hdr->size); + + d_printf(1, dev, "firmware %s #%zd@%08zx: BCF header " + "type:vendor:id 0x%x:%x:%x v%u.%u (%u/%u B) built %08x\n", + i2400m->fw_name, index, offset, + module_type, module_vendor, module_id, + major_version, minor_version, header_len, size, date); + + /* Hard errors */ + if (major_version != 1) { + dev_err(dev, "firmware %s #%zd@%08zx: major header version " + "v%u.%u not supported\n", + i2400m->fw_name, index, offset, + major_version, minor_version); + return -EBADF; + } + + if (module_type != 6) { /* built for the right hardware? */ + dev_err(dev, "firmware %s #%zd@%08zx: unexpected module " + "type 0x%x; aborting\n", + i2400m->fw_name, index, offset, + module_type); + return -EBADF; + } + + if (module_vendor != 0x8086) { + dev_err(dev, "firmware %s #%zd@%08zx: unexpected module " + "vendor 0x%x; aborting\n", + i2400m->fw_name, index, offset, module_vendor); + return -EBADF; + } + + if (date < 0x20080300) + dev_warn(dev, "firmware %s #%zd@%08zx: build date %08x " + "too old; unsupported\n", + i2400m->fw_name, index, offset, date); + return 0; +} + + +/* + * Run consistency tests on the firmware file and load up headers + * + * Check for the firmware being made for the i2400m device, + * etc...These checks are mostly informative, as the device will make + * them too; but the driver's response is more informative on what + * went wrong. + * + * This will also look at all the headers present on the firmware + * file, and update i2400m->fw_hdrs to point to them. + */ +static +int i2400m_fw_check(struct i2400m *i2400m, const void *bcf, size_t bcf_size) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + size_t headers = 0; + const struct i2400m_bcf_hdr *bcf_hdr; + const void *itr, *next, *top; + size_t slots = 0, used_slots = 0; + + for (itr = bcf, top = itr + bcf_size; + itr < top; + headers++, itr = next) { + size_t leftover, offset, header_len, size; + + leftover = top - itr; + offset = itr - bcf; + if (leftover <= sizeof(*bcf_hdr)) { + dev_err(dev, "firmware %s: %zu B left at @%zx, " + "not enough for BCF header\n", + i2400m->fw_name, leftover, offset); + break; + } + bcf_hdr = itr; + /* Only the first header is supposed to be followed by + * payload */ + header_len = sizeof(u32) * le32_to_cpu(bcf_hdr->header_len); + size = sizeof(u32) * le32_to_cpu(bcf_hdr->size); + if (headers == 0) + next = itr + size; + else + next = itr + header_len; + + result = i2400m_fw_hdr_check(i2400m, bcf_hdr, headers, offset); + if (result < 0) + continue; + if (used_slots + 1 >= slots) { + /* +1 -> we need to account for the one we'll + * occupy and at least an extra one for + * always being NULL */ + result = i2400m_zrealloc_2x( + (void **) &i2400m->fw_hdrs, &slots, + sizeof(i2400m->fw_hdrs[0]), + GFP_KERNEL); + if (result < 0) + goto error_zrealloc; + } + i2400m->fw_hdrs[used_slots] = bcf_hdr; + used_slots++; + } + if (headers == 0) { + dev_err(dev, "firmware %s: no usable headers found\n", + i2400m->fw_name); + result = -EBADF; + } else + result = 0; +error_zrealloc: + return result; +} + + +/* + * Match a barker to a BCF header module ID + * + * The device sends a barker which tells the firmware loader which + * header in the BCF file has to be used. This does the matching. + */ +static +unsigned i2400m_bcf_hdr_match(struct i2400m *i2400m, + const struct i2400m_bcf_hdr *bcf_hdr) +{ + u32 barker = le32_to_cpu(i2400m->barker->data[0]) + & 0x7fffffff; + u32 module_id = le32_to_cpu(bcf_hdr->module_id) + & 0x7fffffff; /* high bit used for something else */ + + /* special case for 5x50 */ + if (barker == I2400M_SBOOT_BARKER && module_id == 0) + return 1; + if (module_id == barker) + return 1; + return 0; +} + +static +const struct i2400m_bcf_hdr *i2400m_bcf_hdr_find(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_bcf_hdr **bcf_itr, *bcf_hdr; + unsigned i = 0; + u32 barker = le32_to_cpu(i2400m->barker->data[0]); + + d_printf(2, dev, "finding BCF header for barker %08x\n", barker); + if (barker == I2400M_NBOOT_BARKER) { + bcf_hdr = i2400m->fw_hdrs[0]; + d_printf(1, dev, "using BCF header #%u/%08x for non-signed " + "barker\n", 0, le32_to_cpu(bcf_hdr->module_id)); + return bcf_hdr; + } + for (bcf_itr = i2400m->fw_hdrs; *bcf_itr != NULL; bcf_itr++, i++) { + bcf_hdr = *bcf_itr; + if (i2400m_bcf_hdr_match(i2400m, bcf_hdr)) { + d_printf(1, dev, "hit on BCF hdr #%u/%08x\n", + i, le32_to_cpu(bcf_hdr->module_id)); + return bcf_hdr; + } else + d_printf(1, dev, "miss on BCF hdr #%u/%08x\n", + i, le32_to_cpu(bcf_hdr->module_id)); + } + dev_err(dev, "cannot find a matching BCF header for barker %08x\n", + barker); + return NULL; +} + + +/* + * Download the firmware to the device + * + * @i2400m: device descriptor + * @bcf: pointer to loaded (and minimally verified for consistency) + * firmware + * @bcf_size: size of the @bcf buffer (header plus payloads) + * + * The process for doing this is described in this file's header. + * + * Note we only reinitialize boot-mode if the flags say so. Some hw + * iterations need it, some don't. In any case, if we loop, we always + * need to reinitialize the boot room, hence the flags modification. + */ +static +int i2400m_fw_dnload(struct i2400m *i2400m, const struct i2400m_bcf_hdr *bcf, + size_t fw_size, enum i2400m_bri flags) +{ + int ret = 0; + struct device *dev = i2400m_dev(i2400m); + int count = i2400m->bus_bm_retries; + const struct i2400m_bcf_hdr *bcf_hdr; + size_t bcf_size; + + d_fnstart(5, dev, "(i2400m %p bcf %p fw size %zu)\n", + i2400m, bcf, fw_size); + i2400m->boot_mode = 1; + wmb(); /* Make sure other readers see it */ +hw_reboot: + if (count-- == 0) { + ret = -ERESTARTSYS; + dev_err(dev, "device rebooted too many times, aborting\n"); + goto error_too_many_reboots; + } + if (flags & I2400M_BRI_MAC_REINIT) { + ret = i2400m_bootrom_init(i2400m, flags); + if (ret < 0) { + dev_err(dev, "bootrom init failed: %d\n", ret); + goto error_bootrom_init; + } + } + flags |= I2400M_BRI_MAC_REINIT; + + /* + * Initialize the download, push the bytes to the device and + * then jump to the new firmware. Note @ret is passed with the + * offset of the jump instruction to _dnload_finalize() + * + * Note we need to use the BCF header in the firmware image + * that matches the barker that the device sent when it + * rebooted, so it has to be passed along. + */ + ret = -EBADF; + bcf_hdr = i2400m_bcf_hdr_find(i2400m); + if (bcf_hdr == NULL) + goto error_bcf_hdr_find; + + ret = i2400m_dnload_init(i2400m, bcf_hdr); + if (ret == -ERESTARTSYS) + goto error_dev_rebooted; + if (ret < 0) + goto error_dnload_init; + + /* + * bcf_size refers to one header size plus the fw sections size + * indicated by the header,ie. if there are other extended headers + * at the tail, they are not counted + */ + bcf_size = sizeof(u32) * le32_to_cpu(bcf_hdr->size); + ret = i2400m_dnload_bcf(i2400m, bcf, bcf_size); + if (ret == -ERESTARTSYS) + goto error_dev_rebooted; + if (ret < 0) { + dev_err(dev, "fw %s: download failed: %d\n", + i2400m->fw_name, ret); + goto error_dnload_bcf; + } + + ret = i2400m_dnload_finalize(i2400m, bcf_hdr, bcf, ret); + if (ret == -ERESTARTSYS) + goto error_dev_rebooted; + if (ret < 0) { + dev_err(dev, "fw %s: " + "download finalization failed: %d\n", + i2400m->fw_name, ret); + goto error_dnload_finalize; + } + + d_printf(2, dev, "fw %s successfully uploaded\n", + i2400m->fw_name); + i2400m->boot_mode = 0; + wmb(); /* Make sure i2400m_msg_to_dev() sees boot_mode */ +error_dnload_finalize: +error_dnload_bcf: +error_dnload_init: +error_bcf_hdr_find: +error_bootrom_init: +error_too_many_reboots: + d_fnend(5, dev, "(i2400m %p bcf %p size %zu) = %d\n", + i2400m, bcf, fw_size, ret); + return ret; + +error_dev_rebooted: + dev_err(dev, "device rebooted, %d tries left\n", count); + /* we got the notification already, no need to wait for it again */ + flags |= I2400M_BRI_SOFT; + goto hw_reboot; +} + +static +int i2400m_fw_bootstrap(struct i2400m *i2400m, const struct firmware *fw, + enum i2400m_bri flags) +{ + int ret; + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_bcf_hdr *bcf; /* Firmware data */ + + d_fnstart(5, dev, "(i2400m %p)\n", i2400m); + bcf = (void *) fw->data; + ret = i2400m_fw_check(i2400m, bcf, fw->size); + if (ret >= 0) + ret = i2400m_fw_dnload(i2400m, bcf, fw->size, flags); + if (ret < 0) + dev_err(dev, "%s: cannot use: %d, skipping\n", + i2400m->fw_name, ret); + kfree(i2400m->fw_hdrs); + i2400m->fw_hdrs = NULL; + d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, ret); + return ret; +} + + +/* Refcounted container for firmware data */ +struct i2400m_fw { + struct kref kref; + const struct firmware *fw; +}; + + +static +void i2400m_fw_destroy(struct kref *kref) +{ + struct i2400m_fw *i2400m_fw = + container_of(kref, struct i2400m_fw, kref); + release_firmware(i2400m_fw->fw); + kfree(i2400m_fw); +} + + +static +struct i2400m_fw *i2400m_fw_get(struct i2400m_fw *i2400m_fw) +{ + if (i2400m_fw != NULL && i2400m_fw != (void *) ~0) + kref_get(&i2400m_fw->kref); + return i2400m_fw; +} + + +static +void i2400m_fw_put(struct i2400m_fw *i2400m_fw) +{ + kref_put(&i2400m_fw->kref, i2400m_fw_destroy); +} + + +/** + * i2400m_dev_bootstrap - Bring the device to a known state and upload firmware + * + * @i2400m: device descriptor + * + * Returns: >= 0 if ok, < 0 errno code on error. + * + * This sets up the firmware upload environment, loads the firmware + * file from disk, verifies and then calls the firmware upload process + * per se. + * + * Can be called either from probe, or after a warm reset. Can not be + * called from within an interrupt. All the flow in this code is + * single-threade; all I/Os are synchronous. + */ +int i2400m_dev_bootstrap(struct i2400m *i2400m, enum i2400m_bri flags) +{ + int ret, itr; + struct device *dev = i2400m_dev(i2400m); + struct i2400m_fw *i2400m_fw; + const struct firmware *fw; + const char *fw_name; + + d_fnstart(5, dev, "(i2400m %p)\n", i2400m); + + ret = -ENODEV; + spin_lock(&i2400m->rx_lock); + i2400m_fw = i2400m_fw_get(i2400m->fw_cached); + spin_unlock(&i2400m->rx_lock); + if (i2400m_fw == (void *) ~0) { + dev_err(dev, "can't load firmware now!"); + goto out; + } else if (i2400m_fw != NULL) { + dev_info(dev, "firmware %s: loading from cache\n", + i2400m->fw_name); + ret = i2400m_fw_bootstrap(i2400m, i2400m_fw->fw, flags); + i2400m_fw_put(i2400m_fw); + goto out; + } + + /* Load firmware files to memory. */ + for (itr = 0, ret = -ENOENT; ; itr++) { + fw_name = i2400m->bus_fw_names[itr]; + if (fw_name == NULL) { + dev_err(dev, "Could not find a usable firmware image\n"); + break; + } + d_printf(1, dev, "trying firmware %s (%d)\n", fw_name, itr); + ret = request_firmware(&fw, fw_name, dev); + if (ret < 0) { + dev_err(dev, "fw %s: cannot load file: %d\n", + fw_name, ret); + continue; + } + i2400m->fw_name = fw_name; + ret = i2400m_fw_bootstrap(i2400m, fw, flags); + release_firmware(fw); + if (ret >= 0) /* firmware loaded successfully */ + break; + i2400m->fw_name = NULL; + } +out: + d_fnend(5, dev, "(i2400m %p) = %d\n", i2400m, ret); + return ret; +} +EXPORT_SYMBOL_GPL(i2400m_dev_bootstrap); + + +void i2400m_fw_cache(struct i2400m *i2400m) +{ + int result; + struct i2400m_fw *i2400m_fw; + struct device *dev = i2400m_dev(i2400m); + + /* if there is anything there, free it -- now, this'd be weird */ + spin_lock(&i2400m->rx_lock); + i2400m_fw = i2400m->fw_cached; + spin_unlock(&i2400m->rx_lock); + if (i2400m_fw != NULL && i2400m_fw != (void *) ~0) { + i2400m_fw_put(i2400m_fw); + WARN(1, "%s:%u: still cached fw still present?\n", + __func__, __LINE__); + } + + if (i2400m->fw_name == NULL) { + dev_err(dev, "firmware n/a: can't cache\n"); + i2400m_fw = (void *) ~0; + goto out; + } + + i2400m_fw = kzalloc(sizeof(*i2400m_fw), GFP_ATOMIC); + if (i2400m_fw == NULL) + goto out; + kref_init(&i2400m_fw->kref); + result = request_firmware(&i2400m_fw->fw, i2400m->fw_name, dev); + if (result < 0) { + dev_err(dev, "firmware %s: failed to cache: %d\n", + i2400m->fw_name, result); + kfree(i2400m_fw); + i2400m_fw = (void *) ~0; + } else + dev_info(dev, "firmware %s: cached\n", i2400m->fw_name); +out: + spin_lock(&i2400m->rx_lock); + i2400m->fw_cached = i2400m_fw; + spin_unlock(&i2400m->rx_lock); +} + + +void i2400m_fw_uncache(struct i2400m *i2400m) +{ + struct i2400m_fw *i2400m_fw; + + spin_lock(&i2400m->rx_lock); + i2400m_fw = i2400m->fw_cached; + i2400m->fw_cached = NULL; + spin_unlock(&i2400m->rx_lock); + + if (i2400m_fw != NULL && i2400m_fw != (void *) ~0) + i2400m_fw_put(i2400m_fw); +} + diff --git a/drivers/staging/wimax/i2400m/i2400m-usb.h b/drivers/staging/wimax/i2400m/i2400m-usb.h new file mode 100644 index 000000000000..eff4f464a23e --- /dev/null +++ b/drivers/staging/wimax/i2400m/i2400m-usb.h @@ -0,0 +1,275 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * USB-specific i2400m driver definitions + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Inaky Perez-Gonzalez + * Yanir Lubetkin + * - Initial implementation + * + * + * This driver implements the bus-specific part of the i2400m for + * USB. Check i2400m.h for a generic driver description. + * + * ARCHITECTURE + * + * This driver listens to notifications sent from the notification + * endpoint (in usb-notif.c); when data is ready to read, the code in + * there schedules a read from the device (usb-rx.c) and then passes + * the data to the generic RX code (rx.c). + * + * When the generic driver needs to send data (network or control), it + * queues up in the TX FIFO (tx.c) and that will notify the driver + * through the i2400m->bus_tx_kick() callback + * (usb-tx.c:i2400mu_bus_tx_kick) which will send the items in the + * FIFO queue. + * + * This driver, as well, implements the USB-specific ops for the generic + * driver to be able to setup/teardown communication with the device + * [i2400m_bus_dev_start() and i2400m_bus_dev_stop()], reseting the + * device [i2400m_bus_reset()] and performing firmware upload + * [i2400m_bus_bm_cmd() and i2400_bus_bm_wait_for_ack()]. + */ + +#ifndef __I2400M_USB_H__ +#define __I2400M_USB_H__ + +#include "i2400m.h" +#include + + +/* + * Error Density Count: cheapo error density (over time) counter + * + * Originally by Reinette Chatre + * + * Embed an 'struct edc' somewhere. Each time there is a soft or + * retryable error, call edc_inc() and check if the error top + * watermark has been reached. + */ +enum { + EDC_MAX_ERRORS = 10, + EDC_ERROR_TIMEFRAME = HZ, +}; + +/* error density counter */ +struct edc { + unsigned long timestart; + u16 errorcount; +}; + +struct i2400m_endpoint_cfg { + unsigned char bulk_out; + unsigned char notification; + unsigned char reset_cold; + unsigned char bulk_in; +}; + +static inline void edc_init(struct edc *edc) +{ + edc->timestart = jiffies; +} + +/** + * edc_inc - report a soft error and check if we are over the watermark + * + * @edc: pointer to error density counter. + * @max_err: maximum number of errors we can accept over the timeframe + * @timeframe: length of the timeframe (in jiffies). + * + * Returns: !0 1 if maximum acceptable errors per timeframe has been + * exceeded. 0 otherwise. + * + * This is way to determine if the number of acceptable errors per time + * period has been exceeded. It is not accurate as there are cases in which + * this scheme will not work, for example if there are periodic occurrences + * of errors that straddle updates to the start time. This scheme is + * sufficient for our usage. + * + * To use, embed a 'struct edc' somewhere, initialize it with + * edc_init() and when an error hits: + * + * if (do_something_fails_with_a_soft_error) { + * if (edc_inc(&my->edc, MAX_ERRORS, MAX_TIMEFRAME)) + * Ops, hard error, do something about it + * else + * Retry or ignore, depending on whatever + * } + */ +static inline int edc_inc(struct edc *edc, u16 max_err, u16 timeframe) +{ + unsigned long now; + + now = jiffies; + if (time_after(now, edc->timestart + timeframe)) { + edc->errorcount = 1; + edc->timestart = now; + } else if (++edc->errorcount > max_err) { + edc->errorcount = 0; + edc->timestart = now; + return 1; + } + return 0; +} + +/* Host-Device interface for USB */ +enum { + I2400M_USB_BOOT_RETRIES = 3, + I2400MU_MAX_NOTIFICATION_LEN = 256, + I2400MU_BLK_SIZE = 16, + I2400MU_PL_SIZE_MAX = 0x3EFF, + + /* Device IDs */ + USB_DEVICE_ID_I6050 = 0x0186, + USB_DEVICE_ID_I6050_2 = 0x0188, + USB_DEVICE_ID_I6150 = 0x07d6, + USB_DEVICE_ID_I6150_2 = 0x07d7, + USB_DEVICE_ID_I6150_3 = 0x07d9, + USB_DEVICE_ID_I6250 = 0x0187, +}; + + +/** + * struct i2400mu - descriptor for a USB connected i2400m + * + * @i2400m: bus-generic i2400m implementation; has to be first (see + * it's documentation in i2400m.h). + * + * @usb_dev: pointer to our USB device + * + * @usb_iface: pointer to our USB interface + * + * @urb_edc: error density counter; used to keep a density-on-time tab + * on how many soft (retryable or ignorable) errors we get. If we + * go over the threshold, we consider the bus transport is failing + * too much and reset. + * + * @notif_urb: URB for receiving notifications from the device. + * + * @tx_kthread: thread we use for data TX. We use a thread because in + * order to do deep power saving and put the device to sleep, we + * need to call usb_autopm_*() [blocking functions]. + * + * @tx_wq: waitqueue for the TX kthread to sleep when there is no data + * to be sent; when more data is available, it is woken up by + * i2400mu_bus_tx_kick(). + * + * @rx_kthread: thread we use for data RX. We use a thread because in + * order to do deep power saving and put the device to sleep, we + * need to call usb_autopm_*() [blocking functions]. + * + * @rx_wq: waitqueue for the RX kthread to sleep when there is no data + * to receive. When data is available, it is woken up by + * usb-notif.c:i2400mu_notification_grok(). + * + * @rx_pending_count: number of rx-data-ready notifications that were + * still not handled by the RX kthread. + * + * @rx_size: current RX buffer size that is being used. + * + * @rx_size_acc: accumulator of the sizes of the previous read + * transactions. + * + * @rx_size_cnt: number of read transactions accumulated in + * @rx_size_acc. + * + * @do_autopm: disable(0)/enable(>0) calling the + * usb_autopm_get/put_interface() barriers when executing + * commands. See doc in i2400mu_suspend() for more information. + * + * @rx_size_auto_shrink: if true, the rx_size is shrunk + * automatically based on the average size of the received + * transactions. This allows the receive code to allocate smaller + * chunks of memory and thus reduce pressure on the memory + * allocator by not wasting so much space. By default it is + * enabled. + * + * @debugfs_dentry: hookup for debugfs files. + * These have to be in a separate directory, a child of + * (wimax_dev->debugfs_dentry) so they can be removed when the + * module unloads, as we don't keep each dentry. + */ +struct i2400mu { + struct i2400m i2400m; /* FIRST! See doc */ + + struct usb_device *usb_dev; + struct usb_interface *usb_iface; + struct edc urb_edc; /* Error density counter */ + struct i2400m_endpoint_cfg endpoint_cfg; + + struct urb *notif_urb; + struct task_struct *tx_kthread; + wait_queue_head_t tx_wq; + + struct task_struct *rx_kthread; + wait_queue_head_t rx_wq; + atomic_t rx_pending_count; + size_t rx_size, rx_size_acc, rx_size_cnt; + atomic_t do_autopm; + u8 rx_size_auto_shrink; + + struct dentry *debugfs_dentry; + unsigned i6050:1; /* 1 if this is a 6050 based SKU */ +}; + + +static inline +void i2400mu_init(struct i2400mu *i2400mu) +{ + i2400m_init(&i2400mu->i2400m); + edc_init(&i2400mu->urb_edc); + init_waitqueue_head(&i2400mu->tx_wq); + atomic_set(&i2400mu->rx_pending_count, 0); + init_waitqueue_head(&i2400mu->rx_wq); + i2400mu->rx_size = PAGE_SIZE - sizeof(struct skb_shared_info); + atomic_set(&i2400mu->do_autopm, 1); + i2400mu->rx_size_auto_shrink = 1; +} + +int i2400mu_notification_setup(struct i2400mu *); +void i2400mu_notification_release(struct i2400mu *); + +int i2400mu_rx_setup(struct i2400mu *); +void i2400mu_rx_release(struct i2400mu *); +void i2400mu_rx_kick(struct i2400mu *); + +int i2400mu_tx_setup(struct i2400mu *); +void i2400mu_tx_release(struct i2400mu *); +void i2400mu_bus_tx_kick(struct i2400m *); + +ssize_t i2400mu_bus_bm_cmd_send(struct i2400m *, + const struct i2400m_bootrom_header *, size_t, + int); +ssize_t i2400mu_bus_bm_wait_for_ack(struct i2400m *, + struct i2400m_bootrom_header *, size_t); +#endif /* #ifndef __I2400M_USB_H__ */ diff --git a/drivers/staging/wimax/i2400m/i2400m.h b/drivers/staging/wimax/i2400m/i2400m.h new file mode 100644 index 000000000000..de22cc6f2c5c --- /dev/null +++ b/drivers/staging/wimax/i2400m/i2400m.h @@ -0,0 +1,970 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * Declarations for bus-generic internal APIs + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Inaky Perez-Gonzalez + * Yanir Lubetkin + * - Initial implementation + * + * + * GENERAL DRIVER ARCHITECTURE + * + * The i2400m driver is split in the following two major parts: + * + * - bus specific driver + * - bus generic driver (this part) + * + * The bus specific driver sets up stuff specific to the bus the + * device is connected to (USB, PCI, tam-tam...non-authoritative + * nor binding list) which is basically the device-model management + * (probe/disconnect, etc), moving data from device to kernel and + * back, doing the power saving details and reseting the device. + * + * For details on each bus-specific driver, see it's include file, + * i2400m-BUSNAME.h + * + * The bus-generic functionality break up is: + * + * - Firmware upload: fw.c - takes care of uploading firmware to the + * device. bus-specific driver just needs to provides a way to + * execute boot-mode commands and to reset the device. + * + * - RX handling: rx.c - receives data from the bus-specific code and + * feeds it to the network or WiMAX stack or uses it to modify + * the driver state. bus-specific driver only has to receive + * frames and pass them to this module. + * + * - TX handling: tx.c - manages the TX FIFO queue and provides means + * for the bus-specific TX code to pull data from the FIFO + * queue. bus-specific code just pulls frames from this module + * to sends them to the device. + * + * - netdev glue: netdev.c - interface with Linux networking + * stack. Pass around data frames, and configure when the + * device is up and running or shutdown (through ifconfig up / + * down). Bus-generic only. + * + * - control ops: control.c - implements various commands for + * controlling the device. bus-generic only. + * + * - device model glue: driver.c - implements helpers for the + * device-model glue done by the bus-specific layer + * (setup/release the driver resources), turning the device on + * and off, handling the device reboots/resets and a few simple + * WiMAX stack ops. + * + * Code is also broken up in linux-glue / device-glue. + * + * Linux glue contains functions that deal mostly with gluing with the + * rest of the Linux kernel. + * + * Device-glue are functions that deal mostly with the way the device + * does things and talk the device's language. + * + * device-glue code is licensed BSD so other open source OSes can take + * it to implement their drivers. + * + * + * APIs AND HEADER FILES + * + * This bus generic code exports three APIs: + * + * - HDI (host-device interface) definitions common to all busses + * (include/linux/wimax/i2400m.h); these can be also used by user + * space code. + * - internal API for the bus-generic code + * - external API for the bus-specific drivers + * + * + * LIFE CYCLE: + * + * When the bus-specific driver probes, it allocates a network device + * with enough space for it's data structue, that must contain a + * &struct i2400m at the top. + * + * On probe, it needs to fill the i2400m members marked as [fill], as + * well as i2400m->wimax_dev.net_dev and call i2400m_setup(). The + * i2400m driver will only register with the WiMAX and network stacks; + * the only access done to the device is to read the MAC address so we + * can register a network device. + * + * The high-level call flow is: + * + * bus_probe() + * i2400m_setup() + * i2400m->bus_setup() + * boot rom initialization / read mac addr + * network / WiMAX stacks registration + * i2400m_dev_start() + * i2400m->bus_dev_start() + * i2400m_dev_initialize() + * + * The reverse applies for a disconnect() call: + * + * bus_disconnect() + * i2400m_release() + * i2400m_dev_stop() + * i2400m_dev_shutdown() + * i2400m->bus_dev_stop() + * network / WiMAX stack unregistration + * i2400m->bus_release() + * + * At this point, control and data communications are possible. + * + * While the device is up, it might reset. The bus-specific driver has + * to catch that situation and call i2400m_dev_reset_handle() to deal + * with it (reset the internal driver structures and go back to square + * one). + */ + +#ifndef __I2400M_H__ +#define __I2400M_H__ + +#include +#include +#include +#include +#include +#include "../net-wimax.h" +#include "linux-wimax-i2400m.h" +#include + +enum { +/* netdev interface */ + /* + * Out of NWG spec (R1_v1.2.2), 3.3.3 ASN Bearer Plane MTU Size + * + * The MTU is 1400 or less + */ + I2400M_MAX_MTU = 1400, +}; + +/* Misc constants */ +enum { + /* Size of the Boot Mode Command buffer */ + I2400M_BM_CMD_BUF_SIZE = 16 * 1024, + I2400M_BM_ACK_BUF_SIZE = 256, +}; + +enum { + /* Maximum number of bus reset can be retried */ + I2400M_BUS_RESET_RETRIES = 3, +}; + +/** + * struct i2400m_poke_table - Hardware poke table for the Intel 2400m + * + * This structure will be used to create a device specific poke table + * to put the device in a consistent state at boot time. + * + * @address: The device address to poke + * + * @data: The data value to poke to the device address + * + */ +struct i2400m_poke_table{ + __le32 address; + __le32 data; +}; + +#define I2400M_FW_POKE(a, d) { \ + .address = cpu_to_le32(a), \ + .data = cpu_to_le32(d) \ +} + + +/** + * i2400m_reset_type - methods to reset a device + * + * @I2400M_RT_WARM: Reset without device disconnection, device handles + * are kept valid but state is back to power on, with firmware + * re-uploaded. + * @I2400M_RT_COLD: Tell the device to disconnect itself from the bus + * and reconnect. Renders all device handles invalid. + * @I2400M_RT_BUS: Tells the bus to reset the device; last measure + * used when both types above don't work. + */ +enum i2400m_reset_type { + I2400M_RT_WARM, /* first measure */ + I2400M_RT_COLD, /* second measure */ + I2400M_RT_BUS, /* call in artillery */ +}; + +struct i2400m_reset_ctx; +struct i2400m_roq; +struct i2400m_barker_db; + +/** + * struct i2400m - descriptor for an Intel 2400m + * + * Members marked with [fill] must be filled out/initialized before + * calling i2400m_setup(). + * + * Note the @bus_setup/@bus_release, @bus_dev_start/@bus_dev_release + * call pairs are very much doing almost the same, and depending on + * the underlying bus, some stuff has to be put in one or the + * other. The idea of setup/release is that they setup the minimal + * amount needed for loading firmware, where us dev_start/stop setup + * the rest needed to do full data/control traffic. + * + * @bus_tx_block_size: [fill] USB imposes a 16 block size, but other + * busses will differ. So we have a tx_blk_size variable that the + * bus layer sets to tell the engine how much of that we need. + * + * @bus_tx_room_min: [fill] Minimum room required while allocating + * TX queue's buffer space for message header. USB requires + * 16 bytes. Refer to bus specific driver code for details. + * + * @bus_pl_size_max: [fill] Maximum payload size. + * + * @bus_setup: [optional fill] Function called by the bus-generic code + * [i2400m_setup()] to setup the basic bus-specific communications + * to the the device needed to load firmware. See LIFE CYCLE above. + * + * NOTE: Doesn't need to upload the firmware, as that is taken + * care of by the bus-generic code. + * + * @bus_release: [optional fill] Function called by the bus-generic + * code [i2400m_release()] to shutdown the basic bus-specific + * communications to the the device needed to load firmware. See + * LIFE CYCLE above. + * + * This function does not need to reset the device, just tear down + * all the host resources created to handle communication with + * the device. + * + * @bus_dev_start: [optional fill] Function called by the bus-generic + * code [i2400m_dev_start()] to do things needed to start the + * device. See LIFE CYCLE above. + * + * NOTE: Doesn't need to upload the firmware, as that is taken + * care of by the bus-generic code. + * + * @bus_dev_stop: [optional fill] Function called by the bus-generic + * code [i2400m_dev_stop()] to do things needed for stopping the + * device. See LIFE CYCLE above. + * + * This function does not need to reset the device, just tear down + * all the host resources created to handle communication with + * the device. + * + * @bus_tx_kick: [fill] Function called by the bus-generic code to let + * the bus-specific code know that there is data available in the + * TX FIFO for transmission to the device. + * + * This function cannot sleep. + * + * @bus_reset: [fill] Function called by the bus-generic code to reset + * the device in in various ways. Doesn't need to wait for the + * reset to finish. + * + * If warm or cold reset fail, this function is expected to do a + * bus-specific reset (eg: USB reset) to get the device to a + * working state (even if it implies device disconecction). + * + * Note the warm reset is used by the firmware uploader to + * reinitialize the device. + * + * IMPORTANT: this is called very early in the device setup + * process, so it cannot rely on common infrastructure being laid + * out. + * + * IMPORTANT: don't call reset on RT_BUS with i2400m->init_mutex + * held, as the .pre/.post reset handlers will deadlock. + * + * @bus_bm_retries: [fill] How many times shall a firmware upload / + * device initialization be retried? Different models of the same + * device might need different values, hence it is set by the + * bus-specific driver. Note this value is used in two places, + * i2400m_fw_dnload() and __i2400m_dev_start(); they won't become + * multiplicative (__i2400m_dev_start() calling N times + * i2400m_fw_dnload() and this trying N times to download the + * firmware), as if __i2400m_dev_start() only retries if the + * firmware crashed while initializing the device (not in a + * general case). + * + * @bus_bm_cmd_send: [fill] Function called to send a boot-mode + * command. Flags are defined in 'enum i2400m_bm_cmd_flags'. This + * is synchronous and has to return 0 if ok or < 0 errno code in + * any error condition. + * + * @bus_bm_wait_for_ack: [fill] Function called to wait for a + * boot-mode notification (that can be a response to a previously + * issued command or an asynchronous one). Will read until all the + * indicated size is read or timeout. Reading more or less data + * than asked for is an error condition. Return 0 if ok, < 0 errno + * code on error. + * + * The caller to this function will check if the response is a + * barker that indicates the device going into reset mode. + * + * @bus_fw_names: [fill] a NULL-terminated array with the names of the + * firmware images to try loading. This is made a list so we can + * support backward compatibility of firmware releases (eg: if we + * can't find the default v1.4, we try v1.3). In general, the name + * should be i2400m-fw-X-VERSION.sbcf, where X is the bus name. + * The list is tried in order and the first one that loads is + * used. The fw loader will set i2400m->fw_name to point to the + * active firmware image. + * + * @bus_bm_mac_addr_impaired: [fill] Set to true if the device's MAC + * address provided in boot mode is kind of broken and needs to + * be re-read later on. + * + * @bus_bm_pokes_table: [fill/optional] A table of device addresses + * and values that will be poked at device init time to move the + * device to the correct state for the type of boot/firmware being + * used. This table MUST be terminated with (0x000000, + * 0x00000000) or bad things will happen. + * + * + * @wimax_dev: WiMAX generic device for linkage into the kernel WiMAX + * stack. Due to the way a net_device is allocated, we need to + * force this to be the first field so that we can get from + * netdev_priv() the right pointer. + * + * @updown: the device is up and ready for transmitting control and + * data packets. This implies @ready (communication infrastructure + * with the device is ready) and the device's firmware has been + * loaded and the device initialized. + * + * Write to it only inside a i2400m->init_mutex protected area + * followed with a wmb(); rmb() before accesing (unless locked + * inside i2400m->init_mutex). Read access can be loose like that + * [just using rmb()] because the paths that use this also do + * other error checks later on. + * + * @ready: Communication infrastructure with the device is ready, data + * frames can start to be passed around (this is lighter than + * using the WiMAX state for certain hot paths). + * + * Write to it only inside a i2400m->init_mutex protected area + * followed with a wmb(); rmb() before accesing (unless locked + * inside i2400m->init_mutex). Read access can be loose like that + * [just using rmb()] because the paths that use this also do + * other error checks later on. + * + * @rx_reorder: 1 if RX reordering is enabled; this can only be + * set at probe time. + * + * @state: device's state (as reported by it) + * + * @state_wq: waitqueue that is woken up whenever the state changes + * + * @tx_lock: spinlock to protect TX members + * + * @tx_buf: FIFO buffer for TX; we queue data here + * + * @tx_in: FIFO index for incoming data. Note this doesn't wrap around + * and it is always greater than @tx_out. + * + * @tx_out: FIFO index for outgoing data + * + * @tx_msg: current TX message that is active in the FIFO for + * appending payloads. + * + * @tx_sequence: current sequence number for TX messages from the + * device to the host. + * + * @tx_msg_size: size of the current message being transmitted by the + * bus-specific code. + * + * @tx_pl_num: total number of payloads sent + * + * @tx_pl_max: maximum number of payloads sent in a TX message + * + * @tx_pl_min: minimum number of payloads sent in a TX message + * + * @tx_num: number of TX messages sent + * + * @tx_size_acc: number of bytes in all TX messages sent + * (this is different to net_dev's statistics as it also counts + * control messages). + * + * @tx_size_min: smallest TX message sent. + * + * @tx_size_max: biggest TX message sent. + * + * @rx_lock: spinlock to protect RX members and rx_roq_refcount. + * + * @rx_pl_num: total number of payloads received + * + * @rx_pl_max: maximum number of payloads received in a RX message + * + * @rx_pl_min: minimum number of payloads received in a RX message + * + * @rx_num: number of RX messages received + * + * @rx_size_acc: number of bytes in all RX messages received + * (this is different to net_dev's statistics as it also counts + * control messages). + * + * @rx_size_min: smallest RX message received. + * + * @rx_size_max: buggest RX message received. + * + * @rx_roq: RX ReOrder queues. (fw >= v1.4) When packets are received + * out of order, the device will ask the driver to hold certain + * packets until the ones that are received out of order can be + * delivered. Then the driver can release them to the host. See + * drivers/net/i2400m/rx.c for details. + * + * @rx_roq_refcount: refcount rx_roq. This refcounts any access to + * rx_roq thus preventing rx_roq being destroyed when rx_roq + * is being accessed. rx_roq_refcount is protected by rx_lock. + * + * @rx_reports: reports received from the device that couldn't be + * processed because the driver wasn't still ready; when ready, + * they are pulled from here and chewed. + * + * @rx_reports_ws: Work struct used to kick a scan of the RX reports + * list and to process each. + * + * @src_mac_addr: MAC address used to make ethernet packets be coming + * from. This is generated at i2400m_setup() time and used during + * the life cycle of the instance. See i2400m_fake_eth_header(). + * + * @init_mutex: Mutex used for serializing the device bringup + * sequence; this way if the device reboots in the middle, we + * don't try to do a bringup again while we are tearing down the + * one that failed. + * + * Can't reuse @msg_mutex because from within the bringup sequence + * we need to send messages to the device and thus use @msg_mutex. + * + * @msg_mutex: mutex used to send control commands to the device (we + * only allow one at a time, per host-device interface design). + * + * @msg_completion: used to wait for an ack to a control command sent + * to the device. + * + * @ack_skb: used to store the actual ack to a control command if the + * reception of the command was successful. Otherwise, a ERR_PTR() + * errno code that indicates what failed with the ack reception. + * + * Only valid after @msg_completion is woken up. Only updateable + * if @msg_completion is armed. Only touched by + * i2400m_msg_to_dev(). + * + * Protected by @rx_lock. In theory the command execution flow is + * sequential, but in case the device sends an out-of-phase or + * very delayed response, we need to avoid it trampling current + * execution. + * + * @bm_cmd_buf: boot mode command buffer for composing firmware upload + * commands. + * + * USB can't r/w to stack, vmalloc, etc...as well, we end up + * having to alloc/free a lot to compose commands, so we use these + * for stagging and not having to realloc all the time. + * + * This assumes the code always runs serialized. Only one thread + * can call i2400m_bm_cmd() at the same time. + * + * @bm_ack_buf: boot mode acknoledge buffer for staging reception of + * responses to commands. + * + * See @bm_cmd_buf. + * + * @work_queue: work queue for processing device reports. This + * workqueue cannot be used for processing TX or RX to the device, + * as from it we'll process device reports, which might require + * further communication with the device. + * + * @debugfs_dentry: hookup for debugfs files. + * These have to be in a separate directory, a child of + * (wimax_dev->debugfs_dentry) so they can be removed when the + * module unloads, as we don't keep each dentry. + * + * @fw_name: name of the firmware image that is currently being used. + * + * @fw_version: version of the firmware interface, Major.minor, + * encoded in the high word and low word (major << 16 | minor). + * + * @fw_hdrs: NULL terminated array of pointers to the firmware + * headers. This is only available during firmware load time. + * + * @fw_cached: Used to cache firmware when the system goes to + * suspend/standby/hibernation (as on resume we can't read it). If + * NULL, no firmware was cached, read it. If ~0, you can't read + * any firmware files (the system still didn't come out of suspend + * and failed to cache one), so abort; otherwise, a valid cached + * firmware to be used. Access to this variable is protected by + * the spinlock i2400m->rx_lock. + * + * @barker: barker type that the device uses; this is initialized by + * i2400m_is_boot_barker() the first time it is called. Then it + * won't change during the life cycle of the device and every time + * a boot barker is received, it is just verified for it being the + * same. + * + * @pm_notifier: used to register for PM events + * + * @bus_reset_retries: counter for the number of bus resets attempted for + * this boot. It's not for tracking the number of bus resets during + * the whole driver life cycle (from insmod to rmmod) but for the + * number of dev_start() executed until dev_start() returns a success + * (ie: a good boot means a dev_stop() followed by a successful + * dev_start()). dev_reset_handler() increments this counter whenever + * it is triggering a bus reset. It checks this counter to decide if a + * subsequent bus reset should be retried. dev_reset_handler() retries + * the bus reset until dev_start() succeeds or the counter reaches + * I2400M_BUS_RESET_RETRIES. The counter is cleared to 0 in + * dev_reset_handle() when dev_start() returns a success, + * ie: a successul boot is completed. + * + * @alive: flag to denote if the device *should* be alive. This flag is + * everything like @updown (see doc for @updown) except reflecting + * the device state *we expect* rather than the actual state as denoted + * by @updown. It is set 1 whenever @updown is set 1 in dev_start(). + * Then the device is expected to be alive all the time + * (i2400m->alive remains 1) until the driver is removed. Therefore + * all the device reboot events detected can be still handled properly + * by either dev_reset_handle() or .pre_reset/.post_reset as long as + * the driver presents. It is set 0 along with @updown in dev_stop(). + * + * @error_recovery: flag to denote if we are ready to take an error recovery. + * 0 for ready to take an error recovery; 1 for not ready. It is + * initialized to 1 while probe() since we don't tend to take any error + * recovery during probe(). It is decremented by 1 whenever dev_start() + * succeeds to indicate we are ready to take error recovery from now on. + * It is checked every time we wanna schedule an error recovery. If an + * error recovery is already in place (error_recovery was set 1), we + * should not schedule another one until the last one is done. + */ +struct i2400m { + struct wimax_dev wimax_dev; /* FIRST! See doc */ + + unsigned updown:1; /* Network device is up or down */ + unsigned boot_mode:1; /* is the device in boot mode? */ + unsigned sboot:1; /* signed or unsigned fw boot */ + unsigned ready:1; /* Device comm infrastructure ready */ + unsigned rx_reorder:1; /* RX reorder is enabled */ + u8 trace_msg_from_user; /* echo rx msgs to 'trace' pipe */ + /* typed u8 so /sys/kernel/debug/u8 can tweak */ + enum i2400m_system_state state; + wait_queue_head_t state_wq; /* Woken up when on state updates */ + + size_t bus_tx_block_size; + size_t bus_tx_room_min; + size_t bus_pl_size_max; + unsigned bus_bm_retries; + + int (*bus_setup)(struct i2400m *); + int (*bus_dev_start)(struct i2400m *); + void (*bus_dev_stop)(struct i2400m *); + void (*bus_release)(struct i2400m *); + void (*bus_tx_kick)(struct i2400m *); + int (*bus_reset)(struct i2400m *, enum i2400m_reset_type); + ssize_t (*bus_bm_cmd_send)(struct i2400m *, + const struct i2400m_bootrom_header *, + size_t, int flags); + ssize_t (*bus_bm_wait_for_ack)(struct i2400m *, + struct i2400m_bootrom_header *, size_t); + const char **bus_fw_names; + unsigned bus_bm_mac_addr_impaired:1; + const struct i2400m_poke_table *bus_bm_pokes_table; + + spinlock_t tx_lock; /* protect TX state */ + void *tx_buf; + size_t tx_in, tx_out; + struct i2400m_msg_hdr *tx_msg; + size_t tx_sequence, tx_msg_size; + /* TX stats */ + unsigned tx_pl_num, tx_pl_max, tx_pl_min, + tx_num, tx_size_acc, tx_size_min, tx_size_max; + + /* RX stuff */ + /* protect RX state and rx_roq_refcount */ + spinlock_t rx_lock; + unsigned rx_pl_num, rx_pl_max, rx_pl_min, + rx_num, rx_size_acc, rx_size_min, rx_size_max; + struct i2400m_roq *rx_roq; /* access is refcounted */ + struct kref rx_roq_refcount; /* refcount access to rx_roq */ + u8 src_mac_addr[ETH_HLEN]; + struct list_head rx_reports; /* under rx_lock! */ + struct work_struct rx_report_ws; + + struct mutex msg_mutex; /* serialize command execution */ + struct completion msg_completion; + struct sk_buff *ack_skb; /* protected by rx_lock */ + + void *bm_ack_buf; /* for receiving acks over USB */ + void *bm_cmd_buf; /* for issuing commands over USB */ + + struct workqueue_struct *work_queue; + + struct mutex init_mutex; /* protect bringup seq */ + struct i2400m_reset_ctx *reset_ctx; /* protected by init_mutex */ + + struct work_struct wake_tx_ws; + struct sk_buff *wake_tx_skb; + + struct work_struct reset_ws; + const char *reset_reason; + + struct work_struct recovery_ws; + + struct dentry *debugfs_dentry; + const char *fw_name; /* name of the current firmware image */ + unsigned long fw_version; /* version of the firmware interface */ + const struct i2400m_bcf_hdr **fw_hdrs; + struct i2400m_fw *fw_cached; /* protected by rx_lock */ + struct i2400m_barker_db *barker; + + struct notifier_block pm_notifier; + + /* counting bus reset retries in this boot */ + atomic_t bus_reset_retries; + + /* if the device is expected to be alive */ + unsigned alive; + + /* 0 if we are ready for error recovery; 1 if not ready */ + atomic_t error_recovery; + +}; + + +/* + * Bus-generic internal APIs + * ------------------------- + */ + +static inline +struct i2400m *wimax_dev_to_i2400m(struct wimax_dev *wimax_dev) +{ + return container_of(wimax_dev, struct i2400m, wimax_dev); +} + +static inline +struct i2400m *net_dev_to_i2400m(struct net_device *net_dev) +{ + return wimax_dev_to_i2400m(netdev_priv(net_dev)); +} + +/* + * Boot mode support + */ + +/** + * i2400m_bm_cmd_flags - flags to i2400m_bm_cmd() + * + * @I2400M_BM_CMD_RAW: send the command block as-is, without doing any + * extra processing for adding CRC. + */ +enum i2400m_bm_cmd_flags { + I2400M_BM_CMD_RAW = 1 << 2, +}; + +/** + * i2400m_bri - Boot-ROM indicators + * + * Flags for i2400m_bootrom_init() and i2400m_dev_bootstrap() [which + * are passed from things like i2400m_setup()]. Can be combined with + * |. + * + * @I2400M_BRI_SOFT: The device rebooted already and a reboot + * barker received, proceed directly to ack the boot sequence. + * @I2400M_BRI_NO_REBOOT: Do not reboot the device and proceed + * directly to wait for a reboot barker from the device. + * @I2400M_BRI_MAC_REINIT: We need to reinitialize the boot + * rom after reading the MAC address. This is quite a dirty hack, + * if you ask me -- the device requires the bootrom to be + * initialized after reading the MAC address. + */ +enum i2400m_bri { + I2400M_BRI_SOFT = 1 << 1, + I2400M_BRI_NO_REBOOT = 1 << 2, + I2400M_BRI_MAC_REINIT = 1 << 3, +}; + +void i2400m_bm_cmd_prepare(struct i2400m_bootrom_header *); +int i2400m_dev_bootstrap(struct i2400m *, enum i2400m_bri); +int i2400m_read_mac_addr(struct i2400m *); +int i2400m_bootrom_init(struct i2400m *, enum i2400m_bri); +int i2400m_is_boot_barker(struct i2400m *, const void *, size_t); +static inline +int i2400m_is_d2h_barker(const void *buf) +{ + const __le32 *barker = buf; + return le32_to_cpu(*barker) == I2400M_D2H_MSG_BARKER; +} +void i2400m_unknown_barker(struct i2400m *, const void *, size_t); + +/* Make/grok boot-rom header commands */ + +static inline +__le32 i2400m_brh_command(enum i2400m_brh_opcode opcode, unsigned use_checksum, + unsigned direct_access) +{ + return cpu_to_le32( + I2400M_BRH_SIGNATURE + | (direct_access ? I2400M_BRH_DIRECT_ACCESS : 0) + | I2400M_BRH_RESPONSE_REQUIRED /* response always required */ + | (use_checksum ? I2400M_BRH_USE_CHECKSUM : 0) + | (opcode & I2400M_BRH_OPCODE_MASK)); +} + +static inline +void i2400m_brh_set_opcode(struct i2400m_bootrom_header *hdr, + enum i2400m_brh_opcode opcode) +{ + hdr->command = cpu_to_le32( + (le32_to_cpu(hdr->command) & ~I2400M_BRH_OPCODE_MASK) + | (opcode & I2400M_BRH_OPCODE_MASK)); +} + +static inline +unsigned i2400m_brh_get_opcode(const struct i2400m_bootrom_header *hdr) +{ + return le32_to_cpu(hdr->command) & I2400M_BRH_OPCODE_MASK; +} + +static inline +unsigned i2400m_brh_get_response(const struct i2400m_bootrom_header *hdr) +{ + return (le32_to_cpu(hdr->command) & I2400M_BRH_RESPONSE_MASK) + >> I2400M_BRH_RESPONSE_SHIFT; +} + +static inline +unsigned i2400m_brh_get_use_checksum(const struct i2400m_bootrom_header *hdr) +{ + return le32_to_cpu(hdr->command) & I2400M_BRH_USE_CHECKSUM; +} + +static inline +unsigned i2400m_brh_get_response_required( + const struct i2400m_bootrom_header *hdr) +{ + return le32_to_cpu(hdr->command) & I2400M_BRH_RESPONSE_REQUIRED; +} + +static inline +unsigned i2400m_brh_get_direct_access(const struct i2400m_bootrom_header *hdr) +{ + return le32_to_cpu(hdr->command) & I2400M_BRH_DIRECT_ACCESS; +} + +static inline +unsigned i2400m_brh_get_signature(const struct i2400m_bootrom_header *hdr) +{ + return (le32_to_cpu(hdr->command) & I2400M_BRH_SIGNATURE_MASK) + >> I2400M_BRH_SIGNATURE_SHIFT; +} + + +/* + * Driver / device setup and internal functions + */ +void i2400m_init(struct i2400m *); +int i2400m_reset(struct i2400m *, enum i2400m_reset_type); +void i2400m_netdev_setup(struct net_device *net_dev); +int i2400m_sysfs_setup(struct device_driver *); +void i2400m_sysfs_release(struct device_driver *); +int i2400m_tx_setup(struct i2400m *); +void i2400m_wake_tx_work(struct work_struct *); +void i2400m_tx_release(struct i2400m *); + +int i2400m_rx_setup(struct i2400m *); +void i2400m_rx_release(struct i2400m *); + +void i2400m_fw_cache(struct i2400m *); +void i2400m_fw_uncache(struct i2400m *); + +void i2400m_net_rx(struct i2400m *, struct sk_buff *, unsigned, const void *, + int); +void i2400m_net_erx(struct i2400m *, struct sk_buff *, enum i2400m_cs); +void i2400m_net_wake_stop(struct i2400m *); +enum i2400m_pt; +int i2400m_tx(struct i2400m *, const void *, size_t, enum i2400m_pt); + +#ifdef CONFIG_DEBUG_FS +void i2400m_debugfs_add(struct i2400m *); +void i2400m_debugfs_rm(struct i2400m *); +#else +static inline void i2400m_debugfs_add(struct i2400m *i2400m) {} +static inline void i2400m_debugfs_rm(struct i2400m *i2400m) {} +#endif + +/* Initialize/shutdown the device */ +int i2400m_dev_initialize(struct i2400m *); +void i2400m_dev_shutdown(struct i2400m *); + +extern struct attribute_group i2400m_dev_attr_group; + + +/* HDI message's payload description handling */ + +static inline +size_t i2400m_pld_size(const struct i2400m_pld *pld) +{ + return I2400M_PLD_SIZE_MASK & le32_to_cpu(pld->val); +} + +static inline +enum i2400m_pt i2400m_pld_type(const struct i2400m_pld *pld) +{ + return (I2400M_PLD_TYPE_MASK & le32_to_cpu(pld->val)) + >> I2400M_PLD_TYPE_SHIFT; +} + +static inline +void i2400m_pld_set(struct i2400m_pld *pld, size_t size, + enum i2400m_pt type) +{ + pld->val = cpu_to_le32( + ((type << I2400M_PLD_TYPE_SHIFT) & I2400M_PLD_TYPE_MASK) + | (size & I2400M_PLD_SIZE_MASK)); +} + + +/* + * API for the bus-specific drivers + * -------------------------------- + */ + +static inline +struct i2400m *i2400m_get(struct i2400m *i2400m) +{ + dev_hold(i2400m->wimax_dev.net_dev); + return i2400m; +} + +static inline +void i2400m_put(struct i2400m *i2400m) +{ + dev_put(i2400m->wimax_dev.net_dev); +} + +int i2400m_dev_reset_handle(struct i2400m *, const char *); +int i2400m_pre_reset(struct i2400m *); +int i2400m_post_reset(struct i2400m *); +void i2400m_error_recovery(struct i2400m *); + +/* + * _setup()/_release() are called by the probe/disconnect functions of + * the bus-specific drivers. + */ +int i2400m_setup(struct i2400m *, enum i2400m_bri bm_flags); +void i2400m_release(struct i2400m *); + +int i2400m_rx(struct i2400m *, struct sk_buff *); +struct i2400m_msg_hdr *i2400m_tx_msg_get(struct i2400m *, size_t *); +void i2400m_tx_msg_sent(struct i2400m *); + + +/* + * Utility functions + */ + +static inline +struct device *i2400m_dev(struct i2400m *i2400m) +{ + return i2400m->wimax_dev.net_dev->dev.parent; +} + +int i2400m_msg_check_status(const struct i2400m_l3l4_hdr *, char *, size_t); +int i2400m_msg_size_check(struct i2400m *, const struct i2400m_l3l4_hdr *, + size_t); +struct sk_buff *i2400m_msg_to_dev(struct i2400m *, const void *, size_t); +void i2400m_msg_to_dev_cancel_wait(struct i2400m *, int); +void i2400m_report_hook(struct i2400m *, const struct i2400m_l3l4_hdr *, + size_t); +void i2400m_report_hook_work(struct work_struct *); +int i2400m_cmd_enter_powersave(struct i2400m *); +int i2400m_cmd_exit_idle(struct i2400m *); +struct sk_buff *i2400m_get_device_info(struct i2400m *); +int i2400m_firmware_check(struct i2400m *); +int i2400m_set_idle_timeout(struct i2400m *, unsigned); + +static inline +struct usb_endpoint_descriptor *usb_get_epd(struct usb_interface *iface, int ep) +{ + return &iface->cur_altsetting->endpoint[ep].desc; +} + +int i2400m_op_rfkill_sw_toggle(struct wimax_dev *, enum wimax_rf_state); +void i2400m_report_tlv_rf_switches_status(struct i2400m *, + const struct i2400m_tlv_rf_switches_status *); + +/* + * Helpers for firmware backwards compatibility + * + * As we aim to support at least the firmware version that was + * released with the previous kernel/driver release, some code will be + * conditionally executed depending on the firmware version. On each + * release, the code to support fw releases past the last two ones + * will be purged. + * + * By making it depend on this macros, it is easier to keep it a tab + * on what has to go and what not. + */ +static inline +unsigned i2400m_le_v1_3(struct i2400m *i2400m) +{ + /* running fw is lower or v1.3 */ + return i2400m->fw_version <= 0x00090001; +} + +static inline +unsigned i2400m_ge_v1_4(struct i2400m *i2400m) +{ + /* running fw is higher or v1.4 */ + return i2400m->fw_version >= 0x00090002; +} + + +/* + * Do a millisecond-sleep for allowing wireshark to dump all the data + * packets. Used only for debugging. + */ +static inline +void __i2400m_msleep(unsigned ms) +{ +#if 1 +#else + msleep(ms); +#endif +} + + +/* module initialization helpers */ +int i2400m_barker_db_init(const char *); +void i2400m_barker_db_exit(void); + + + +#endif /* #ifndef __I2400M_H__ */ diff --git a/drivers/staging/wimax/i2400m/linux-wimax-i2400m.h b/drivers/staging/wimax/i2400m/linux-wimax-i2400m.h new file mode 100644 index 000000000000..fd198bc24a3c --- /dev/null +++ b/drivers/staging/wimax/i2400m/linux-wimax-i2400m.h @@ -0,0 +1,572 @@ +/* + * Intel Wireless WiMax Connection 2400m + * Host-Device protocol interface definitions + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Inaky Perez-Gonzalez + * - Initial implementation + * + * + * This header defines the data structures and constants used to + * communicate with the device. + * + * BOOTMODE/BOOTROM/FIRMWARE UPLOAD PROTOCOL + * + * The firmware upload protocol is quite simple and only requires a + * handful of commands. See drivers/net/wimax/i2400m/fw.c for more + * details. + * + * The BCF data structure is for the firmware file header. + * + * + * THE DATA / CONTROL PROTOCOL + * + * This is the normal protocol spoken with the device once the + * firmware is uploaded. It transports data payloads and control + * messages back and forth. + * + * It consists 'messages' that pack one or more payloads each. The + * format is described in detail in drivers/net/wimax/i2400m/rx.c and + * tx.c. + * + * + * THE L3L4 PROTOCOL + * + * The term L3L4 refers to Layer 3 (the device), Layer 4 (the + * driver/host software). + * + * This is the control protocol used by the host to control the i2400m + * device (scan, connect, disconnect...). This is sent to / received + * as control frames. These frames consist of a header and zero or + * more TLVs with information. We call each control frame a "message". + * + * Each message is composed of: + * + * HEADER + * [TLV0 + PAYLOAD0] + * [TLV1 + PAYLOAD1] + * [...] + * [TLVN + PAYLOADN] + * + * The HEADER is defined by 'struct i2400m_l3l4_hdr'. The payloads are + * defined by a TLV structure (Type Length Value) which is a 'header' + * (struct i2400m_tlv_hdr) and then the payload. + * + * All integers are represented as Little Endian. + * + * - REQUESTS AND EVENTS + * + * The requests can be clasified as follows: + * + * COMMAND: implies a request from the host to the device requesting + * an action being performed. The device will reply with a + * message (with the same type as the command), status and + * no (TLV) payload. Execution of a command might cause + * events (of different type) to be sent later on as + * device's state changes. + * + * GET/SET: similar to COMMAND, but will not cause other + * EVENTs. The reply, in the case of GET, will contain + * TLVs with the requested information. + * + * EVENT: asynchronous messages sent from the device, maybe as a + * consequence of previous COMMANDs but disassociated from + * them. + * + * Only one request might be pending at the same time (ie: don't + * parallelize nor post another GET request before the previous + * COMMAND has been acknowledged with it's corresponding reply by the + * device). + * + * The different requests and their formats are described below: + * + * I2400M_MT_* Message types + * I2400M_MS_* Message status (for replies, events) + * i2400m_tlv_* TLVs + * + * data types are named 'struct i2400m_msg_OPNAME', OPNAME matching the + * operation. + */ + +#ifndef __LINUX__WIMAX__I2400M_H__ +#define __LINUX__WIMAX__I2400M_H__ + +#include +#include + +/* + * Host Device Interface (HDI) common to all busses + */ + +/* Boot-mode (firmware upload mode) commands */ + +/* Header for the firmware file */ +struct i2400m_bcf_hdr { + __le32 module_type; + __le32 header_len; + __le32 header_version; + __le32 module_id; + __le32 module_vendor; + __le32 date; /* BCD YYYMMDD */ + __le32 size; /* in dwords */ + __le32 key_size; /* in dwords */ + __le32 modulus_size; /* in dwords */ + __le32 exponent_size; /* in dwords */ + __u8 reserved[88]; +} __attribute__ ((packed)); + +/* Boot mode opcodes */ +enum i2400m_brh_opcode { + I2400M_BRH_READ = 1, + I2400M_BRH_WRITE = 2, + I2400M_BRH_JUMP = 3, + I2400M_BRH_SIGNED_JUMP = 8, + I2400M_BRH_HASH_PAYLOAD_ONLY = 9, +}; + +/* Boot mode command masks and stuff */ +enum i2400m_brh { + I2400M_BRH_SIGNATURE = 0xcbbc0000, + I2400M_BRH_SIGNATURE_MASK = 0xffff0000, + I2400M_BRH_SIGNATURE_SHIFT = 16, + I2400M_BRH_OPCODE_MASK = 0x0000000f, + I2400M_BRH_RESPONSE_MASK = 0x000000f0, + I2400M_BRH_RESPONSE_SHIFT = 4, + I2400M_BRH_DIRECT_ACCESS = 0x00000400, + I2400M_BRH_RESPONSE_REQUIRED = 0x00000200, + I2400M_BRH_USE_CHECKSUM = 0x00000100, +}; + + +/** + * i2400m_bootrom_header - Header for a boot-mode command + * + * @cmd: the above command descriptor + * @target_addr: where on the device memory should the action be performed. + * @data_size: for read/write, amount of data to be read/written + * @block_checksum: checksum value (if applicable) + * @payload: the beginning of data attached to this header + */ +struct i2400m_bootrom_header { + __le32 command; /* Compose with enum i2400_brh */ + __le32 target_addr; + __le32 data_size; + __le32 block_checksum; + char payload[0]; +} __attribute__ ((packed)); + + +/* + * Data / control protocol + */ + +/* Packet types for the host-device interface */ +enum i2400m_pt { + I2400M_PT_DATA = 0, + I2400M_PT_CTRL, + I2400M_PT_TRACE, /* For device debug */ + I2400M_PT_RESET_WARM, /* device reset */ + I2400M_PT_RESET_COLD, /* USB[transport] reset, like reconnect */ + I2400M_PT_EDATA, /* Extended RX data */ + I2400M_PT_ILLEGAL +}; + + +/* + * Payload for a data packet + * + * This is prefixed to each and every outgoing DATA type. + */ +struct i2400m_pl_data_hdr { + __le32 reserved; +} __attribute__((packed)); + + +/* + * Payload for an extended data packet + * + * New in fw v1.4 + * + * @reorder: if this payload has to be reorder or not (and how) + * @cs: the type of data in the packet, as defined per (802.16e + * T11.13.19.1). Currently only 2 (IPv4 packet) supported. + * + * This is prefixed to each and every INCOMING DATA packet. + */ +struct i2400m_pl_edata_hdr { + __le32 reorder; /* bits defined in i2400m_ro */ + __u8 cs; + __u8 reserved[11]; +} __attribute__((packed)); + +enum i2400m_cs { + I2400M_CS_IPV4_0 = 0, + I2400M_CS_IPV4 = 2, +}; + +enum i2400m_ro { + I2400M_RO_NEEDED = 0x01, + I2400M_RO_TYPE = 0x03, + I2400M_RO_TYPE_SHIFT = 1, + I2400M_RO_CIN = 0x0f, + I2400M_RO_CIN_SHIFT = 4, + I2400M_RO_FBN = 0x07ff, + I2400M_RO_FBN_SHIFT = 8, + I2400M_RO_SN = 0x07ff, + I2400M_RO_SN_SHIFT = 21, +}; + +enum i2400m_ro_type { + I2400M_RO_TYPE_RESET = 0, + I2400M_RO_TYPE_PACKET, + I2400M_RO_TYPE_WS, + I2400M_RO_TYPE_PACKET_WS, +}; + + +/* Misc constants */ +enum { + I2400M_PL_ALIGN = 16, /* Payload data size alignment */ + I2400M_PL_SIZE_MAX = 0x3EFF, + I2400M_MAX_PLS_IN_MSG = 60, + /* protocol barkers: sync sequences; for notifications they + * are sent in groups of four. */ + I2400M_H2D_PREVIEW_BARKER = 0xcafe900d, + I2400M_COLD_RESET_BARKER = 0xc01dc01d, + I2400M_WARM_RESET_BARKER = 0x50f750f7, + I2400M_NBOOT_BARKER = 0xdeadbeef, + I2400M_SBOOT_BARKER = 0x0ff1c1a1, + I2400M_SBOOT_BARKER_6050 = 0x80000001, + I2400M_ACK_BARKER = 0xfeedbabe, + I2400M_D2H_MSG_BARKER = 0xbeefbabe, +}; + + +/* + * Hardware payload descriptor + * + * Bitfields encoded in a struct to enforce typing semantics. + * + * Look in rx.c and tx.c for a full description of the format. + */ +struct i2400m_pld { + __le32 val; +} __attribute__ ((packed)); + +#define I2400M_PLD_SIZE_MASK 0x00003fff +#define I2400M_PLD_TYPE_SHIFT 16 +#define I2400M_PLD_TYPE_MASK 0x000f0000 + +/* + * Header for a TX message or RX message + * + * @barker: preamble + * @size: used for management of the FIFO queue buffer; before + * sending, this is converted to be a real preamble. This + * indicates the real size of the TX message that starts at this + * point. If the highest bit is set, then this message is to be + * skipped. + * @sequence: sequence number of this message + * @offset: offset where the message itself starts -- see the comments + * in the file header about message header and payload descriptor + * alignment. + * @num_pls: number of payloads in this message + * @padding: amount of padding bytes at the end of the message to make + * it be of block-size aligned + * + * Look in rx.c and tx.c for a full description of the format. + */ +struct i2400m_msg_hdr { + union { + __le32 barker; + __u32 size; /* same size type as barker!! */ + }; + union { + __le32 sequence; + __u32 offset; /* same size type as barker!! */ + }; + __le16 num_pls; + __le16 rsv1; + __le16 padding; + __le16 rsv2; + struct i2400m_pld pld[0]; +} __attribute__ ((packed)); + + + +/* + * L3/L4 control protocol + */ + +enum { + /* Interface version */ + I2400M_L3L4_VERSION = 0x0100, +}; + +/* Message types */ +enum i2400m_mt { + I2400M_MT_RESERVED = 0x0000, + I2400M_MT_INVALID = 0xffff, + I2400M_MT_REPORT_MASK = 0x8000, + + I2400M_MT_GET_SCAN_RESULT = 0x4202, + I2400M_MT_SET_SCAN_PARAM = 0x4402, + I2400M_MT_CMD_RF_CONTROL = 0x4602, + I2400M_MT_CMD_SCAN = 0x4603, + I2400M_MT_CMD_CONNECT = 0x4604, + I2400M_MT_CMD_DISCONNECT = 0x4605, + I2400M_MT_CMD_EXIT_IDLE = 0x4606, + I2400M_MT_GET_LM_VERSION = 0x5201, + I2400M_MT_GET_DEVICE_INFO = 0x5202, + I2400M_MT_GET_LINK_STATUS = 0x5203, + I2400M_MT_GET_STATISTICS = 0x5204, + I2400M_MT_GET_STATE = 0x5205, + I2400M_MT_GET_MEDIA_STATUS = 0x5206, + I2400M_MT_SET_INIT_CONFIG = 0x5404, + I2400M_MT_CMD_INIT = 0x5601, + I2400M_MT_CMD_TERMINATE = 0x5602, + I2400M_MT_CMD_MODE_OF_OP = 0x5603, + I2400M_MT_CMD_RESET_DEVICE = 0x5604, + I2400M_MT_CMD_MONITOR_CONTROL = 0x5605, + I2400M_MT_CMD_ENTER_POWERSAVE = 0x5606, + I2400M_MT_GET_TLS_OPERATION_RESULT = 0x6201, + I2400M_MT_SET_EAP_SUCCESS = 0x6402, + I2400M_MT_SET_EAP_FAIL = 0x6403, + I2400M_MT_SET_EAP_KEY = 0x6404, + I2400M_MT_CMD_SEND_EAP_RESPONSE = 0x6602, + I2400M_MT_REPORT_SCAN_RESULT = 0xc002, + I2400M_MT_REPORT_STATE = 0xd002, + I2400M_MT_REPORT_POWERSAVE_READY = 0xd005, + I2400M_MT_REPORT_EAP_REQUEST = 0xe002, + I2400M_MT_REPORT_EAP_RESTART = 0xe003, + I2400M_MT_REPORT_ALT_ACCEPT = 0xe004, + I2400M_MT_REPORT_KEY_REQUEST = 0xe005, +}; + + +/* + * Message Ack Status codes + * + * When a message is replied-to, this status is reported. + */ +enum i2400m_ms { + I2400M_MS_DONE_OK = 0, + I2400M_MS_DONE_IN_PROGRESS = 1, + I2400M_MS_INVALID_OP = 2, + I2400M_MS_BAD_STATE = 3, + I2400M_MS_ILLEGAL_VALUE = 4, + I2400M_MS_MISSING_PARAMS = 5, + I2400M_MS_VERSION_ERROR = 6, + I2400M_MS_ACCESSIBILITY_ERROR = 7, + I2400M_MS_BUSY = 8, + I2400M_MS_CORRUPTED_TLV = 9, + I2400M_MS_UNINITIALIZED = 10, + I2400M_MS_UNKNOWN_ERROR = 11, + I2400M_MS_PRODUCTION_ERROR = 12, + I2400M_MS_NO_RF = 13, + I2400M_MS_NOT_READY_FOR_POWERSAVE = 14, + I2400M_MS_THERMAL_CRITICAL = 15, + I2400M_MS_MAX +}; + + +/** + * i2400m_tlv - enumeration of the different types of TLVs + * + * TLVs stand for type-length-value and are the header for a payload + * composed of almost anything. Each payload has a type assigned + * and a length. + */ +enum i2400m_tlv { + I2400M_TLV_L4_MESSAGE_VERSIONS = 129, + I2400M_TLV_SYSTEM_STATE = 141, + I2400M_TLV_MEDIA_STATUS = 161, + I2400M_TLV_RF_OPERATION = 162, + I2400M_TLV_RF_STATUS = 163, + I2400M_TLV_DEVICE_RESET_TYPE = 132, + I2400M_TLV_CONFIG_IDLE_PARAMETERS = 601, + I2400M_TLV_CONFIG_IDLE_TIMEOUT = 611, + I2400M_TLV_CONFIG_D2H_DATA_FORMAT = 614, + I2400M_TLV_CONFIG_DL_HOST_REORDER = 615, +}; + + +struct i2400m_tlv_hdr { + __le16 type; + __le16 length; /* payload's */ + __u8 pl[0]; +} __attribute__((packed)); + + +struct i2400m_l3l4_hdr { + __le16 type; + __le16 length; /* payload's */ + __le16 version; + __le16 resv1; + __le16 status; + __le16 resv2; + struct i2400m_tlv_hdr pl[0]; +} __attribute__((packed)); + + +/** + * i2400m_system_state - different states of the device + */ +enum i2400m_system_state { + I2400M_SS_UNINITIALIZED = 1, + I2400M_SS_INIT, + I2400M_SS_READY, + I2400M_SS_SCAN, + I2400M_SS_STANDBY, + I2400M_SS_CONNECTING, + I2400M_SS_WIMAX_CONNECTED, + I2400M_SS_DATA_PATH_CONNECTED, + I2400M_SS_IDLE, + I2400M_SS_DISCONNECTING, + I2400M_SS_OUT_OF_ZONE, + I2400M_SS_SLEEPACTIVE, + I2400M_SS_PRODUCTION, + I2400M_SS_CONFIG, + I2400M_SS_RF_OFF, + I2400M_SS_RF_SHUTDOWN, + I2400M_SS_DEVICE_DISCONNECT, + I2400M_SS_MAX, +}; + + +/** + * i2400m_tlv_system_state - report on the state of the system + * + * @state: see enum i2400m_system_state + */ +struct i2400m_tlv_system_state { + struct i2400m_tlv_hdr hdr; + __le32 state; +} __attribute__((packed)); + + +struct i2400m_tlv_l4_message_versions { + struct i2400m_tlv_hdr hdr; + __le16 major; + __le16 minor; + __le16 branch; + __le16 reserved; +} __attribute__((packed)); + + +struct i2400m_tlv_detailed_device_info { + struct i2400m_tlv_hdr hdr; + __u8 reserved1[400]; + __u8 mac_address[ETH_ALEN]; + __u8 reserved2[2]; +} __attribute__((packed)); + + +enum i2400m_rf_switch_status { + I2400M_RF_SWITCH_ON = 1, + I2400M_RF_SWITCH_OFF = 2, +}; + +struct i2400m_tlv_rf_switches_status { + struct i2400m_tlv_hdr hdr; + __u8 sw_rf_switch; /* 1 ON, 2 OFF */ + __u8 hw_rf_switch; /* 1 ON, 2 OFF */ + __u8 reserved[2]; +} __attribute__((packed)); + + +enum { + i2400m_rf_operation_on = 1, + i2400m_rf_operation_off = 2 +}; + +struct i2400m_tlv_rf_operation { + struct i2400m_tlv_hdr hdr; + __le32 status; /* 1 ON, 2 OFF */ +} __attribute__((packed)); + + +enum i2400m_tlv_reset_type { + I2400M_RESET_TYPE_COLD = 1, + I2400M_RESET_TYPE_WARM +}; + +struct i2400m_tlv_device_reset_type { + struct i2400m_tlv_hdr hdr; + __le32 reset_type; +} __attribute__((packed)); + + +struct i2400m_tlv_config_idle_parameters { + struct i2400m_tlv_hdr hdr; + __le32 idle_timeout; /* 100 to 300000 ms [5min], 100 increments + * 0 disabled */ + __le32 idle_paging_interval; /* frames */ +} __attribute__((packed)); + + +enum i2400m_media_status { + I2400M_MEDIA_STATUS_LINK_UP = 1, + I2400M_MEDIA_STATUS_LINK_DOWN, + I2400M_MEDIA_STATUS_LINK_RENEW, +}; + +struct i2400m_tlv_media_status { + struct i2400m_tlv_hdr hdr; + __le32 media_status; +} __attribute__((packed)); + + +/* New in v1.4 */ +struct i2400m_tlv_config_idle_timeout { + struct i2400m_tlv_hdr hdr; + __le32 timeout; /* 100 to 300000 ms [5min], 100 increments + * 0 disabled */ +} __attribute__((packed)); + +/* New in v1.4 -- for backward compat, will be removed */ +struct i2400m_tlv_config_d2h_data_format { + struct i2400m_tlv_hdr hdr; + __u8 format; /* 0 old format, 1 enhanced */ + __u8 reserved[3]; +} __attribute__((packed)); + +/* New in v1.4 */ +struct i2400m_tlv_config_dl_host_reorder { + struct i2400m_tlv_hdr hdr; + __u8 reorder; /* 0 disabled, 1 enabled */ + __u8 reserved[3]; +} __attribute__((packed)); + + +#endif /* #ifndef __LINUX__WIMAX__I2400M_H__ */ diff --git a/drivers/staging/wimax/i2400m/netdev.c b/drivers/staging/wimax/i2400m/netdev.c new file mode 100644 index 000000000000..a7fcbceb6e6b --- /dev/null +++ b/drivers/staging/wimax/i2400m/netdev.c @@ -0,0 +1,603 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Wireless WiMAX Connection 2400m + * Glue with the networking stack + * + * Copyright (C) 2007 Intel Corporation + * Yanir Lubetkin + * Inaky Perez-Gonzalez + * + * This implements an ethernet device for the i2400m. + * + * We fake being an ethernet device to simplify the support from user + * space and from the other side. The world is (sadly) configured to + * take in only Ethernet devices... + * + * Because of this, when using firmwares <= v1.3, there is an + * copy-each-rxed-packet overhead on the RX path. Each IP packet has + * to be reallocated to add an ethernet header (as there is no space + * in what we get from the device). This is a known drawback and + * firmwares >= 1.4 add header space that can be used to insert the + * ethernet header without having to reallocate and copy. + * + * TX error handling is tricky; because we have to FIFO/queue the + * buffers for transmission (as the hardware likes it aggregated), we + * just give the skb to the TX subsystem and by the time it is + * transmitted, we have long forgotten about it. So we just don't care + * too much about it. + * + * Note that when the device is in idle mode with the basestation, we + * need to negotiate coming back up online. That involves negotiation + * and possible user space interaction. Thus, we defer to a workqueue + * to do all that. By default, we only queue a single packet and drop + * the rest, as potentially the time to go back from idle to normal is + * long. + * + * ROADMAP + * + * i2400m_open Called on ifconfig up + * i2400m_stop Called on ifconfig down + * + * i2400m_hard_start_xmit Called by the network stack to send a packet + * i2400m_net_wake_tx Wake up device from basestation-IDLE & TX + * i2400m_wake_tx_work + * i2400m_cmd_exit_idle + * i2400m_tx + * i2400m_net_tx TX a data frame + * i2400m_tx + * + * i2400m_change_mtu Called on ifconfig mtu XXX + * + * i2400m_tx_timeout Called when the device times out + * + * i2400m_net_rx Called by the RX code when a data frame is + * available (firmware <= 1.3) + * i2400m_net_erx Called by the RX code when a data frame is + * available (firmware >= 1.4). + * i2400m_netdev_setup Called to setup all the netdev stuff from + * alloc_netdev. + */ +#include +#include +#include +#include +#include +#include "i2400m.h" + + +#define D_SUBMODULE netdev +#include "debug-levels.h" + +enum { +/* netdev interface */ + /* 20 secs? yep, this is the maximum timeout that the device + * might take to get out of IDLE / negotiate it with the base + * station. We add 1sec for good measure. */ + I2400M_TX_TIMEOUT = 21 * HZ, + /* + * Experimentation has determined that, 20 to be a good value + * for minimizing the jitter in the throughput. + */ + I2400M_TX_QLEN = 20, +}; + + +static +int i2400m_open(struct net_device *net_dev) +{ + int result; + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(net_dev %p [i2400m %p])\n", net_dev, i2400m); + /* Make sure we wait until init is complete... */ + mutex_lock(&i2400m->init_mutex); + if (i2400m->updown) + result = 0; + else + result = -EBUSY; + mutex_unlock(&i2400m->init_mutex); + d_fnend(3, dev, "(net_dev %p [i2400m %p]) = %d\n", + net_dev, i2400m, result); + return result; +} + + +static +int i2400m_stop(struct net_device *net_dev) +{ + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(net_dev %p [i2400m %p])\n", net_dev, i2400m); + i2400m_net_wake_stop(i2400m); + d_fnend(3, dev, "(net_dev %p [i2400m %p]) = 0\n", net_dev, i2400m); + return 0; +} + + +/* + * Wake up the device and transmit a held SKB, then restart the net queue + * + * When the device goes into basestation-idle mode, we need to tell it + * to exit that mode; it will negotiate with the base station, user + * space may have to intervene to rehandshake crypto and then tell us + * when it is ready to transmit the packet we have "queued". Still we + * need to give it sometime after it reports being ok. + * + * On error, there is not much we can do. If the error was on TX, we + * still wake the queue up to see if the next packet will be luckier. + * + * If _cmd_exit_idle() fails...well, it could be many things; most + * commonly it is that something else took the device out of IDLE mode + * (for example, the base station). In that case we get an -EILSEQ and + * we are just going to ignore that one. If the device is back to + * connected, then fine -- if it is someother state, the packet will + * be dropped anyway. + */ +void i2400m_wake_tx_work(struct work_struct *ws) +{ + int result; + struct i2400m *i2400m = container_of(ws, struct i2400m, wake_tx_ws); + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *skb; + unsigned long flags; + + spin_lock_irqsave(&i2400m->tx_lock, flags); + skb = i2400m->wake_tx_skb; + i2400m->wake_tx_skb = NULL; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + + d_fnstart(3, dev, "(ws %p i2400m %p skb %p)\n", ws, i2400m, skb); + result = -EINVAL; + if (skb == NULL) { + dev_err(dev, "WAKE&TX: skb disappeared!\n"); + goto out_put; + } + /* If we have, somehow, lost the connection after this was + * queued, don't do anything; this might be the device got + * reset or just disconnected. */ + if (unlikely(!netif_carrier_ok(net_dev))) + goto out_kfree; + result = i2400m_cmd_exit_idle(i2400m); + if (result == -EILSEQ) + result = 0; + if (result < 0) { + dev_err(dev, "WAKE&TX: device didn't get out of idle: " + "%d - resetting\n", result); + i2400m_reset(i2400m, I2400M_RT_BUS); + goto error; + } + result = wait_event_timeout(i2400m->state_wq, + i2400m->state != I2400M_SS_IDLE, + net_dev->watchdog_timeo - HZ/2); + if (result == 0) + result = -ETIMEDOUT; + if (result < 0) { + dev_err(dev, "WAKE&TX: error waiting for device to exit IDLE: " + "%d - resetting\n", result); + i2400m_reset(i2400m, I2400M_RT_BUS); + goto error; + } + msleep(20); /* device still needs some time or it drops it */ + result = i2400m_tx(i2400m, skb->data, skb->len, I2400M_PT_DATA); +error: + netif_wake_queue(net_dev); +out_kfree: + kfree_skb(skb); /* refcount transferred by _hard_start_xmit() */ +out_put: + i2400m_put(i2400m); + d_fnend(3, dev, "(ws %p i2400m %p skb %p) = void [%d]\n", + ws, i2400m, skb, result); +} + + +/* + * Prepare the data payload TX header + * + * The i2400m expects a 4 byte header in front of a data packet. + * + * Because we pretend to be an ethernet device, this packet comes with + * an ethernet header. Pull it and push our header. + */ +static +void i2400m_tx_prep_header(struct sk_buff *skb) +{ + struct i2400m_pl_data_hdr *pl_hdr; + skb_pull(skb, ETH_HLEN); + pl_hdr = skb_push(skb, sizeof(*pl_hdr)); + pl_hdr->reserved = 0; +} + + + +/* + * Cleanup resources acquired during i2400m_net_wake_tx() + * + * This is called by __i2400m_dev_stop and means we have to make sure + * the workqueue is flushed from any pending work. + */ +void i2400m_net_wake_stop(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *wake_tx_skb; + unsigned long flags; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + /* + * See i2400m_hard_start_xmit(), references are taken there and + * here we release them if the packet was still pending. + */ + cancel_work_sync(&i2400m->wake_tx_ws); + + spin_lock_irqsave(&i2400m->tx_lock, flags); + wake_tx_skb = i2400m->wake_tx_skb; + i2400m->wake_tx_skb = NULL; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + + if (wake_tx_skb) { + i2400m_put(i2400m); + kfree_skb(wake_tx_skb); + } + + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} + + +/* + * TX an skb to an idle device + * + * When the device is in basestation-idle mode, we need to wake it up + * and then TX. So we queue a work_struct for doing so. + * + * We need to get an extra ref for the skb (so it is not dropped), as + * well as be careful not to queue more than one request (won't help + * at all). If more than one request comes or there are errors, we + * just drop the packets (see i2400m_hard_start_xmit()). + */ +static +int i2400m_net_wake_tx(struct i2400m *i2400m, struct net_device *net_dev, + struct sk_buff *skb) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + unsigned long flags; + + d_fnstart(3, dev, "(skb %p net_dev %p)\n", skb, net_dev); + if (net_ratelimit()) { + d_printf(3, dev, "WAKE&NETTX: " + "skb %p sending %d bytes to radio\n", + skb, skb->len); + d_dump(4, dev, skb->data, skb->len); + } + /* We hold a ref count for i2400m and skb, so when + * stopping() the device, we need to cancel that work + * and if pending, release those resources. */ + result = 0; + spin_lock_irqsave(&i2400m->tx_lock, flags); + if (!i2400m->wake_tx_skb) { + netif_stop_queue(net_dev); + i2400m_get(i2400m); + i2400m->wake_tx_skb = skb_get(skb); /* transfer ref count */ + i2400m_tx_prep_header(skb); + result = schedule_work(&i2400m->wake_tx_ws); + WARN_ON(result == 0); + } + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + if (result == 0) { + /* Yes, this happens even if we stopped the + * queue -- blame the queue disciplines that + * queue without looking -- I guess there is a reason + * for that. */ + if (net_ratelimit()) + d_printf(1, dev, "NETTX: device exiting idle, " + "dropping skb %p, queue running %d\n", + skb, netif_queue_stopped(net_dev)); + result = -EBUSY; + } + d_fnend(3, dev, "(skb %p net_dev %p) = %d\n", skb, net_dev, result); + return result; +} + + +/* + * Transmit a packet to the base station on behalf of the network stack. + * + * Returns: 0 if ok, < 0 errno code on error. + * + * We need to pull the ethernet header and add the hardware header, + * which is currently set to all zeroes and reserved. + */ +static +int i2400m_net_tx(struct i2400m *i2400m, struct net_device *net_dev, + struct sk_buff *skb) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p net_dev %p skb %p)\n", + i2400m, net_dev, skb); + /* FIXME: check eth hdr, only IPv4 is routed by the device as of now */ + netif_trans_update(net_dev); + i2400m_tx_prep_header(skb); + d_printf(3, dev, "NETTX: skb %p sending %d bytes to radio\n", + skb, skb->len); + d_dump(4, dev, skb->data, skb->len); + result = i2400m_tx(i2400m, skb->data, skb->len, I2400M_PT_DATA); + d_fnend(3, dev, "(i2400m %p net_dev %p skb %p) = %d\n", + i2400m, net_dev, skb, result); + return result; +} + + +/* + * Transmit a packet to the base station on behalf of the network stack + * + * + * Returns: NETDEV_TX_OK (always, even in case of error) + * + * In case of error, we just drop it. Reasons: + * + * - we add a hw header to each skb, and if the network stack + * retries, we have no way to know if that skb has it or not. + * + * - network protocols have their own drop-recovery mechanisms + * + * - there is not much else we can do + * + * If the device is idle, we need to wake it up; that is an operation + * that will sleep. See i2400m_net_wake_tx() for details. + */ +static +netdev_tx_t i2400m_hard_start_xmit(struct sk_buff *skb, + struct net_device *net_dev) +{ + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + struct device *dev = i2400m_dev(i2400m); + int result = -1; + + d_fnstart(3, dev, "(skb %p net_dev %p)\n", skb, net_dev); + + if (skb_cow_head(skb, 0)) + goto drop; + + if (i2400m->state == I2400M_SS_IDLE) + result = i2400m_net_wake_tx(i2400m, net_dev, skb); + else + result = i2400m_net_tx(i2400m, net_dev, skb); + if (result < 0) { +drop: + net_dev->stats.tx_dropped++; + } else { + net_dev->stats.tx_packets++; + net_dev->stats.tx_bytes += skb->len; + } + dev_kfree_skb(skb); + d_fnend(3, dev, "(skb %p net_dev %p) = %d\n", skb, net_dev, result); + return NETDEV_TX_OK; +} + + +static +void i2400m_tx_timeout(struct net_device *net_dev, unsigned int txqueue) +{ + /* + * We might want to kick the device + * + * There is not much we can do though, as the device requires + * that we send the data aggregated. By the time we receive + * this, there might be data pending to be sent or not... + */ + net_dev->stats.tx_errors++; +} + + +/* + * Create a fake ethernet header + * + * For emulating an ethernet device, every received IP header has to + * be prefixed with an ethernet header. Fake it with the given + * protocol. + */ +static +void i2400m_rx_fake_eth_header(struct net_device *net_dev, + void *_eth_hdr, __be16 protocol) +{ + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + struct ethhdr *eth_hdr = _eth_hdr; + + memcpy(eth_hdr->h_dest, net_dev->dev_addr, sizeof(eth_hdr->h_dest)); + memcpy(eth_hdr->h_source, i2400m->src_mac_addr, + sizeof(eth_hdr->h_source)); + eth_hdr->h_proto = protocol; +} + + +/* + * i2400m_net_rx - pass a network packet to the stack + * + * @i2400m: device instance + * @skb_rx: the skb where the buffer pointed to by @buf is + * @i: 1 if payload is the only one + * @buf: pointer to the buffer containing the data + * @len: buffer's length + * + * This is only used now for the v1.3 firmware. It will be deprecated + * in >= 2.6.31. + * + * Note that due to firmware limitations, we don't have space to add + * an ethernet header, so we need to copy each packet. Firmware + * versions >= v1.4 fix this [see i2400m_net_erx()]. + * + * We just clone the skb and set it up so that it's skb->data pointer + * points to "buf" and it's length. + * + * Note that if the payload is the last (or the only one) in a + * multi-payload message, we don't clone the SKB but just reuse it. + * + * This function is normally run from a thread context. However, we + * still use netif_rx() instead of netif_receive_skb() as was + * recommended in the mailing list. Reason is in some stress tests + * when sending/receiving a lot of data we seem to hit a softlock in + * the kernel's TCP implementation [aroudn tcp_delay_timer()]. Using + * netif_rx() took care of the issue. + * + * This is, of course, still open to do more research on why running + * with netif_receive_skb() hits this softlock. FIXME. + * + * FIXME: currently we don't do any efforts at distinguishing if what + * we got was an IPv4 or IPv6 header, to setup the protocol field + * correctly. + */ +void i2400m_net_rx(struct i2400m *i2400m, struct sk_buff *skb_rx, + unsigned i, const void *buf, int buf_len) +{ + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *skb; + + d_fnstart(2, dev, "(i2400m %p buf %p buf_len %d)\n", + i2400m, buf, buf_len); + if (i) { + skb = skb_get(skb_rx); + d_printf(2, dev, "RX: reusing first payload skb %p\n", skb); + skb_pull(skb, buf - (void *) skb->data); + skb_trim(skb, (void *) skb_end_pointer(skb) - buf); + } else { + /* Yes, this is bad -- a lot of overhead -- see + * comments at the top of the file */ + skb = __netdev_alloc_skb(net_dev, buf_len, GFP_KERNEL); + if (skb == NULL) { + dev_err(dev, "NETRX: no memory to realloc skb\n"); + net_dev->stats.rx_dropped++; + goto error_skb_realloc; + } + skb_put_data(skb, buf, buf_len); + } + i2400m_rx_fake_eth_header(i2400m->wimax_dev.net_dev, + skb->data - ETH_HLEN, + cpu_to_be16(ETH_P_IP)); + skb_set_mac_header(skb, -ETH_HLEN); + skb->dev = i2400m->wimax_dev.net_dev; + skb->protocol = htons(ETH_P_IP); + net_dev->stats.rx_packets++; + net_dev->stats.rx_bytes += buf_len; + d_printf(3, dev, "NETRX: receiving %d bytes to network stack\n", + buf_len); + d_dump(4, dev, buf, buf_len); + netif_rx_ni(skb); /* see notes in function header */ +error_skb_realloc: + d_fnend(2, dev, "(i2400m %p buf %p buf_len %d) = void\n", + i2400m, buf, buf_len); +} + + +/* + * i2400m_net_erx - pass a network packet to the stack (extended version) + * + * @i2400m: device descriptor + * @skb: the skb where the packet is - the skb should be set to point + * at the IP packet; this function will add ethernet headers if + * needed. + * @cs: packet type + * + * This is only used now for firmware >= v1.4. Note it is quite + * similar to i2400m_net_rx() (used only for v1.3 firmware). + * + * This function is normally run from a thread context. However, we + * still use netif_rx() instead of netif_receive_skb() as was + * recommended in the mailing list. Reason is in some stress tests + * when sending/receiving a lot of data we seem to hit a softlock in + * the kernel's TCP implementation [aroudn tcp_delay_timer()]. Using + * netif_rx() took care of the issue. + * + * This is, of course, still open to do more research on why running + * with netif_receive_skb() hits this softlock. FIXME. + */ +void i2400m_net_erx(struct i2400m *i2400m, struct sk_buff *skb, + enum i2400m_cs cs) +{ + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(2, dev, "(i2400m %p skb %p [%u] cs %d)\n", + i2400m, skb, skb->len, cs); + switch(cs) { + case I2400M_CS_IPV4_0: + case I2400M_CS_IPV4: + i2400m_rx_fake_eth_header(i2400m->wimax_dev.net_dev, + skb->data - ETH_HLEN, + cpu_to_be16(ETH_P_IP)); + skb_set_mac_header(skb, -ETH_HLEN); + skb->dev = i2400m->wimax_dev.net_dev; + skb->protocol = htons(ETH_P_IP); + net_dev->stats.rx_packets++; + net_dev->stats.rx_bytes += skb->len; + break; + default: + dev_err(dev, "ERX: BUG? CS type %u unsupported\n", cs); + goto error; + + } + d_printf(3, dev, "ERX: receiving %d bytes to the network stack\n", + skb->len); + d_dump(4, dev, skb->data, skb->len); + netif_rx_ni(skb); /* see notes in function header */ +error: + d_fnend(2, dev, "(i2400m %p skb %p [%u] cs %d) = void\n", + i2400m, skb, skb->len, cs); +} + +static const struct net_device_ops i2400m_netdev_ops = { + .ndo_open = i2400m_open, + .ndo_stop = i2400m_stop, + .ndo_start_xmit = i2400m_hard_start_xmit, + .ndo_tx_timeout = i2400m_tx_timeout, +}; + +static void i2400m_get_drvinfo(struct net_device *net_dev, + struct ethtool_drvinfo *info) +{ + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + + strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); + strlcpy(info->fw_version, i2400m->fw_name ? : "", + sizeof(info->fw_version)); + if (net_dev->dev.parent) + strlcpy(info->bus_info, dev_name(net_dev->dev.parent), + sizeof(info->bus_info)); +} + +static const struct ethtool_ops i2400m_ethtool_ops = { + .get_drvinfo = i2400m_get_drvinfo, + .get_link = ethtool_op_get_link, +}; + +/** + * i2400m_netdev_setup - Setup setup @net_dev's i2400m private data + * + * Called by alloc_netdev() + */ +void i2400m_netdev_setup(struct net_device *net_dev) +{ + d_fnstart(3, NULL, "(net_dev %p)\n", net_dev); + ether_setup(net_dev); + net_dev->mtu = I2400M_MAX_MTU; + net_dev->min_mtu = 0; + net_dev->max_mtu = I2400M_MAX_MTU; + net_dev->tx_queue_len = I2400M_TX_QLEN; + net_dev->features = + NETIF_F_VLAN_CHALLENGED + | NETIF_F_HIGHDMA; + net_dev->flags = + IFF_NOARP /* i2400m is apure IP device */ + & (~IFF_BROADCAST /* i2400m is P2P */ + & ~IFF_MULTICAST); + net_dev->watchdog_timeo = I2400M_TX_TIMEOUT; + net_dev->netdev_ops = &i2400m_netdev_ops; + net_dev->ethtool_ops = &i2400m_ethtool_ops; + d_fnend(3, NULL, "(net_dev %p) = void\n", net_dev); +} +EXPORT_SYMBOL_GPL(i2400m_netdev_setup); + diff --git a/drivers/staging/wimax/i2400m/op-rfkill.c b/drivers/staging/wimax/i2400m/op-rfkill.c new file mode 100644 index 000000000000..fbddf2e18c14 --- /dev/null +++ b/drivers/staging/wimax/i2400m/op-rfkill.c @@ -0,0 +1,196 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Wireless WiMAX Connection 2400m + * Implement backend for the WiMAX stack rfkill support + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + * + * The WiMAX kernel stack integrates into RF-Kill and keeps the + * switches's status. We just need to: + * + * - report changes in the HW RF Kill switch [with + * wimax_rfkill_{sw,hw}_report(), which happens when we detect those + * indications coming through hardware reports]. We also do it on + * initialization to let the stack know the initial HW state. + * + * - implement indications from the stack to change the SW RF Kill + * switch (coming from sysfs, the wimax stack or user space). + */ +#include "i2400m.h" +#include "linux-wimax-i2400m.h" +#include + + + +#define D_SUBMODULE rfkill +#include "debug-levels.h" + +/* + * Return true if the i2400m radio is in the requested wimax_rf_state state + * + */ +static +int i2400m_radio_is(struct i2400m *i2400m, enum wimax_rf_state state) +{ + if (state == WIMAX_RF_OFF) + return i2400m->state == I2400M_SS_RF_OFF + || i2400m->state == I2400M_SS_RF_SHUTDOWN; + else if (state == WIMAX_RF_ON) + /* state == WIMAX_RF_ON */ + return i2400m->state != I2400M_SS_RF_OFF + && i2400m->state != I2400M_SS_RF_SHUTDOWN; + else { + BUG(); + return -EINVAL; /* shut gcc warnings on certain arches */ + } +} + + +/* + * WiMAX stack operation: implement SW RFKill toggling + * + * @wimax_dev: device descriptor + * @skb: skb where the message has been received; skb->data is + * expected to point to the message payload. + * @genl_info: passed by the generic netlink layer + * + * Generic Netlink will call this function when a message is sent from + * userspace to change the software RF-Kill switch status. + * + * This function will set the device's software RF-Kill switch state to + * match what is requested. + * + * NOTE: the i2400m has a strict state machine; we can only set the + * RF-Kill switch when it is on, the HW RF-Kill is on and the + * device is initialized. So we ignore errors steaming from not + * being in the right state (-EILSEQ). + */ +int i2400m_op_rfkill_sw_toggle(struct wimax_dev *wimax_dev, + enum wimax_rf_state state) +{ + int result; + struct i2400m *i2400m = wimax_dev_to_i2400m(wimax_dev); + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *ack_skb; + struct { + struct i2400m_l3l4_hdr hdr; + struct i2400m_tlv_rf_operation sw_rf; + } __packed *cmd; + char strerr[32]; + + d_fnstart(4, dev, "(wimax_dev %p state %d)\n", wimax_dev, state); + + result = -ENOMEM; + cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); + if (cmd == NULL) + goto error_alloc; + cmd->hdr.type = cpu_to_le16(I2400M_MT_CMD_RF_CONTROL); + cmd->hdr.length = sizeof(cmd->sw_rf); + cmd->hdr.version = cpu_to_le16(I2400M_L3L4_VERSION); + cmd->sw_rf.hdr.type = cpu_to_le16(I2400M_TLV_RF_OPERATION); + cmd->sw_rf.hdr.length = cpu_to_le16(sizeof(cmd->sw_rf.status)); + switch (state) { + case WIMAX_RF_OFF: /* RFKILL ON, radio OFF */ + cmd->sw_rf.status = cpu_to_le32(2); + break; + case WIMAX_RF_ON: /* RFKILL OFF, radio ON */ + cmd->sw_rf.status = cpu_to_le32(1); + break; + default: + BUG(); + } + + ack_skb = i2400m_msg_to_dev(i2400m, cmd, sizeof(*cmd)); + result = PTR_ERR(ack_skb); + if (IS_ERR(ack_skb)) { + dev_err(dev, "Failed to issue 'RF Control' command: %d\n", + result); + goto error_msg_to_dev; + } + result = i2400m_msg_check_status(wimax_msg_data(ack_skb), + strerr, sizeof(strerr)); + if (result < 0) { + dev_err(dev, "'RF Control' (0x%04x) command failed: %d - %s\n", + I2400M_MT_CMD_RF_CONTROL, result, strerr); + goto error_cmd; + } + + /* Now we wait for the state to change to RADIO_OFF or RADIO_ON */ + result = wait_event_timeout( + i2400m->state_wq, i2400m_radio_is(i2400m, state), + 5 * HZ); + if (result == 0) + result = -ETIMEDOUT; + if (result < 0) + dev_err(dev, "Error waiting for device to toggle RF state: " + "%d\n", result); + result = 0; +error_cmd: + kfree_skb(ack_skb); +error_msg_to_dev: +error_alloc: + d_fnend(4, dev, "(wimax_dev %p state %d) = %d\n", + wimax_dev, state, result); + kfree(cmd); + return result; +} + + +/* + * Inform the WiMAX stack of changes in the RF Kill switches reported + * by the device + * + * @i2400m: device descriptor + * @rfss: TLV for RF Switches status; already validated + * + * NOTE: the reports on RF switch status cannot be trusted + * or used until the device is in a state of RADIO_OFF + * or greater. + */ +void i2400m_report_tlv_rf_switches_status( + struct i2400m *i2400m, + const struct i2400m_tlv_rf_switches_status *rfss) +{ + struct device *dev = i2400m_dev(i2400m); + enum i2400m_rf_switch_status hw, sw; + enum wimax_st wimax_state; + + sw = le32_to_cpu(rfss->sw_rf_switch); + hw = le32_to_cpu(rfss->hw_rf_switch); + + d_fnstart(3, dev, "(i2400m %p rfss %p [hw %u sw %u])\n", + i2400m, rfss, hw, sw); + /* We only process rw switch evens when the device has been + * fully initialized */ + wimax_state = wimax_state_get(&i2400m->wimax_dev); + if (wimax_state < WIMAX_ST_RADIO_OFF) { + d_printf(3, dev, "ignoring RF switches report, state %u\n", + wimax_state); + goto out; + } + switch (sw) { + case I2400M_RF_SWITCH_ON: /* RF Kill disabled (radio on) */ + wimax_report_rfkill_sw(&i2400m->wimax_dev, WIMAX_RF_ON); + break; + case I2400M_RF_SWITCH_OFF: /* RF Kill enabled (radio off) */ + wimax_report_rfkill_sw(&i2400m->wimax_dev, WIMAX_RF_OFF); + break; + default: + dev_err(dev, "HW BUG? Unknown RF SW state 0x%x\n", sw); + } + + switch (hw) { + case I2400M_RF_SWITCH_ON: /* RF Kill disabled (radio on) */ + wimax_report_rfkill_hw(&i2400m->wimax_dev, WIMAX_RF_ON); + break; + case I2400M_RF_SWITCH_OFF: /* RF Kill enabled (radio off) */ + wimax_report_rfkill_hw(&i2400m->wimax_dev, WIMAX_RF_OFF); + break; + default: + dev_err(dev, "HW BUG? Unknown RF HW state 0x%x\n", hw); + } +out: + d_fnend(3, dev, "(i2400m %p rfss %p [hw %u sw %u]) = void\n", + i2400m, rfss, hw, sw); +} diff --git a/drivers/staging/wimax/i2400m/rx.c b/drivers/staging/wimax/i2400m/rx.c new file mode 100644 index 000000000000..c9fb619a9e01 --- /dev/null +++ b/drivers/staging/wimax/i2400m/rx.c @@ -0,0 +1,1395 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * Handle incoming traffic and deliver it to the control or data planes + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * - Initial implementation + * Inaky Perez-Gonzalez + * - Use skb_clone(), break up processing in chunks + * - Split transport/device specific + * - Make buffer size dynamic to exert less memory pressure + * - RX reorder support + * + * This handles the RX path. + * + * We receive an RX message from the bus-specific driver, which + * contains one or more payloads that have potentially different + * destinataries (data or control paths). + * + * So we just take that payload from the transport specific code in + * the form of an skb, break it up in chunks (a cloned skb each in the + * case of network packets) and pass it to netdev or to the + * command/ack handler (and from there to the WiMAX stack). + * + * PROTOCOL FORMAT + * + * The format of the buffer is: + * + * HEADER (struct i2400m_msg_hdr) + * PAYLOAD DESCRIPTOR 0 (struct i2400m_pld) + * PAYLOAD DESCRIPTOR 1 + * ... + * PAYLOAD DESCRIPTOR N + * PAYLOAD 0 (raw bytes) + * PAYLOAD 1 + * ... + * PAYLOAD N + * + * See tx.c for a deeper description on alignment requirements and + * other fun facts of it. + * + * DATA PACKETS + * + * In firmwares <= v1.3, data packets have no header for RX, but they + * do for TX (currently unused). + * + * In firmware >= 1.4, RX packets have an extended header (16 + * bytes). This header conveys information for management of host + * reordering of packets (the device offloads storage of the packets + * for reordering to the host). Read below for more information. + * + * The header is used as dummy space to emulate an ethernet header and + * thus be able to act as an ethernet device without having to reallocate. + * + * DATA RX REORDERING + * + * Starting in firmware v1.4, the device can deliver packets for + * delivery with special reordering information; this allows it to + * more effectively do packet management when some frames were lost in + * the radio traffic. + * + * Thus, for RX packets that come out of order, the device gives the + * driver enough information to queue them properly and then at some + * point, the signal to deliver the whole (or part) of the queued + * packets to the networking stack. There are 16 such queues. + * + * This only happens when a packet comes in with the "need reorder" + * flag set in the RX header. When such bit is set, the following + * operations might be indicated: + * + * - reset queue: send all queued packets to the OS + * + * - queue: queue a packet + * + * - update ws: update the queue's window start and deliver queued + * packets that meet the criteria + * + * - queue & update ws: queue a packet, update the window start and + * deliver queued packets that meet the criteria + * + * (delivery criteria: the packet's [normalized] sequence number is + * lower than the new [normalized] window start). + * + * See the i2400m_roq_*() functions for details. + * + * ROADMAP + * + * i2400m_rx + * i2400m_rx_msg_hdr_check + * i2400m_rx_pl_descr_check + * i2400m_rx_payload + * i2400m_net_rx + * i2400m_rx_edata + * i2400m_net_erx + * i2400m_roq_reset + * i2400m_net_erx + * i2400m_roq_queue + * __i2400m_roq_queue + * i2400m_roq_update_ws + * __i2400m_roq_update_ws + * i2400m_net_erx + * i2400m_roq_queue_update_ws + * __i2400m_roq_queue + * __i2400m_roq_update_ws + * i2400m_net_erx + * i2400m_rx_ctl + * i2400m_msg_size_check + * i2400m_report_hook_work [in a workqueue] + * i2400m_report_hook + * wimax_msg_to_user + * i2400m_rx_ctl_ack + * wimax_msg_to_user_alloc + * i2400m_rx_trace + * i2400m_msg_size_check + * wimax_msg + */ +#include +#include +#include +#include +#include +#include +#include +#include "i2400m.h" + + +#define D_SUBMODULE rx +#include "debug-levels.h" + +static int i2400m_rx_reorder_disabled; /* 0 (rx reorder enabled) by default */ +module_param_named(rx_reorder_disabled, i2400m_rx_reorder_disabled, int, 0644); +MODULE_PARM_DESC(rx_reorder_disabled, + "If true, RX reordering will be disabled."); + +struct i2400m_report_hook_args { + struct sk_buff *skb_rx; + const struct i2400m_l3l4_hdr *l3l4_hdr; + size_t size; + struct list_head list_node; +}; + + +/* + * Execute i2400m_report_hook in a workqueue + * + * Goes over the list of queued reports in i2400m->rx_reports and + * processes them. + * + * NOTE: refcounts on i2400m are not needed because we flush the + * workqueue this runs on (i2400m->work_queue) before destroying + * i2400m. + */ +void i2400m_report_hook_work(struct work_struct *ws) +{ + struct i2400m *i2400m = container_of(ws, struct i2400m, rx_report_ws); + struct device *dev = i2400m_dev(i2400m); + struct i2400m_report_hook_args *args, *args_next; + LIST_HEAD(list); + unsigned long flags; + + while (1) { + spin_lock_irqsave(&i2400m->rx_lock, flags); + list_splice_init(&i2400m->rx_reports, &list); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + if (list_empty(&list)) + break; + else + d_printf(1, dev, "processing queued reports\n"); + list_for_each_entry_safe(args, args_next, &list, list_node) { + d_printf(2, dev, "processing queued report %p\n", args); + i2400m_report_hook(i2400m, args->l3l4_hdr, args->size); + kfree_skb(args->skb_rx); + list_del(&args->list_node); + kfree(args); + } + } +} + + +/* + * Flush the list of queued reports + */ +static +void i2400m_report_hook_flush(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + struct i2400m_report_hook_args *args, *args_next; + LIST_HEAD(list); + unsigned long flags; + + d_printf(1, dev, "flushing queued reports\n"); + spin_lock_irqsave(&i2400m->rx_lock, flags); + list_splice_init(&i2400m->rx_reports, &list); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + list_for_each_entry_safe(args, args_next, &list, list_node) { + d_printf(2, dev, "flushing queued report %p\n", args); + kfree_skb(args->skb_rx); + list_del(&args->list_node); + kfree(args); + } +} + + +/* + * Queue a report for later processing + * + * @i2400m: device descriptor + * @skb_rx: skb that contains the payload (for reference counting) + * @l3l4_hdr: pointer to the control + * @size: size of the message + */ +static +void i2400m_report_hook_queue(struct i2400m *i2400m, struct sk_buff *skb_rx, + const void *l3l4_hdr, size_t size) +{ + struct device *dev = i2400m_dev(i2400m); + unsigned long flags; + struct i2400m_report_hook_args *args; + + args = kzalloc(sizeof(*args), GFP_NOIO); + if (args) { + args->skb_rx = skb_get(skb_rx); + args->l3l4_hdr = l3l4_hdr; + args->size = size; + spin_lock_irqsave(&i2400m->rx_lock, flags); + list_add_tail(&args->list_node, &i2400m->rx_reports); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + d_printf(2, dev, "queued report %p\n", args); + rmb(); /* see i2400m->ready's documentation */ + if (likely(i2400m->ready)) /* only send if up */ + queue_work(i2400m->work_queue, &i2400m->rx_report_ws); + } else { + if (printk_ratelimit()) + dev_err(dev, "%s:%u: Can't allocate %zu B\n", + __func__, __LINE__, sizeof(*args)); + } +} + + +/* + * Process an ack to a command + * + * @i2400m: device descriptor + * @payload: pointer to message + * @size: size of the message + * + * Pass the acknodledgment (in an skb) to the thread that is waiting + * for it in i2400m->msg_completion. + * + * We need to coordinate properly with the thread waiting for the + * ack. Check if it is waiting or if it is gone. We loose the spinlock + * to avoid allocating on atomic contexts (yeah, could use GFP_ATOMIC, + * but this is not so speed critical). + */ +static +void i2400m_rx_ctl_ack(struct i2400m *i2400m, + const void *payload, size_t size) +{ + struct device *dev = i2400m_dev(i2400m); + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + unsigned long flags; + struct sk_buff *ack_skb; + + /* Anyone waiting for an answer? */ + spin_lock_irqsave(&i2400m->rx_lock, flags); + if (i2400m->ack_skb != ERR_PTR(-EINPROGRESS)) { + dev_err(dev, "Huh? reply to command with no waiters\n"); + goto error_no_waiter; + } + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + + ack_skb = wimax_msg_alloc(wimax_dev, NULL, payload, size, GFP_KERNEL); + + /* Check waiter didn't time out waiting for the answer... */ + spin_lock_irqsave(&i2400m->rx_lock, flags); + if (i2400m->ack_skb != ERR_PTR(-EINPROGRESS)) { + d_printf(1, dev, "Huh? waiter for command reply cancelled\n"); + goto error_waiter_cancelled; + } + if (IS_ERR(ack_skb)) + dev_err(dev, "CMD/GET/SET ack: cannot allocate SKB\n"); + i2400m->ack_skb = ack_skb; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + complete(&i2400m->msg_completion); + return; + +error_waiter_cancelled: + if (!IS_ERR(ack_skb)) + kfree_skb(ack_skb); +error_no_waiter: + spin_unlock_irqrestore(&i2400m->rx_lock, flags); +} + + +/* + * Receive and process a control payload + * + * @i2400m: device descriptor + * @skb_rx: skb that contains the payload (for reference counting) + * @payload: pointer to message + * @size: size of the message + * + * There are two types of control RX messages: reports (asynchronous, + * like your every day interrupts) and 'acks' (reponses to a command, + * get or set request). + * + * If it is a report, we run hooks on it (to extract information for + * things we need to do in the driver) and then pass it over to the + * WiMAX stack to send it to user space. + * + * NOTE: report processing is done in a workqueue specific to the + * generic driver, to avoid deadlocks in the system. + * + * If it is not a report, it is an ack to a previously executed + * command, set or get, so wake up whoever is waiting for it from + * i2400m_msg_to_dev(). i2400m_rx_ctl_ack() takes care of that. + * + * Note that the sizes we pass to other functions from here are the + * sizes of the _l3l4_hdr + payload, not full buffer sizes, as we have + * verified in _msg_size_check() that they are congruent. + * + * For reports: We can't clone the original skb where the data is + * because we need to send this up via netlink; netlink has to add + * headers and we can't overwrite what's preceding the payload...as + * it is another message. So we just dup them. + */ +static +void i2400m_rx_ctl(struct i2400m *i2400m, struct sk_buff *skb_rx, + const void *payload, size_t size) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_l3l4_hdr *l3l4_hdr = payload; + unsigned msg_type; + + result = i2400m_msg_size_check(i2400m, l3l4_hdr, size); + if (result < 0) { + dev_err(dev, "HW BUG? device sent a bad message: %d\n", + result); + goto error_check; + } + msg_type = le16_to_cpu(l3l4_hdr->type); + d_printf(1, dev, "%s 0x%04x: %zu bytes\n", + msg_type & I2400M_MT_REPORT_MASK ? "REPORT" : "CMD/SET/GET", + msg_type, size); + d_dump(2, dev, l3l4_hdr, size); + if (msg_type & I2400M_MT_REPORT_MASK) { + /* + * Process each report + * + * - has to be ran serialized as well + * + * - the handling might force the execution of + * commands. That might cause reentrancy issues with + * bus-specific subdrivers and workqueues, so the we + * run it in a separate workqueue. + * + * - when the driver is not yet ready to handle them, + * they are queued and at some point the queue is + * restarted [NOTE: we can't queue SKBs directly, as + * this might be a piece of a SKB, not the whole + * thing, and this is cheaper than cloning the + * SKB]. + * + * Note we don't do refcounting for the device + * structure; this is because before destroying + * 'i2400m', we make sure to flush the + * i2400m->work_queue, so there are no issues. + */ + i2400m_report_hook_queue(i2400m, skb_rx, l3l4_hdr, size); + if (unlikely(i2400m->trace_msg_from_user)) + wimax_msg(&i2400m->wimax_dev, "echo", + l3l4_hdr, size, GFP_KERNEL); + result = wimax_msg(&i2400m->wimax_dev, NULL, l3l4_hdr, size, + GFP_KERNEL); + if (result < 0) + dev_err(dev, "error sending report to userspace: %d\n", + result); + } else /* an ack to a CMD, GET or SET */ + i2400m_rx_ctl_ack(i2400m, payload, size); +error_check: + return; +} + + +/* + * Receive and send up a trace + * + * @i2400m: device descriptor + * @skb_rx: skb that contains the trace (for reference counting) + * @payload: pointer to trace message inside the skb + * @size: size of the message + * + * THe i2400m might produce trace information (diagnostics) and we + * send them through a different kernel-to-user pipe (to avoid + * clogging it). + * + * As in i2400m_rx_ctl(), we can't clone the original skb where the + * data is because we need to send this up via netlink; netlink has to + * add headers and we can't overwrite what's preceding the + * payload...as it is another message. So we just dup them. + */ +static +void i2400m_rx_trace(struct i2400m *i2400m, + const void *payload, size_t size) +{ + int result; + struct device *dev = i2400m_dev(i2400m); + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + const struct i2400m_l3l4_hdr *l3l4_hdr = payload; + unsigned msg_type; + + result = i2400m_msg_size_check(i2400m, l3l4_hdr, size); + if (result < 0) { + dev_err(dev, "HW BUG? device sent a bad trace message: %d\n", + result); + goto error_check; + } + msg_type = le16_to_cpu(l3l4_hdr->type); + d_printf(1, dev, "Trace %s 0x%04x: %zu bytes\n", + msg_type & I2400M_MT_REPORT_MASK ? "REPORT" : "CMD/SET/GET", + msg_type, size); + d_dump(2, dev, l3l4_hdr, size); + result = wimax_msg(wimax_dev, "trace", l3l4_hdr, size, GFP_KERNEL); + if (result < 0) + dev_err(dev, "error sending trace to userspace: %d\n", + result); +error_check: + return; +} + + +/* + * Reorder queue data stored on skb->cb while the skb is queued in the + * reorder queues. + */ +struct i2400m_roq_data { + unsigned sn; /* Serial number for the skb */ + enum i2400m_cs cs; /* packet type for the skb */ +}; + + +/* + * ReOrder Queue + * + * @ws: Window Start; sequence number where the current window start + * is for this queue + * @queue: the skb queue itself + * @log: circular ring buffer used to log information about the + * reorder process in this queue that can be displayed in case of + * error to help diagnose it. + * + * This is the head for a list of skbs. In the skb->cb member of the + * skb when queued here contains a 'struct i2400m_roq_data' were we + * store the sequence number (sn) and the cs (packet type) coming from + * the RX payload header from the device. + */ +struct i2400m_roq +{ + unsigned ws; + struct sk_buff_head queue; + struct i2400m_roq_log *log; +}; + + +static +void __i2400m_roq_init(struct i2400m_roq *roq) +{ + roq->ws = 0; + skb_queue_head_init(&roq->queue); +} + + +static +unsigned __i2400m_roq_index(struct i2400m *i2400m, struct i2400m_roq *roq) +{ + return ((unsigned long) roq - (unsigned long) i2400m->rx_roq) + / sizeof(*roq); +} + + +/* + * Normalize a sequence number based on the queue's window start + * + * nsn = (sn - ws) % 2048 + * + * Note that if @sn < @roq->ws, we still need a positive number; %'s + * sign is implementation specific, so we normalize it by adding 2048 + * to bring it to be positive. + */ +static +unsigned __i2400m_roq_nsn(struct i2400m_roq *roq, unsigned sn) +{ + int r; + r = ((int) sn - (int) roq->ws) % 2048; + if (r < 0) + r += 2048; + return r; +} + + +/* + * Circular buffer to keep the last N reorder operations + * + * In case something fails, dumb then to try to come up with what + * happened. + */ +enum { + I2400M_ROQ_LOG_LENGTH = 32, +}; + +struct i2400m_roq_log { + struct i2400m_roq_log_entry { + enum i2400m_ro_type type; + unsigned ws, count, sn, nsn, new_ws; + } entry[I2400M_ROQ_LOG_LENGTH]; + unsigned in, out; +}; + + +/* Print a log entry */ +static +void i2400m_roq_log_entry_print(struct i2400m *i2400m, unsigned index, + unsigned e_index, + struct i2400m_roq_log_entry *e) +{ + struct device *dev = i2400m_dev(i2400m); + + switch(e->type) { + case I2400M_RO_TYPE_RESET: + dev_err(dev, "q#%d reset ws %u cnt %u sn %u/%u" + " - new nws %u\n", + index, e->ws, e->count, e->sn, e->nsn, e->new_ws); + break; + case I2400M_RO_TYPE_PACKET: + dev_err(dev, "q#%d queue ws %u cnt %u sn %u/%u\n", + index, e->ws, e->count, e->sn, e->nsn); + break; + case I2400M_RO_TYPE_WS: + dev_err(dev, "q#%d update_ws ws %u cnt %u sn %u/%u" + " - new nws %u\n", + index, e->ws, e->count, e->sn, e->nsn, e->new_ws); + break; + case I2400M_RO_TYPE_PACKET_WS: + dev_err(dev, "q#%d queue_update_ws ws %u cnt %u sn %u/%u" + " - new nws %u\n", + index, e->ws, e->count, e->sn, e->nsn, e->new_ws); + break; + default: + dev_err(dev, "q#%d BUG? entry %u - unknown type %u\n", + index, e_index, e->type); + break; + } +} + + +static +void i2400m_roq_log_add(struct i2400m *i2400m, + struct i2400m_roq *roq, enum i2400m_ro_type type, + unsigned ws, unsigned count, unsigned sn, + unsigned nsn, unsigned new_ws) +{ + struct i2400m_roq_log_entry *e; + unsigned cnt_idx; + int index = __i2400m_roq_index(i2400m, roq); + + /* if we run out of space, we eat from the end */ + if (roq->log->in - roq->log->out == I2400M_ROQ_LOG_LENGTH) + roq->log->out++; + cnt_idx = roq->log->in++ % I2400M_ROQ_LOG_LENGTH; + e = &roq->log->entry[cnt_idx]; + + e->type = type; + e->ws = ws; + e->count = count; + e->sn = sn; + e->nsn = nsn; + e->new_ws = new_ws; + + if (d_test(1)) + i2400m_roq_log_entry_print(i2400m, index, cnt_idx, e); +} + + +/* Dump all the entries in the FIFO and reinitialize it */ +static +void i2400m_roq_log_dump(struct i2400m *i2400m, struct i2400m_roq *roq) +{ + unsigned cnt, cnt_idx; + struct i2400m_roq_log_entry *e; + int index = __i2400m_roq_index(i2400m, roq); + + BUG_ON(roq->log->out > roq->log->in); + for (cnt = roq->log->out; cnt < roq->log->in; cnt++) { + cnt_idx = cnt % I2400M_ROQ_LOG_LENGTH; + e = &roq->log->entry[cnt_idx]; + i2400m_roq_log_entry_print(i2400m, index, cnt_idx, e); + memset(e, 0, sizeof(*e)); + } + roq->log->in = roq->log->out = 0; +} + + +/* + * Backbone for the queuing of an skb (by normalized sequence number) + * + * @i2400m: device descriptor + * @roq: reorder queue where to add + * @skb: the skb to add + * @sn: the sequence number of the skb + * @nsn: the normalized sequence number of the skb (pre-computed by the + * caller from the @sn and @roq->ws). + * + * We try first a couple of quick cases: + * + * - the queue is empty + * - the skb would be appended to the queue + * + * These will be the most common operations. + * + * If these fail, then we have to do a sorted insertion in the queue, + * which is the slowest path. + * + * We don't have to acquire a reference count as we are going to own it. + */ +static +void __i2400m_roq_queue(struct i2400m *i2400m, struct i2400m_roq *roq, + struct sk_buff *skb, unsigned sn, unsigned nsn) +{ + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *skb_itr; + struct i2400m_roq_data *roq_data_itr, *roq_data; + unsigned nsn_itr; + + d_fnstart(4, dev, "(i2400m %p roq %p skb %p sn %u nsn %u)\n", + i2400m, roq, skb, sn, nsn); + + roq_data = (struct i2400m_roq_data *) &skb->cb; + BUILD_BUG_ON(sizeof(*roq_data) > sizeof(skb->cb)); + roq_data->sn = sn; + d_printf(3, dev, "ERX: roq %p [ws %u] nsn %d sn %u\n", + roq, roq->ws, nsn, roq_data->sn); + + /* Queues will be empty on not-so-bad environments, so try + * that first */ + if (skb_queue_empty(&roq->queue)) { + d_printf(2, dev, "ERX: roq %p - first one\n", roq); + __skb_queue_head(&roq->queue, skb); + goto out; + } + /* Now try append, as most of the operations will be that */ + skb_itr = skb_peek_tail(&roq->queue); + roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; + nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); + /* NSN bounds assumed correct (checked when it was queued) */ + if (nsn >= nsn_itr) { + d_printf(2, dev, "ERX: roq %p - appended after %p (nsn %d sn %u)\n", + roq, skb_itr, nsn_itr, roq_data_itr->sn); + __skb_queue_tail(&roq->queue, skb); + goto out; + } + /* None of the fast paths option worked. Iterate to find the + * right spot where to insert the packet; we know the queue is + * not empty, so we are not the first ones; we also know we + * are not going to be the last ones. The list is sorted, so + * we have to insert before the the first guy with an nsn_itr + * greater that our nsn. */ + skb_queue_walk(&roq->queue, skb_itr) { + roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; + nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); + /* NSN bounds assumed correct (checked when it was queued) */ + if (nsn_itr > nsn) { + d_printf(2, dev, "ERX: roq %p - queued before %p " + "(nsn %d sn %u)\n", roq, skb_itr, nsn_itr, + roq_data_itr->sn); + __skb_queue_before(&roq->queue, skb_itr, skb); + goto out; + } + } + /* If we get here, that is VERY bad -- print info to help + * diagnose and crash it */ + dev_err(dev, "SW BUG? failed to insert packet\n"); + dev_err(dev, "ERX: roq %p [ws %u] skb %p nsn %d sn %u\n", + roq, roq->ws, skb, nsn, roq_data->sn); + skb_queue_walk(&roq->queue, skb_itr) { + roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; + nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); + /* NSN bounds assumed correct (checked when it was queued) */ + dev_err(dev, "ERX: roq %p skb_itr %p nsn %d sn %u\n", + roq, skb_itr, nsn_itr, roq_data_itr->sn); + } + BUG(); +out: + d_fnend(4, dev, "(i2400m %p roq %p skb %p sn %u nsn %d) = void\n", + i2400m, roq, skb, sn, nsn); +} + + +/* + * Backbone for the update window start operation + * + * @i2400m: device descriptor + * @roq: Reorder queue + * @sn: New sequence number + * + * Updates the window start of a queue; when doing so, it must deliver + * to the networking stack all the queued skb's whose normalized + * sequence number is lower than the new normalized window start. + */ +static +unsigned __i2400m_roq_update_ws(struct i2400m *i2400m, struct i2400m_roq *roq, + unsigned sn) +{ + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *skb_itr, *tmp_itr; + struct i2400m_roq_data *roq_data_itr; + unsigned new_nws, nsn_itr; + + new_nws = __i2400m_roq_nsn(roq, sn); + /* + * For type 2(update_window_start) rx messages, there is no + * need to check if the normalized sequence number is greater 1023. + * Simply insert and deliver all packets to the host up to the + * window start. + */ + skb_queue_walk_safe(&roq->queue, skb_itr, tmp_itr) { + roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; + nsn_itr = __i2400m_roq_nsn(roq, roq_data_itr->sn); + /* NSN bounds assumed correct (checked when it was queued) */ + if (nsn_itr < new_nws) { + d_printf(2, dev, "ERX: roq %p - release skb %p " + "(nsn %u/%u new nws %u)\n", + roq, skb_itr, nsn_itr, roq_data_itr->sn, + new_nws); + __skb_unlink(skb_itr, &roq->queue); + i2400m_net_erx(i2400m, skb_itr, roq_data_itr->cs); + } + else + break; /* rest of packets all nsn_itr > nws */ + } + roq->ws = sn; + return new_nws; +} + + +/* + * Reset a queue + * + * @i2400m: device descriptor + * @cin: Queue Index + * + * Deliver all the packets and reset the window-start to zero. Name is + * kind of misleading. + */ +static +void i2400m_roq_reset(struct i2400m *i2400m, struct i2400m_roq *roq) +{ + struct device *dev = i2400m_dev(i2400m); + struct sk_buff *skb_itr, *tmp_itr; + struct i2400m_roq_data *roq_data_itr; + + d_fnstart(2, dev, "(i2400m %p roq %p)\n", i2400m, roq); + i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_RESET, + roq->ws, skb_queue_len(&roq->queue), + ~0, ~0, 0); + skb_queue_walk_safe(&roq->queue, skb_itr, tmp_itr) { + roq_data_itr = (struct i2400m_roq_data *) &skb_itr->cb; + d_printf(2, dev, "ERX: roq %p - release skb %p (sn %u)\n", + roq, skb_itr, roq_data_itr->sn); + __skb_unlink(skb_itr, &roq->queue); + i2400m_net_erx(i2400m, skb_itr, roq_data_itr->cs); + } + roq->ws = 0; + d_fnend(2, dev, "(i2400m %p roq %p) = void\n", i2400m, roq); +} + + +/* + * Queue a packet + * + * @i2400m: device descriptor + * @cin: Queue Index + * @skb: containing the packet data + * @fbn: First block number of the packet in @skb + * @lbn: Last block number of the packet in @skb + * + * The hardware is asking the driver to queue a packet for later + * delivery to the networking stack. + */ +static +void i2400m_roq_queue(struct i2400m *i2400m, struct i2400m_roq *roq, + struct sk_buff * skb, unsigned lbn) +{ + struct device *dev = i2400m_dev(i2400m); + unsigned nsn, len; + + d_fnstart(2, dev, "(i2400m %p roq %p skb %p lbn %u) = void\n", + i2400m, roq, skb, lbn); + len = skb_queue_len(&roq->queue); + nsn = __i2400m_roq_nsn(roq, lbn); + if (unlikely(nsn >= 1024)) { + dev_err(dev, "SW BUG? queue nsn %d (lbn %u ws %u)\n", + nsn, lbn, roq->ws); + i2400m_roq_log_dump(i2400m, roq); + i2400m_reset(i2400m, I2400M_RT_WARM); + } else { + __i2400m_roq_queue(i2400m, roq, skb, lbn, nsn); + i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_PACKET, + roq->ws, len, lbn, nsn, ~0); + } + d_fnend(2, dev, "(i2400m %p roq %p skb %p lbn %u) = void\n", + i2400m, roq, skb, lbn); +} + + +/* + * Update the window start in a reorder queue and deliver all skbs + * with a lower window start + * + * @i2400m: device descriptor + * @roq: Reorder queue + * @sn: New sequence number + */ +static +void i2400m_roq_update_ws(struct i2400m *i2400m, struct i2400m_roq *roq, + unsigned sn) +{ + struct device *dev = i2400m_dev(i2400m); + unsigned old_ws, nsn, len; + + d_fnstart(2, dev, "(i2400m %p roq %p sn %u)\n", i2400m, roq, sn); + old_ws = roq->ws; + len = skb_queue_len(&roq->queue); + nsn = __i2400m_roq_update_ws(i2400m, roq, sn); + i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_WS, + old_ws, len, sn, nsn, roq->ws); + d_fnstart(2, dev, "(i2400m %p roq %p sn %u) = void\n", i2400m, roq, sn); +} + + +/* + * Queue a packet and update the window start + * + * @i2400m: device descriptor + * @cin: Queue Index + * @skb: containing the packet data + * @fbn: First block number of the packet in @skb + * @sn: Last block number of the packet in @skb + * + * Note that unlike i2400m_roq_update_ws(), which sets the new window + * start to @sn, in here we'll set it to @sn + 1. + */ +static +void i2400m_roq_queue_update_ws(struct i2400m *i2400m, struct i2400m_roq *roq, + struct sk_buff * skb, unsigned sn) +{ + struct device *dev = i2400m_dev(i2400m); + unsigned nsn, old_ws, len; + + d_fnstart(2, dev, "(i2400m %p roq %p skb %p sn %u)\n", + i2400m, roq, skb, sn); + len = skb_queue_len(&roq->queue); + nsn = __i2400m_roq_nsn(roq, sn); + /* + * For type 3(queue_update_window_start) rx messages, there is no + * need to check if the normalized sequence number is greater 1023. + * Simply insert and deliver all packets to the host up to the + * window start. + */ + old_ws = roq->ws; + /* If the queue is empty, don't bother as we'd queue + * it and immediately unqueue it -- just deliver it. + */ + if (len == 0) { + struct i2400m_roq_data *roq_data; + roq_data = (struct i2400m_roq_data *) &skb->cb; + i2400m_net_erx(i2400m, skb, roq_data->cs); + } else + __i2400m_roq_queue(i2400m, roq, skb, sn, nsn); + + __i2400m_roq_update_ws(i2400m, roq, sn + 1); + i2400m_roq_log_add(i2400m, roq, I2400M_RO_TYPE_PACKET_WS, + old_ws, len, sn, nsn, roq->ws); + + d_fnend(2, dev, "(i2400m %p roq %p skb %p sn %u) = void\n", + i2400m, roq, skb, sn); +} + + +/* + * This routine destroys the memory allocated for rx_roq, when no + * other thread is accessing it. Access to rx_roq is refcounted by + * rx_roq_refcount, hence memory allocated must be destroyed when + * rx_roq_refcount becomes zero. This routine gets executed when + * rx_roq_refcount becomes zero. + */ +static void i2400m_rx_roq_destroy(struct kref *ref) +{ + unsigned itr; + struct i2400m *i2400m + = container_of(ref, struct i2400m, rx_roq_refcount); + for (itr = 0; itr < I2400M_RO_CIN + 1; itr++) + __skb_queue_purge(&i2400m->rx_roq[itr].queue); + kfree(i2400m->rx_roq[0].log); + kfree(i2400m->rx_roq); + i2400m->rx_roq = NULL; +} + +/* + * Receive and send up an extended data packet + * + * @i2400m: device descriptor + * @skb_rx: skb that contains the extended data packet + * @single_last: 1 if the payload is the only one or the last one of + * the skb. + * @payload: pointer to the packet's data inside the skb + * @size: size of the payload + * + * Starting in v1.4 of the i2400m's firmware, the device can send data + * packets to the host in an extended format that; this incudes a 16 + * byte header (struct i2400m_pl_edata_hdr). Using this header's space + * we can fake ethernet headers for ethernet device emulation without + * having to copy packets around. + * + * This function handles said path. + * + * + * Receive and send up an extended data packet that requires no reordering + * + * @i2400m: device descriptor + * @skb_rx: skb that contains the extended data packet + * @single_last: 1 if the payload is the only one or the last one of + * the skb. + * @payload: pointer to the packet's data (past the actual extended + * data payload header). + * @size: size of the payload + * + * Pass over to the networking stack a data packet that might have + * reordering requirements. + * + * This needs to the decide if the skb in which the packet is + * contained can be reused or if it needs to be cloned. Then it has to + * be trimmed in the edges so that the beginning is the space for eth + * header and then pass it to i2400m_net_erx() for the stack + * + * Assumes the caller has verified the sanity of the payload (size, + * etc) already. + */ +static +void i2400m_rx_edata(struct i2400m *i2400m, struct sk_buff *skb_rx, + unsigned single_last, const void *payload, size_t size) +{ + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_pl_edata_hdr *hdr = payload; + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + struct sk_buff *skb; + enum i2400m_cs cs; + u32 reorder; + unsigned ro_needed, ro_type, ro_cin, ro_sn; + struct i2400m_roq *roq; + struct i2400m_roq_data *roq_data; + unsigned long flags; + + BUILD_BUG_ON(ETH_HLEN > sizeof(*hdr)); + + d_fnstart(2, dev, "(i2400m %p skb_rx %p single %u payload %p " + "size %zu)\n", i2400m, skb_rx, single_last, payload, size); + if (size < sizeof(*hdr)) { + dev_err(dev, "ERX: HW BUG? message with short header (%zu " + "vs %zu bytes expected)\n", size, sizeof(*hdr)); + goto error; + } + + if (single_last) { + skb = skb_get(skb_rx); + d_printf(3, dev, "ERX: skb %p reusing\n", skb); + } else { + skb = skb_clone(skb_rx, GFP_KERNEL); + if (skb == NULL) { + dev_err(dev, "ERX: no memory to clone skb\n"); + net_dev->stats.rx_dropped++; + goto error_skb_clone; + } + d_printf(3, dev, "ERX: skb %p cloned from %p\n", skb, skb_rx); + } + /* now we have to pull and trim so that the skb points to the + * beginning of the IP packet; the netdev part will add the + * ethernet header as needed - we know there is enough space + * because we checked in i2400m_rx_edata(). */ + skb_pull(skb, payload + sizeof(*hdr) - (void *) skb->data); + skb_trim(skb, (void *) skb_end_pointer(skb) - payload - sizeof(*hdr)); + + reorder = le32_to_cpu(hdr->reorder); + ro_needed = reorder & I2400M_RO_NEEDED; + cs = hdr->cs; + if (ro_needed) { + ro_type = (reorder >> I2400M_RO_TYPE_SHIFT) & I2400M_RO_TYPE; + ro_cin = (reorder >> I2400M_RO_CIN_SHIFT) & I2400M_RO_CIN; + ro_sn = (reorder >> I2400M_RO_SN_SHIFT) & I2400M_RO_SN; + + spin_lock_irqsave(&i2400m->rx_lock, flags); + if (i2400m->rx_roq == NULL) { + kfree_skb(skb); /* rx_roq is already destroyed */ + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + goto error; + } + roq = &i2400m->rx_roq[ro_cin]; + kref_get(&i2400m->rx_roq_refcount); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + + roq_data = (struct i2400m_roq_data *) &skb->cb; + roq_data->sn = ro_sn; + roq_data->cs = cs; + d_printf(2, dev, "ERX: reorder needed: " + "type %u cin %u [ws %u] sn %u/%u len %zuB\n", + ro_type, ro_cin, roq->ws, ro_sn, + __i2400m_roq_nsn(roq, ro_sn), size); + d_dump(2, dev, payload, size); + switch(ro_type) { + case I2400M_RO_TYPE_RESET: + i2400m_roq_reset(i2400m, roq); + kfree_skb(skb); /* no data here */ + break; + case I2400M_RO_TYPE_PACKET: + i2400m_roq_queue(i2400m, roq, skb, ro_sn); + break; + case I2400M_RO_TYPE_WS: + i2400m_roq_update_ws(i2400m, roq, ro_sn); + kfree_skb(skb); /* no data here */ + break; + case I2400M_RO_TYPE_PACKET_WS: + i2400m_roq_queue_update_ws(i2400m, roq, skb, ro_sn); + break; + default: + dev_err(dev, "HW BUG? unknown reorder type %u\n", ro_type); + } + + spin_lock_irqsave(&i2400m->rx_lock, flags); + kref_put(&i2400m->rx_roq_refcount, i2400m_rx_roq_destroy); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + } + else + i2400m_net_erx(i2400m, skb, cs); +error_skb_clone: +error: + d_fnend(2, dev, "(i2400m %p skb_rx %p single %u payload %p " + "size %zu) = void\n", i2400m, skb_rx, single_last, payload, size); +} + + +/* + * Act on a received payload + * + * @i2400m: device instance + * @skb_rx: skb where the transaction was received + * @single_last: 1 this is the only payload or the last one (so the + * skb can be reused instead of cloned). + * @pld: payload descriptor + * @payload: payload data + * + * Upon reception of a payload, look at its guts in the payload + * descriptor and decide what to do with it. If it is a single payload + * skb or if the last skb is a data packet, the skb will be referenced + * and modified (so it doesn't have to be cloned). + */ +static +void i2400m_rx_payload(struct i2400m *i2400m, struct sk_buff *skb_rx, + unsigned single_last, const struct i2400m_pld *pld, + const void *payload) +{ + struct device *dev = i2400m_dev(i2400m); + size_t pl_size = i2400m_pld_size(pld); + enum i2400m_pt pl_type = i2400m_pld_type(pld); + + d_printf(7, dev, "RX: received payload type %u, %zu bytes\n", + pl_type, pl_size); + d_dump(8, dev, payload, pl_size); + + switch (pl_type) { + case I2400M_PT_DATA: + d_printf(3, dev, "RX: data payload %zu bytes\n", pl_size); + i2400m_net_rx(i2400m, skb_rx, single_last, payload, pl_size); + break; + case I2400M_PT_CTRL: + i2400m_rx_ctl(i2400m, skb_rx, payload, pl_size); + break; + case I2400M_PT_TRACE: + i2400m_rx_trace(i2400m, payload, pl_size); + break; + case I2400M_PT_EDATA: + d_printf(3, dev, "ERX: data payload %zu bytes\n", pl_size); + i2400m_rx_edata(i2400m, skb_rx, single_last, payload, pl_size); + break; + default: /* Anything else shouldn't come to the host */ + if (printk_ratelimit()) + dev_err(dev, "RX: HW BUG? unexpected payload type %u\n", + pl_type); + } +} + + +/* + * Check a received transaction's message header + * + * @i2400m: device descriptor + * @msg_hdr: message header + * @buf_size: size of the received buffer + * + * Check that the declarations done by a RX buffer message header are + * sane and consistent with the amount of data that was received. + */ +static +int i2400m_rx_msg_hdr_check(struct i2400m *i2400m, + const struct i2400m_msg_hdr *msg_hdr, + size_t buf_size) +{ + int result = -EIO; + struct device *dev = i2400m_dev(i2400m); + if (buf_size < sizeof(*msg_hdr)) { + dev_err(dev, "RX: HW BUG? message with short header (%zu " + "vs %zu bytes expected)\n", buf_size, sizeof(*msg_hdr)); + goto error; + } + if (msg_hdr->barker != cpu_to_le32(I2400M_D2H_MSG_BARKER)) { + dev_err(dev, "RX: HW BUG? message received with unknown " + "barker 0x%08x (buf_size %zu bytes)\n", + le32_to_cpu(msg_hdr->barker), buf_size); + goto error; + } + if (msg_hdr->num_pls == 0) { + dev_err(dev, "RX: HW BUG? zero payload packets in message\n"); + goto error; + } + if (le16_to_cpu(msg_hdr->num_pls) > I2400M_MAX_PLS_IN_MSG) { + dev_err(dev, "RX: HW BUG? message contains more payload " + "than maximum; ignoring.\n"); + goto error; + } + result = 0; +error: + return result; +} + + +/* + * Check a payload descriptor against the received data + * + * @i2400m: device descriptor + * @pld: payload descriptor + * @pl_itr: offset (in bytes) in the received buffer the payload is + * located + * @buf_size: size of the received buffer + * + * Given a payload descriptor (part of a RX buffer), check it is sane + * and that the data it declares fits in the buffer. + */ +static +int i2400m_rx_pl_descr_check(struct i2400m *i2400m, + const struct i2400m_pld *pld, + size_t pl_itr, size_t buf_size) +{ + int result = -EIO; + struct device *dev = i2400m_dev(i2400m); + size_t pl_size = i2400m_pld_size(pld); + enum i2400m_pt pl_type = i2400m_pld_type(pld); + + if (pl_size > i2400m->bus_pl_size_max) { + dev_err(dev, "RX: HW BUG? payload @%zu: size %zu is " + "bigger than maximum %zu; ignoring message\n", + pl_itr, pl_size, i2400m->bus_pl_size_max); + goto error; + } + if (pl_itr + pl_size > buf_size) { /* enough? */ + dev_err(dev, "RX: HW BUG? payload @%zu: size %zu " + "goes beyond the received buffer " + "size (%zu bytes); ignoring message\n", + pl_itr, pl_size, buf_size); + goto error; + } + if (pl_type >= I2400M_PT_ILLEGAL) { + dev_err(dev, "RX: HW BUG? illegal payload type %u; " + "ignoring message\n", pl_type); + goto error; + } + result = 0; +error: + return result; +} + + +/** + * i2400m_rx - Receive a buffer of data from the device + * + * @i2400m: device descriptor + * @skb: skbuff where the data has been received + * + * Parse in a buffer of data that contains an RX message sent from the + * device. See the file header for the format. Run all checks on the + * buffer header, then run over each payload's descriptors, verify + * their consistency and act on each payload's contents. If + * everything is successful, update the device's statistics. + * + * Note: You need to set the skb to contain only the length of the + * received buffer; for that, use skb_trim(skb, RECEIVED_SIZE). + * + * Returns: + * + * 0 if ok, < 0 errno on error + * + * If ok, this function owns now the skb and the caller DOESN'T have + * to run kfree_skb() on it. However, on error, the caller still owns + * the skb and it is responsible for releasing it. + */ +int i2400m_rx(struct i2400m *i2400m, struct sk_buff *skb) +{ + int i, result; + struct device *dev = i2400m_dev(i2400m); + const struct i2400m_msg_hdr *msg_hdr; + size_t pl_itr, pl_size; + unsigned long flags; + unsigned num_pls, single_last, skb_len; + + skb_len = skb->len; + d_fnstart(4, dev, "(i2400m %p skb %p [size %u])\n", + i2400m, skb, skb_len); + msg_hdr = (void *) skb->data; + result = i2400m_rx_msg_hdr_check(i2400m, msg_hdr, skb_len); + if (result < 0) + goto error_msg_hdr_check; + result = -EIO; + num_pls = le16_to_cpu(msg_hdr->num_pls); + /* Check payload descriptor(s) */ + pl_itr = struct_size(msg_hdr, pld, num_pls); + pl_itr = ALIGN(pl_itr, I2400M_PL_ALIGN); + if (pl_itr > skb_len) { /* got all the payload descriptors? */ + dev_err(dev, "RX: HW BUG? message too short (%u bytes) for " + "%u payload descriptors (%zu each, total %zu)\n", + skb_len, num_pls, sizeof(msg_hdr->pld[0]), pl_itr); + goto error_pl_descr_short; + } + /* Walk each payload payload--check we really got it */ + for (i = 0; i < num_pls; i++) { + /* work around old gcc warnings */ + pl_size = i2400m_pld_size(&msg_hdr->pld[i]); + result = i2400m_rx_pl_descr_check(i2400m, &msg_hdr->pld[i], + pl_itr, skb_len); + if (result < 0) + goto error_pl_descr_check; + single_last = num_pls == 1 || i == num_pls - 1; + i2400m_rx_payload(i2400m, skb, single_last, &msg_hdr->pld[i], + skb->data + pl_itr); + pl_itr += ALIGN(pl_size, I2400M_PL_ALIGN); + cond_resched(); /* Don't monopolize */ + } + kfree_skb(skb); + /* Update device statistics */ + spin_lock_irqsave(&i2400m->rx_lock, flags); + i2400m->rx_pl_num += i; + if (i > i2400m->rx_pl_max) + i2400m->rx_pl_max = i; + if (i < i2400m->rx_pl_min) + i2400m->rx_pl_min = i; + i2400m->rx_num++; + i2400m->rx_size_acc += skb_len; + if (skb_len < i2400m->rx_size_min) + i2400m->rx_size_min = skb_len; + if (skb_len > i2400m->rx_size_max) + i2400m->rx_size_max = skb_len; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); +error_pl_descr_check: +error_pl_descr_short: +error_msg_hdr_check: + d_fnend(4, dev, "(i2400m %p skb %p [size %u]) = %d\n", + i2400m, skb, skb_len, result); + return result; +} +EXPORT_SYMBOL_GPL(i2400m_rx); + + +void i2400m_unknown_barker(struct i2400m *i2400m, + const void *buf, size_t size) +{ + struct device *dev = i2400m_dev(i2400m); + char prefix[64]; + const __le32 *barker = buf; + dev_err(dev, "RX: HW BUG? unknown barker %08x, " + "dropping %zu bytes\n", le32_to_cpu(*barker), size); + snprintf(prefix, sizeof(prefix), "%s %s: ", + dev_driver_string(dev), dev_name(dev)); + if (size > 64) { + print_hex_dump(KERN_ERR, prefix, DUMP_PREFIX_OFFSET, + 8, 4, buf, 64, 0); + printk(KERN_ERR "%s... (only first 64 bytes " + "dumped)\n", prefix); + } else + print_hex_dump(KERN_ERR, prefix, DUMP_PREFIX_OFFSET, + 8, 4, buf, size, 0); +} +EXPORT_SYMBOL(i2400m_unknown_barker); + + +/* + * Initialize the RX queue and infrastructure + * + * This sets up all the RX reordering infrastructures, which will not + * be used if reordering is not enabled or if the firmware does not + * support it. The device is told to do reordering in + * i2400m_dev_initialize(), where it also looks at the value of the + * i2400m->rx_reorder switch before taking a decission. + * + * Note we allocate the roq queues in one chunk and the actual logging + * support for it (logging) in another one and then we setup the + * pointers from the first to the last. + */ +int i2400m_rx_setup(struct i2400m *i2400m) +{ + int result = 0; + + i2400m->rx_reorder = i2400m_rx_reorder_disabled? 0 : 1; + if (i2400m->rx_reorder) { + unsigned itr; + struct i2400m_roq_log *rd; + + result = -ENOMEM; + + i2400m->rx_roq = kcalloc(I2400M_RO_CIN + 1, + sizeof(i2400m->rx_roq[0]), GFP_KERNEL); + if (i2400m->rx_roq == NULL) + goto error_roq_alloc; + + rd = kcalloc(I2400M_RO_CIN + 1, sizeof(*i2400m->rx_roq[0].log), + GFP_KERNEL); + if (rd == NULL) { + result = -ENOMEM; + goto error_roq_log_alloc; + } + + for(itr = 0; itr < I2400M_RO_CIN + 1; itr++) { + __i2400m_roq_init(&i2400m->rx_roq[itr]); + i2400m->rx_roq[itr].log = &rd[itr]; + } + kref_init(&i2400m->rx_roq_refcount); + } + return 0; + +error_roq_log_alloc: + kfree(i2400m->rx_roq); +error_roq_alloc: + return result; +} + + +/* Tear down the RX queue and infrastructure */ +void i2400m_rx_release(struct i2400m *i2400m) +{ + unsigned long flags; + + if (i2400m->rx_reorder) { + spin_lock_irqsave(&i2400m->rx_lock, flags); + kref_put(&i2400m->rx_roq_refcount, i2400m_rx_roq_destroy); + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + } + /* at this point, nothing can be received... */ + i2400m_report_hook_flush(i2400m); +} diff --git a/drivers/staging/wimax/i2400m/sysfs.c b/drivers/staging/wimax/i2400m/sysfs.c new file mode 100644 index 000000000000..895ee265909b --- /dev/null +++ b/drivers/staging/wimax/i2400m/sysfs.c @@ -0,0 +1,65 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Wireless WiMAX Connection 2400m + * Sysfs interfaces to show driver and device information + * + * Copyright (C) 2007 Intel Corporation + * Inaky Perez-Gonzalez + */ + +#include +#include +#include +#include +#include "i2400m.h" + + +#define D_SUBMODULE sysfs +#include "debug-levels.h" + + +/* + * Set the idle timeout (msecs) + * + * FIXME: eventually this should be a common WiMAX stack method, but + * would like to wait to see how other devices manage it. + */ +static +ssize_t i2400m_idle_timeout_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t size) +{ + ssize_t result; + struct i2400m *i2400m = net_dev_to_i2400m(to_net_dev(dev)); + unsigned val; + + result = -EINVAL; + if (sscanf(buf, "%u\n", &val) != 1) + goto error_no_unsigned; + if (val != 0 && (val < 100 || val > 300000 || val % 100 != 0)) { + dev_err(dev, "idle_timeout: %u: invalid msecs specification; " + "valid values are 0, 100-300000 in 100 increments\n", + val); + goto error_bad_value; + } + result = i2400m_set_idle_timeout(i2400m, val); + if (result >= 0) + result = size; +error_no_unsigned: +error_bad_value: + return result; +} + +static +DEVICE_ATTR_WO(i2400m_idle_timeout); + +static +struct attribute *i2400m_dev_attrs[] = { + &dev_attr_i2400m_idle_timeout.attr, + NULL, +}; + +struct attribute_group i2400m_dev_attr_group = { + .name = NULL, /* we want them in the same directory */ + .attrs = i2400m_dev_attrs, +}; diff --git a/drivers/staging/wimax/i2400m/tx.c b/drivers/staging/wimax/i2400m/tx.c new file mode 100644 index 000000000000..1255302e251e --- /dev/null +++ b/drivers/staging/wimax/i2400m/tx.c @@ -0,0 +1,1011 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * Generic (non-bus specific) TX handling + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * - Initial implementation + * + * Intel Corporation + * Inaky Perez-Gonzalez + * - Rewritten to use a single FIFO to lower the memory allocation + * pressure and optimize cache hits when copying to the queue, as + * well as splitting out bus-specific code. + * + * + * Implements data transmission to the device; this is done through a + * software FIFO, as data/control frames can be coalesced (while the + * device is reading the previous tx transaction, others accumulate). + * + * A FIFO is used because at the end it is resource-cheaper that trying + * to implement scatter/gather over USB. As well, most traffic is going + * to be download (vs upload). + * + * The format for sending/receiving data to/from the i2400m is + * described in detail in rx.c:PROTOCOL FORMAT. In here we implement + * the transmission of that. This is split between a bus-independent + * part that just prepares everything and a bus-specific part that + * does the actual transmission over the bus to the device (in the + * bus-specific driver). + * + * + * The general format of a device-host transaction is MSG-HDR, PLD1, + * PLD2...PLDN, PL1, PL2,...PLN, PADDING. + * + * Because we need the send payload descriptors and then payloads and + * because it is kind of expensive to do scatterlists in USB (one URB + * per node), it becomes cheaper to append all the data to a FIFO + * (copying to a FIFO potentially in cache is cheaper). + * + * Then the bus-specific code takes the parts of that FIFO that are + * written and passes them to the device. + * + * So the concepts to keep in mind there are: + * + * We use a FIFO to queue the data in a linear buffer. We first append + * a MSG-HDR, space for I2400M_TX_PLD_MAX payload descriptors and then + * go appending payloads until we run out of space or of payload + * descriptors. Then we append padding to make the whole transaction a + * multiple of i2400m->bus_tx_block_size (as defined by the bus layer). + * + * - A TX message: a combination of a message header, payload + * descriptors and payloads. + * + * Open: it is marked as active (i2400m->tx_msg is valid) and we + * can keep adding payloads to it. + * + * Closed: we are not appending more payloads to this TX message + * (exahusted space in the queue, too many payloads or + * whichever). We have appended padding so the whole message + * length is aligned to i2400m->bus_tx_block_size (as set by the + * bus/transport layer). + * + * - Most of the time we keep a TX message open to which we append + * payloads. + * + * - If we are going to append and there is no more space (we are at + * the end of the FIFO), we close the message, mark the rest of the + * FIFO space unusable (skip_tail), create a new message at the + * beginning of the FIFO (if there is space) and append the message + * there. + * + * This is because we need to give linear TX messages to the bus + * engine. So we don't write a message to the remaining FIFO space + * until the tail and continue at the head of it. + * + * - We overload one of the fields in the message header to use it as + * 'size' of the TX message, so we can iterate over them. It also + * contains a flag that indicates if we have to skip it or not. + * When we send the buffer, we update that to its real on-the-wire + * value. + * + * - The MSG-HDR PLD1...PLD2 stuff has to be a size multiple of 16. + * + * It follows that if MSG-HDR says we have N messages, the whole + * header + descriptors is 16 + 4*N; for those to be a multiple of + * 16, it follows that N can be 4, 8, 12, ... (32, 48, 64, 80... + * bytes). + * + * So if we have only 1 payload, we have to submit a header that in + * all truth has space for 4. + * + * The implication is that we reserve space for 12 (64 bytes); but + * if we fill up only (eg) 2, our header becomes 32 bytes only. So + * the TX engine has to shift those 32 bytes of msg header and 2 + * payloads and padding so that right after it the payloads start + * and the TX engine has to know about that. + * + * It is cheaper to move the header up than the whole payloads down. + * + * We do this in i2400m_tx_close(). See 'i2400m_msg_hdr->offset'. + * + * - Each payload has to be size-padded to 16 bytes; before appending + * it, we just do it. + * + * - The whole message has to be padded to i2400m->bus_tx_block_size; + * we do this at close time. Thus, when reserving space for the + * payload, we always make sure there is also free space for this + * padding that sooner or later will happen. + * + * When we append a message, we tell the bus specific code to kick in + * TXs. It will TX (in parallel) until the buffer is exhausted--hence + * the lockin we do. The TX code will only send a TX message at the + * time (which remember, might contain more than one payload). Of + * course, when the bus-specific driver attempts to TX a message that + * is still open, it gets closed first. + * + * Gee, this is messy; well a picture. In the example below we have a + * partially full FIFO, with a closed message ready to be delivered + * (with a moved message header to make sure it is size-aligned to + * 16), TAIL room that was unusable (and thus is marked with a message + * header that says 'skip this') and at the head of the buffer, an + * incomplete message with a couple of payloads. + * + * N ___________________________________________________ + * | | + * | TAIL room | + * | | + * | msg_hdr to skip (size |= 0x80000) | + * |---------------------------------------------------|------- + * | | /|\ + * | | | + * | TX message padding | | + * | | | + * | | | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -| | + * | | | + * | payload 1 | | + * | | N * tx_block_size + * | | | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -| | + * | | | + * | payload 1 | | + * | | | + * | | | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -|- -|- - - - + * | padding 3 /|\ | | /|\ + * | padding 2 | | | | + * | pld 1 32 bytes (2 * 16) | | | + * | pld 0 | | | | + * | moved msg_hdr \|/ | \|/ | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -|- - - | + * | | _PLD_SIZE + * | unused | | + * | | | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -| | + * | msg_hdr (size X) [this message is closed] | \|/ + * |===================================================|========== <=== OUT + * | | + * | | + * | | + * | Free rooom | + * | | + * | | + * | | + * | | + * | | + * | | + * | | + * | | + * | | + * |===================================================|========== <=== IN + * | | + * | | + * | | + * | | + * | payload 1 | + * | | + * | | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -| + * | | + * | payload 0 | + * | | + * | | + * |- - - - - - - - - - - - - - - - - - - - - - - - - -| + * | pld 11 /|\ | + * | ... | | + * | pld 1 64 bytes (2 * 16) | + * | pld 0 | | + * | msg_hdr (size X) \|/ [message is open] | + * 0 --------------------------------------------------- + * + * + * ROADMAP + * + * i2400m_tx_setup() Called by i2400m_setup + * i2400m_tx_release() Called by i2400m_release() + * + * i2400m_tx() Called to send data or control frames + * i2400m_tx_fifo_push() Allocates append-space in the FIFO + * i2400m_tx_new() Opens a new message in the FIFO + * i2400m_tx_fits() Checks if a new payload fits in the message + * i2400m_tx_close() Closes an open message in the FIFO + * i2400m_tx_skip_tail() Marks unusable FIFO tail space + * i2400m->bus_tx_kick() + * + * Now i2400m->bus_tx_kick() is the the bus-specific driver backend + * implementation; that would do: + * + * i2400m->bus_tx_kick() + * i2400m_tx_msg_get() Gets first message ready to go + * ...sends it... + * i2400m_tx_msg_sent() Ack the message is sent; repeat from + * _tx_msg_get() until it returns NULL + * (FIFO empty). + */ +#include +#include +#include +#include "i2400m.h" + + +#define D_SUBMODULE tx +#include "debug-levels.h" + +enum { + /** + * TX Buffer size + * + * Doc says maximum transaction is 16KiB. If we had 16KiB en + * route and 16KiB being queued, it boils down to needing + * 32KiB. + * 32KiB is insufficient for 1400 MTU, hence increasing + * tx buffer size to 64KiB. + */ + I2400M_TX_BUF_SIZE = 65536, + /** + * Message header and payload descriptors have to be 16 + * aligned (16 + 4 * N = 16 * M). If we take that average sent + * packets are MTU size (~1400-~1500) it follows that we could + * fit at most 10-11 payloads in one transaction. To meet the + * alignment requirement, that means we need to leave space + * for 12 (64 bytes). To simplify, we leave space for that. If + * at the end there are less, we pad up to the nearest + * multiple of 16. + */ + /* + * According to Intel Wimax i3200, i5x50 and i6x50 specification + * documents, the maximum number of payloads per message can be + * up to 60. Increasing the number of payloads to 60 per message + * helps to accommodate smaller payloads in a single transaction. + */ + I2400M_TX_PLD_MAX = 60, + I2400M_TX_PLD_SIZE = sizeof(struct i2400m_msg_hdr) + + I2400M_TX_PLD_MAX * sizeof(struct i2400m_pld), + I2400M_TX_SKIP = 0x80000000, + /* + * According to Intel Wimax i3200, i5x50 and i6x50 specification + * documents, the maximum size of each message can be up to 16KiB. + */ + I2400M_TX_MSG_SIZE = 16384, +}; + +#define TAIL_FULL ((void *)~(unsigned long)NULL) + +/* + * Calculate how much tail room is available + * + * Note the trick here. This path is ONLY caleed for Case A (see + * i2400m_tx_fifo_push() below), where we have: + * + * Case A + * N ___________ + * | tail room | + * | | + * |<- IN ->| + * | | + * | data | + * | | + * |<- OUT ->| + * | | + * | head room | + * 0 ----------- + * + * When calculating the tail_room, tx_in might get to be zero if + * i2400m->tx_in is right at the end of the buffer (really full + * buffer) if there is no head room. In this case, tail_room would be + * I2400M_TX_BUF_SIZE, although it is actually zero. Hence the final + * mod (%) operation. However, when doing this kind of optimization, + * i2400m->tx_in being zero would fail, so we treat is an a special + * case. + */ +static inline +size_t __i2400m_tx_tail_room(struct i2400m *i2400m) +{ + size_t tail_room; + size_t tx_in; + + if (unlikely(i2400m->tx_in == 0)) + return I2400M_TX_BUF_SIZE; + tx_in = i2400m->tx_in % I2400M_TX_BUF_SIZE; + tail_room = I2400M_TX_BUF_SIZE - tx_in; + tail_room %= I2400M_TX_BUF_SIZE; + return tail_room; +} + + +/* + * Allocate @size bytes in the TX fifo, return a pointer to it + * + * @i2400m: device descriptor + * @size: size of the buffer we need to allocate + * @padding: ensure that there is at least this many bytes of free + * contiguous space in the fifo. This is needed because later on + * we might need to add padding. + * @try_head: specify either to allocate head room or tail room space + * in the TX FIFO. This boolean is required to avoids a system hang + * due to an infinite loop caused by i2400m_tx_fifo_push(). + * The caller must always try to allocate tail room space first by + * calling this routine with try_head = 0. In case if there + * is not enough tail room space but there is enough head room space, + * (i2400m_tx_fifo_push() returns TAIL_FULL) try to allocate head + * room space, by calling this routine again with try_head = 1. + * + * Returns: + * + * Pointer to the allocated space. NULL if there is no + * space. TAIL_FULL if there is no space at the tail but there is at + * the head (Case B below). + * + * These are the two basic cases we need to keep an eye for -- it is + * much better explained in linux/kernel/kfifo.c, but this code + * basically does the same. No rocket science here. + * + * Case A Case B + * N ___________ ___________ + * | tail room | | data | + * | | | | + * |<- IN ->| |<- OUT ->| + * | | | | + * | data | | room | + * | | | | + * |<- OUT ->| |<- IN ->| + * | | | | + * | head room | | data | + * 0 ----------- ----------- + * + * We allocate only *contiguous* space. + * + * We can allocate only from 'room'. In Case B, it is simple; in case + * A, we only try from the tail room; if it is not enough, we just + * fail and return TAIL_FULL and let the caller figure out if we wants to + * skip the tail room and try to allocate from the head. + * + * There is a corner case, wherein i2400m_tx_new() can get into + * an infinite loop calling i2400m_tx_fifo_push(). + * In certain situations, tx_in would have reached on the top of TX FIFO + * and i2400m_tx_tail_room() returns 0, as described below: + * + * N ___________ tail room is zero + * |<- IN ->| + * | | + * | | + * | | + * | data | + * |<- OUT ->| + * | | + * | | + * | head room | + * 0 ----------- + * During such a time, where tail room is zero in the TX FIFO and if there + * is a request to add a payload to TX FIFO, which calls: + * i2400m_tx() + * ->calls i2400m_tx_close() + * ->calls i2400m_tx_skip_tail() + * goto try_new; + * ->calls i2400m_tx_new() + * |----> [try_head:] + * infinite loop | ->calls i2400m_tx_fifo_push() + * | if (tail_room < needed) + * | if (head_room => needed) + * | return TAIL_FULL; + * |<---- goto try_head; + * + * i2400m_tx() calls i2400m_tx_close() to close the message, since there + * is no tail room to accommodate the payload and calls + * i2400m_tx_skip_tail() to skip the tail space. Now i2400m_tx() calls + * i2400m_tx_new() to allocate space for new message header calling + * i2400m_tx_fifo_push() that returns TAIL_FULL, since there is no tail space + * to accommodate the message header, but there is enough head space. + * The i2400m_tx_new() keeps re-retrying by calling i2400m_tx_fifo_push() + * ending up in a loop causing system freeze. + * + * This corner case is avoided by using a try_head boolean, + * as an argument to i2400m_tx_fifo_push(). + * + * Note: + * + * Assumes i2400m->tx_lock is taken, and we use that as a barrier + * + * The indexes keep increasing and we reset them to zero when we + * pop data off the queue + */ +static +void *i2400m_tx_fifo_push(struct i2400m *i2400m, size_t size, + size_t padding, bool try_head) +{ + struct device *dev = i2400m_dev(i2400m); + size_t room, tail_room, needed_size; + void *ptr; + + needed_size = size + padding; + room = I2400M_TX_BUF_SIZE - (i2400m->tx_in - i2400m->tx_out); + if (room < needed_size) { /* this takes care of Case B */ + d_printf(2, dev, "fifo push %zu/%zu: no space\n", + size, padding); + return NULL; + } + /* Is there space at the tail? */ + tail_room = __i2400m_tx_tail_room(i2400m); + if (!try_head && tail_room < needed_size) { + /* + * If the tail room space is not enough to push the message + * in the TX FIFO, then there are two possibilities: + * 1. There is enough head room space to accommodate + * this message in the TX FIFO. + * 2. There is not enough space in the head room and + * in tail room of the TX FIFO to accommodate the message. + * In the case (1), return TAIL_FULL so that the caller + * can figure out, if the caller wants to push the message + * into the head room space. + * In the case (2), return NULL, indicating that the TX FIFO + * cannot accommodate the message. + */ + if (room - tail_room >= needed_size) { + d_printf(2, dev, "fifo push %zu/%zu: tail full\n", + size, padding); + return TAIL_FULL; /* There might be head space */ + } else { + d_printf(2, dev, "fifo push %zu/%zu: no head space\n", + size, padding); + return NULL; /* There is no space */ + } + } + ptr = i2400m->tx_buf + i2400m->tx_in % I2400M_TX_BUF_SIZE; + d_printf(2, dev, "fifo push %zu/%zu: at @%zu\n", size, padding, + i2400m->tx_in % I2400M_TX_BUF_SIZE); + i2400m->tx_in += size; + return ptr; +} + + +/* + * Mark the tail of the FIFO buffer as 'to-skip' + * + * We should never hit the BUG_ON() because all the sizes we push to + * the FIFO are padded to be a multiple of 16 -- the size of *msg + * (I2400M_PL_PAD for the payloads, I2400M_TX_PLD_SIZE for the + * header). + * + * Tail room can get to be zero if a message was opened when there was + * space only for a header. _tx_close() will mark it as to-skip (as it + * will have no payloads) and there will be no more space to flush, so + * nothing has to be done here. This is probably cheaper than ensuring + * in _tx_new() that there is some space for payloads...as we could + * always possibly hit the same problem if the payload wouldn't fit. + * + * Note: + * + * Assumes i2400m->tx_lock is taken, and we use that as a barrier + * + * This path is only taken for Case A FIFO situations [see + * i2400m_tx_fifo_push()] + */ +static +void i2400m_tx_skip_tail(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + size_t tx_in = i2400m->tx_in % I2400M_TX_BUF_SIZE; + size_t tail_room = __i2400m_tx_tail_room(i2400m); + struct i2400m_msg_hdr *msg = i2400m->tx_buf + tx_in; + if (unlikely(tail_room == 0)) + return; + BUG_ON(tail_room < sizeof(*msg)); + msg->size = tail_room | I2400M_TX_SKIP; + d_printf(2, dev, "skip tail: skipping %zu bytes @%zu\n", + tail_room, tx_in); + i2400m->tx_in += tail_room; +} + + +/* + * Check if a skb will fit in the TX queue's current active TX + * message (if there are still descriptors left unused). + * + * Returns: + * 0 if the message won't fit, 1 if it will. + * + * Note: + * + * Assumes a TX message is active (i2400m->tx_msg). + * + * Assumes i2400m->tx_lock is taken, and we use that as a barrier + */ +static +unsigned i2400m_tx_fits(struct i2400m *i2400m) +{ + struct i2400m_msg_hdr *msg_hdr = i2400m->tx_msg; + return le16_to_cpu(msg_hdr->num_pls) < I2400M_TX_PLD_MAX; + +} + + +/* + * Start a new TX message header in the queue. + * + * Reserve memory from the base FIFO engine and then just initialize + * the message header. + * + * We allocate the biggest TX message header we might need (one that'd + * fit I2400M_TX_PLD_MAX payloads) -- when it is closed it will be + * 'ironed it out' and the unneeded parts removed. + * + * NOTE: + * + * Assumes that the previous message is CLOSED (eg: either + * there was none or 'i2400m_tx_close()' was called on it). + * + * Assumes i2400m->tx_lock is taken, and we use that as a barrier + */ +static +void i2400m_tx_new(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + struct i2400m_msg_hdr *tx_msg; + bool try_head = false; + BUG_ON(i2400m->tx_msg != NULL); + /* + * In certain situations, TX queue might have enough space to + * accommodate the new message header I2400M_TX_PLD_SIZE, but + * might not have enough space to accommodate the payloads. + * Adding bus_tx_room_min padding while allocating a new TX message + * increases the possibilities of including at least one payload of the + * size <= bus_tx_room_min. + */ +try_head: + tx_msg = i2400m_tx_fifo_push(i2400m, I2400M_TX_PLD_SIZE, + i2400m->bus_tx_room_min, try_head); + if (tx_msg == NULL) + goto out; + else if (tx_msg == TAIL_FULL) { + i2400m_tx_skip_tail(i2400m); + d_printf(2, dev, "new TX message: tail full, trying head\n"); + try_head = true; + goto try_head; + } + memset(tx_msg, 0, I2400M_TX_PLD_SIZE); + tx_msg->size = I2400M_TX_PLD_SIZE; +out: + i2400m->tx_msg = tx_msg; + d_printf(2, dev, "new TX message: %p @%zu\n", + tx_msg, (void *) tx_msg - i2400m->tx_buf); +} + + +/* + * Finalize the current TX message header + * + * Sets the message header to be at the proper location depending on + * how many descriptors we have (check documentation at the file's + * header for more info on that). + * + * Appends padding bytes to make sure the whole TX message (counting + * from the 'relocated' message header) is aligned to + * tx_block_size. We assume the _append() code has left enough space + * in the FIFO for that. If there are no payloads, just pass, as it + * won't be transferred. + * + * The amount of padding bytes depends on how many payloads are in the + * TX message, as the "msg header and payload descriptors" will be + * shifted up in the buffer. + */ +static +void i2400m_tx_close(struct i2400m *i2400m) +{ + struct device *dev = i2400m_dev(i2400m); + struct i2400m_msg_hdr *tx_msg = i2400m->tx_msg; + struct i2400m_msg_hdr *tx_msg_moved; + size_t aligned_size, padding, hdr_size; + void *pad_buf; + unsigned num_pls; + + if (tx_msg->size & I2400M_TX_SKIP) /* a skipper? nothing to do */ + goto out; + num_pls = le16_to_cpu(tx_msg->num_pls); + /* We can get this situation when a new message was started + * and there was no space to add payloads before hitting the + tail (and taking padding into consideration). */ + if (num_pls == 0) { + tx_msg->size |= I2400M_TX_SKIP; + goto out; + } + /* Relocate the message header + * + * Find the current header size, align it to 16 and if we need + * to move it so the tail is next to the payloads, move it and + * set the offset. + * + * If it moved, this header is good only for transmission; the + * original one (it is kept if we moved) is still used to + * figure out where the next TX message starts (and where the + * offset to the moved header is). + */ + hdr_size = struct_size(tx_msg, pld, le16_to_cpu(tx_msg->num_pls)); + hdr_size = ALIGN(hdr_size, I2400M_PL_ALIGN); + tx_msg->offset = I2400M_TX_PLD_SIZE - hdr_size; + tx_msg_moved = (void *) tx_msg + tx_msg->offset; + memmove(tx_msg_moved, tx_msg, hdr_size); + tx_msg_moved->size -= tx_msg->offset; + /* + * Now figure out how much we have to add to the (moved!) + * message so the size is a multiple of i2400m->bus_tx_block_size. + */ + aligned_size = ALIGN(tx_msg_moved->size, i2400m->bus_tx_block_size); + padding = aligned_size - tx_msg_moved->size; + if (padding > 0) { + pad_buf = i2400m_tx_fifo_push(i2400m, padding, 0, 0); + if (WARN_ON(pad_buf == NULL || pad_buf == TAIL_FULL)) { + /* This should not happen -- append should verify + * there is always space left at least to append + * tx_block_size */ + dev_err(dev, + "SW BUG! Possible data leakage from memory the " + "device should not read for padding - " + "size %lu aligned_size %zu tx_buf %p in " + "%zu out %zu\n", + (unsigned long) tx_msg_moved->size, + aligned_size, i2400m->tx_buf, i2400m->tx_in, + i2400m->tx_out); + } else + memset(pad_buf, 0xad, padding); + } + tx_msg_moved->padding = cpu_to_le16(padding); + tx_msg_moved->size += padding; + if (tx_msg != tx_msg_moved) + tx_msg->size += padding; +out: + i2400m->tx_msg = NULL; +} + + +/** + * i2400m_tx - send the data in a buffer to the device + * + * @buf: pointer to the buffer to transmit + * + * @buf_len: buffer size + * + * @pl_type: type of the payload we are sending. + * + * Returns: + * 0 if ok, < 0 errno code on error (-ENOSPC, if there is no more + * room for the message in the queue). + * + * Appends the buffer to the TX FIFO and notifies the bus-specific + * part of the driver that there is new data ready to transmit. + * Once this function returns, the buffer has been copied, so it can + * be reused. + * + * The steps followed to append are explained in detail in the file + * header. + * + * Whenever we write to a message, we increase msg->size, so it + * reflects exactly how big the message is. This is needed so that if + * we concatenate two messages before they can be sent, the code that + * sends the messages can find the boundaries (and it will replace the + * size with the real barker before sending). + * + * Note: + * + * Cold and warm reset payloads need to be sent as a single + * payload, so we handle that. + */ +int i2400m_tx(struct i2400m *i2400m, const void *buf, size_t buf_len, + enum i2400m_pt pl_type) +{ + int result = -ENOSPC; + struct device *dev = i2400m_dev(i2400m); + unsigned long flags; + size_t padded_len; + void *ptr; + bool try_head = false; + unsigned is_singleton = pl_type == I2400M_PT_RESET_WARM + || pl_type == I2400M_PT_RESET_COLD; + + d_fnstart(3, dev, "(i2400m %p skb %p [%zu bytes] pt %u)\n", + i2400m, buf, buf_len, pl_type); + padded_len = ALIGN(buf_len, I2400M_PL_ALIGN); + d_printf(5, dev, "padded_len %zd buf_len %zd\n", padded_len, buf_len); + /* If there is no current TX message, create one; if the + * current one is out of payload slots or we have a singleton, + * close it and start a new one */ + spin_lock_irqsave(&i2400m->tx_lock, flags); + /* If tx_buf is NULL, device is shutdown */ + if (i2400m->tx_buf == NULL) { + result = -ESHUTDOWN; + goto error_tx_new; + } +try_new: + if (unlikely(i2400m->tx_msg == NULL)) + i2400m_tx_new(i2400m); + else if (unlikely(!i2400m_tx_fits(i2400m) + || (is_singleton && i2400m->tx_msg->num_pls != 0))) { + d_printf(2, dev, "closing TX message (fits %u singleton " + "%u num_pls %u)\n", i2400m_tx_fits(i2400m), + is_singleton, i2400m->tx_msg->num_pls); + i2400m_tx_close(i2400m); + i2400m_tx_new(i2400m); + } + if (i2400m->tx_msg == NULL) + goto error_tx_new; + /* + * Check if this skb will fit in the TX queue's current active + * TX message. The total message size must not exceed the maximum + * size of each message I2400M_TX_MSG_SIZE. If it exceeds, + * close the current message and push this skb into the new message. + */ + if (i2400m->tx_msg->size + padded_len > I2400M_TX_MSG_SIZE) { + d_printf(2, dev, "TX: message too big, going new\n"); + i2400m_tx_close(i2400m); + i2400m_tx_new(i2400m); + } + if (i2400m->tx_msg == NULL) + goto error_tx_new; + /* So we have a current message header; now append space for + * the message -- if there is not enough, try the head */ + ptr = i2400m_tx_fifo_push(i2400m, padded_len, + i2400m->bus_tx_block_size, try_head); + if (ptr == TAIL_FULL) { /* Tail is full, try head */ + d_printf(2, dev, "pl append: tail full\n"); + i2400m_tx_close(i2400m); + i2400m_tx_skip_tail(i2400m); + try_head = true; + goto try_new; + } else if (ptr == NULL) { /* All full */ + result = -ENOSPC; + d_printf(2, dev, "pl append: all full\n"); + } else { /* Got space, copy it, set padding */ + struct i2400m_msg_hdr *tx_msg = i2400m->tx_msg; + unsigned num_pls = le16_to_cpu(tx_msg->num_pls); + memcpy(ptr, buf, buf_len); + memset(ptr + buf_len, 0xad, padded_len - buf_len); + i2400m_pld_set(&tx_msg->pld[num_pls], buf_len, pl_type); + d_printf(3, dev, "pld 0x%08x (type 0x%1x len 0x%04zx\n", + le32_to_cpu(tx_msg->pld[num_pls].val), + pl_type, buf_len); + tx_msg->num_pls = le16_to_cpu(num_pls+1); + tx_msg->size += padded_len; + d_printf(2, dev, "TX: appended %zu b (up to %u b) pl #%u\n", + padded_len, tx_msg->size, num_pls+1); + d_printf(2, dev, + "TX: appended hdr @%zu %zu b pl #%u @%zu %zu/%zu b\n", + (void *)tx_msg - i2400m->tx_buf, (size_t)tx_msg->size, + num_pls+1, ptr - i2400m->tx_buf, buf_len, padded_len); + result = 0; + if (is_singleton) + i2400m_tx_close(i2400m); + } +error_tx_new: + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + /* kick in most cases, except when the TX subsys is down, as + * it might free space */ + if (likely(result != -ESHUTDOWN)) + i2400m->bus_tx_kick(i2400m); + d_fnend(3, dev, "(i2400m %p skb %p [%zu bytes] pt %u) = %d\n", + i2400m, buf, buf_len, pl_type, result); + return result; +} +EXPORT_SYMBOL_GPL(i2400m_tx); + + +/** + * i2400m_tx_msg_get - Get the first TX message in the FIFO to start sending it + * + * @i2400m: device descriptors + * @bus_size: where to place the size of the TX message + * + * Called by the bus-specific driver to get the first TX message at + * the FIF that is ready for transmission. + * + * It sets the state in @i2400m to indicate the bus-specific driver is + * transferring that message (i2400m->tx_msg_size). + * + * Once the transfer is completed, call i2400m_tx_msg_sent(). + * + * Notes: + * + * The size of the TX message to be transmitted might be smaller than + * that of the TX message in the FIFO (in case the header was + * shorter). Hence, we copy it in @bus_size, for the bus layer to + * use. We keep the message's size in i2400m->tx_msg_size so that + * when the bus later is done transferring we know how much to + * advance the fifo. + * + * We collect statistics here as all the data is available and we + * assume it is going to work [see i2400m_tx_msg_sent()]. + */ +struct i2400m_msg_hdr *i2400m_tx_msg_get(struct i2400m *i2400m, + size_t *bus_size) +{ + struct device *dev = i2400m_dev(i2400m); + struct i2400m_msg_hdr *tx_msg, *tx_msg_moved; + unsigned long flags, pls; + + d_fnstart(3, dev, "(i2400m %p bus_size %p)\n", i2400m, bus_size); + spin_lock_irqsave(&i2400m->tx_lock, flags); + tx_msg_moved = NULL; + if (i2400m->tx_buf == NULL) + goto out_unlock; +skip: + tx_msg_moved = NULL; + if (i2400m->tx_in == i2400m->tx_out) { /* Empty FIFO? */ + i2400m->tx_in = 0; + i2400m->tx_out = 0; + d_printf(2, dev, "TX: FIFO empty: resetting\n"); + goto out_unlock; + } + tx_msg = i2400m->tx_buf + i2400m->tx_out % I2400M_TX_BUF_SIZE; + if (tx_msg->size & I2400M_TX_SKIP) { /* skip? */ + d_printf(2, dev, "TX: skip: msg @%zu (%zu b)\n", + i2400m->tx_out % I2400M_TX_BUF_SIZE, + (size_t) tx_msg->size & ~I2400M_TX_SKIP); + i2400m->tx_out += tx_msg->size & ~I2400M_TX_SKIP; + goto skip; + } + + if (tx_msg->num_pls == 0) { /* No payloads? */ + if (tx_msg == i2400m->tx_msg) { /* open, we are done */ + d_printf(2, dev, + "TX: FIFO empty: open msg w/o payloads @%zu\n", + (void *) tx_msg - i2400m->tx_buf); + tx_msg = NULL; + goto out_unlock; + } else { /* closed, skip it */ + d_printf(2, dev, + "TX: skip msg w/o payloads @%zu (%zu b)\n", + (void *) tx_msg - i2400m->tx_buf, + (size_t) tx_msg->size); + i2400m->tx_out += tx_msg->size & ~I2400M_TX_SKIP; + goto skip; + } + } + if (tx_msg == i2400m->tx_msg) /* open msg? */ + i2400m_tx_close(i2400m); + + /* Now we have a valid TX message (with payloads) to TX */ + tx_msg_moved = (void *) tx_msg + tx_msg->offset; + i2400m->tx_msg_size = tx_msg->size; + *bus_size = tx_msg_moved->size; + d_printf(2, dev, "TX: pid %d msg hdr at @%zu offset +@%zu " + "size %zu bus_size %zu\n", + current->pid, (void *) tx_msg - i2400m->tx_buf, + (size_t) tx_msg->offset, (size_t) tx_msg->size, + (size_t) tx_msg_moved->size); + tx_msg_moved->barker = le32_to_cpu(I2400M_H2D_PREVIEW_BARKER); + tx_msg_moved->sequence = le32_to_cpu(i2400m->tx_sequence++); + + pls = le32_to_cpu(tx_msg_moved->num_pls); + i2400m->tx_pl_num += pls; /* Update stats */ + if (pls > i2400m->tx_pl_max) + i2400m->tx_pl_max = pls; + if (pls < i2400m->tx_pl_min) + i2400m->tx_pl_min = pls; + i2400m->tx_num++; + i2400m->tx_size_acc += *bus_size; + if (*bus_size < i2400m->tx_size_min) + i2400m->tx_size_min = *bus_size; + if (*bus_size > i2400m->tx_size_max) + i2400m->tx_size_max = *bus_size; +out_unlock: + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + d_fnstart(3, dev, "(i2400m %p bus_size %p [%zu]) = %p\n", + i2400m, bus_size, *bus_size, tx_msg_moved); + return tx_msg_moved; +} +EXPORT_SYMBOL_GPL(i2400m_tx_msg_get); + + +/** + * i2400m_tx_msg_sent - indicate the transmission of a TX message + * + * @i2400m: device descriptor + * + * Called by the bus-specific driver when a message has been sent; + * this pops it from the FIFO; and as there is space, start the queue + * in case it was stopped. + * + * Should be called even if the message send failed and we are + * dropping this TX message. + */ +void i2400m_tx_msg_sent(struct i2400m *i2400m) +{ + unsigned n; + unsigned long flags; + struct device *dev = i2400m_dev(i2400m); + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + spin_lock_irqsave(&i2400m->tx_lock, flags); + if (i2400m->tx_buf == NULL) + goto out_unlock; + i2400m->tx_out += i2400m->tx_msg_size; + d_printf(2, dev, "TX: sent %zu b\n", (size_t) i2400m->tx_msg_size); + i2400m->tx_msg_size = 0; + BUG_ON(i2400m->tx_out > i2400m->tx_in); + /* level them FIFO markers off */ + n = i2400m->tx_out / I2400M_TX_BUF_SIZE; + i2400m->tx_out %= I2400M_TX_BUF_SIZE; + i2400m->tx_in -= n * I2400M_TX_BUF_SIZE; +out_unlock: + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} +EXPORT_SYMBOL_GPL(i2400m_tx_msg_sent); + + +/** + * i2400m_tx_setup - Initialize the TX queue and infrastructure + * + * Make sure we reset the TX sequence to zero, as when this function + * is called, the firmware has been just restarted. Same rational + * for tx_in, tx_out, tx_msg_size and tx_msg. We reset them since + * the memory for TX queue is reallocated. + */ +int i2400m_tx_setup(struct i2400m *i2400m) +{ + int result = 0; + void *tx_buf; + unsigned long flags; + + /* Do this here only once -- can't do on + * i2400m_hard_start_xmit() as we'll cause race conditions if + * the WS was scheduled on another CPU */ + INIT_WORK(&i2400m->wake_tx_ws, i2400m_wake_tx_work); + + tx_buf = kmalloc(I2400M_TX_BUF_SIZE, GFP_ATOMIC); + if (tx_buf == NULL) { + result = -ENOMEM; + goto error_kmalloc; + } + + /* + * Fail the build if we can't fit at least two maximum size messages + * on the TX FIFO [one being delivered while one is constructed]. + */ + BUILD_BUG_ON(2 * I2400M_TX_MSG_SIZE > I2400M_TX_BUF_SIZE); + spin_lock_irqsave(&i2400m->tx_lock, flags); + i2400m->tx_sequence = 0; + i2400m->tx_in = 0; + i2400m->tx_out = 0; + i2400m->tx_msg_size = 0; + i2400m->tx_msg = NULL; + i2400m->tx_buf = tx_buf; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + /* Huh? the bus layer has to define this... */ + BUG_ON(i2400m->bus_tx_block_size == 0); +error_kmalloc: + return result; + +} + + +/** + * i2400m_tx_release - Tear down the TX queue and infrastructure + */ +void i2400m_tx_release(struct i2400m *i2400m) +{ + unsigned long flags; + spin_lock_irqsave(&i2400m->tx_lock, flags); + kfree(i2400m->tx_buf); + i2400m->tx_buf = NULL; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); +} diff --git a/drivers/staging/wimax/i2400m/usb-debug-levels.h b/drivers/staging/wimax/i2400m/usb-debug-levels.h new file mode 100644 index 000000000000..8fd0111560f6 --- /dev/null +++ b/drivers/staging/wimax/i2400m/usb-debug-levels.h @@ -0,0 +1,28 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Intel Wireless WiMAX Connection 2400m + * Debug levels control file for the i2400m-usb module + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + */ +#ifndef __debug_levels__h__ +#define __debug_levels__h__ + +/* Maximum compile and run time debug level for all submodules */ +#define D_MODULENAME i2400m_usb +#define D_MASTER CONFIG_WIMAX_I2400M_DEBUG_LEVEL + +#include "../linux-wimax-debug.h" + +/* List of all the enabled modules */ +enum d_module { + D_SUBMODULE_DECLARE(usb), + D_SUBMODULE_DECLARE(fw), + D_SUBMODULE_DECLARE(notif), + D_SUBMODULE_DECLARE(rx), + D_SUBMODULE_DECLARE(tx), +}; + + +#endif /* #ifndef __debug_levels__h__ */ diff --git a/drivers/staging/wimax/i2400m/usb-fw.c b/drivers/staging/wimax/i2400m/usb-fw.c new file mode 100644 index 000000000000..27ab233650d5 --- /dev/null +++ b/drivers/staging/wimax/i2400m/usb-fw.c @@ -0,0 +1,365 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * Firmware uploader's USB specifics + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * Inaky Perez-Gonzalez + * - Initial implementation + * + * Inaky Perez-Gonzalez + * - bus generic/specific split + * + * THE PROCEDURE + * + * See fw.c for the generic description of this procedure. + * + * This file implements only the USB specifics. It boils down to how + * to send a command and waiting for an acknowledgement from the + * device. + * + * This code (and process) is single threaded. It assumes it is the + * only thread poking around (guaranteed by fw.c). + * + * COMMAND EXECUTION + * + * A write URB is posted with the buffer to the bulk output endpoint. + * + * ACK RECEPTION + * + * We just post a URB to the notification endpoint and wait for + * data. We repeat until we get all the data we expect (as indicated + * by the call from the bus generic code). + * + * The data is not read from the bulk in endpoint for boot mode. + * + * ROADMAP + * + * i2400mu_bus_bm_cmd_send + * i2400m_bm_cmd_prepare... + * i2400mu_tx_bulk_out + * + * i2400mu_bus_bm_wait_for_ack + * i2400m_notif_submit + */ +#include +#include +#include "i2400m-usb.h" + + +#define D_SUBMODULE fw +#include "usb-debug-levels.h" + + +/* + * Synchronous write to the device + * + * Takes care of updating EDC counts and thus, handle device errors. + */ +static +ssize_t i2400mu_tx_bulk_out(struct i2400mu *i2400mu, void *buf, size_t buf_size) +{ + int result; + struct device *dev = &i2400mu->usb_iface->dev; + int len; + struct usb_endpoint_descriptor *epd; + int pipe, do_autopm = 1; + + result = usb_autopm_get_interface(i2400mu->usb_iface); + if (result < 0) { + dev_err(dev, "BM-CMD: can't get autopm: %d\n", result); + do_autopm = 0; + } + epd = usb_get_epd(i2400mu->usb_iface, i2400mu->endpoint_cfg.bulk_out); + pipe = usb_sndbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); +retry: + result = usb_bulk_msg(i2400mu->usb_dev, pipe, buf, buf_size, &len, 200); + switch (result) { + case 0: + if (len != buf_size) { + dev_err(dev, "BM-CMD: short write (%u B vs %zu " + "expected)\n", len, buf_size); + result = -EIO; + break; + } + result = len; + break; + case -EPIPE: + /* + * Stall -- maybe the device is choking with our + * requests. Clear it and give it some time. If they + * happen to often, it might be another symptom, so we + * reset. + * + * No error handling for usb_clear_halt(0; if it + * works, the retry works; if it fails, this switch + * does the error handling for us. + */ + if (edc_inc(&i2400mu->urb_edc, + 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "BM-CMD: too many stalls in " + "URB; resetting device\n"); + usb_queue_reset_device(i2400mu->usb_iface); + } else { + usb_clear_halt(i2400mu->usb_dev, pipe); + msleep(10); /* give the device some time */ + goto retry; + } + fallthrough; + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* just ignore it */ + case -ESHUTDOWN: /* and exit */ + case -ECONNRESET: + result = -ESHUTDOWN; + break; + case -ETIMEDOUT: /* bah... */ + break; + default: /* any other? */ + if (edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "BM-CMD: maximum errors in " + "URB exceeded; resetting device\n"); + usb_queue_reset_device(i2400mu->usb_iface); + result = -ENODEV; + break; + } + dev_err(dev, "BM-CMD: URB error %d, retrying\n", + result); + goto retry; + } + if (do_autopm) + usb_autopm_put_interface(i2400mu->usb_iface); + return result; +} + + +/* + * Send a boot-mode command over the bulk-out pipe + * + * Command can be a raw command, which requires no preparation (and + * which might not even be following the command format). Checks that + * the right amount of data was transferred. + * + * To satisfy USB requirements (no onstack, vmalloc or in data segment + * buffers), we copy the command to i2400m->bm_cmd_buf and send it from + * there. + * + * @flags: pass thru from i2400m_bm_cmd() + * @return: cmd_size if ok, < 0 errno code on error. + */ +ssize_t i2400mu_bus_bm_cmd_send(struct i2400m *i2400m, + const struct i2400m_bootrom_header *_cmd, + size_t cmd_size, int flags) +{ + ssize_t result; + struct device *dev = i2400m_dev(i2400m); + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + int opcode = _cmd == NULL ? -1 : i2400m_brh_get_opcode(_cmd); + struct i2400m_bootrom_header *cmd; + size_t cmd_size_a = ALIGN(cmd_size, 16); /* USB restriction */ + + d_fnstart(8, dev, "(i2400m %p cmd %p size %zu)\n", + i2400m, _cmd, cmd_size); + result = -E2BIG; + if (cmd_size > I2400M_BM_CMD_BUF_SIZE) + goto error_too_big; + if (_cmd != i2400m->bm_cmd_buf) + memmove(i2400m->bm_cmd_buf, _cmd, cmd_size); + cmd = i2400m->bm_cmd_buf; + if (cmd_size_a > cmd_size) /* Zero pad space */ + memset(i2400m->bm_cmd_buf + cmd_size, 0, cmd_size_a - cmd_size); + if ((flags & I2400M_BM_CMD_RAW) == 0) { + if (WARN_ON(i2400m_brh_get_response_required(cmd) == 0)) + dev_warn(dev, "SW BUG: response_required == 0\n"); + i2400m_bm_cmd_prepare(cmd); + } + result = i2400mu_tx_bulk_out(i2400mu, i2400m->bm_cmd_buf, cmd_size); + if (result < 0) { + dev_err(dev, "boot-mode cmd %d: cannot send: %zd\n", + opcode, result); + goto error_cmd_send; + } + if (result != cmd_size) { /* all was transferred? */ + dev_err(dev, "boot-mode cmd %d: incomplete transfer " + "(%zd vs %zu submitted)\n", opcode, result, cmd_size); + result = -EIO; + goto error_cmd_size; + } +error_cmd_size: +error_cmd_send: +error_too_big: + d_fnend(8, dev, "(i2400m %p cmd %p size %zu) = %zd\n", + i2400m, _cmd, cmd_size, result); + return result; +} + + +static +void __i2400mu_bm_notif_cb(struct urb *urb) +{ + complete(urb->context); +} + + +/* + * submit a read to the notification endpoint + * + * @i2400m: device descriptor + * @urb: urb to use + * @completion: completion variable to complete when done + * + * Data is always read to i2400m->bm_ack_buf + */ +static +int i2400mu_notif_submit(struct i2400mu *i2400mu, struct urb *urb, + struct completion *completion) +{ + struct i2400m *i2400m = &i2400mu->i2400m; + struct usb_endpoint_descriptor *epd; + int pipe; + + epd = usb_get_epd(i2400mu->usb_iface, + i2400mu->endpoint_cfg.notification); + pipe = usb_rcvintpipe(i2400mu->usb_dev, epd->bEndpointAddress); + usb_fill_int_urb(urb, i2400mu->usb_dev, pipe, + i2400m->bm_ack_buf, I2400M_BM_ACK_BUF_SIZE, + __i2400mu_bm_notif_cb, completion, + epd->bInterval); + return usb_submit_urb(urb, GFP_KERNEL); +} + + +/* + * Read an ack from the notification endpoint + * + * @i2400m: + * @_ack: pointer to where to store the read data + * @ack_size: how many bytes we should read + * + * Returns: < 0 errno code on error; otherwise, amount of received bytes. + * + * Submits a notification read, appends the read data to the given ack + * buffer and then repeats (until @ack_size bytes have been + * received). + */ +ssize_t i2400mu_bus_bm_wait_for_ack(struct i2400m *i2400m, + struct i2400m_bootrom_header *_ack, + size_t ack_size) +{ + ssize_t result = -ENOMEM; + struct device *dev = i2400m_dev(i2400m); + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + struct urb notif_urb; + void *ack = _ack; + size_t offset, len; + long val; + int do_autopm = 1; + DECLARE_COMPLETION_ONSTACK(notif_completion); + + d_fnstart(8, dev, "(i2400m %p ack %p size %zu)\n", + i2400m, ack, ack_size); + BUG_ON(_ack == i2400m->bm_ack_buf); + result = usb_autopm_get_interface(i2400mu->usb_iface); + if (result < 0) { + dev_err(dev, "BM-ACK: can't get autopm: %d\n", (int) result); + do_autopm = 0; + } + usb_init_urb(¬if_urb); /* ready notifications */ + usb_get_urb(¬if_urb); + offset = 0; + while (offset < ack_size) { + init_completion(¬if_completion); + result = i2400mu_notif_submit(i2400mu, ¬if_urb, + ¬if_completion); + if (result < 0) + goto error_notif_urb_submit; + val = wait_for_completion_interruptible_timeout( + ¬if_completion, HZ); + if (val == 0) { + result = -ETIMEDOUT; + usb_kill_urb(¬if_urb); /* Timedout */ + goto error_notif_wait; + } + if (val == -ERESTARTSYS) { + result = -EINTR; /* Interrupted */ + usb_kill_urb(¬if_urb); + goto error_notif_wait; + } + result = notif_urb.status; /* How was the ack? */ + switch (result) { + case 0: + break; + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* just ignore it */ + case -ESHUTDOWN: /* and exit */ + case -ECONNRESET: + result = -ESHUTDOWN; + goto error_dev_gone; + default: /* any other? */ + usb_kill_urb(¬if_urb); /* Timedout */ + if (edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) + goto error_exceeded; + dev_err(dev, "BM-ACK: URB error %d, " + "retrying\n", notif_urb.status); + continue; /* retry */ + } + if (notif_urb.actual_length == 0) { + d_printf(6, dev, "ZLP received, retrying\n"); + continue; + } + /* Got data, append it to the buffer */ + len = min(ack_size - offset, (size_t) notif_urb.actual_length); + memcpy(ack + offset, i2400m->bm_ack_buf, len); + offset += len; + } + result = offset; +error_notif_urb_submit: +error_notif_wait: +error_dev_gone: +out: + if (do_autopm) + usb_autopm_put_interface(i2400mu->usb_iface); + d_fnend(8, dev, "(i2400m %p ack %p size %zu) = %ld\n", + i2400m, ack, ack_size, (long) result); + usb_put_urb(¬if_urb); + return result; + +error_exceeded: + dev_err(dev, "bm: maximum errors in notification URB exceeded; " + "resetting device\n"); + usb_queue_reset_device(i2400mu->usb_iface); + goto out; +} diff --git a/drivers/staging/wimax/i2400m/usb-notif.c b/drivers/staging/wimax/i2400m/usb-notif.c new file mode 100644 index 000000000000..5d429f816125 --- /dev/null +++ b/drivers/staging/wimax/i2400m/usb-notif.c @@ -0,0 +1,258 @@ +/* + * Intel Wireless WiMAX Connection 2400m over USB + * Notification handling + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * Inaky Perez-Gonzalez + * - Initial implementation + * + * + * The notification endpoint is active when the device is not in boot + * mode; in here we just read and get notifications; based on those, + * we act to either reinitialize the device after a reboot or to + * submit a RX request. + * + * ROADMAP + * + * i2400mu_usb_notification_setup() + * + * i2400mu_usb_notification_release() + * + * i2400mu_usb_notification_cb() Called when a URB is ready + * i2400mu_notif_grok() + * i2400m_is_boot_barker() + * i2400m_dev_reset_handle() + * i2400mu_rx_kick() + */ +#include +#include +#include "i2400m-usb.h" + + +#define D_SUBMODULE notif +#include "usb-debug-levels.h" + + +static const +__le32 i2400m_ZERO_BARKER[4] = { 0, 0, 0, 0 }; + + +/* + * Process a received notification + * + * In normal operation mode, we can only receive two types of payloads + * on the notification endpoint: + * + * - a reboot barker, we do a bootstrap (the device has reseted). + * + * - a block of zeroes: there is pending data in the IN endpoint + */ +static +int i2400mu_notification_grok(struct i2400mu *i2400mu, const void *buf, + size_t buf_len) +{ + int ret; + struct device *dev = &i2400mu->usb_iface->dev; + struct i2400m *i2400m = &i2400mu->i2400m; + + d_fnstart(4, dev, "(i2400m %p buf %p buf_len %zu)\n", + i2400mu, buf, buf_len); + ret = -EIO; + if (buf_len < sizeof(i2400m_ZERO_BARKER)) + /* Not a bug, just ignore */ + goto error_bad_size; + ret = 0; + if (!memcmp(i2400m_ZERO_BARKER, buf, sizeof(i2400m_ZERO_BARKER))) { + i2400mu_rx_kick(i2400mu); + goto out; + } + ret = i2400m_is_boot_barker(i2400m, buf, buf_len); + if (unlikely(ret >= 0)) + ret = i2400m_dev_reset_handle(i2400m, "device rebooted"); + else /* Unknown or unexpected data in the notif message */ + i2400m_unknown_barker(i2400m, buf, buf_len); +error_bad_size: +out: + d_fnend(4, dev, "(i2400m %p buf %p buf_len %zu) = %d\n", + i2400mu, buf, buf_len, ret); + return ret; +} + + +/* + * URB callback for the notification endpoint + * + * @urb: the urb received from the notification endpoint + * + * This function will just process the USB side of the transaction, + * checking everything is fine, pass the processing to + * i2400m_notification_grok() and resubmit the URB. + */ +static +void i2400mu_notification_cb(struct urb *urb) +{ + int ret; + struct i2400mu *i2400mu = urb->context; + struct device *dev = &i2400mu->usb_iface->dev; + + d_fnstart(4, dev, "(urb %p status %d actual_length %d)\n", + urb, urb->status, urb->actual_length); + ret = urb->status; + switch (ret) { + case 0: + ret = i2400mu_notification_grok(i2400mu, urb->transfer_buffer, + urb->actual_length); + if (ret == -EIO && edc_inc(&i2400mu->urb_edc, EDC_MAX_ERRORS, + EDC_ERROR_TIMEFRAME)) + goto error_exceeded; + if (ret == -ENOMEM) /* uff...power cycle? shutdown? */ + goto error_exceeded; + break; + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* ditto */ + case -ESHUTDOWN: /* URB killed */ + case -ECONNRESET: /* disconnection */ + goto out; /* Notify around */ + default: /* Some error? */ + if (edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) + goto error_exceeded; + dev_err(dev, "notification: URB error %d, retrying\n", + urb->status); + } + usb_mark_last_busy(i2400mu->usb_dev); + ret = usb_submit_urb(i2400mu->notif_urb, GFP_ATOMIC); + switch (ret) { + case 0: + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* ditto */ + case -ESHUTDOWN: /* URB killed */ + case -ECONNRESET: /* disconnection */ + break; /* just ignore */ + default: /* Some error? */ + dev_err(dev, "notification: cannot submit URB: %d\n", ret); + goto error_submit; + } + d_fnend(4, dev, "(urb %p status %d actual_length %d) = void\n", + urb, urb->status, urb->actual_length); + return; + +error_exceeded: + dev_err(dev, "maximum errors in notification URB exceeded; " + "resetting device\n"); +error_submit: + usb_queue_reset_device(i2400mu->usb_iface); +out: + d_fnend(4, dev, "(urb %p status %d actual_length %d) = void\n", + urb, urb->status, urb->actual_length); +} + + +/* + * setup the notification endpoint + * + * @i2400m: device descriptor + * + * This procedure prepares the notification urb and handler for receiving + * unsolicited barkers from the device. + */ +int i2400mu_notification_setup(struct i2400mu *i2400mu) +{ + struct device *dev = &i2400mu->usb_iface->dev; + int usb_pipe, ret = 0; + struct usb_endpoint_descriptor *epd; + char *buf; + + d_fnstart(4, dev, "(i2400m %p)\n", i2400mu); + buf = kmalloc(I2400MU_MAX_NOTIFICATION_LEN, GFP_KERNEL | GFP_DMA); + if (buf == NULL) { + ret = -ENOMEM; + goto error_buf_alloc; + } + + i2400mu->notif_urb = usb_alloc_urb(0, GFP_KERNEL); + if (!i2400mu->notif_urb) { + ret = -ENOMEM; + goto error_alloc_urb; + } + epd = usb_get_epd(i2400mu->usb_iface, + i2400mu->endpoint_cfg.notification); + usb_pipe = usb_rcvintpipe(i2400mu->usb_dev, epd->bEndpointAddress); + usb_fill_int_urb(i2400mu->notif_urb, i2400mu->usb_dev, usb_pipe, + buf, I2400MU_MAX_NOTIFICATION_LEN, + i2400mu_notification_cb, i2400mu, epd->bInterval); + ret = usb_submit_urb(i2400mu->notif_urb, GFP_KERNEL); + if (ret != 0) { + dev_err(dev, "notification: cannot submit URB: %d\n", ret); + goto error_submit; + } + d_fnend(4, dev, "(i2400m %p) = %d\n", i2400mu, ret); + return ret; + +error_submit: + usb_free_urb(i2400mu->notif_urb); +error_alloc_urb: + kfree(buf); +error_buf_alloc: + d_fnend(4, dev, "(i2400m %p) = %d\n", i2400mu, ret); + return ret; +} + + +/* + * Tear down of the notification mechanism + * + * @i2400m: device descriptor + * + * Kill the interrupt endpoint urb, free any allocated resources. + * + * We need to check if we have done it before as for example, + * _suspend() call this; if after a suspend() we get a _disconnect() + * (as the case is when hibernating), nothing bad happens. + */ +void i2400mu_notification_release(struct i2400mu *i2400mu) +{ + struct device *dev = &i2400mu->usb_iface->dev; + + d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); + if (i2400mu->notif_urb != NULL) { + usb_kill_urb(i2400mu->notif_urb); + kfree(i2400mu->notif_urb->transfer_buffer); + usb_free_urb(i2400mu->notif_urb); + i2400mu->notif_urb = NULL; + } + d_fnend(4, dev, "(i2400mu %p)\n", i2400mu); +} diff --git a/drivers/staging/wimax/i2400m/usb-rx.c b/drivers/staging/wimax/i2400m/usb-rx.c new file mode 100644 index 000000000000..5b64bda7d9e7 --- /dev/null +++ b/drivers/staging/wimax/i2400m/usb-rx.c @@ -0,0 +1,462 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * USB RX handling + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * - Initial implementation + * Inaky Perez-Gonzalez + * - Use skb_clone(), break up processing in chunks + * - Split transport/device specific + * - Make buffer size dynamic to exert less memory pressure + * + * + * This handles the RX path on USB. + * + * When a notification is received that says 'there is RX data ready', + * we call i2400mu_rx_kick(); that wakes up the RX kthread, which + * reads a buffer from USB and passes it to i2400m_rx() in the generic + * handling code. The RX buffer has an specific format that is + * described in rx.c. + * + * We use a kernel thread in a loop because: + * + * - we want to be able to call the USB power management get/put + * functions (blocking) before each transaction. + * + * - We might get a lot of notifications and we don't want to submit + * a zillion reads; by serializing, we are throttling. + * + * - RX data processing can get heavy enough so that it is not + * appropriate for doing it in the USB callback; thus we run it in a + * process context. + * + * We provide a read buffer of an arbitrary size (short of a page); if + * the callback reports -EOVERFLOW, it means it was too small, so we + * just double the size and retry (being careful to append, as + * sometimes the device provided some data). Every now and then we + * check if the average packet size is smaller than the current packet + * size and if so, we halve it. At the end, the size of the + * preallocated buffer should be following the average received + * transaction size, adapting dynamically to it. + * + * ROADMAP + * + * i2400mu_rx_kick() Called from notif.c when we get a + * 'data ready' notification + * i2400mu_rxd() Kernel RX daemon + * i2400mu_rx() Receive USB data + * i2400m_rx() Send data to generic i2400m RX handling + * + * i2400mu_rx_setup() called from i2400mu_bus_dev_start() + * + * i2400mu_rx_release() called from i2400mu_bus_dev_stop() + */ +#include +#include +#include +#include "i2400m-usb.h" + + +#define D_SUBMODULE rx +#include "usb-debug-levels.h" + +/* + * Dynamic RX size + * + * We can't let the rx_size be a multiple of 512 bytes (the RX + * endpoint's max packet size). On some USB host controllers (we + * haven't been able to fully characterize which), if the device is + * about to send (for example) X bytes and we only post a buffer to + * receive n*512, it will fail to mark that as babble (so that + * i2400mu_rx() [case -EOVERFLOW] can resize the buffer and get the + * rest). + * + * So on growing or shrinking, if it is a multiple of the + * maxpacketsize, we remove some (instead of incresing some, so in a + * buddy allocator we try to waste less space). + * + * Note we also need a hook for this on i2400mu_rx() -- when we do the + * first read, we are sure we won't hit this spot because + * i240mm->rx_size has been set properly. However, if we have to + * double because of -EOVERFLOW, when we launch the read to get the + * rest of the data, we *have* to make sure that also is not a + * multiple of the max_pkt_size. + */ + +static +size_t i2400mu_rx_size_grow(struct i2400mu *i2400mu) +{ + struct device *dev = &i2400mu->usb_iface->dev; + size_t rx_size; + const size_t max_pkt_size = 512; + + rx_size = 2 * i2400mu->rx_size; + if (rx_size % max_pkt_size == 0) { + rx_size -= 8; + d_printf(1, dev, + "RX: expected size grew to %zu [adjusted -8] " + "from %zu\n", + rx_size, i2400mu->rx_size); + } else + d_printf(1, dev, + "RX: expected size grew to %zu from %zu\n", + rx_size, i2400mu->rx_size); + return rx_size; +} + + +static +void i2400mu_rx_size_maybe_shrink(struct i2400mu *i2400mu) +{ + const size_t max_pkt_size = 512; + struct device *dev = &i2400mu->usb_iface->dev; + + if (unlikely(i2400mu->rx_size_cnt >= 100 + && i2400mu->rx_size_auto_shrink)) { + size_t avg_rx_size = + i2400mu->rx_size_acc / i2400mu->rx_size_cnt; + size_t new_rx_size = i2400mu->rx_size / 2; + if (avg_rx_size < new_rx_size) { + if (new_rx_size % max_pkt_size == 0) { + new_rx_size -= 8; + d_printf(1, dev, + "RX: expected size shrank to %zu " + "[adjusted -8] from %zu\n", + new_rx_size, i2400mu->rx_size); + } else + d_printf(1, dev, + "RX: expected size shrank to %zu " + "from %zu\n", + new_rx_size, i2400mu->rx_size); + i2400mu->rx_size = new_rx_size; + i2400mu->rx_size_cnt = 0; + i2400mu->rx_size_acc = i2400mu->rx_size; + } + } +} + +/* + * Receive a message with payloads from the USB bus into an skb + * + * @i2400mu: USB device descriptor + * @rx_skb: skb where to place the received message + * + * Deals with all the USB-specifics of receiving, dynamically + * increasing the buffer size if so needed. Returns the payload in the + * skb, ready to process. On a zero-length packet, we retry. + * + * On soft USB errors, we retry (until they become too frequent and + * then are promoted to hard); on hard USB errors, we reset the + * device. On other errors (skb realloacation, we just drop it and + * hope for the next invocation to solve it). + * + * Returns: pointer to the skb if ok, ERR_PTR on error. + * NOTE: this function might realloc the skb (if it is too small), + * so always update with the one returned. + * ERR_PTR() is < 0 on error. + * Will return NULL if it cannot reallocate -- this can be + * considered a transient retryable error. + */ +static +struct sk_buff *i2400mu_rx(struct i2400mu *i2400mu, struct sk_buff *rx_skb) +{ + int result = 0; + struct device *dev = &i2400mu->usb_iface->dev; + int usb_pipe, read_size, rx_size, do_autopm; + struct usb_endpoint_descriptor *epd; + const size_t max_pkt_size = 512; + + d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); + do_autopm = atomic_read(&i2400mu->do_autopm); + result = do_autopm ? + usb_autopm_get_interface(i2400mu->usb_iface) : 0; + if (result < 0) { + dev_err(dev, "RX: can't get autopm: %d\n", result); + do_autopm = 0; + } + epd = usb_get_epd(i2400mu->usb_iface, i2400mu->endpoint_cfg.bulk_in); + usb_pipe = usb_rcvbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); +retry: + rx_size = skb_end_pointer(rx_skb) - rx_skb->data - rx_skb->len; + if (unlikely(rx_size % max_pkt_size == 0)) { + rx_size -= 8; + d_printf(1, dev, "RX: rx_size adapted to %d [-8]\n", rx_size); + } + result = usb_bulk_msg( + i2400mu->usb_dev, usb_pipe, rx_skb->data + rx_skb->len, + rx_size, &read_size, 200); + usb_mark_last_busy(i2400mu->usb_dev); + switch (result) { + case 0: + if (read_size == 0) + goto retry; /* ZLP, just resubmit */ + skb_put(rx_skb, read_size); + break; + case -EPIPE: + /* + * Stall -- maybe the device is choking with our + * requests. Clear it and give it some time. If they + * happen to often, it might be another symptom, so we + * reset. + * + * No error handling for usb_clear_halt(0; if it + * works, the retry works; if it fails, this switch + * does the error handling for us. + */ + if (edc_inc(&i2400mu->urb_edc, + 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "BM-CMD: too many stalls in " + "URB; resetting device\n"); + goto do_reset; + } + usb_clear_halt(i2400mu->usb_dev, usb_pipe); + msleep(10); /* give the device some time */ + goto retry; + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* just ignore it */ + case -ESHUTDOWN: + case -ECONNRESET: + break; + case -EOVERFLOW: { /* too small, reallocate */ + struct sk_buff *new_skb; + rx_size = i2400mu_rx_size_grow(i2400mu); + if (rx_size <= (1 << 16)) /* cap it */ + i2400mu->rx_size = rx_size; + else if (printk_ratelimit()) { + dev_err(dev, "BUG? rx_size up to %d\n", rx_size); + result = -EINVAL; + goto out; + } + skb_put(rx_skb, read_size); + new_skb = skb_copy_expand(rx_skb, 0, rx_size - rx_skb->len, + GFP_KERNEL); + if (new_skb == NULL) { + kfree_skb(rx_skb); + rx_skb = NULL; + goto out; /* drop it...*/ + } + kfree_skb(rx_skb); + rx_skb = new_skb; + i2400mu->rx_size_cnt = 0; + i2400mu->rx_size_acc = i2400mu->rx_size; + d_printf(1, dev, "RX: size changed to %d, received %d, " + "copied %d, capacity %ld\n", + rx_size, read_size, rx_skb->len, + (long) skb_end_offset(new_skb)); + goto retry; + } + /* In most cases, it happens due to the hardware scheduling a + * read when there was no data - unfortunately, we have no way + * to tell this timeout from a USB timeout. So we just ignore + * it. */ + case -ETIMEDOUT: + dev_err(dev, "RX: timeout: %d\n", result); + result = 0; + break; + default: /* Any error */ + if (edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) + goto error_reset; + dev_err(dev, "RX: error receiving URB: %d, retrying\n", result); + goto retry; + } +out: + if (do_autopm) + usb_autopm_put_interface(i2400mu->usb_iface); + d_fnend(4, dev, "(i2400mu %p) = %p\n", i2400mu, rx_skb); + return rx_skb; + +error_reset: + dev_err(dev, "RX: maximum errors in URB exceeded; " + "resetting device\n"); +do_reset: + usb_queue_reset_device(i2400mu->usb_iface); + rx_skb = ERR_PTR(result); + goto out; +} + + +/* + * Kernel thread for USB reception of data + * + * This thread waits for a kick; once kicked, it will allocate an skb + * and receive a single message to it from USB (using + * i2400mu_rx()). Once received, it is passed to the generic i2400m RX + * code for processing. + * + * When done processing, it runs some dirty statistics to verify if + * the last 100 messages received were smaller than half of the + * current RX buffer size. In that case, the RX buffer size is + * halved. This will helps lowering the pressure on the memory + * allocator. + * + * Hard errors force the thread to exit. + */ +static +int i2400mu_rxd(void *_i2400mu) +{ + int result = 0; + struct i2400mu *i2400mu = _i2400mu; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = &i2400mu->usb_iface->dev; + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + size_t pending; + int rx_size; + struct sk_buff *rx_skb; + unsigned long flags; + + d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); + spin_lock_irqsave(&i2400m->rx_lock, flags); + BUG_ON(i2400mu->rx_kthread != NULL); + i2400mu->rx_kthread = current; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + while (1) { + d_printf(2, dev, "RX: waiting for messages\n"); + pending = 0; + wait_event_interruptible( + i2400mu->rx_wq, + (kthread_should_stop() /* check this first! */ + || (pending = atomic_read(&i2400mu->rx_pending_count))) + ); + if (kthread_should_stop()) + break; + if (pending == 0) + continue; + rx_size = i2400mu->rx_size; + d_printf(2, dev, "RX: reading up to %d bytes\n", rx_size); + rx_skb = __netdev_alloc_skb(net_dev, rx_size, GFP_KERNEL); + if (rx_skb == NULL) { + dev_err(dev, "RX: can't allocate skb [%d bytes]\n", + rx_size); + msleep(50); /* give it some time? */ + continue; + } + + /* Receive the message with the payloads */ + rx_skb = i2400mu_rx(i2400mu, rx_skb); + result = PTR_ERR(rx_skb); + if (IS_ERR(rx_skb)) + goto out; + atomic_dec(&i2400mu->rx_pending_count); + if (rx_skb == NULL || rx_skb->len == 0) { + /* some "ignorable" condition */ + kfree_skb(rx_skb); + continue; + } + + /* Deliver the message to the generic i2400m code */ + i2400mu->rx_size_cnt++; + i2400mu->rx_size_acc += rx_skb->len; + result = i2400m_rx(i2400m, rx_skb); + if (result == -EIO + && edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + goto error_reset; + } + + /* Maybe adjust RX buffer size */ + i2400mu_rx_size_maybe_shrink(i2400mu); + } + result = 0; +out: + spin_lock_irqsave(&i2400m->rx_lock, flags); + i2400mu->rx_kthread = NULL; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + d_fnend(4, dev, "(i2400mu %p) = %d\n", i2400mu, result); + return result; + +error_reset: + dev_err(dev, "RX: maximum errors in received buffer exceeded; " + "resetting device\n"); + usb_queue_reset_device(i2400mu->usb_iface); + goto out; +} + + +/* + * Start reading from the device + * + * @i2400m: device instance + * + * Notify the RX thread that there is data pending. + */ +void i2400mu_rx_kick(struct i2400mu *i2400mu) +{ + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = &i2400mu->usb_iface->dev; + + d_fnstart(3, dev, "(i2400mu %p)\n", i2400m); + atomic_inc(&i2400mu->rx_pending_count); + wake_up_all(&i2400mu->rx_wq); + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} + + +int i2400mu_rx_setup(struct i2400mu *i2400mu) +{ + int result = 0; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = &i2400mu->usb_iface->dev; + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + struct task_struct *kthread; + + kthread = kthread_run(i2400mu_rxd, i2400mu, "%s-rx", + wimax_dev->name); + /* the kthread function sets i2400mu->rx_thread */ + if (IS_ERR(kthread)) { + result = PTR_ERR(kthread); + dev_err(dev, "RX: cannot start thread: %d\n", result); + } + return result; +} + + +void i2400mu_rx_release(struct i2400mu *i2400mu) +{ + unsigned long flags; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = i2400m_dev(i2400m); + struct task_struct *kthread; + + spin_lock_irqsave(&i2400m->rx_lock, flags); + kthread = i2400mu->rx_kthread; + i2400mu->rx_kthread = NULL; + spin_unlock_irqrestore(&i2400m->rx_lock, flags); + if (kthread) + kthread_stop(kthread); + else + d_printf(1, dev, "RX: kthread had already exited\n"); +} + diff --git a/drivers/staging/wimax/i2400m/usb-tx.c b/drivers/staging/wimax/i2400m/usb-tx.c new file mode 100644 index 000000000000..3ba9d70cca1b --- /dev/null +++ b/drivers/staging/wimax/i2400m/usb-tx.c @@ -0,0 +1,273 @@ +/* + * Intel Wireless WiMAX Connection 2400m + * USB specific TX handling + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Yanir Lubetkin + * - Initial implementation + * Inaky Perez-Gonzalez + * - Split transport/device specific + * + * + * Takes the TX messages in the i2400m's driver TX FIFO and sends them + * to the device until there are no more. + * + * If we fail sending the message, we just drop it. There isn't much + * we can do at this point. We could also retry, but the USB stack has + * already retried and still failed, so there is not much of a + * point. As well, most of the traffic is network, which has recovery + * methods for dropped packets. + * + * For sending we just obtain a FIFO buffer to send, send it to the + * USB bulk out, tell the TX FIFO code we have sent it; query for + * another one, etc... until done. + * + * We use a thread so we can call usb_autopm_enable() and + * usb_autopm_disable() for each transaction; this way when the device + * goes idle, it will suspend. It also has less overhead than a + * dedicated workqueue, as it is being used for a single task. + * + * ROADMAP + * + * i2400mu_tx_setup() + * i2400mu_tx_release() + * + * i2400mu_bus_tx_kick() - Called by the tx.c code when there + * is new data in the FIFO. + * i2400mu_txd() + * i2400m_tx_msg_get() + * i2400m_tx_msg_sent() + */ +#include "i2400m-usb.h" + + +#define D_SUBMODULE tx +#include "usb-debug-levels.h" + + +/* + * Get the next TX message in the TX FIFO and send it to the device + * + * Note that any iteration consumes a message to be sent, no matter if + * it succeeds or fails (we have no real way to retry or complain). + * + * Return: 0 if ok, < 0 errno code on hard error. + */ +static +int i2400mu_tx(struct i2400mu *i2400mu, struct i2400m_msg_hdr *tx_msg, + size_t tx_msg_size) +{ + int result = 0; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = &i2400mu->usb_iface->dev; + int usb_pipe, sent_size, do_autopm; + struct usb_endpoint_descriptor *epd; + + d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); + do_autopm = atomic_read(&i2400mu->do_autopm); + result = do_autopm ? + usb_autopm_get_interface(i2400mu->usb_iface) : 0; + if (result < 0) { + dev_err(dev, "TX: can't get autopm: %d\n", result); + do_autopm = 0; + } + epd = usb_get_epd(i2400mu->usb_iface, i2400mu->endpoint_cfg.bulk_out); + usb_pipe = usb_sndbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); +retry: + result = usb_bulk_msg(i2400mu->usb_dev, usb_pipe, + tx_msg, tx_msg_size, &sent_size, 200); + usb_mark_last_busy(i2400mu->usb_dev); + switch (result) { + case 0: + if (sent_size != tx_msg_size) { /* Too short? drop it */ + dev_err(dev, "TX: short write (%d B vs %zu " + "expected)\n", sent_size, tx_msg_size); + result = -EIO; + } + break; + case -EPIPE: + /* + * Stall -- maybe the device is choking with our + * requests. Clear it and give it some time. If they + * happen to often, it might be another symptom, so we + * reset. + * + * No error handling for usb_clear_halt(0; if it + * works, the retry works; if it fails, this switch + * does the error handling for us. + */ + if (edc_inc(&i2400mu->urb_edc, + 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "BM-CMD: too many stalls in " + "URB; resetting device\n"); + usb_queue_reset_device(i2400mu->usb_iface); + } else { + usb_clear_halt(i2400mu->usb_dev, usb_pipe); + msleep(10); /* give the device some time */ + goto retry; + } + fallthrough; + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* just ignore it */ + case -ESHUTDOWN: /* and exit */ + case -ECONNRESET: + result = -ESHUTDOWN; + break; + default: /* Some error? */ + if (edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "TX: maximum errors in URB " + "exceeded; resetting device\n"); + usb_queue_reset_device(i2400mu->usb_iface); + } else { + dev_err(dev, "TX: cannot send URB; retrying. " + "tx_msg @%zu %zu B [%d sent]: %d\n", + (void *) tx_msg - i2400m->tx_buf, + tx_msg_size, sent_size, result); + goto retry; + } + } + if (do_autopm) + usb_autopm_put_interface(i2400mu->usb_iface); + d_fnend(4, dev, "(i2400mu %p) = result\n", i2400mu); + return result; +} + + +/* + * Get the next TX message in the TX FIFO and send it to the device + * + * Note we exit the loop if i2400mu_tx() fails; that function only + * fails on hard error (failing to tx a buffer not being one of them, + * see its doc). + * + * Return: 0 + */ +static +int i2400mu_txd(void *_i2400mu) +{ + struct i2400mu *i2400mu = _i2400mu; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = &i2400mu->usb_iface->dev; + struct i2400m_msg_hdr *tx_msg; + size_t tx_msg_size; + unsigned long flags; + + d_fnstart(4, dev, "(i2400mu %p)\n", i2400mu); + + spin_lock_irqsave(&i2400m->tx_lock, flags); + BUG_ON(i2400mu->tx_kthread != NULL); + i2400mu->tx_kthread = current; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + + while (1) { + d_printf(2, dev, "TX: waiting for messages\n"); + tx_msg = NULL; + wait_event_interruptible( + i2400mu->tx_wq, + (kthread_should_stop() /* check this first! */ + || (tx_msg = i2400m_tx_msg_get(i2400m, &tx_msg_size))) + ); + if (kthread_should_stop()) + break; + WARN_ON(tx_msg == NULL); /* should not happen...*/ + d_printf(2, dev, "TX: submitting %zu bytes\n", tx_msg_size); + d_dump(5, dev, tx_msg, tx_msg_size); + /* Yeah, we ignore errors ... not much we can do */ + i2400mu_tx(i2400mu, tx_msg, tx_msg_size); + i2400m_tx_msg_sent(i2400m); /* ack it, advance the FIFO */ + } + + spin_lock_irqsave(&i2400m->tx_lock, flags); + i2400mu->tx_kthread = NULL; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + + d_fnend(4, dev, "(i2400mu %p)\n", i2400mu); + return 0; +} + + +/* + * i2400m TX engine notifies us that there is data in the FIFO ready + * for TX + * + * If there is a URB in flight, don't do anything; when it finishes, + * it will see there is data in the FIFO and send it. Else, just + * submit a write. + */ +void i2400mu_bus_tx_kick(struct i2400m *i2400m) +{ + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + struct device *dev = &i2400mu->usb_iface->dev; + + d_fnstart(3, dev, "(i2400m %p) = void\n", i2400m); + wake_up_all(&i2400mu->tx_wq); + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} + + +int i2400mu_tx_setup(struct i2400mu *i2400mu) +{ + int result = 0; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = &i2400mu->usb_iface->dev; + struct wimax_dev *wimax_dev = &i2400m->wimax_dev; + struct task_struct *kthread; + + kthread = kthread_run(i2400mu_txd, i2400mu, "%s-tx", + wimax_dev->name); + /* the kthread function sets i2400mu->tx_thread */ + if (IS_ERR(kthread)) { + result = PTR_ERR(kthread); + dev_err(dev, "TX: cannot start thread: %d\n", result); + } + return result; +} + +void i2400mu_tx_release(struct i2400mu *i2400mu) +{ + unsigned long flags; + struct i2400m *i2400m = &i2400mu->i2400m; + struct device *dev = i2400m_dev(i2400m); + struct task_struct *kthread; + + spin_lock_irqsave(&i2400m->tx_lock, flags); + kthread = i2400mu->tx_kthread; + i2400mu->tx_kthread = NULL; + spin_unlock_irqrestore(&i2400m->tx_lock, flags); + if (kthread) + kthread_stop(kthread); + else + d_printf(1, dev, "TX: kthread had already exited\n"); +} diff --git a/drivers/staging/wimax/i2400m/usb.c b/drivers/staging/wimax/i2400m/usb.c new file mode 100644 index 000000000000..3b84dd7b5567 --- /dev/null +++ b/drivers/staging/wimax/i2400m/usb.c @@ -0,0 +1,764 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Intel Wireless WiMAX Connection 2400m + * Linux driver model glue for USB device, reset & fw upload + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + * Yanir Lubetkin + * + * See i2400m-usb.h for a general description of this driver. + * + * This file implements driver model glue, and hook ups for the + * generic driver to implement the bus-specific functions (device + * communication setup/tear down, firmware upload and resetting). + * + * ROADMAP + * + * i2400mu_probe() + * alloc_netdev()... + * i2400mu_netdev_setup() + * i2400mu_init() + * i2400m_netdev_setup() + * i2400m_setup()... + * + * i2400mu_disconnect + * i2400m_release() + * free_netdev() + * + * i2400mu_suspend() + * i2400m_cmd_enter_powersave() + * i2400mu_notification_release() + * + * i2400mu_resume() + * i2400mu_notification_setup() + * + * i2400mu_bus_dev_start() Called by i2400m_dev_start() [who is + * i2400mu_tx_setup() called by i2400m_setup()] + * i2400mu_rx_setup() + * i2400mu_notification_setup() + * + * i2400mu_bus_dev_stop() Called by i2400m_dev_stop() [who is + * i2400mu_notification_release() called by i2400m_release()] + * i2400mu_rx_release() + * i2400mu_tx_release() + * + * i2400mu_bus_reset() Called by i2400m_reset + * __i2400mu_reset() + * __i2400mu_send_barker() + * usb_reset_device() + */ +#include "i2400m-usb.h" +#include "linux-wimax-i2400m.h" +#include +#include +#include + + +#define D_SUBMODULE usb +#include "usb-debug-levels.h" + +static char i2400mu_debug_params[128]; +module_param_string(debug, i2400mu_debug_params, sizeof(i2400mu_debug_params), + 0644); +MODULE_PARM_DESC(debug, + "String of space-separated NAME:VALUE pairs, where NAMEs " + "are the different debug submodules and VALUE are the " + "initial debug value to set."); + +/* Our firmware file name */ +static const char *i2400mu_bus_fw_names_5x50[] = { +#define I2400MU_FW_FILE_NAME_v1_5 "i2400m-fw-usb-1.5.sbcf" + I2400MU_FW_FILE_NAME_v1_5, +#define I2400MU_FW_FILE_NAME_v1_4 "i2400m-fw-usb-1.4.sbcf" + I2400MU_FW_FILE_NAME_v1_4, + NULL, +}; + + +static const char *i2400mu_bus_fw_names_6050[] = { +#define I6050U_FW_FILE_NAME_v1_5 "i6050-fw-usb-1.5.sbcf" + I6050U_FW_FILE_NAME_v1_5, + NULL, +}; + + +static +int i2400mu_bus_dev_start(struct i2400m *i2400m) +{ + int result; + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + struct device *dev = &i2400mu->usb_iface->dev; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + result = i2400mu_tx_setup(i2400mu); + if (result < 0) + goto error_usb_tx_setup; + result = i2400mu_rx_setup(i2400mu); + if (result < 0) + goto error_usb_rx_setup; + result = i2400mu_notification_setup(i2400mu); + if (result < 0) + goto error_notif_setup; + d_fnend(3, dev, "(i2400m %p) = %d\n", i2400m, result); + return result; + +error_notif_setup: + i2400mu_rx_release(i2400mu); +error_usb_rx_setup: + i2400mu_tx_release(i2400mu); +error_usb_tx_setup: + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); + return result; +} + + +static +void i2400mu_bus_dev_stop(struct i2400m *i2400m) +{ + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + struct device *dev = &i2400mu->usb_iface->dev; + + d_fnstart(3, dev, "(i2400m %p)\n", i2400m); + i2400mu_notification_release(i2400mu); + i2400mu_rx_release(i2400mu); + i2400mu_tx_release(i2400mu); + d_fnend(3, dev, "(i2400m %p) = void\n", i2400m); +} + + +/* + * Sends a barker buffer to the device + * + * This helper will allocate a kmalloced buffer and use it to transmit + * (then free it). Reason for this is that other arches cannot use + * stack/vmalloc/text areas for DMA transfers. + * + * Error recovery here is simpler: anything is considered a hard error + * and will move the reset code to use a last-resort bus-based reset. + */ +static +int __i2400mu_send_barker(struct i2400mu *i2400mu, + const __le32 *barker, + size_t barker_size, + unsigned endpoint) +{ + struct usb_endpoint_descriptor *epd = NULL; + int pipe, actual_len, ret; + struct device *dev = &i2400mu->usb_iface->dev; + void *buffer; + int do_autopm = 1; + + ret = usb_autopm_get_interface(i2400mu->usb_iface); + if (ret < 0) { + dev_err(dev, "RESET: can't get autopm: %d\n", ret); + do_autopm = 0; + } + ret = -ENOMEM; + buffer = kmalloc(barker_size, GFP_KERNEL); + if (buffer == NULL) + goto error_kzalloc; + epd = usb_get_epd(i2400mu->usb_iface, endpoint); + pipe = usb_sndbulkpipe(i2400mu->usb_dev, epd->bEndpointAddress); + memcpy(buffer, barker, barker_size); +retry: + ret = usb_bulk_msg(i2400mu->usb_dev, pipe, buffer, barker_size, + &actual_len, 200); + switch (ret) { + case 0: + if (actual_len != barker_size) { /* Too short? drop it */ + dev_err(dev, "E: %s: short write (%d B vs %zu " + "expected)\n", + __func__, actual_len, barker_size); + ret = -EIO; + } + break; + case -EPIPE: + /* + * Stall -- maybe the device is choking with our + * requests. Clear it and give it some time. If they + * happen to often, it might be another symptom, so we + * reset. + * + * No error handling for usb_clear_halt(0; if it + * works, the retry works; if it fails, this switch + * does the error handling for us. + */ + if (edc_inc(&i2400mu->urb_edc, + 10 * EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "E: %s: too many stalls in " + "URB; resetting device\n", __func__); + usb_queue_reset_device(i2400mu->usb_iface); + /* fallthrough */ + } else { + usb_clear_halt(i2400mu->usb_dev, pipe); + msleep(10); /* give the device some time */ + goto retry; + } + fallthrough; + case -EINVAL: /* while removing driver */ + case -ENODEV: /* dev disconnect ... */ + case -ENOENT: /* just ignore it */ + case -ESHUTDOWN: /* and exit */ + case -ECONNRESET: + ret = -ESHUTDOWN; + break; + default: /* Some error? */ + if (edc_inc(&i2400mu->urb_edc, + EDC_MAX_ERRORS, EDC_ERROR_TIMEFRAME)) { + dev_err(dev, "E: %s: maximum errors in URB " + "exceeded; resetting device\n", + __func__); + usb_queue_reset_device(i2400mu->usb_iface); + } else { + dev_warn(dev, "W: %s: cannot send URB: %d\n", + __func__, ret); + goto retry; + } + } + kfree(buffer); +error_kzalloc: + if (do_autopm) + usb_autopm_put_interface(i2400mu->usb_iface); + return ret; +} + + +/* + * Reset a device at different levels (warm, cold or bus) + * + * @i2400m: device descriptor + * @reset_type: soft, warm or bus reset (I2400M_RT_WARM/SOFT/BUS) + * + * Warm and cold resets get a USB reset if they fail. + * + * Warm reset: + * + * The device will be fully reset internally, but won't be + * disconnected from the USB bus (so no reenumeration will + * happen). Firmware upload will be necessary. + * + * The device will send a reboot barker in the notification endpoint + * that will trigger the driver to reinitialize the state + * automatically from notif.c:i2400m_notification_grok() into + * i2400m_dev_bootstrap_delayed(). + * + * Cold and bus (USB) reset: + * + * The device will be fully reset internally, disconnected from the + * USB bus an a reenumeration will happen. Firmware upload will be + * necessary. Thus, we don't do any locking or struct + * reinitialization, as we are going to be fully disconnected and + * reenumerated. + * + * Note we need to return -ENODEV if a warm reset was requested and we + * had to resort to a bus reset. See i2400m_op_reset(), wimax_reset() + * and wimax_dev->op_reset. + * + * WARNING: no driver state saved/fixed + */ +static +int i2400mu_bus_reset(struct i2400m *i2400m, enum i2400m_reset_type rt) +{ + int result; + struct i2400mu *i2400mu = + container_of(i2400m, struct i2400mu, i2400m); + struct device *dev = i2400m_dev(i2400m); + static const __le32 i2400m_WARM_BOOT_BARKER[4] = { + cpu_to_le32(I2400M_WARM_RESET_BARKER), + cpu_to_le32(I2400M_WARM_RESET_BARKER), + cpu_to_le32(I2400M_WARM_RESET_BARKER), + cpu_to_le32(I2400M_WARM_RESET_BARKER), + }; + static const __le32 i2400m_COLD_BOOT_BARKER[4] = { + cpu_to_le32(I2400M_COLD_RESET_BARKER), + cpu_to_le32(I2400M_COLD_RESET_BARKER), + cpu_to_le32(I2400M_COLD_RESET_BARKER), + cpu_to_le32(I2400M_COLD_RESET_BARKER), + }; + + d_fnstart(3, dev, "(i2400m %p rt %u)\n", i2400m, rt); + if (rt == I2400M_RT_WARM) + result = __i2400mu_send_barker( + i2400mu, i2400m_WARM_BOOT_BARKER, + sizeof(i2400m_WARM_BOOT_BARKER), + i2400mu->endpoint_cfg.bulk_out); + else if (rt == I2400M_RT_COLD) + result = __i2400mu_send_barker( + i2400mu, i2400m_COLD_BOOT_BARKER, + sizeof(i2400m_COLD_BOOT_BARKER), + i2400mu->endpoint_cfg.reset_cold); + else if (rt == I2400M_RT_BUS) { + result = usb_reset_device(i2400mu->usb_dev); + switch (result) { + case 0: + case -EINVAL: /* device is gone */ + case -ENODEV: + case -ENOENT: + case -ESHUTDOWN: + result = 0; + break; /* We assume the device is disconnected */ + default: + dev_err(dev, "USB reset failed (%d), giving up!\n", + result); + } + } else { + result = -EINVAL; /* shut gcc up in certain arches */ + BUG(); + } + if (result < 0 + && result != -EINVAL /* device is gone */ + && rt != I2400M_RT_BUS) { + /* + * Things failed -- resort to lower level reset, that + * we queue in another context; the reason for this is + * that the pre and post reset functionality requires + * the i2400m->init_mutex; RT_WARM and RT_COLD can + * come from areas where i2400m->init_mutex is taken. + */ + dev_err(dev, "%s reset failed (%d); trying USB reset\n", + rt == I2400M_RT_WARM ? "warm" : "cold", result); + usb_queue_reset_device(i2400mu->usb_iface); + result = -ENODEV; + } + d_fnend(3, dev, "(i2400m %p rt %u) = %d\n", i2400m, rt, result); + return result; +} + +static void i2400mu_get_drvinfo(struct net_device *net_dev, + struct ethtool_drvinfo *info) +{ + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + struct usb_device *udev = i2400mu->usb_dev; + + strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); + strlcpy(info->fw_version, i2400m->fw_name ? : "", + sizeof(info->fw_version)); + usb_make_path(udev, info->bus_info, sizeof(info->bus_info)); +} + +static const struct ethtool_ops i2400mu_ethtool_ops = { + .get_drvinfo = i2400mu_get_drvinfo, + .get_link = ethtool_op_get_link, +}; + +static +void i2400mu_netdev_setup(struct net_device *net_dev) +{ + struct i2400m *i2400m = net_dev_to_i2400m(net_dev); + struct i2400mu *i2400mu = container_of(i2400m, struct i2400mu, i2400m); + i2400mu_init(i2400mu); + i2400m_netdev_setup(net_dev); + net_dev->ethtool_ops = &i2400mu_ethtool_ops; +} + + +/* + * Debug levels control; see debug.h + */ +struct d_level D_LEVEL[] = { + D_SUBMODULE_DEFINE(usb), + D_SUBMODULE_DEFINE(fw), + D_SUBMODULE_DEFINE(notif), + D_SUBMODULE_DEFINE(rx), + D_SUBMODULE_DEFINE(tx), +}; +size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); + +static +void i2400mu_debugfs_add(struct i2400mu *i2400mu) +{ + struct dentry *dentry = i2400mu->i2400m.wimax_dev.debugfs_dentry; + + dentry = debugfs_create_dir("i2400m-usb", dentry); + i2400mu->debugfs_dentry = dentry; + + d_level_register_debugfs("dl_", usb, dentry); + d_level_register_debugfs("dl_", fw, dentry); + d_level_register_debugfs("dl_", notif, dentry); + d_level_register_debugfs("dl_", rx, dentry); + d_level_register_debugfs("dl_", tx, dentry); + + /* Don't touch these if you don't know what you are doing */ + debugfs_create_u8("rx_size_auto_shrink", 0600, dentry, + &i2400mu->rx_size_auto_shrink); + + debugfs_create_size_t("rx_size", 0600, dentry, &i2400mu->rx_size); +} + + +static struct device_type i2400mu_type = { + .name = "wimax", +}; + +/* + * Probe a i2400m interface and register it + * + * @iface: USB interface to link to + * @id: USB class/subclass/protocol id + * @returns: 0 if ok, < 0 errno code on error. + * + * Alloc a net device, initialize the bus-specific details and then + * calls the bus-generic initialization routine. That will register + * the wimax and netdev devices, upload the firmware [using + * _bus_bm_*()], call _bus_dev_start() to finalize the setup of the + * communication with the device and then will start to talk to it to + * finnish setting it up. + */ +static +int i2400mu_probe(struct usb_interface *iface, + const struct usb_device_id *id) +{ + int result; + struct net_device *net_dev; + struct device *dev = &iface->dev; + struct i2400m *i2400m; + struct i2400mu *i2400mu; + struct usb_device *usb_dev = interface_to_usbdev(iface); + + if (iface->cur_altsetting->desc.bNumEndpoints < 4) + return -ENODEV; + + if (usb_dev->speed != USB_SPEED_HIGH) + dev_err(dev, "device not connected as high speed\n"); + + /* Allocate instance [calls i2400m_netdev_setup() on it]. */ + result = -ENOMEM; + net_dev = alloc_netdev(sizeof(*i2400mu), "wmx%d", NET_NAME_UNKNOWN, + i2400mu_netdev_setup); + if (net_dev == NULL) { + dev_err(dev, "no memory for network device instance\n"); + goto error_alloc_netdev; + } + SET_NETDEV_DEV(net_dev, dev); + SET_NETDEV_DEVTYPE(net_dev, &i2400mu_type); + i2400m = net_dev_to_i2400m(net_dev); + i2400mu = container_of(i2400m, struct i2400mu, i2400m); + i2400m->wimax_dev.net_dev = net_dev; + i2400mu->usb_dev = usb_get_dev(usb_dev); + i2400mu->usb_iface = iface; + usb_set_intfdata(iface, i2400mu); + + i2400m->bus_tx_block_size = I2400MU_BLK_SIZE; + /* + * Room required in the Tx queue for USB message to accommodate + * a smallest payload while allocating header space is 16 bytes. + * Adding this room for the new tx message increases the + * possibilities of including any payload with size <= 16 bytes. + */ + i2400m->bus_tx_room_min = I2400MU_BLK_SIZE; + i2400m->bus_pl_size_max = I2400MU_PL_SIZE_MAX; + i2400m->bus_setup = NULL; + i2400m->bus_dev_start = i2400mu_bus_dev_start; + i2400m->bus_dev_stop = i2400mu_bus_dev_stop; + i2400m->bus_release = NULL; + i2400m->bus_tx_kick = i2400mu_bus_tx_kick; + i2400m->bus_reset = i2400mu_bus_reset; + i2400m->bus_bm_retries = I2400M_USB_BOOT_RETRIES; + i2400m->bus_bm_cmd_send = i2400mu_bus_bm_cmd_send; + i2400m->bus_bm_wait_for_ack = i2400mu_bus_bm_wait_for_ack; + i2400m->bus_bm_mac_addr_impaired = 0; + + switch (id->idProduct) { + case USB_DEVICE_ID_I6050: + case USB_DEVICE_ID_I6050_2: + case USB_DEVICE_ID_I6150: + case USB_DEVICE_ID_I6150_2: + case USB_DEVICE_ID_I6150_3: + case USB_DEVICE_ID_I6250: + i2400mu->i6050 = 1; + break; + default: + break; + } + + if (i2400mu->i6050) { + i2400m->bus_fw_names = i2400mu_bus_fw_names_6050; + i2400mu->endpoint_cfg.bulk_out = 0; + i2400mu->endpoint_cfg.notification = 3; + i2400mu->endpoint_cfg.reset_cold = 2; + i2400mu->endpoint_cfg.bulk_in = 1; + } else { + i2400m->bus_fw_names = i2400mu_bus_fw_names_5x50; + i2400mu->endpoint_cfg.bulk_out = 0; + i2400mu->endpoint_cfg.notification = 1; + i2400mu->endpoint_cfg.reset_cold = 2; + i2400mu->endpoint_cfg.bulk_in = 3; + } +#ifdef CONFIG_PM + iface->needs_remote_wakeup = 1; /* autosuspend (15s delay) */ + device_init_wakeup(dev, 1); + pm_runtime_set_autosuspend_delay(&usb_dev->dev, 15000); + usb_enable_autosuspend(usb_dev); +#endif + + result = i2400m_setup(i2400m, I2400M_BRI_MAC_REINIT); + if (result < 0) { + dev_err(dev, "cannot setup device: %d\n", result); + goto error_setup; + } + i2400mu_debugfs_add(i2400mu); + return 0; + +error_setup: + usb_set_intfdata(iface, NULL); + usb_put_dev(i2400mu->usb_dev); + free_netdev(net_dev); +error_alloc_netdev: + return result; +} + + +/* + * Disconnect a i2400m from the system. + * + * i2400m_stop() has been called before, so al the rx and tx contexts + * have been taken down already. Make sure the queue is stopped, + * unregister netdev and i2400m, free and kill. + */ +static +void i2400mu_disconnect(struct usb_interface *iface) +{ + struct i2400mu *i2400mu = usb_get_intfdata(iface); + struct i2400m *i2400m = &i2400mu->i2400m; + struct net_device *net_dev = i2400m->wimax_dev.net_dev; + struct device *dev = &iface->dev; + + d_fnstart(3, dev, "(iface %p i2400m %p)\n", iface, i2400m); + + debugfs_remove_recursive(i2400mu->debugfs_dentry); + i2400m_release(i2400m); + usb_set_intfdata(iface, NULL); + usb_put_dev(i2400mu->usb_dev); + free_netdev(net_dev); + d_fnend(3, dev, "(iface %p i2400m %p) = void\n", iface, i2400m); +} + + +/* + * Get the device ready for USB port or system standby and hibernation + * + * USB port and system standby are handled the same. + * + * When the system hibernates, the USB device is powered down and then + * up, so we don't really have to do much here, as it will be seen as + * a reconnect. Still for simplicity we consider this case the same as + * suspend, so that the device has a chance to do notify the base + * station (if connected). + * + * So at the end, the three cases require common handling. + * + * If at the time of this call the device's firmware is not loaded, + * nothing has to be done. Note we can be "loose" about not reading + * i2400m->updown under i2400m->init_mutex. If it happens to change + * inmediately, other parts of the call flow will fail and effectively + * catch it. + * + * If the firmware is loaded, we need to: + * + * - tell the device to go into host interface power save mode, wait + * for it to ack + * + * This is quite more interesting than it is; we need to execute a + * command, but this time, we don't want the code in usb-{tx,rx}.c + * to call the usb_autopm_get/put_interface() barriers as it'd + * deadlock, so we need to decrement i2400mu->do_autopm, that acts + * as a poor man's semaphore. Ugly, but it works. + * + * As well, the device might refuse going to sleep for whichever + * reason. In this case we just fail. For system suspend/hibernate, + * we *can't* fail. We check PMSG_IS_AUTO to see if the + * suspend call comes from the USB stack or from the system and act + * in consequence. + * + * - stop the notification endpoint polling + */ +static +int i2400mu_suspend(struct usb_interface *iface, pm_message_t pm_msg) +{ + int result = 0; + struct device *dev = &iface->dev; + struct i2400mu *i2400mu = usb_get_intfdata(iface); + unsigned is_autosuspend = 0; + struct i2400m *i2400m = &i2400mu->i2400m; + +#ifdef CONFIG_PM + if (PMSG_IS_AUTO(pm_msg)) + is_autosuspend = 1; +#endif + + d_fnstart(3, dev, "(iface %p pm_msg %u)\n", iface, pm_msg.event); + rmb(); /* see i2400m->updown's documentation */ + if (i2400m->updown == 0) + goto no_firmware; + if (i2400m->state == I2400M_SS_DATA_PATH_CONNECTED && is_autosuspend) { + /* ugh -- the device is connected and this suspend + * request is an autosuspend one (not a system standby + * / hibernate). + * + * The only way the device can go to standby is if the + * link with the base station is in IDLE mode; that + * were the case, we'd be in status + * I2400M_SS_CONNECTED_IDLE. But we are not. + * + * If we *tell* him to go power save now, it'll reset + * as a precautionary measure, so if this is an + * autosuspend thing, say no and it'll come back + * later, when the link is IDLE + */ + result = -EBADF; + d_printf(1, dev, "fw up, link up, not-idle, autosuspend: " + "not entering powersave\n"); + goto error_not_now; + } + d_printf(1, dev, "fw up: entering powersave\n"); + atomic_dec(&i2400mu->do_autopm); + result = i2400m_cmd_enter_powersave(i2400m); + atomic_inc(&i2400mu->do_autopm); + if (result < 0 && !is_autosuspend) { + /* System suspend, can't fail */ + dev_err(dev, "failed to suspend, will reset on resume\n"); + result = 0; + } + if (result < 0) + goto error_enter_powersave; + i2400mu_notification_release(i2400mu); + d_printf(1, dev, "powersave requested\n"); +error_enter_powersave: +error_not_now: +no_firmware: + d_fnend(3, dev, "(iface %p pm_msg %u) = %d\n", + iface, pm_msg.event, result); + return result; +} + + +static +int i2400mu_resume(struct usb_interface *iface) +{ + int ret = 0; + struct device *dev = &iface->dev; + struct i2400mu *i2400mu = usb_get_intfdata(iface); + struct i2400m *i2400m = &i2400mu->i2400m; + + d_fnstart(3, dev, "(iface %p)\n", iface); + rmb(); /* see i2400m->updown's documentation */ + if (i2400m->updown == 0) { + d_printf(1, dev, "fw was down, no resume needed\n"); + goto out; + } + d_printf(1, dev, "fw was up, resuming\n"); + i2400mu_notification_setup(i2400mu); + /* USB has flow control, so we don't need to give it time to + * come back; otherwise, we'd use something like a get-state + * command... */ +out: + d_fnend(3, dev, "(iface %p) = %d\n", iface, ret); + return ret; +} + + +static +int i2400mu_reset_resume(struct usb_interface *iface) +{ + int result; + struct device *dev = &iface->dev; + struct i2400mu *i2400mu = usb_get_intfdata(iface); + struct i2400m *i2400m = &i2400mu->i2400m; + + d_fnstart(3, dev, "(iface %p)\n", iface); + result = i2400m_dev_reset_handle(i2400m, "device reset on resume"); + d_fnend(3, dev, "(iface %p) = %d\n", iface, result); + return result < 0 ? result : 0; +} + + +/* + * Another driver or user space is triggering a reset on the device + * which contains the interface passed as an argument. Cease IO and + * save any device state you need to restore. + * + * If you need to allocate memory here, use GFP_NOIO or GFP_ATOMIC, if + * you are in atomic context. + */ +static +int i2400mu_pre_reset(struct usb_interface *iface) +{ + struct i2400mu *i2400mu = usb_get_intfdata(iface); + return i2400m_pre_reset(&i2400mu->i2400m); +} + + +/* + * The reset has completed. Restore any saved device state and begin + * using the device again. + * + * If you need to allocate memory here, use GFP_NOIO or GFP_ATOMIC, if + * you are in atomic context. + */ +static +int i2400mu_post_reset(struct usb_interface *iface) +{ + struct i2400mu *i2400mu = usb_get_intfdata(iface); + return i2400m_post_reset(&i2400mu->i2400m); +} + + +static +struct usb_device_id i2400mu_id_table[] = { + { USB_DEVICE(0x8086, USB_DEVICE_ID_I6050) }, + { USB_DEVICE(0x8086, USB_DEVICE_ID_I6050_2) }, + { USB_DEVICE(0x8087, USB_DEVICE_ID_I6150) }, + { USB_DEVICE(0x8087, USB_DEVICE_ID_I6150_2) }, + { USB_DEVICE(0x8087, USB_DEVICE_ID_I6150_3) }, + { USB_DEVICE(0x8086, USB_DEVICE_ID_I6250) }, + { USB_DEVICE(0x8086, 0x0181) }, + { USB_DEVICE(0x8086, 0x1403) }, + { USB_DEVICE(0x8086, 0x1405) }, + { USB_DEVICE(0x8086, 0x0180) }, + { USB_DEVICE(0x8086, 0x0182) }, + { USB_DEVICE(0x8086, 0x1406) }, + { USB_DEVICE(0x8086, 0x1403) }, + { }, +}; +MODULE_DEVICE_TABLE(usb, i2400mu_id_table); + + +static +struct usb_driver i2400mu_driver = { + .name = KBUILD_MODNAME, + .suspend = i2400mu_suspend, + .resume = i2400mu_resume, + .reset_resume = i2400mu_reset_resume, + .probe = i2400mu_probe, + .disconnect = i2400mu_disconnect, + .pre_reset = i2400mu_pre_reset, + .post_reset = i2400mu_post_reset, + .id_table = i2400mu_id_table, + .supports_autosuspend = 1, +}; + +static +int __init i2400mu_driver_init(void) +{ + d_parse_params(D_LEVEL, D_LEVEL_SIZE, i2400mu_debug_params, + "i2400m_usb.debug"); + return usb_register(&i2400mu_driver); +} +module_init(i2400mu_driver_init); + + +static +void __exit i2400mu_driver_exit(void) +{ + usb_deregister(&i2400mu_driver); +} +module_exit(i2400mu_driver_exit); + +MODULE_AUTHOR("Intel Corporation "); +MODULE_DESCRIPTION("Driver for USB based Intel Wireless WiMAX Connection 2400M " + "(5x50 & 6050)"); +MODULE_LICENSE("GPL"); +MODULE_FIRMWARE(I2400MU_FW_FILE_NAME_v1_5); +MODULE_FIRMWARE(I6050U_FW_FILE_NAME_v1_5); diff --git a/drivers/staging/wimax/id-table.c b/drivers/staging/wimax/id-table.c new file mode 100644 index 000000000000..0e6f4aa87bc9 --- /dev/null +++ b/drivers/staging/wimax/id-table.c @@ -0,0 +1,130 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * Mappping of generic netlink family IDs to net devices + * + * Copyright (C) 2005-2006 Intel Corporation + * Inaky Perez-Gonzalez + * + * We assign a single generic netlink family ID to each device (to + * simplify lookup). + * + * We need a way to map family ID to a wimax_dev pointer. + * + * The idea is to use a very simple lookup. Using a netlink attribute + * with (for example) the interface name implies a heavier search over + * all the network devices; seemed kind of a waste given that we know + * we are looking for a WiMAX device and that most systems will have + * just a single WiMAX adapter. + * + * We put all the WiMAX devices in the system in a linked list and + * match the generic link family ID against the list. + * + * By using a linked list, the case of a single adapter in the system + * becomes (almost) no overhead, while still working for many more. If + * it ever goes beyond two, I'll be surprised. + */ +#include +#include +#include +#include +#include "linux-wimax.h" +#include "wimax-internal.h" + + +#define D_SUBMODULE id_table +#include "debug-levels.h" + + +static DEFINE_SPINLOCK(wimax_id_table_lock); +static struct list_head wimax_id_table = LIST_HEAD_INIT(wimax_id_table); + + +/* + * wimax_id_table_add - add a gennetlink familiy ID / wimax_dev mapping + * + * @wimax_dev: WiMAX device descriptor to associate to the Generic + * Netlink family ID. + * + * Look for an empty spot in the ID table; if none found, double the + * table's size and get the first spot. + */ +void wimax_id_table_add(struct wimax_dev *wimax_dev) +{ + d_fnstart(3, NULL, "(wimax_dev %p)\n", wimax_dev); + spin_lock(&wimax_id_table_lock); + list_add(&wimax_dev->id_table_node, &wimax_id_table); + spin_unlock(&wimax_id_table_lock); + d_fnend(3, NULL, "(wimax_dev %p)\n", wimax_dev); +} + + +/* + * wimax_get_netdev_by_info - lookup a wimax_dev from the gennetlink info + * + * The generic netlink family ID has been filled out in the + * nlmsghdr->nlmsg_type field, so we pull it from there, look it up in + * the mapping table and reference the wimax_dev. + * + * When done, the reference should be dropped with + * 'dev_put(wimax_dev->net_dev)'. + */ +struct wimax_dev *wimax_dev_get_by_genl_info( + struct genl_info *info, int ifindex) +{ + struct wimax_dev *wimax_dev = NULL; + + d_fnstart(3, NULL, "(info %p ifindex %d)\n", info, ifindex); + spin_lock(&wimax_id_table_lock); + list_for_each_entry(wimax_dev, &wimax_id_table, id_table_node) { + if (wimax_dev->net_dev->ifindex == ifindex) { + dev_hold(wimax_dev->net_dev); + goto found; + } + } + wimax_dev = NULL; + d_printf(1, NULL, "wimax: no devices found with ifindex %d\n", + ifindex); +found: + spin_unlock(&wimax_id_table_lock); + d_fnend(3, NULL, "(info %p ifindex %d) = %p\n", + info, ifindex, wimax_dev); + return wimax_dev; +} + + +/* + * wimax_id_table_rm - Remove a gennetlink familiy ID / wimax_dev mapping + * + * @id: family ID to remove from the table + */ +void wimax_id_table_rm(struct wimax_dev *wimax_dev) +{ + spin_lock(&wimax_id_table_lock); + list_del_init(&wimax_dev->id_table_node); + spin_unlock(&wimax_id_table_lock); +} + + +/* + * Release the gennetlink family id / mapping table + * + * On debug, verify that the table is empty upon removal. We want the + * code always compiled, to ensure it doesn't bit rot. It will be + * compiled out if CONFIG_BUG is disabled. + */ +void wimax_id_table_release(void) +{ + struct wimax_dev *wimax_dev; + +#ifndef CONFIG_BUG + return; +#endif + spin_lock(&wimax_id_table_lock); + list_for_each_entry(wimax_dev, &wimax_id_table, id_table_node) { + pr_err("BUG: %s wimax_dev %p ifindex %d not cleared\n", + __func__, wimax_dev, wimax_dev->net_dev->ifindex); + WARN_ON(1); + } + spin_unlock(&wimax_id_table_lock); +} diff --git a/drivers/staging/wimax/linux-wimax-debug.h b/drivers/staging/wimax/linux-wimax-debug.h new file mode 100644 index 000000000000..5b5ec405143b --- /dev/null +++ b/drivers/staging/wimax/linux-wimax-debug.h @@ -0,0 +1,491 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Linux WiMAX + * Collection of tools to manage debug operations. + * + * Copyright (C) 2005-2007 Intel Corporation + * Inaky Perez-Gonzalez + * + * Don't #include this file directly, read on! + * + * EXECUTING DEBUGGING ACTIONS OR NOT + * + * The main thing this framework provides is decission power to take a + * debug action (like printing a message) if the current debug level + * allows it. + * + * The decission power is at two levels: at compile-time (what does + * not make it is compiled out) and at run-time. The run-time + * selection is done per-submodule (as they are declared by the user + * of the framework). + * + * A call to d_test(L) (L being the target debug level) returns true + * if the action should be taken because the current debug levels + * allow it (both compile and run time). + * + * It follows that a call to d_test() that can be determined to be + * always false at compile time will get the code depending on it + * compiled out by optimization. + * + * DEBUG LEVELS + * + * It is up to the caller to define how much a debugging level is. + * + * Convention sets 0 as "no debug" (so an action marked as debug level 0 + * will always be taken). The increasing debug levels are used for + * increased verbosity. + * + * USAGE + * + * Group the code in modules and submodules inside each module [which + * in most cases maps to Linux modules and .c files that compose + * those]. + * + * For each module, there is: + * + * - a MODULENAME (single word, legal C identifier) + * + * - a debug-levels.h header file that declares the list of + * submodules and that is included by all .c files that use + * the debugging tools. The file name can be anything. + * + * - some (optional) .c code to manipulate the runtime debug levels + * through debugfs. + * + * The debug-levels.h file would look like: + * + * #ifndef __debug_levels__h__ + * #define __debug_levels__h__ + * + * #define D_MODULENAME modulename + * #define D_MASTER 10 + * + * #include "linux-wimax-debug.h" + * + * enum d_module { + * D_SUBMODULE_DECLARE(submodule_1), + * D_SUBMODULE_DECLARE(submodule_2), + * ... + * D_SUBMODULE_DECLARE(submodule_N) + * }; + * + * #endif + * + * D_MASTER is the maximum compile-time debug level; any debug actions + * above this will be out. D_MODULENAME is the module name (legal C + * identifier), which has to be unique for each module (to avoid + * namespace collisions during linkage). Note those #defines need to + * be done before #including debug.h + * + * We declare N different submodules whose debug level can be + * independently controlled during runtime. + * + * In a .c file of the module (and only in one of them), define the + * following code: + * + * struct d_level D_LEVEL[] = { + * D_SUBMODULE_DEFINE(submodule_1), + * D_SUBMODULE_DEFINE(submodule_2), + * ... + * D_SUBMODULE_DEFINE(submodule_N), + * }; + * size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); + * + * Externs for d_level_MODULENAME and d_level_size_MODULENAME are used + * and declared in this file using the D_LEVEL and D_LEVEL_SIZE macros + * #defined also in this file. + * + * To manipulate from user space the levels, create a debugfs dentry + * and then register each submodule with: + * + * d_level_register_debugfs("PREFIX_", submodule_X, parent); + * + * Where PREFIX_ is a name of your chosing. This will create debugfs + * file with a single numeric value that can be use to tweak it. To + * remove the entires, just use debugfs_remove_recursive() on 'parent'. + * + * NOTE: remember that even if this will show attached to some + * particular instance of a device, the settings are *global*. + * + * On each submodule (for example, .c files), the debug infrastructure + * should be included like this: + * + * #define D_SUBMODULE submodule_x // matches one in debug-levels.h + * #include "debug-levels.h" + * + * after #including all your include files. + * + * Now you can use the d_*() macros below [d_test(), d_fnstart(), + * d_fnend(), d_printf(), d_dump()]. + * + * If their debug level is greater than D_MASTER, they will be + * compiled out. + * + * If their debug level is lower or equal than D_MASTER but greater + * than the current debug level of their submodule, they'll be + * ignored. + * + * Otherwise, the action will be performed. + */ +#ifndef __debug__h__ +#define __debug__h__ + +#include +#include + +struct device; + +/* Backend stuff */ + +/* + * Debug backend: generate a message header from a 'struct device' + * + * @head: buffer where to place the header + * @head_size: length of @head + * @dev: pointer to device used to generate a header from. If NULL, + * an empty ("") header is generated. + */ +static inline +void __d_head(char *head, size_t head_size, + struct device *dev) +{ + if (dev == NULL) + head[0] = 0; + else if ((unsigned long)dev < 4096) { + printk(KERN_ERR "E: Corrupt dev %p\n", dev); + WARN_ON(1); + } else + snprintf(head, head_size, "%s %s: ", + dev_driver_string(dev), dev_name(dev)); +} + + +/* + * Debug backend: log some message if debugging is enabled + * + * @l: intended debug level + * @tag: tag to prefix the message with + * @dev: 'struct device' associated to this message + * @f: printf-like format and arguments + * + * Note this is optimized out if it doesn't pass the compile-time + * check; however, it is *always* compiled. This is useful to make + * sure the printf-like formats and variables are always checked and + * they don't get bit rot if you have all the debugging disabled. + */ +#define _d_printf(l, tag, dev, f, a...) \ +do { \ + char head[64]; \ + if (!d_test(l)) \ + break; \ + __d_head(head, sizeof(head), dev); \ + printk(KERN_ERR "%s%s%s: " f, head, __func__, tag, ##a); \ +} while (0) + + +/* + * CPP syntactic sugar to generate A_B like symbol names when one of + * the arguments is a preprocessor #define. + */ +#define __D_PASTE__(varname, modulename) varname##_##modulename +#define __D_PASTE(varname, modulename) (__D_PASTE__(varname, modulename)) +#define _D_SUBMODULE_INDEX(_name) (D_SUBMODULE_DECLARE(_name)) + + +/* + * Store a submodule's runtime debug level and name + */ +struct d_level { + u8 level; + const char *name; +}; + + +/* + * List of available submodules and their debug levels + * + * We call them d_level_MODULENAME and d_level_size_MODULENAME; the + * macros D_LEVEL and D_LEVEL_SIZE contain the name already for + * convenience. + * + * This array and the size are defined on some .c file that is part of + * the current module. + */ +#define D_LEVEL __D_PASTE(d_level, D_MODULENAME) +#define D_LEVEL_SIZE __D_PASTE(d_level_size, D_MODULENAME) + +extern struct d_level D_LEVEL[]; +extern size_t D_LEVEL_SIZE; + + +/* + * Frontend stuff + * + * + * Stuff you need to declare prior to using the actual "debug" actions + * (defined below). + */ + +#ifndef D_MODULENAME +#error D_MODULENAME is not defined in your debug-levels.h file +/** + * D_MODULE - Name of the current module + * + * #define in your module's debug-levels.h, making sure it is + * unique. This has to be a legal C identifier. + */ +#define D_MODULENAME undefined_modulename +#endif + + +#ifndef D_MASTER +#warning D_MASTER not defined, but debug.h included! [see docs] +/** + * D_MASTER - Compile time maximum debug level + * + * #define in your debug-levels.h file to the maximum debug level the + * runtime code will be allowed to have. This allows you to provide a + * main knob. + * + * Anything above that level will be optimized out of the compile. + * + * Defaults to zero (no debug code compiled in). + * + * Maximum one definition per module (at the debug-levels.h file). + */ +#define D_MASTER 0 +#endif + +#ifndef D_SUBMODULE +#error D_SUBMODULE not defined, but debug.h included! [see docs] +/** + * D_SUBMODULE - Name of the current submodule + * + * #define in your submodule .c file before #including debug-levels.h + * to the name of the current submodule as previously declared and + * defined with D_SUBMODULE_DECLARE() (in your module's + * debug-levels.h) and D_SUBMODULE_DEFINE(). + * + * This is used to provide runtime-control over the debug levels. + * + * Maximum one per .c file! Can be shared among different .c files + * (meaning they belong to the same submodule categorization). + */ +#define D_SUBMODULE undefined_module +#endif + + +/** + * D_SUBMODULE_DECLARE - Declare a submodule for runtime debug level control + * + * @_name: name of the submodule, restricted to the chars that make up a + * valid C identifier ([a-zA-Z0-9_]). + * + * Declare in the module's debug-levels.h header file as: + * + * enum d_module { + * D_SUBMODULE_DECLARE(submodule_1), + * D_SUBMODULE_DECLARE(submodule_2), + * D_SUBMODULE_DECLARE(submodule_3), + * }; + * + * Some corresponding .c file needs to have a matching + * D_SUBMODULE_DEFINE(). + */ +#define D_SUBMODULE_DECLARE(_name) __D_SUBMODULE_##_name + + +/** + * D_SUBMODULE_DEFINE - Define a submodule for runtime debug level control + * + * @_name: name of the submodule, restricted to the chars that make up a + * valid C identifier ([a-zA-Z0-9_]). + * + * Use once per module (in some .c file) as: + * + * static + * struct d_level d_level_SUBMODULENAME[] = { + * D_SUBMODULE_DEFINE(submodule_1), + * D_SUBMODULE_DEFINE(submodule_2), + * D_SUBMODULE_DEFINE(submodule_3), + * }; + * size_t d_level_size_SUBDMODULENAME = ARRAY_SIZE(d_level_SUBDMODULENAME); + * + * Matching D_SUBMODULE_DECLARE()s have to be present in a + * debug-levels.h header file. + */ +#define D_SUBMODULE_DEFINE(_name) \ +[__D_SUBMODULE_##_name] = { \ + .level = 0, \ + .name = #_name \ +} + + + +/* The actual "debug" operations */ + + +/** + * d_test - Returns true if debugging should be enabled + * + * @l: intended debug level (unsigned) + * + * If the master debug switch is enabled and the current settings are + * higher or equal to the requested level, then debugging + * output/actions should be enabled. + * + * NOTE: + * + * This needs to be coded so that it can be evaluated in compile + * time; this is why the ugly BUG_ON() is placed in there, so the + * D_MASTER evaluation compiles all out if it is compile-time false. + */ +#define d_test(l) \ +({ \ + unsigned __l = l; /* type enforcer */ \ + (D_MASTER) >= __l \ + && ({ \ + BUG_ON(_D_SUBMODULE_INDEX(D_SUBMODULE) >= D_LEVEL_SIZE);\ + D_LEVEL[_D_SUBMODULE_INDEX(D_SUBMODULE)].level >= __l; \ + }); \ +}) + + +/** + * d_fnstart - log message at function start if debugging enabled + * + * @l: intended debug level + * @_dev: 'struct device' pointer, NULL if none (for context) + * @f: printf-like format and arguments + */ +#define d_fnstart(l, _dev, f, a...) _d_printf(l, " FNSTART", _dev, f, ## a) + + +/** + * d_fnend - log message at function end if debugging enabled + * + * @l: intended debug level + * @_dev: 'struct device' pointer, NULL if none (for context) + * @f: printf-like format and arguments + */ +#define d_fnend(l, _dev, f, a...) _d_printf(l, " FNEND", _dev, f, ## a) + + +/** + * d_printf - log message if debugging enabled + * + * @l: intended debug level + * @_dev: 'struct device' pointer, NULL if none (for context) + * @f: printf-like format and arguments + */ +#define d_printf(l, _dev, f, a...) _d_printf(l, "", _dev, f, ## a) + + +/** + * d_dump - log buffer hex dump if debugging enabled + * + * @l: intended debug level + * @_dev: 'struct device' pointer, NULL if none (for context) + * @f: printf-like format and arguments + */ +#define d_dump(l, dev, ptr, size) \ +do { \ + char head[64]; \ + if (!d_test(l)) \ + break; \ + __d_head(head, sizeof(head), dev); \ + print_hex_dump(KERN_ERR, head, 0, 16, 1, \ + ((void *) ptr), (size), 0); \ +} while (0) + + +/** + * Export a submodule's debug level over debugfs as PREFIXSUBMODULE + * + * @prefix: string to prefix the name with + * @submodule: name of submodule (not a string, just the name) + * @dentry: debugfs parent dentry + * + * For removing, just use debugfs_remove_recursive() on the parent. + */ +#define d_level_register_debugfs(prefix, name, parent) \ +({ \ + debugfs_create_u8( \ + prefix #name, 0600, parent, \ + &(D_LEVEL[__D_SUBMODULE_ ## name].level)); \ +}) + + +static inline +void d_submodule_set(struct d_level *d_level, size_t d_level_size, + const char *submodule, u8 level, const char *tag) +{ + struct d_level *itr, *top; + int index = -1; + + for (itr = d_level, top = itr + d_level_size; itr < top; itr++) { + index++; + if (itr->name == NULL) { + printk(KERN_ERR "%s: itr->name NULL?? (%p, #%d)\n", + tag, itr, index); + continue; + } + if (!strcmp(itr->name, submodule)) { + itr->level = level; + return; + } + } + printk(KERN_ERR "%s: unknown submodule %s\n", tag, submodule); +} + + +/** + * d_parse_params - Parse a string with debug parameters from the + * command line + * + * @d_level: level structure (D_LEVEL) + * @d_level_size: number of items in the level structure + * (D_LEVEL_SIZE). + * @_params: string with the parameters; this is a space (not tab!) + * separated list of NAME:VALUE, where value is the debug level + * and NAME is the name of the submodule. + * @tag: string for error messages (example: MODULE.ARGNAME). + */ +static inline +void d_parse_params(struct d_level *d_level, size_t d_level_size, + const char *_params, const char *tag) +{ + char submodule[130], *params, *params_orig, *token, *colon; + unsigned level, tokens; + + if (_params == NULL) + return; + params_orig = kstrdup(_params, GFP_KERNEL); + params = params_orig; + while (1) { + token = strsep(¶ms, " "); + if (token == NULL) + break; + if (*token == '\0') /* eat joint spaces */ + continue; + /* kernel's sscanf %s eats until whitespace, so we + * replace : by \n so it doesn't get eaten later by + * strsep */ + colon = strchr(token, ':'); + if (colon != NULL) + *colon = '\n'; + tokens = sscanf(token, "%s\n%u", submodule, &level); + if (colon != NULL) + *colon = ':'; /* set back, for error messages */ + if (tokens == 2) + d_submodule_set(d_level, d_level_size, + submodule, level, tag); + else + printk(KERN_ERR "%s: can't parse '%s' as a " + "SUBMODULE:LEVEL (%d tokens)\n", + tag, token, tokens); + } + kfree(params_orig); +} + +#endif /* #ifndef __debug__h__ */ diff --git a/drivers/staging/wimax/linux-wimax.h b/drivers/staging/wimax/linux-wimax.h new file mode 100644 index 000000000000..9f6b77af2f6d --- /dev/null +++ b/drivers/staging/wimax/linux-wimax.h @@ -0,0 +1,239 @@ +/* + * Linux WiMax + * API for user space + * + * + * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * + * Intel Corporation + * Inaky Perez-Gonzalez + * - Initial implementation + * + * + * This file declares the user/kernel protocol that is spoken over + * Generic Netlink, as well as any type declaration that is to be used + * by kernel and user space. + * + * It is intended for user space to clone it verbatim to use it as a + * primary reference for definitions. + * + * Stuff intended for kernel usage as well as full protocol and stack + * documentation is rooted in include/net/wimax.h. + */ + +#ifndef __LINUX__WIMAX_H__ +#define __LINUX__WIMAX_H__ + +#include + +enum { + /** + * Version of the interface (unsigned decimal, MMm, max 25.5) + * M - Major: change if removing or modifying an existing call. + * m - minor: change when adding a new call + */ + WIMAX_GNL_VERSION = 01, + /* Generic NetLink attributes */ + WIMAX_GNL_ATTR_INVALID = 0x00, + WIMAX_GNL_ATTR_MAX = 10, +}; + + +/* + * Generic NetLink operations + * + * Most of these map to an API call; _OP_ stands for operation, _RP_ + * for reply and _RE_ for report (aka: signal). + */ +enum { + WIMAX_GNL_OP_MSG_FROM_USER, /* User to kernel message */ + WIMAX_GNL_OP_MSG_TO_USER, /* Kernel to user message */ + WIMAX_GNL_OP_RFKILL, /* Run wimax_rfkill() */ + WIMAX_GNL_OP_RESET, /* Run wimax_rfkill() */ + WIMAX_GNL_RE_STATE_CHANGE, /* Report: status change */ + WIMAX_GNL_OP_STATE_GET, /* Request for current state */ +}; + + +/* Message from user / to user */ +enum { + WIMAX_GNL_MSG_IFIDX = 1, + WIMAX_GNL_MSG_PIPE_NAME, + WIMAX_GNL_MSG_DATA, +}; + + +/* + * wimax_rfkill() + * + * The state of the radio (ON/OFF) is mapped to the rfkill subsystem's + * switch state (DISABLED/ENABLED). + */ +enum wimax_rf_state { + WIMAX_RF_OFF = 0, /* Radio is off, rfkill on/enabled */ + WIMAX_RF_ON = 1, /* Radio is on, rfkill off/disabled */ + WIMAX_RF_QUERY = 2, +}; + +/* Attributes */ +enum { + WIMAX_GNL_RFKILL_IFIDX = 1, + WIMAX_GNL_RFKILL_STATE, +}; + + +/* Attributes for wimax_reset() */ +enum { + WIMAX_GNL_RESET_IFIDX = 1, +}; + +/* Attributes for wimax_state_get() */ +enum { + WIMAX_GNL_STGET_IFIDX = 1, +}; + +/* + * Attributes for the Report State Change + * + * For now we just have the old and new states; new attributes might + * be added later on. + */ +enum { + WIMAX_GNL_STCH_IFIDX = 1, + WIMAX_GNL_STCH_STATE_OLD, + WIMAX_GNL_STCH_STATE_NEW, +}; + + +/** + * enum wimax_st - The different states of a WiMAX device + * @__WIMAX_ST_NULL: The device structure has been allocated and zeroed, + * but still wimax_dev_add() hasn't been called. There is no state. + * + * @WIMAX_ST_DOWN: The device has been registered with the WiMAX and + * networking stacks, but it is not initialized (normally that is + * done with 'ifconfig DEV up' [or equivalent], which can upload + * firmware and enable communications with the device). + * In this state, the device is powered down and using as less + * power as possible. + * This state is the default after a call to wimax_dev_add(). It + * is ok to have drivers move directly to %WIMAX_ST_UNINITIALIZED + * or %WIMAX_ST_RADIO_OFF in _probe() after the call to + * wimax_dev_add(). + * It is recommended that the driver leaves this state when + * calling 'ifconfig DEV up' and enters it back on 'ifconfig DEV + * down'. + * + * @__WIMAX_ST_QUIESCING: The device is being torn down, so no API + * operations are allowed to proceed except the ones needed to + * complete the device clean up process. + * + * @WIMAX_ST_UNINITIALIZED: [optional] Communication with the device + * is setup, but the device still requires some configuration + * before being operational. + * Some WiMAX API calls might work. + * + * @WIMAX_ST_RADIO_OFF: The device is fully up; radio is off (wether + * by hardware or software switches). + * It is recommended to always leave the device in this state + * after initialization. + * + * @WIMAX_ST_READY: The device is fully up and radio is on. + * + * @WIMAX_ST_SCANNING: [optional] The device has been instructed to + * scan. In this state, the device cannot be actively connected to + * a network. + * + * @WIMAX_ST_CONNECTING: The device is connecting to a network. This + * state exists because in some devices, the connect process can + * include a number of negotiations between user space, kernel + * space and the device. User space needs to know what the device + * is doing. If the connect sequence in a device is atomic and + * fast, the device can transition directly to CONNECTED + * + * @WIMAX_ST_CONNECTED: The device is connected to a network. + * + * @__WIMAX_ST_INVALID: This is an invalid state used to mark the + * maximum numeric value of states. + * + * Description: + * + * Transitions from one state to another one are atomic and can only + * be caused in kernel space with wimax_state_change(). To read the + * state, use wimax_state_get(). + * + * States starting with __ are internal and shall not be used or + * referred to by drivers or userspace. They look ugly, but that's the + * point -- if any use is made non-internal to the stack, it is easier + * to catch on review. + * + * All API operations [with well defined exceptions] will take the + * device mutex before starting and then check the state. If the state + * is %__WIMAX_ST_NULL, %WIMAX_ST_DOWN, %WIMAX_ST_UNINITIALIZED or + * %__WIMAX_ST_QUIESCING, it will drop the lock and quit with + * -%EINVAL, -%ENOMEDIUM, -%ENOTCONN or -%ESHUTDOWN. + * + * The order of the definitions is important, so we can do numerical + * comparisons (eg: < %WIMAX_ST_RADIO_OFF means the device is not ready + * to operate). + */ +/* + * The allowed state transitions are described in the table below + * (states in rows can go to states in columns where there is an X): + * + * UNINI RADIO READY SCAN CONNEC CONNEC + * NULL DOWN QUIESCING TIALIZED OFF NING TING TED + * NULL - x + * DOWN - x x x + * QUIESCING x - + * UNINITIALIZED x - x + * RADIO_OFF x - x + * READY x x - x x x + * SCANNING x x x - x x + * CONNECTING x x x x - x + * CONNECTED x x x - + * + * This table not available in kernel-doc because the formatting messes it up. + */ + enum wimax_st { + __WIMAX_ST_NULL = 0, + WIMAX_ST_DOWN, + __WIMAX_ST_QUIESCING, + WIMAX_ST_UNINITIALIZED, + WIMAX_ST_RADIO_OFF, + WIMAX_ST_READY, + WIMAX_ST_SCANNING, + WIMAX_ST_CONNECTING, + WIMAX_ST_CONNECTED, + __WIMAX_ST_INVALID /* Always keep last */ +}; + + +#endif /* #ifndef __LINUX__WIMAX_H__ */ diff --git a/drivers/staging/wimax/net-wimax.h b/drivers/staging/wimax/net-wimax.h new file mode 100644 index 000000000000..f578e345e2bd --- /dev/null +++ b/drivers/staging/wimax/net-wimax.h @@ -0,0 +1,503 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Linux WiMAX + * Kernel space API for accessing WiMAX devices + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + * + * The WiMAX stack provides an API for controlling and managing the + * system's WiMAX devices. This API affects the control plane; the + * data plane is accessed via the network stack (netdev). + * + * Parts of the WiMAX stack API and notifications are exported to + * user space via Generic Netlink. In user space, libwimax (part of + * the wimax-tools package) provides a shim layer for accessing those + * calls. + * + * The API is standarized for all WiMAX devices and different drivers + * implement the backend support for it. However, device-specific + * messaging pipes are provided that can be used to issue commands and + * receive notifications in free form. + * + * Currently the messaging pipes are the only means of control as it + * is not known (due to the lack of more devices in the market) what + * will be a good abstraction layer. Expect this to change as more + * devices show in the market. This API is designed to be growable in + * order to address this problem. + * + * USAGE + * + * Embed a `struct wimax_dev` at the beginning of the device's + * private structure, initialize and register it. For details, see + * `struct wimax_dev`s documentation. + * + * Once this is done, wimax-tools's libwimaxll can be used to + * communicate with the driver from user space. You user space + * application does not have to forcibily use libwimaxll and can talk + * the generic netlink protocol directly if desired. + * + * Remember this is a very low level API that will to provide all of + * WiMAX features. Other daemons and services running in user space + * are the expected clients of it. They offer a higher level API that + * applications should use (an example of this is the Intel's WiMAX + * Network Service for the i2400m). + * + * DESIGN + * + * Although not set on final stone, this very basic interface is + * mostly completed. Remember this is meant to grow as new common + * operations are decided upon. New operations will be added to the + * interface, intent being on keeping backwards compatibility as much + * as possible. + * + * This layer implements a set of calls to control a WiMAX device, + * exposing a frontend to the rest of the kernel and user space (via + * generic netlink) and a backend implementation in the driver through + * function pointers. + * + * WiMAX devices have a state, and a kernel-only API allows the + * drivers to manipulate that state. State transitions are atomic, and + * only some of them are allowed (see `enum wimax_st`). + * + * Most API calls will set the state automatically; in most cases + * drivers have to only report state changes due to external + * conditions. + * + * All API operations are 'atomic', serialized through a mutex in the + * `struct wimax_dev`. + * + * EXPORTING TO USER SPACE THROUGH GENERIC NETLINK + * + * The API is exported to user space using generic netlink (other + * methods can be added as needed). + * + * There is a Generic Netlink Family named "WiMAX", where interfaces + * supporting the WiMAX interface receive commands and broadcast their + * signals over a multicast group named "msg". + * + * Mapping to the source/destination interface is done by an interface + * index attribute. + * + * For user-to-kernel traffic (commands) we use a function call + * marshalling mechanism, where a message X with attributes A, B, C + * sent from user space to kernel space means executing the WiMAX API + * call wimax_X(A, B, C), sending the results back as a message. + * + * Kernel-to-user (notifications or signals) communication is sent + * over multicast groups. This allows to have multiple applications + * monitoring them. + * + * Each command/signal gets assigned it's own attribute policy. This + * way the validator will verify that all the attributes in there are + * only the ones that should be for each command/signal. Thing of an + * attribute mapping to a type+argumentname for each command/signal. + * + * If we had a single policy for *all* commands/signals, after running + * the validator we'd have to check "does this attribute belong in + * here"? for each one. It can be done manually, but it's just easier + * to have the validator do that job with multiple policies. As well, + * it makes it easier to later expand each command/signal signature + * without affecting others and keeping the namespace more or less + * sane. Not that it is too complicated, but it makes it even easier. + * + * No state information is maintained in the kernel for each user + * space connection (the connection is stateless). + * + * TESTING FOR THE INTERFACE AND VERSIONING + * + * If network interface X is a WiMAX device, there will be a Generic + * Netlink family named "WiMAX X" and the device will present a + * "wimax" directory in it's network sysfs directory + * (/sys/class/net/DEVICE/wimax) [used by HAL]. + * + * The inexistence of any of these means the device does not support + * this WiMAX API. + * + * By querying the generic netlink controller, versioning information + * and the multicast groups available can be found. Applications using + * the interface can either rely on that or use the generic netlink + * controller to figure out which generic netlink commands/signals are + * supported. + * + * NOTE: this versioning is a last resort to avoid hard + * incompatibilities. It is the intention of the design of this + * stack not to introduce backward incompatible changes. + * + * The version code has to fit in one byte (restrictions imposed by + * generic netlink); we use `version / 10` for the major version and + * `version % 10` for the minor. This gives 9 minors for each major + * and 25 majors. + * + * The version change protocol is as follow: + * + * - Major versions: needs to be increased if an existing message/API + * call is changed or removed. Doesn't need to be changed if a new + * message is added. + * + * - Minor version: needs to be increased if new messages/API calls are + * being added or some other consideration that doesn't impact the + * user-kernel interface too much (like some kind of bug fix) and + * that is kind of left up in the air to common sense. + * + * User space code should not try to work if the major version it was + * compiled for differs from what the kernel offers. As well, if the + * minor version of the kernel interface is lower than the one user + * space is expecting (the one it was compiled for), the kernel + * might be missing API calls; user space shall be ready to handle + * said condition. Use the generic netlink controller operations to + * find which ones are supported and which not. + * + * libwimaxll:wimaxll_open() takes care of checking versions. + * + * THE OPERATIONS: + * + * Each operation is defined in its on file (drivers/net/wimax/op-*.c) + * for clarity. The parts needed for an operation are: + * + * - a function pointer in `struct wimax_dev`: optional, as the + * operation might be implemented by the stack and not by the + * driver. + * + * All function pointers are named wimax_dev->op_*(), and drivers + * must implement them except where noted otherwise. + * + * - When exported to user space, a `struct nla_policy` to define the + * attributes of the generic netlink command and a `struct genl_ops` + * to define the operation. + * + * All the declarations for the operation codes (WIMAX_GNL_OP_) + * and generic netlink attributes (WIMAX_GNL__*) are declared in + * include/linux/wimax.h; this file is intended to be cloned by user + * space to gain access to those declarations. + * + * A few caveats to remember: + * + * - Need to define attribute numbers starting in 1; otherwise it + * fails. + * + * - the `struct genl_family` requires a maximum attribute id; when + * defining the `struct nla_policy` for each message, it has to have + * an array size of WIMAX_GNL_ATTR_MAX+1. + * + * The op_*() function pointers will not be called if the wimax_dev is + * in a state <= %WIMAX_ST_UNINITIALIZED. The exception is: + * + * - op_reset: can be called at any time after wimax_dev_add() has + * been called. + * + * THE PIPE INTERFACE: + * + * This interface is kept intentionally simple. The driver can send + * and receive free-form messages to/from user space through a + * pipe. See drivers/net/wimax/op-msg.c for details. + * + * The kernel-to-user messages are sent with + * wimax_msg(). user-to-kernel messages are delivered via + * wimax_dev->op_msg_from_user(). + * + * RFKILL: + * + * RFKILL support is built into the wimax_dev layer; the driver just + * needs to call wimax_report_rfkill_{hw,sw}() to inform of changes in + * the hardware or software RF kill switches. When the stack wants to + * turn the radio off, it will call wimax_dev->op_rfkill_sw_toggle(), + * which the driver implements. + * + * User space can set the software RF Kill switch by calling + * wimax_rfkill(). + * + * The code for now only supports devices that don't require polling; + * If the device needs to be polled, create a self-rearming delayed + * work struct for polling or look into adding polled support to the + * WiMAX stack. + * + * When initializing the hardware (_probe), after calling + * wimax_dev_add(), query the device for it's RF Kill switches status + * and feed it back to the WiMAX stack using + * wimax_report_rfkill_{hw,sw}(). If any switch is missing, always + * report it as ON. + * + * NOTE: the wimax stack uses an inverted terminology to that of the + * RFKILL subsystem: + * + * - ON: radio is ON, RFKILL is DISABLED or OFF. + * - OFF: radio is OFF, RFKILL is ENABLED or ON. + * + * MISCELLANEOUS OPS: + * + * wimax_reset() can be used to reset the device to power on state; by + * default it issues a warm reset that maintains the same device + * node. If that is not possible, it falls back to a cold reset + * (device reconnect). The driver implements the backend to this + * through wimax_dev->op_reset(). + */ + +#ifndef __NET__WIMAX_H__ +#define __NET__WIMAX_H__ + +#include "linux-wimax.h" +#include +#include + +struct net_device; +struct genl_info; +struct wimax_dev; + +/** + * struct wimax_dev - Generic WiMAX device + * + * @net_dev: [fill] Pointer to the &struct net_device this WiMAX + * device implements. + * + * @op_msg_from_user: [fill] Driver-specific operation to + * handle a raw message from user space to the driver. The + * driver can send messages to user space using with + * wimax_msg_to_user(). + * + * @op_rfkill_sw_toggle: [fill] Driver-specific operation to act on + * userspace (or any other agent) requesting the WiMAX device to + * change the RF Kill software switch (WIMAX_RF_ON or + * WIMAX_RF_OFF). + * If such hardware support is not present, it is assumed the + * radio cannot be switched off and it is always on (and the stack + * will error out when trying to switch it off). In such case, + * this function pointer can be left as NULL. + * + * @op_reset: [fill] Driver specific operation to reset the + * device. + * This operation should always attempt first a warm reset that + * does not disconnect the device from the bus and return 0. + * If that fails, it should resort to some sort of cold or bus + * reset (even if it implies a bus disconnection and device + * disappearance). In that case, -ENODEV should be returned to + * indicate the device is gone. + * This operation has to be synchronous, and return only when the + * reset is complete. In case of having had to resort to bus/cold + * reset implying a device disconnection, the call is allowed to + * return immediately. + * NOTE: wimax_dev->mutex is NOT locked when this op is being + * called; however, wimax_dev->mutex_reset IS locked to ensure + * serialization of calls to wimax_reset(). + * See wimax_reset()'s documentation. + * + * @name: [fill] A way to identify this device. We need to register a + * name with many subsystems (rfkill, workqueue creation, etc). + * We can't use the network device name as that + * might change and in some instances we don't know it yet (until + * we don't call register_netdev()). So we generate an unique one + * using the driver name and device bus id, place it here and use + * it across the board. Recommended naming: + * DRIVERNAME-BUSNAME:BUSID (dev->bus->name, dev->bus_id). + * + * @id_table_node: [private] link to the list of wimax devices kept by + * id-table.c. Protected by it's own spinlock. + * + * @mutex: [private] Serializes all concurrent access and execution of + * operations. + * + * @mutex_reset: [private] Serializes reset operations. Needs to be a + * different mutex because as part of the reset operation, the + * driver has to call back into the stack to do things such as + * state change, that require wimax_dev->mutex. + * + * @state: [private] Current state of the WiMAX device. + * + * @rfkill: [private] integration into the RF-Kill infrastructure. + * + * @rf_sw: [private] State of the software radio switch (OFF/ON) + * + * @rf_hw: [private] State of the hardware radio switch (OFF/ON) + * + * @debugfs_dentry: [private] Used to hook up a debugfs entry. This + * shows up in the debugfs root as wimax\:DEVICENAME. + * + * Description: + * This structure defines a common interface to access all WiMAX + * devices from different vendors and provides a common API as well as + * a free-form device-specific messaging channel. + * + * Usage: + * 1. Embed a &struct wimax_dev at *the beginning* the network + * device structure so that netdev_priv() points to it. + * + * 2. memset() it to zero + * + * 3. Initialize with wimax_dev_init(). This will leave the WiMAX + * device in the %__WIMAX_ST_NULL state. + * + * 4. Fill all the fields marked with [fill]; once called + * wimax_dev_add(), those fields CANNOT be modified. + * + * 5. Call wimax_dev_add() *after* registering the network + * device. This will leave the WiMAX device in the %WIMAX_ST_DOWN + * state. + * Protect the driver's net_device->open() against succeeding if + * the wimax device state is lower than %WIMAX_ST_DOWN. + * + * 6. Select when the device is going to be turned on/initialized; + * for example, it could be initialized on 'ifconfig up' (when the + * netdev op 'open()' is called on the driver). + * + * When the device is initialized (at `ifconfig up` time, or right + * after calling wimax_dev_add() from _probe(), make sure the + * following steps are taken + * + * a. Move the device to %WIMAX_ST_UNINITIALIZED. This is needed so + * some API calls that shouldn't work until the device is ready + * can be blocked. + * + * b. Initialize the device. Make sure to turn the SW radio switch + * off and move the device to state %WIMAX_ST_RADIO_OFF when + * done. When just initialized, a device should be left in RADIO + * OFF state until user space devices to turn it on. + * + * c. Query the device for the state of the hardware rfkill switch + * and call wimax_rfkill_report_hw() and wimax_rfkill_report_sw() + * as needed. See below. + * + * wimax_dev_rm() undoes before unregistering the network device. Once + * wimax_dev_add() is called, the driver can get called on the + * wimax_dev->op_* function pointers + * + * CONCURRENCY: + * + * The stack provides a mutex for each device that will disallow API + * calls happening concurrently; thus, op calls into the driver + * through the wimax_dev->op*() function pointers will always be + * serialized and *never* concurrent. + * + * For locking, take wimax_dev->mutex is taken; (most) operations in + * the API have to check for wimax_dev_is_ready() to return 0 before + * continuing (this is done internally). + * + * REFERENCE COUNTING: + * + * The WiMAX device is reference counted by the associated network + * device. The only operation that can be used to reference the device + * is wimax_dev_get_by_genl_info(), and the reference it acquires has + * to be released with dev_put(wimax_dev->net_dev). + * + * RFKILL: + * + * At startup, both HW and SW radio switchess are assumed to be off. + * + * At initialization time [after calling wimax_dev_add()], have the + * driver query the device for the status of the software and hardware + * RF kill switches and call wimax_report_rfkill_hw() and + * wimax_rfkill_report_sw() to indicate their state. If any is + * missing, just call it to indicate it is ON (radio always on). + * + * Whenever the driver detects a change in the state of the RF kill + * switches, it should call wimax_report_rfkill_hw() or + * wimax_report_rfkill_sw() to report it to the stack. + */ +struct wimax_dev { + struct net_device *net_dev; + struct list_head id_table_node; + struct mutex mutex; /* Protects all members and API calls */ + struct mutex mutex_reset; + enum wimax_st state; + + int (*op_msg_from_user)(struct wimax_dev *wimax_dev, + const char *, + const void *, size_t, + const struct genl_info *info); + int (*op_rfkill_sw_toggle)(struct wimax_dev *wimax_dev, + enum wimax_rf_state); + int (*op_reset)(struct wimax_dev *wimax_dev); + + struct rfkill *rfkill; + unsigned int rf_hw; + unsigned int rf_sw; + char name[32]; + + struct dentry *debugfs_dentry; +}; + + + +/* + * WiMAX stack public API for device drivers + * ----------------------------------------- + * + * These functions are not exported to user space. + */ +void wimax_dev_init(struct wimax_dev *); +int wimax_dev_add(struct wimax_dev *, struct net_device *); +void wimax_dev_rm(struct wimax_dev *); + +static inline +struct wimax_dev *net_dev_to_wimax(struct net_device *net_dev) +{ + return netdev_priv(net_dev); +} + +static inline +struct device *wimax_dev_to_dev(struct wimax_dev *wimax_dev) +{ + return wimax_dev->net_dev->dev.parent; +} + +void wimax_state_change(struct wimax_dev *, enum wimax_st); +enum wimax_st wimax_state_get(struct wimax_dev *); + +/* + * Radio Switch state reporting. + * + * enum wimax_rf_state is declared in linux/wimax.h so the exports + * to user space can use it. + */ +void wimax_report_rfkill_hw(struct wimax_dev *, enum wimax_rf_state); +void wimax_report_rfkill_sw(struct wimax_dev *, enum wimax_rf_state); + + +/* + * Free-form messaging to/from user space + * + * Sending a message: + * + * wimax_msg(wimax_dev, pipe_name, buf, buf_size, GFP_KERNEL); + * + * Broken up: + * + * skb = wimax_msg_alloc(wimax_dev, pipe_name, buf_size, GFP_KERNEL); + * ...fill up skb... + * wimax_msg_send(wimax_dev, pipe_name, skb); + * + * Be sure not to modify skb->data in the middle (ie: don't use + * skb_push()/skb_pull()/skb_reserve() on the skb). + * + * "pipe_name" is any string, that can be interpreted as the name of + * the pipe or recipient; the interpretation of it is driver + * specific, so the recipient can multiplex it as wished. It can be + * NULL, it won't be used - an example is using a "diagnostics" tag to + * send diagnostics information that a device-specific diagnostics + * tool would be interested in. + */ +struct sk_buff *wimax_msg_alloc(struct wimax_dev *, const char *, const void *, + size_t, gfp_t); +int wimax_msg_send(struct wimax_dev *, struct sk_buff *); +int wimax_msg(struct wimax_dev *, const char *, const void *, size_t, gfp_t); + +const void *wimax_msg_data_len(struct sk_buff *, size_t *); +const void *wimax_msg_data(struct sk_buff *); +ssize_t wimax_msg_len(struct sk_buff *); + + +/* + * WiMAX stack user space API + * -------------------------- + * + * This API is what gets exported to user space for general + * operations. As well, they can be called from within the kernel, + * (with a properly referenced `struct wimax_dev`). + * + * Properly referenced means: the 'struct net_device' that embeds the + * device's control structure and (as such) the 'struct wimax_dev' is + * referenced by the caller. + */ +int wimax_rfkill(struct wimax_dev *, enum wimax_rf_state); +int wimax_reset(struct wimax_dev *); + +#endif /* #ifndef __NET__WIMAX_H__ */ diff --git a/drivers/staging/wimax/op-msg.c b/drivers/staging/wimax/op-msg.c new file mode 100644 index 000000000000..e20ac7d84e82 --- /dev/null +++ b/drivers/staging/wimax/op-msg.c @@ -0,0 +1,391 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * Generic messaging interface between userspace and driver/device + * + * Copyright (C) 2007-2008 Intel Corporation + * Inaky Perez-Gonzalez + * + * This implements a direct communication channel between user space and + * the driver/device, by which free form messages can be sent back and + * forth. + * + * This is intended for device-specific features, vendor quirks, etc. + * + * See include/net/wimax.h + * + * GENERIC NETLINK ENCODING AND CAPACITY + * + * A destination "pipe name" is added to each message; it is up to the + * drivers to assign or use those names (if using them at all). + * + * Messages are encoded as a binary netlink attribute using nla_put() + * using type NLA_UNSPEC (as some versions of libnl still in + * deployment don't yet understand NLA_BINARY). + * + * The maximum capacity of this transport is PAGESIZE per message (so + * the actual payload will be bit smaller depending on the + * netlink/generic netlink attributes and headers). + * + * RECEPTION OF MESSAGES + * + * When a message is received from user space, it is passed verbatim + * to the driver calling wimax_dev->op_msg_from_user(). The return + * value from this function is passed back to user space as an ack + * over the generic netlink protocol. + * + * The stack doesn't do any processing or interpretation of these + * messages. + * + * SENDING MESSAGES + * + * Messages can be sent with wimax_msg(). + * + * If the message delivery needs to happen on a different context to + * that of its creation, wimax_msg_alloc() can be used to get a + * pointer to the message that can be delivered later on with + * wimax_msg_send(). + * + * ROADMAP + * + * wimax_gnl_doit_msg_from_user() Process a message from user space + * wimax_dev_get_by_genl_info() + * wimax_dev->op_msg_from_user() Delivery of message to the driver + * + * wimax_msg() Send a message to user space + * wimax_msg_alloc() + * wimax_msg_send() + */ +#include +#include +#include +#include +#include "linux-wimax.h" +#include +#include +#include "wimax-internal.h" + + +#define D_SUBMODULE op_msg +#include "debug-levels.h" + + +/** + * wimax_msg_alloc - Create a new skb for sending a message to userspace + * + * @wimax_dev: WiMAX device descriptor + * @pipe_name: "named pipe" the message will be sent to + * @msg: pointer to the message data to send + * @size: size of the message to send (in bytes), including the header. + * @gfp_flags: flags for memory allocation. + * + * Returns: %0 if ok, negative errno code on error + * + * Description: + * + * Allocates an skb that will contain the message to send to user + * space over the messaging pipe and initializes it, copying the + * payload. + * + * Once this call is done, you can deliver it with + * wimax_msg_send(). + * + * IMPORTANT: + * + * Don't use skb_push()/skb_pull()/skb_reserve() on the skb, as + * wimax_msg_send() depends on skb->data being placed at the + * beginning of the user message. + * + * Unlike other WiMAX stack calls, this call can be used way early, + * even before wimax_dev_add() is called, as long as the + * wimax_dev->net_dev pointer is set to point to a proper + * net_dev. This is so that drivers can use it early in case they need + * to send stuff around or communicate with user space. + */ +struct sk_buff *wimax_msg_alloc(struct wimax_dev *wimax_dev, + const char *pipe_name, + const void *msg, size_t size, + gfp_t gfp_flags) +{ + int result; + struct device *dev = wimax_dev_to_dev(wimax_dev); + size_t msg_size; + void *genl_msg; + struct sk_buff *skb; + + msg_size = nla_total_size(size) + + nla_total_size(sizeof(u32)) + + (pipe_name ? nla_total_size(strlen(pipe_name)) : 0); + result = -ENOMEM; + skb = genlmsg_new(msg_size, gfp_flags); + if (skb == NULL) + goto error_new; + genl_msg = genlmsg_put(skb, 0, 0, &wimax_gnl_family, + 0, WIMAX_GNL_OP_MSG_TO_USER); + if (genl_msg == NULL) { + dev_err(dev, "no memory to create generic netlink message\n"); + goto error_genlmsg_put; + } + result = nla_put_u32(skb, WIMAX_GNL_MSG_IFIDX, + wimax_dev->net_dev->ifindex); + if (result < 0) { + dev_err(dev, "no memory to add ifindex attribute\n"); + goto error_nla_put; + } + if (pipe_name) { + result = nla_put_string(skb, WIMAX_GNL_MSG_PIPE_NAME, + pipe_name); + if (result < 0) { + dev_err(dev, "no memory to add pipe_name attribute\n"); + goto error_nla_put; + } + } + result = nla_put(skb, WIMAX_GNL_MSG_DATA, size, msg); + if (result < 0) { + dev_err(dev, "no memory to add payload (msg %p size %zu) in " + "attribute: %d\n", msg, size, result); + goto error_nla_put; + } + genlmsg_end(skb, genl_msg); + return skb; + +error_nla_put: +error_genlmsg_put: +error_new: + nlmsg_free(skb); + return ERR_PTR(result); +} +EXPORT_SYMBOL_GPL(wimax_msg_alloc); + + +/** + * wimax_msg_data_len - Return a pointer and size of a message's payload + * + * @msg: Pointer to a message created with wimax_msg_alloc() + * @size: Pointer to where to store the message's size + * + * Returns the pointer to the message data. + */ +const void *wimax_msg_data_len(struct sk_buff *msg, size_t *size) +{ + struct nlmsghdr *nlh = (void *) msg->head; + struct nlattr *nla; + + nla = nlmsg_find_attr(nlh, sizeof(struct genlmsghdr), + WIMAX_GNL_MSG_DATA); + if (nla == NULL) { + pr_err("Cannot find attribute WIMAX_GNL_MSG_DATA\n"); + return NULL; + } + *size = nla_len(nla); + return nla_data(nla); +} +EXPORT_SYMBOL_GPL(wimax_msg_data_len); + + +/** + * wimax_msg_data - Return a pointer to a message's payload + * + * @msg: Pointer to a message created with wimax_msg_alloc() + */ +const void *wimax_msg_data(struct sk_buff *msg) +{ + struct nlmsghdr *nlh = (void *) msg->head; + struct nlattr *nla; + + nla = nlmsg_find_attr(nlh, sizeof(struct genlmsghdr), + WIMAX_GNL_MSG_DATA); + if (nla == NULL) { + pr_err("Cannot find attribute WIMAX_GNL_MSG_DATA\n"); + return NULL; + } + return nla_data(nla); +} +EXPORT_SYMBOL_GPL(wimax_msg_data); + + +/** + * wimax_msg_len - Return a message's payload length + * + * @msg: Pointer to a message created with wimax_msg_alloc() + */ +ssize_t wimax_msg_len(struct sk_buff *msg) +{ + struct nlmsghdr *nlh = (void *) msg->head; + struct nlattr *nla; + + nla = nlmsg_find_attr(nlh, sizeof(struct genlmsghdr), + WIMAX_GNL_MSG_DATA); + if (nla == NULL) { + pr_err("Cannot find attribute WIMAX_GNL_MSG_DATA\n"); + return -EINVAL; + } + return nla_len(nla); +} +EXPORT_SYMBOL_GPL(wimax_msg_len); + + +/** + * wimax_msg_send - Send a pre-allocated message to user space + * + * @wimax_dev: WiMAX device descriptor + * + * @skb: &struct sk_buff returned by wimax_msg_alloc(). Note the + * ownership of @skb is transferred to this function. + * + * Returns: 0 if ok, < 0 errno code on error + * + * Description: + * + * Sends a free-form message that was preallocated with + * wimax_msg_alloc() and filled up. + * + * Assumes that once you pass an skb to this function for sending, it + * owns it and will release it when done (on success). + * + * IMPORTANT: + * + * Don't use skb_push()/skb_pull()/skb_reserve() on the skb, as + * wimax_msg_send() depends on skb->data being placed at the + * beginning of the user message. + * + * Unlike other WiMAX stack calls, this call can be used way early, + * even before wimax_dev_add() is called, as long as the + * wimax_dev->net_dev pointer is set to point to a proper + * net_dev. This is so that drivers can use it early in case they need + * to send stuff around or communicate with user space. + */ +int wimax_msg_send(struct wimax_dev *wimax_dev, struct sk_buff *skb) +{ + struct device *dev = wimax_dev_to_dev(wimax_dev); + void *msg = skb->data; + size_t size = skb->len; + might_sleep(); + + d_printf(1, dev, "CTX: wimax msg, %zu bytes\n", size); + d_dump(2, dev, msg, size); + genlmsg_multicast(&wimax_gnl_family, skb, 0, 0, GFP_KERNEL); + d_printf(1, dev, "CTX: genl multicast done\n"); + return 0; +} +EXPORT_SYMBOL_GPL(wimax_msg_send); + + +/** + * wimax_msg - Send a message to user space + * + * @wimax_dev: WiMAX device descriptor (properly referenced) + * @pipe_name: "named pipe" the message will be sent to + * @buf: pointer to the message to send. + * @size: size of the buffer pointed to by @buf (in bytes). + * @gfp_flags: flags for memory allocation. + * + * Returns: %0 if ok, negative errno code on error. + * + * Description: + * + * Sends a free-form message to user space on the device @wimax_dev. + * + * NOTES: + * + * Once the @skb is given to this function, who will own it and will + * release it when done (unless it returns error). + */ +int wimax_msg(struct wimax_dev *wimax_dev, const char *pipe_name, + const void *buf, size_t size, gfp_t gfp_flags) +{ + int result = -ENOMEM; + struct sk_buff *skb; + + skb = wimax_msg_alloc(wimax_dev, pipe_name, buf, size, gfp_flags); + if (IS_ERR(skb)) + result = PTR_ERR(skb); + else + result = wimax_msg_send(wimax_dev, skb); + return result; +} +EXPORT_SYMBOL_GPL(wimax_msg); + +/* + * Relays a message from user space to the driver + * + * The skb is passed to the driver-specific function with the netlink + * and generic netlink headers already stripped. + * + * This call will block while handling/relaying the message. + */ +int wimax_gnl_doit_msg_from_user(struct sk_buff *skb, struct genl_info *info) +{ + int result, ifindex; + struct wimax_dev *wimax_dev; + struct device *dev; + struct nlmsghdr *nlh = info->nlhdr; + char *pipe_name; + void *msg_buf; + size_t msg_len; + + might_sleep(); + d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); + result = -ENODEV; + if (info->attrs[WIMAX_GNL_MSG_IFIDX] == NULL) { + pr_err("WIMAX_GNL_MSG_FROM_USER: can't find IFIDX attribute\n"); + goto error_no_wimax_dev; + } + ifindex = nla_get_u32(info->attrs[WIMAX_GNL_MSG_IFIDX]); + wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); + if (wimax_dev == NULL) + goto error_no_wimax_dev; + dev = wimax_dev_to_dev(wimax_dev); + + /* Unpack arguments */ + result = -EINVAL; + if (info->attrs[WIMAX_GNL_MSG_DATA] == NULL) { + dev_err(dev, "WIMAX_GNL_MSG_FROM_USER: can't find MSG_DATA " + "attribute\n"); + goto error_no_data; + } + msg_buf = nla_data(info->attrs[WIMAX_GNL_MSG_DATA]); + msg_len = nla_len(info->attrs[WIMAX_GNL_MSG_DATA]); + + if (info->attrs[WIMAX_GNL_MSG_PIPE_NAME] == NULL) + pipe_name = NULL; + else { + struct nlattr *attr = info->attrs[WIMAX_GNL_MSG_PIPE_NAME]; + size_t attr_len = nla_len(attr); + /* libnl-1.1 does not yet support NLA_NUL_STRING */ + result = -ENOMEM; + pipe_name = kstrndup(nla_data(attr), attr_len + 1, GFP_KERNEL); + if (pipe_name == NULL) + goto error_alloc; + pipe_name[attr_len] = 0; + } + mutex_lock(&wimax_dev->mutex); + result = wimax_dev_is_ready(wimax_dev); + if (result == -ENOMEDIUM) + result = 0; + if (result < 0) + goto error_not_ready; + result = -ENOSYS; + if (wimax_dev->op_msg_from_user == NULL) + goto error_noop; + + d_printf(1, dev, + "CRX: nlmsghdr len %u type %u flags 0x%04x seq 0x%x pid %u\n", + nlh->nlmsg_len, nlh->nlmsg_type, nlh->nlmsg_flags, + nlh->nlmsg_seq, nlh->nlmsg_pid); + d_printf(1, dev, "CRX: wimax message %zu bytes\n", msg_len); + d_dump(2, dev, msg_buf, msg_len); + + result = wimax_dev->op_msg_from_user(wimax_dev, pipe_name, + msg_buf, msg_len, info); +error_noop: +error_not_ready: + mutex_unlock(&wimax_dev->mutex); +error_alloc: + kfree(pipe_name); +error_no_data: + dev_put(wimax_dev->net_dev); +error_no_wimax_dev: + d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); + return result; +} diff --git a/drivers/staging/wimax/op-reset.c b/drivers/staging/wimax/op-reset.c new file mode 100644 index 000000000000..b3f000cbe112 --- /dev/null +++ b/drivers/staging/wimax/op-reset.c @@ -0,0 +1,108 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * Implement and export a method for resetting a WiMAX device + * + * Copyright (C) 2008 Intel Corporation + * Inaky Perez-Gonzalez + * + * This implements a simple synchronous call to reset a WiMAX device. + * + * Resets aim at being warm, keeping the device handles active; + * however, when that fails, it falls back to a cold reset (that will + * disconnect and reconnect the device). + */ + +#include "net-wimax.h" +#include +#include "linux-wimax.h" +#include +#include +#include "wimax-internal.h" + +#define D_SUBMODULE op_reset +#include "debug-levels.h" + + +/** + * wimax_reset - Reset a WiMAX device + * + * @wimax_dev: WiMAX device descriptor + * + * Returns: + * + * %0 if ok and a warm reset was done (the device still exists in + * the system). + * + * -%ENODEV if a cold/bus reset had to be done (device has + * disconnected and reconnected, so current handle is not valid + * any more). + * + * -%EINVAL if the device is not even registered. + * + * Any other negative error code shall be considered as + * non-recoverable. + * + * Description: + * + * Called when wanting to reset the device for any reason. Device is + * taken back to power on status. + * + * This call blocks; on successful return, the device has completed the + * reset process and is ready to operate. + */ +int wimax_reset(struct wimax_dev *wimax_dev) +{ + int result = -EINVAL; + struct device *dev = wimax_dev_to_dev(wimax_dev); + enum wimax_st state; + + might_sleep(); + d_fnstart(3, dev, "(wimax_dev %p)\n", wimax_dev); + mutex_lock(&wimax_dev->mutex); + dev_hold(wimax_dev->net_dev); + state = wimax_dev->state; + mutex_unlock(&wimax_dev->mutex); + + if (state >= WIMAX_ST_DOWN) { + mutex_lock(&wimax_dev->mutex_reset); + result = wimax_dev->op_reset(wimax_dev); + mutex_unlock(&wimax_dev->mutex_reset); + } + dev_put(wimax_dev->net_dev); + + d_fnend(3, dev, "(wimax_dev %p) = %d\n", wimax_dev, result); + return result; +} +EXPORT_SYMBOL(wimax_reset); + + +/* + * Exporting to user space over generic netlink + * + * Parse the reset command from user space, return error code. + * + * No attributes. + */ +int wimax_gnl_doit_reset(struct sk_buff *skb, struct genl_info *info) +{ + int result, ifindex; + struct wimax_dev *wimax_dev; + + d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); + result = -ENODEV; + if (info->attrs[WIMAX_GNL_RESET_IFIDX] == NULL) { + pr_err("WIMAX_GNL_OP_RFKILL: can't find IFIDX attribute\n"); + goto error_no_wimax_dev; + } + ifindex = nla_get_u32(info->attrs[WIMAX_GNL_RESET_IFIDX]); + wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); + if (wimax_dev == NULL) + goto error_no_wimax_dev; + /* Execute the operation and send the result back to user space */ + result = wimax_reset(wimax_dev); + dev_put(wimax_dev->net_dev); +error_no_wimax_dev: + d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); + return result; +} diff --git a/drivers/staging/wimax/op-rfkill.c b/drivers/staging/wimax/op-rfkill.c new file mode 100644 index 000000000000..78b294481a59 --- /dev/null +++ b/drivers/staging/wimax/op-rfkill.c @@ -0,0 +1,431 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * RF-kill framework integration + * + * Copyright (C) 2008 Intel Corporation + * Inaky Perez-Gonzalez + * + * This integrates into the Linux Kernel rfkill susbystem so that the + * drivers just have to do the bare minimal work, which is providing a + * method to set the software RF-Kill switch and to report changes in + * the software and hardware switch status. + * + * A non-polled generic rfkill device is embedded into the WiMAX + * subsystem's representation of a device. + * + * FIXME: Need polled support? Let drivers provide a poll routine + * and hand it to rfkill ops then? + * + * All device drivers have to do is after wimax_dev_init(), call + * wimax_report_rfkill_hw() and wimax_report_rfkill_sw() to update + * initial state and then every time it changes. See wimax.h:struct + * wimax_dev for more information. + * + * ROADMAP + * + * wimax_gnl_doit_rfkill() User space calling wimax_rfkill() + * wimax_rfkill() Kernel calling wimax_rfkill() + * __wimax_rf_toggle_radio() + * + * wimax_rfkill_set_radio_block() RF-Kill subsystem calling + * __wimax_rf_toggle_radio() + * + * __wimax_rf_toggle_radio() + * wimax_dev->op_rfkill_sw_toggle() Driver backend + * __wimax_state_change() + * + * wimax_report_rfkill_sw() Driver reports state change + * __wimax_state_change() + * + * wimax_report_rfkill_hw() Driver reports state change + * __wimax_state_change() + * + * wimax_rfkill_add() Initialize/shutdown rfkill support + * wimax_rfkill_rm() [called by wimax_dev_add/rm()] + */ + +#include "net-wimax.h" +#include +#include "linux-wimax.h" +#include +#include +#include +#include "wimax-internal.h" + +#define D_SUBMODULE op_rfkill +#include "debug-levels.h" + +/** + * wimax_report_rfkill_hw - Reports changes in the hardware RF switch + * + * @wimax_dev: WiMAX device descriptor + * + * @state: New state of the RF Kill switch. %WIMAX_RF_ON radio on, + * %WIMAX_RF_OFF radio off. + * + * When the device detects a change in the state of thehardware RF + * switch, it must call this function to let the WiMAX kernel stack + * know that the state has changed so it can be properly propagated. + * + * The WiMAX stack caches the state (the driver doesn't need to). As + * well, as the change is propagated it will come back as a request to + * change the software state to mirror the hardware state. + * + * If the device doesn't have a hardware kill switch, just report + * it on initialization as always on (%WIMAX_RF_ON, radio on). + */ +void wimax_report_rfkill_hw(struct wimax_dev *wimax_dev, + enum wimax_rf_state state) +{ + int result; + struct device *dev = wimax_dev_to_dev(wimax_dev); + enum wimax_st wimax_state; + + d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); + BUG_ON(state == WIMAX_RF_QUERY); + BUG_ON(state != WIMAX_RF_ON && state != WIMAX_RF_OFF); + + mutex_lock(&wimax_dev->mutex); + result = wimax_dev_is_ready(wimax_dev); + if (result < 0) + goto error_not_ready; + + if (state != wimax_dev->rf_hw) { + wimax_dev->rf_hw = state; + if (wimax_dev->rf_hw == WIMAX_RF_ON && + wimax_dev->rf_sw == WIMAX_RF_ON) + wimax_state = WIMAX_ST_READY; + else + wimax_state = WIMAX_ST_RADIO_OFF; + + result = rfkill_set_hw_state(wimax_dev->rfkill, + state == WIMAX_RF_OFF); + + __wimax_state_change(wimax_dev, wimax_state); + } +error_not_ready: + mutex_unlock(&wimax_dev->mutex); + d_fnend(3, dev, "(wimax_dev %p state %u) = void [%d]\n", + wimax_dev, state, result); +} +EXPORT_SYMBOL_GPL(wimax_report_rfkill_hw); + + +/** + * wimax_report_rfkill_sw - Reports changes in the software RF switch + * + * @wimax_dev: WiMAX device descriptor + * + * @state: New state of the RF kill switch. %WIMAX_RF_ON radio on, + * %WIMAX_RF_OFF radio off. + * + * Reports changes in the software RF switch state to the WiMAX stack. + * + * The main use is during initialization, so the driver can query the + * device for its current software radio kill switch state and feed it + * to the system. + * + * On the side, the device does not change the software state by + * itself. In practice, this can happen, as the device might decide to + * switch (in software) the radio off for different reasons. + */ +void wimax_report_rfkill_sw(struct wimax_dev *wimax_dev, + enum wimax_rf_state state) +{ + int result; + struct device *dev = wimax_dev_to_dev(wimax_dev); + enum wimax_st wimax_state; + + d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); + BUG_ON(state == WIMAX_RF_QUERY); + BUG_ON(state != WIMAX_RF_ON && state != WIMAX_RF_OFF); + + mutex_lock(&wimax_dev->mutex); + result = wimax_dev_is_ready(wimax_dev); + if (result < 0) + goto error_not_ready; + + if (state != wimax_dev->rf_sw) { + wimax_dev->rf_sw = state; + if (wimax_dev->rf_hw == WIMAX_RF_ON && + wimax_dev->rf_sw == WIMAX_RF_ON) + wimax_state = WIMAX_ST_READY; + else + wimax_state = WIMAX_ST_RADIO_OFF; + __wimax_state_change(wimax_dev, wimax_state); + rfkill_set_sw_state(wimax_dev->rfkill, state == WIMAX_RF_OFF); + } +error_not_ready: + mutex_unlock(&wimax_dev->mutex); + d_fnend(3, dev, "(wimax_dev %p state %u) = void [%d]\n", + wimax_dev, state, result); +} +EXPORT_SYMBOL_GPL(wimax_report_rfkill_sw); + + +/* + * Callback for the RF Kill toggle operation + * + * This function is called by: + * + * - The rfkill subsystem when the RF-Kill key is pressed in the + * hardware and the driver notifies through + * wimax_report_rfkill_hw(). The rfkill subsystem ends up calling back + * here so the software RF Kill switch state is changed to reflect + * the hardware switch state. + * + * - When the user sets the state through sysfs' rfkill/state file + * + * - When the user calls wimax_rfkill(). + * + * This call blocks! + * + * WARNING! When we call rfkill_unregister(), this will be called with + * state 0! + * + * WARNING: wimax_dev must be locked + */ +static +int __wimax_rf_toggle_radio(struct wimax_dev *wimax_dev, + enum wimax_rf_state state) +{ + int result = 0; + struct device *dev = wimax_dev_to_dev(wimax_dev); + enum wimax_st wimax_state; + + might_sleep(); + d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); + if (wimax_dev->rf_sw == state) + goto out_no_change; + if (wimax_dev->op_rfkill_sw_toggle != NULL) + result = wimax_dev->op_rfkill_sw_toggle(wimax_dev, state); + else if (state == WIMAX_RF_OFF) /* No op? can't turn off */ + result = -ENXIO; + else /* No op? can turn on */ + result = 0; /* should never happen tho */ + if (result >= 0) { + result = 0; + wimax_dev->rf_sw = state; + wimax_state = state == WIMAX_RF_ON ? + WIMAX_ST_READY : WIMAX_ST_RADIO_OFF; + __wimax_state_change(wimax_dev, wimax_state); + } +out_no_change: + d_fnend(3, dev, "(wimax_dev %p state %u) = %d\n", + wimax_dev, state, result); + return result; +} + + +/* + * Translate from rfkill state to wimax state + * + * NOTE: Special state handling rules here + * + * Just pretend the call didn't happen if we are in a state where + * we know for sure it cannot be handled (WIMAX_ST_DOWN or + * __WIMAX_ST_QUIESCING). rfkill() needs it to register and + * unregister, as it will run this path. + * + * NOTE: This call will block until the operation is completed. + */ +static int wimax_rfkill_set_radio_block(void *data, bool blocked) +{ + int result; + struct wimax_dev *wimax_dev = data; + struct device *dev = wimax_dev_to_dev(wimax_dev); + enum wimax_rf_state rf_state; + + d_fnstart(3, dev, "(wimax_dev %p blocked %u)\n", wimax_dev, blocked); + rf_state = WIMAX_RF_ON; + if (blocked) + rf_state = WIMAX_RF_OFF; + mutex_lock(&wimax_dev->mutex); + if (wimax_dev->state <= __WIMAX_ST_QUIESCING) + result = 0; + else + result = __wimax_rf_toggle_radio(wimax_dev, rf_state); + mutex_unlock(&wimax_dev->mutex); + d_fnend(3, dev, "(wimax_dev %p blocked %u) = %d\n", + wimax_dev, blocked, result); + return result; +} + +static const struct rfkill_ops wimax_rfkill_ops = { + .set_block = wimax_rfkill_set_radio_block, +}; + +/** + * wimax_rfkill - Set the software RF switch state for a WiMAX device + * + * @wimax_dev: WiMAX device descriptor + * + * @state: New RF state. + * + * Returns: + * + * >= 0 toggle state if ok, < 0 errno code on error. The toggle state + * is returned as a bitmap, bit 0 being the hardware RF state, bit 1 + * the software RF state. + * + * 0 means disabled (%WIMAX_RF_ON, radio on), 1 means enabled radio + * off (%WIMAX_RF_OFF). + * + * Description: + * + * Called by the user when he wants to request the WiMAX radio to be + * switched on (%WIMAX_RF_ON) or off (%WIMAX_RF_OFF). With + * %WIMAX_RF_QUERY, just the current state is returned. + * + * NOTE: + * + * This call will block until the operation is complete. + */ +int wimax_rfkill(struct wimax_dev *wimax_dev, enum wimax_rf_state state) +{ + int result; + struct device *dev = wimax_dev_to_dev(wimax_dev); + + d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); + mutex_lock(&wimax_dev->mutex); + result = wimax_dev_is_ready(wimax_dev); + if (result < 0) { + /* While initializing, < 1.4.3 wimax-tools versions use + * this call to check if the device is a valid WiMAX + * device; so we allow it to proceed always, + * considering the radios are all off. */ + if (result == -ENOMEDIUM && state == WIMAX_RF_QUERY) + result = WIMAX_RF_OFF << 1 | WIMAX_RF_OFF; + goto error_not_ready; + } + switch (state) { + case WIMAX_RF_ON: + case WIMAX_RF_OFF: + result = __wimax_rf_toggle_radio(wimax_dev, state); + if (result < 0) + goto error; + rfkill_set_sw_state(wimax_dev->rfkill, state == WIMAX_RF_OFF); + break; + case WIMAX_RF_QUERY: + break; + default: + result = -EINVAL; + goto error; + } + result = wimax_dev->rf_sw << 1 | wimax_dev->rf_hw; +error: +error_not_ready: + mutex_unlock(&wimax_dev->mutex); + d_fnend(3, dev, "(wimax_dev %p state %u) = %d\n", + wimax_dev, state, result); + return result; +} +EXPORT_SYMBOL(wimax_rfkill); + + +/* + * Register a new WiMAX device's RF Kill support + * + * WARNING: wimax_dev->mutex must be unlocked + */ +int wimax_rfkill_add(struct wimax_dev *wimax_dev) +{ + int result; + struct rfkill *rfkill; + struct device *dev = wimax_dev_to_dev(wimax_dev); + + d_fnstart(3, dev, "(wimax_dev %p)\n", wimax_dev); + /* Initialize RF Kill */ + result = -ENOMEM; + rfkill = rfkill_alloc(wimax_dev->name, dev, RFKILL_TYPE_WIMAX, + &wimax_rfkill_ops, wimax_dev); + if (rfkill == NULL) + goto error_rfkill_allocate; + + d_printf(1, dev, "rfkill %p\n", rfkill); + + wimax_dev->rfkill = rfkill; + + rfkill_init_sw_state(rfkill, 1); + result = rfkill_register(wimax_dev->rfkill); + if (result < 0) + goto error_rfkill_register; + + /* If there is no SW toggle op, SW RFKill is always on */ + if (wimax_dev->op_rfkill_sw_toggle == NULL) + wimax_dev->rf_sw = WIMAX_RF_ON; + + d_fnend(3, dev, "(wimax_dev %p) = 0\n", wimax_dev); + return 0; + +error_rfkill_register: + rfkill_destroy(wimax_dev->rfkill); +error_rfkill_allocate: + d_fnend(3, dev, "(wimax_dev %p) = %d\n", wimax_dev, result); + return result; +} + + +/* + * Deregister a WiMAX device's RF Kill support + * + * Ick, we can't call rfkill_free() after rfkill_unregister()...oh + * well. + * + * WARNING: wimax_dev->mutex must be unlocked + */ +void wimax_rfkill_rm(struct wimax_dev *wimax_dev) +{ + struct device *dev = wimax_dev_to_dev(wimax_dev); + d_fnstart(3, dev, "(wimax_dev %p)\n", wimax_dev); + rfkill_unregister(wimax_dev->rfkill); + rfkill_destroy(wimax_dev->rfkill); + d_fnend(3, dev, "(wimax_dev %p)\n", wimax_dev); +} + + +/* + * Exporting to user space over generic netlink + * + * Parse the rfkill command from user space, return a combination + * value that describe the states of the different toggles. + * + * Only one attribute: the new state requested (on, off or no change, + * just query). + */ + +int wimax_gnl_doit_rfkill(struct sk_buff *skb, struct genl_info *info) +{ + int result, ifindex; + struct wimax_dev *wimax_dev; + struct device *dev; + enum wimax_rf_state new_state; + + d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); + result = -ENODEV; + if (info->attrs[WIMAX_GNL_RFKILL_IFIDX] == NULL) { + pr_err("WIMAX_GNL_OP_RFKILL: can't find IFIDX attribute\n"); + goto error_no_wimax_dev; + } + ifindex = nla_get_u32(info->attrs[WIMAX_GNL_RFKILL_IFIDX]); + wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); + if (wimax_dev == NULL) + goto error_no_wimax_dev; + dev = wimax_dev_to_dev(wimax_dev); + result = -EINVAL; + if (info->attrs[WIMAX_GNL_RFKILL_STATE] == NULL) { + dev_err(dev, "WIMAX_GNL_RFKILL: can't find RFKILL_STATE " + "attribute\n"); + goto error_no_pid; + } + new_state = nla_get_u32(info->attrs[WIMAX_GNL_RFKILL_STATE]); + + /* Execute the operation and send the result back to user space */ + result = wimax_rfkill(wimax_dev, new_state); +error_no_pid: + dev_put(wimax_dev->net_dev); +error_no_wimax_dev: + d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); + return result; +} diff --git a/drivers/staging/wimax/op-state-get.c b/drivers/staging/wimax/op-state-get.c new file mode 100644 index 000000000000..c5bfbed505f5 --- /dev/null +++ b/drivers/staging/wimax/op-state-get.c @@ -0,0 +1,52 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * Implement and export a method for getting a WiMAX device current state + * + * Copyright (C) 2009 Paulius Zaleckas + * + * Based on previous WiMAX core work by: + * Copyright (C) 2008 Intel Corporation + * Inaky Perez-Gonzalez + */ + +#include "net-wimax.h" +#include +#include "linux-wimax.h" +#include +#include "wimax-internal.h" + +#define D_SUBMODULE op_state_get +#include "debug-levels.h" + + +/* + * Exporting to user space over generic netlink + * + * Parse the state get command from user space, return a combination + * value that describe the current state. + * + * No attributes. + */ +int wimax_gnl_doit_state_get(struct sk_buff *skb, struct genl_info *info) +{ + int result, ifindex; + struct wimax_dev *wimax_dev; + + d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); + result = -ENODEV; + if (info->attrs[WIMAX_GNL_STGET_IFIDX] == NULL) { + pr_err("WIMAX_GNL_OP_STATE_GET: can't find IFIDX attribute\n"); + goto error_no_wimax_dev; + } + ifindex = nla_get_u32(info->attrs[WIMAX_GNL_STGET_IFIDX]); + wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); + if (wimax_dev == NULL) + goto error_no_wimax_dev; + /* Execute the operation and send the result back to user space */ + result = wimax_state_get(wimax_dev); + dev_put(wimax_dev->net_dev); +error_no_wimax_dev: + d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); + return result; +} diff --git a/drivers/staging/wimax/stack.c b/drivers/staging/wimax/stack.c new file mode 100644 index 000000000000..ace24a6dfd2d --- /dev/null +++ b/drivers/staging/wimax/stack.c @@ -0,0 +1,616 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Linux WiMAX + * Initialization, addition and removal of wimax devices + * + * Copyright (C) 2005-2006 Intel Corporation + * Inaky Perez-Gonzalez + * + * This implements: + * + * - basic life cycle of 'struct wimax_dev' [wimax_dev_*()]; on + * addition/registration initialize all subfields and allocate + * generic netlink resources for user space communication. On + * removal/unregistration, undo all that. + * + * - device state machine [wimax_state_change()] and support to send + * reports to user space when the state changes + * [wimax_gnl_re_state_change*()]. + * + * See include/net/wimax.h for rationales and design. + * + * ROADMAP + * + * [__]wimax_state_change() Called by drivers to update device's state + * wimax_gnl_re_state_change_alloc() + * wimax_gnl_re_state_change_send() + * + * wimax_dev_init() Init a device + * wimax_dev_add() Register + * wimax_rfkill_add() + * wimax_gnl_add() Register all the generic netlink resources. + * wimax_id_table_add() + * wimax_dev_rm() Unregister + * wimax_id_table_rm() + * wimax_gnl_rm() + * wimax_rfkill_rm() + */ +#include +#include +#include +#include +#include "linux-wimax.h" +#include +#include "wimax-internal.h" + + +#define D_SUBMODULE stack +#include "debug-levels.h" + +static char wimax_debug_params[128]; +module_param_string(debug, wimax_debug_params, sizeof(wimax_debug_params), + 0644); +MODULE_PARM_DESC(debug, + "String of space-separated NAME:VALUE pairs, where NAMEs " + "are the different debug submodules and VALUE are the " + "initial debug value to set."); + +/* + * Authoritative source for the RE_STATE_CHANGE attribute policy + * + * We don't really use it here, but /me likes to keep the definition + * close to where the data is generated. + */ +/* +static const struct nla_policy wimax_gnl_re_status_change[WIMAX_GNL_ATTR_MAX + 1] = { + [WIMAX_GNL_STCH_STATE_OLD] = { .type = NLA_U8 }, + [WIMAX_GNL_STCH_STATE_NEW] = { .type = NLA_U8 }, +}; +*/ + + +/* + * Allocate a Report State Change message + * + * @header: save it, you need it for _send() + * + * Creates and fills a basic state change message; different code + * paths can then add more attributes to the message as needed. + * + * Use wimax_gnl_re_state_change_send() to send the returned skb. + * + * Returns: skb with the genl message if ok, IS_ERR() ptr on error + * with an errno code. + */ +static +struct sk_buff *wimax_gnl_re_state_change_alloc( + struct wimax_dev *wimax_dev, + enum wimax_st new_state, enum wimax_st old_state, + void **header) +{ + int result; + struct device *dev = wimax_dev_to_dev(wimax_dev); + void *data; + struct sk_buff *report_skb; + + d_fnstart(3, dev, "(wimax_dev %p new_state %u old_state %u)\n", + wimax_dev, new_state, old_state); + result = -ENOMEM; + report_skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); + if (report_skb == NULL) { + dev_err(dev, "RE_STCH: can't create message\n"); + goto error_new; + } + /* FIXME: sending a group ID as the seq is wrong */ + data = genlmsg_put(report_skb, 0, wimax_gnl_family.mcgrp_offset, + &wimax_gnl_family, 0, WIMAX_GNL_RE_STATE_CHANGE); + if (data == NULL) { + dev_err(dev, "RE_STCH: can't put data into message\n"); + goto error_put; + } + *header = data; + + result = nla_put_u8(report_skb, WIMAX_GNL_STCH_STATE_OLD, old_state); + if (result < 0) { + dev_err(dev, "RE_STCH: Error adding OLD attr: %d\n", result); + goto error_put; + } + result = nla_put_u8(report_skb, WIMAX_GNL_STCH_STATE_NEW, new_state); + if (result < 0) { + dev_err(dev, "RE_STCH: Error adding NEW attr: %d\n", result); + goto error_put; + } + result = nla_put_u32(report_skb, WIMAX_GNL_STCH_IFIDX, + wimax_dev->net_dev->ifindex); + if (result < 0) { + dev_err(dev, "RE_STCH: Error adding IFINDEX attribute\n"); + goto error_put; + } + d_fnend(3, dev, "(wimax_dev %p new_state %u old_state %u) = %p\n", + wimax_dev, new_state, old_state, report_skb); + return report_skb; + +error_put: + nlmsg_free(report_skb); +error_new: + d_fnend(3, dev, "(wimax_dev %p new_state %u old_state %u) = %d\n", + wimax_dev, new_state, old_state, result); + return ERR_PTR(result); +} + + +/* + * Send a Report State Change message (as created with _alloc). + * + * @report_skb: as returned by wimax_gnl_re_state_change_alloc() + * @header: as returned by wimax_gnl_re_state_change_alloc() + * + * Returns: 0 if ok, < 0 errno code on error. + * + * If the message is NULL, pretend it didn't happen. + */ +static +int wimax_gnl_re_state_change_send( + struct wimax_dev *wimax_dev, struct sk_buff *report_skb, + void *header) +{ + int result = 0; + struct device *dev = wimax_dev_to_dev(wimax_dev); + d_fnstart(3, dev, "(wimax_dev %p report_skb %p)\n", + wimax_dev, report_skb); + if (report_skb == NULL) { + result = -ENOMEM; + goto out; + } + genlmsg_end(report_skb, header); + genlmsg_multicast(&wimax_gnl_family, report_skb, 0, 0, GFP_KERNEL); +out: + d_fnend(3, dev, "(wimax_dev %p report_skb %p) = %d\n", + wimax_dev, report_skb, result); + return result; +} + + +static +void __check_new_state(enum wimax_st old_state, enum wimax_st new_state, + unsigned int allowed_states_bm) +{ + if (WARN_ON(((1 << new_state) & allowed_states_bm) == 0)) { + pr_err("SW BUG! Forbidden state change %u -> %u\n", + old_state, new_state); + } +} + + +/* + * Set the current state of a WiMAX device [unlocking version of + * wimax_state_change(). + */ +void __wimax_state_change(struct wimax_dev *wimax_dev, enum wimax_st new_state) +{ + struct device *dev = wimax_dev_to_dev(wimax_dev); + enum wimax_st old_state = wimax_dev->state; + struct sk_buff *stch_skb; + void *header; + + d_fnstart(3, dev, "(wimax_dev %p new_state %u [old %u])\n", + wimax_dev, new_state, old_state); + + if (WARN_ON(new_state >= __WIMAX_ST_INVALID)) { + dev_err(dev, "SW BUG: requesting invalid state %u\n", + new_state); + goto out; + } + if (old_state == new_state) + goto out; + header = NULL; /* gcc complains? can't grok why */ + stch_skb = wimax_gnl_re_state_change_alloc( + wimax_dev, new_state, old_state, &header); + + /* Verify the state transition and do exit-from-state actions */ + switch (old_state) { + case __WIMAX_ST_NULL: + __check_new_state(old_state, new_state, + 1 << WIMAX_ST_DOWN); + break; + case WIMAX_ST_DOWN: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_UNINITIALIZED + | 1 << WIMAX_ST_RADIO_OFF); + break; + case __WIMAX_ST_QUIESCING: + __check_new_state(old_state, new_state, 1 << WIMAX_ST_DOWN); + break; + case WIMAX_ST_UNINITIALIZED: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_RADIO_OFF); + break; + case WIMAX_ST_RADIO_OFF: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_READY); + break; + case WIMAX_ST_READY: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_RADIO_OFF + | 1 << WIMAX_ST_SCANNING + | 1 << WIMAX_ST_CONNECTING + | 1 << WIMAX_ST_CONNECTED); + break; + case WIMAX_ST_SCANNING: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_RADIO_OFF + | 1 << WIMAX_ST_READY + | 1 << WIMAX_ST_CONNECTING + | 1 << WIMAX_ST_CONNECTED); + break; + case WIMAX_ST_CONNECTING: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_RADIO_OFF + | 1 << WIMAX_ST_READY + | 1 << WIMAX_ST_SCANNING + | 1 << WIMAX_ST_CONNECTED); + break; + case WIMAX_ST_CONNECTED: + __check_new_state(old_state, new_state, + 1 << __WIMAX_ST_QUIESCING + | 1 << WIMAX_ST_RADIO_OFF + | 1 << WIMAX_ST_READY); + netif_tx_disable(wimax_dev->net_dev); + netif_carrier_off(wimax_dev->net_dev); + break; + case __WIMAX_ST_INVALID: + default: + dev_err(dev, "SW BUG: wimax_dev %p is in unknown state %u\n", + wimax_dev, wimax_dev->state); + WARN_ON(1); + goto out; + } + + /* Execute the actions of entry to the new state */ + switch (new_state) { + case __WIMAX_ST_NULL: + dev_err(dev, "SW BUG: wimax_dev %p entering NULL state " + "from %u\n", wimax_dev, wimax_dev->state); + WARN_ON(1); /* Nobody can enter this state */ + break; + case WIMAX_ST_DOWN: + break; + case __WIMAX_ST_QUIESCING: + break; + case WIMAX_ST_UNINITIALIZED: + break; + case WIMAX_ST_RADIO_OFF: + break; + case WIMAX_ST_READY: + break; + case WIMAX_ST_SCANNING: + break; + case WIMAX_ST_CONNECTING: + break; + case WIMAX_ST_CONNECTED: + netif_carrier_on(wimax_dev->net_dev); + netif_wake_queue(wimax_dev->net_dev); + break; + case __WIMAX_ST_INVALID: + default: + BUG(); + } + __wimax_state_set(wimax_dev, new_state); + if (!IS_ERR(stch_skb)) + wimax_gnl_re_state_change_send(wimax_dev, stch_skb, header); +out: + d_fnend(3, dev, "(wimax_dev %p new_state %u [old %u]) = void\n", + wimax_dev, new_state, old_state); +} + + +/** + * wimax_state_change - Set the current state of a WiMAX device + * + * @wimax_dev: WiMAX device descriptor (properly referenced) + * @new_state: New state to switch to + * + * This implements the state changes for the wimax devices. It will + * + * - verify that the state transition is legal (for now it'll just + * print a warning if not) according to the table in + * linux/wimax.h's documentation for 'enum wimax_st'. + * + * - perform the actions needed for leaving the current state and + * whichever are needed for entering the new state. + * + * - issue a report to user space indicating the new state (and an + * optional payload with information about the new state). + * + * NOTE: @wimax_dev must be locked + */ +void wimax_state_change(struct wimax_dev *wimax_dev, enum wimax_st new_state) +{ + /* + * A driver cannot take the wimax_dev out of the + * __WIMAX_ST_NULL state unless by calling wimax_dev_add(). If + * the wimax_dev's state is still NULL, we ignore any request + * to change its state because it means it hasn't been yet + * registered. + * + * There is no need to complain about it, as routines that + * call this might be shared from different code paths that + * are called before or after wimax_dev_add() has done its + * job. + */ + mutex_lock(&wimax_dev->mutex); + if (wimax_dev->state > __WIMAX_ST_NULL) + __wimax_state_change(wimax_dev, new_state); + mutex_unlock(&wimax_dev->mutex); +} +EXPORT_SYMBOL_GPL(wimax_state_change); + + +/** + * wimax_state_get() - Return the current state of a WiMAX device + * + * @wimax_dev: WiMAX device descriptor + * + * Returns: Current state of the device according to its driver. + */ +enum wimax_st wimax_state_get(struct wimax_dev *wimax_dev) +{ + enum wimax_st state; + mutex_lock(&wimax_dev->mutex); + state = wimax_dev->state; + mutex_unlock(&wimax_dev->mutex); + return state; +} +EXPORT_SYMBOL_GPL(wimax_state_get); + + +/** + * wimax_dev_init - initialize a newly allocated instance + * + * @wimax_dev: WiMAX device descriptor to initialize. + * + * Initializes fields of a freshly allocated @wimax_dev instance. This + * function assumes that after allocation, the memory occupied by + * @wimax_dev was zeroed. + */ +void wimax_dev_init(struct wimax_dev *wimax_dev) +{ + INIT_LIST_HEAD(&wimax_dev->id_table_node); + __wimax_state_set(wimax_dev, __WIMAX_ST_NULL); + mutex_init(&wimax_dev->mutex); + mutex_init(&wimax_dev->mutex_reset); +} +EXPORT_SYMBOL_GPL(wimax_dev_init); + +/* + * There are multiple enums reusing the same values, adding + * others is only possible if they use a compatible policy. + */ +static const struct nla_policy wimax_gnl_policy[WIMAX_GNL_ATTR_MAX + 1] = { + /* + * WIMAX_GNL_RESET_IFIDX, WIMAX_GNL_RFKILL_IFIDX, + * WIMAX_GNL_STGET_IFIDX, WIMAX_GNL_MSG_IFIDX + */ + [1] = { .type = NLA_U32, }, + /* + * WIMAX_GNL_RFKILL_STATE, WIMAX_GNL_MSG_PIPE_NAME + */ + [2] = { .type = NLA_U32, }, /* enum wimax_rf_state */ + /* + * WIMAX_GNL_MSG_DATA + */ + [3] = { .type = NLA_UNSPEC, }, /* libnl doesn't grok BINARY yet */ +}; + +static const struct genl_small_ops wimax_gnl_ops[] = { + { + .cmd = WIMAX_GNL_OP_MSG_FROM_USER, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .flags = GENL_ADMIN_PERM, + .doit = wimax_gnl_doit_msg_from_user, + }, + { + .cmd = WIMAX_GNL_OP_RESET, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .flags = GENL_ADMIN_PERM, + .doit = wimax_gnl_doit_reset, + }, + { + .cmd = WIMAX_GNL_OP_RFKILL, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .flags = GENL_ADMIN_PERM, + .doit = wimax_gnl_doit_rfkill, + }, + { + .cmd = WIMAX_GNL_OP_STATE_GET, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .flags = GENL_ADMIN_PERM, + .doit = wimax_gnl_doit_state_get, + }, +}; + + +static +size_t wimax_addr_scnprint(char *addr_str, size_t addr_str_size, + unsigned char *addr, size_t addr_len) +{ + unsigned int cnt, total; + + for (total = cnt = 0; cnt < addr_len; cnt++) + total += scnprintf(addr_str + total, addr_str_size - total, + "%02x%c", addr[cnt], + cnt == addr_len - 1 ? '\0' : ':'); + return total; +} + + +/** + * wimax_dev_add - Register a new WiMAX device + * + * @wimax_dev: WiMAX device descriptor (as embedded in your @net_dev's + * priv data). You must have called wimax_dev_init() on it before. + * + * @net_dev: net device the @wimax_dev is associated with. The + * function expects SET_NETDEV_DEV() and register_netdev() were + * already called on it. + * + * Registers the new WiMAX device, sets up the user-kernel control + * interface (generic netlink) and common WiMAX infrastructure. + * + * Note that the parts that will allow interaction with user space are + * setup at the very end, when the rest is in place, as once that + * happens, the driver might get user space control requests via + * netlink or from debugfs that might translate into calls into + * wimax_dev->op_*(). + */ +int wimax_dev_add(struct wimax_dev *wimax_dev, struct net_device *net_dev) +{ + int result; + struct device *dev = net_dev->dev.parent; + char addr_str[32]; + + d_fnstart(3, dev, "(wimax_dev %p net_dev %p)\n", wimax_dev, net_dev); + + /* Do the RFKILL setup before locking, as RFKILL will call + * into our functions. + */ + wimax_dev->net_dev = net_dev; + result = wimax_rfkill_add(wimax_dev); + if (result < 0) + goto error_rfkill_add; + + /* Set up user-space interaction */ + mutex_lock(&wimax_dev->mutex); + wimax_id_table_add(wimax_dev); + wimax_debugfs_add(wimax_dev); + + __wimax_state_set(wimax_dev, WIMAX_ST_DOWN); + mutex_unlock(&wimax_dev->mutex); + + wimax_addr_scnprint(addr_str, sizeof(addr_str), + net_dev->dev_addr, net_dev->addr_len); + dev_err(dev, "WiMAX interface %s (%s) ready\n", + net_dev->name, addr_str); + d_fnend(3, dev, "(wimax_dev %p net_dev %p) = 0\n", wimax_dev, net_dev); + return 0; + +error_rfkill_add: + d_fnend(3, dev, "(wimax_dev %p net_dev %p) = %d\n", + wimax_dev, net_dev, result); + return result; +} +EXPORT_SYMBOL_GPL(wimax_dev_add); + + +/** + * wimax_dev_rm - Unregister an existing WiMAX device + * + * @wimax_dev: WiMAX device descriptor + * + * Unregisters a WiMAX device previously registered for use with + * wimax_add_rm(). + * + * IMPORTANT! Must call before calling unregister_netdev(). + * + * After this function returns, you will not get any more user space + * control requests (via netlink or debugfs) and thus to wimax_dev->ops. + * + * Reentrancy control is ensured by setting the state to + * %__WIMAX_ST_QUIESCING. rfkill operations coming through + * wimax_*rfkill*() will be stopped by the quiescing state; ops coming + * from the rfkill subsystem will be stopped by the support being + * removed by wimax_rfkill_rm(). + */ +void wimax_dev_rm(struct wimax_dev *wimax_dev) +{ + d_fnstart(3, NULL, "(wimax_dev %p)\n", wimax_dev); + + mutex_lock(&wimax_dev->mutex); + __wimax_state_change(wimax_dev, __WIMAX_ST_QUIESCING); + wimax_debugfs_rm(wimax_dev); + wimax_id_table_rm(wimax_dev); + __wimax_state_change(wimax_dev, WIMAX_ST_DOWN); + mutex_unlock(&wimax_dev->mutex); + wimax_rfkill_rm(wimax_dev); + d_fnend(3, NULL, "(wimax_dev %p) = void\n", wimax_dev); +} +EXPORT_SYMBOL_GPL(wimax_dev_rm); + + +/* Debug framework control of debug levels */ +struct d_level D_LEVEL[] = { + D_SUBMODULE_DEFINE(debugfs), + D_SUBMODULE_DEFINE(id_table), + D_SUBMODULE_DEFINE(op_msg), + D_SUBMODULE_DEFINE(op_reset), + D_SUBMODULE_DEFINE(op_rfkill), + D_SUBMODULE_DEFINE(op_state_get), + D_SUBMODULE_DEFINE(stack), +}; +size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); + + +static const struct genl_multicast_group wimax_gnl_mcgrps[] = { + { .name = "msg", }, +}; + +struct genl_family wimax_gnl_family __ro_after_init = { + .name = "WiMAX", + .version = WIMAX_GNL_VERSION, + .hdrsize = 0, + .maxattr = WIMAX_GNL_ATTR_MAX, + .policy = wimax_gnl_policy, + .module = THIS_MODULE, + .small_ops = wimax_gnl_ops, + .n_small_ops = ARRAY_SIZE(wimax_gnl_ops), + .mcgrps = wimax_gnl_mcgrps, + .n_mcgrps = ARRAY_SIZE(wimax_gnl_mcgrps), +}; + + + +/* Shutdown the wimax stack */ +static +int __init wimax_subsys_init(void) +{ + int result; + + d_fnstart(4, NULL, "()\n"); + d_parse_params(D_LEVEL, D_LEVEL_SIZE, wimax_debug_params, + "wimax.debug"); + + result = genl_register_family(&wimax_gnl_family); + if (unlikely(result < 0)) { + pr_err("cannot register generic netlink family: %d\n", result); + goto error_register_family; + } + + d_fnend(4, NULL, "() = 0\n"); + return 0; + +error_register_family: + d_fnend(4, NULL, "() = %d\n", result); + return result; + +} +module_init(wimax_subsys_init); + + +/* Shutdown the wimax stack */ +static +void __exit wimax_subsys_exit(void) +{ + wimax_id_table_release(); + genl_unregister_family(&wimax_gnl_family); +} +module_exit(wimax_subsys_exit); + +MODULE_AUTHOR("Intel Corporation "); +MODULE_DESCRIPTION("Linux WiMAX stack"); +MODULE_LICENSE("GPL"); diff --git a/drivers/staging/wimax/wimax-internal.h b/drivers/staging/wimax/wimax-internal.h new file mode 100644 index 000000000000..a6b6990642a1 --- /dev/null +++ b/drivers/staging/wimax/wimax-internal.h @@ -0,0 +1,85 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Linux WiMAX + * Internal API for kernel space WiMAX stack + * + * Copyright (C) 2007 Intel Corporation + * Inaky Perez-Gonzalez + * + * This header file is for declarations and definitions internal to + * the WiMAX stack. For public APIs and documentation, see + * include/net/wimax.h and include/linux/wimax.h. + */ + +#ifndef __WIMAX_INTERNAL_H__ +#define __WIMAX_INTERNAL_H__ +#ifdef __KERNEL__ + +#ifdef pr_fmt +#undef pr_fmt +#endif + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include "net-wimax.h" + + +/* + * Decide if a (locked) device is ready for use + * + * Before using the device structure, it must be locked + * (wimax_dev->mutex). As well, most operations need to call this + * function to check if the state is the right one. + * + * An error value will be returned if the state is not the right + * one. In that case, the caller should not attempt to use the device + * and just unlock it. + */ +static inline __must_check +int wimax_dev_is_ready(struct wimax_dev *wimax_dev) +{ + if (wimax_dev->state == __WIMAX_ST_NULL) + return -EINVAL; /* Device is not even registered! */ + if (wimax_dev->state == WIMAX_ST_DOWN) + return -ENOMEDIUM; + if (wimax_dev->state == __WIMAX_ST_QUIESCING) + return -ESHUTDOWN; + return 0; +} + + +static inline +void __wimax_state_set(struct wimax_dev *wimax_dev, enum wimax_st state) +{ + wimax_dev->state = state; +} +void __wimax_state_change(struct wimax_dev *, enum wimax_st); + +#ifdef CONFIG_DEBUG_FS +void wimax_debugfs_add(struct wimax_dev *); +void wimax_debugfs_rm(struct wimax_dev *); +#else +static inline void wimax_debugfs_add(struct wimax_dev *wimax_dev) {} +static inline void wimax_debugfs_rm(struct wimax_dev *wimax_dev) {} +#endif + +void wimax_id_table_add(struct wimax_dev *); +struct wimax_dev *wimax_dev_get_by_genl_info(struct genl_info *, int); +void wimax_id_table_rm(struct wimax_dev *); +void wimax_id_table_release(void); + +int wimax_rfkill_add(struct wimax_dev *); +void wimax_rfkill_rm(struct wimax_dev *); + +/* generic netlink */ +extern struct genl_family wimax_gnl_family; + +/* ops */ +int wimax_gnl_doit_msg_from_user(struct sk_buff *skb, struct genl_info *info); +int wimax_gnl_doit_reset(struct sk_buff *skb, struct genl_info *info); +int wimax_gnl_doit_rfkill(struct sk_buff *skb, struct genl_info *info); +int wimax_gnl_doit_state_get(struct sk_buff *skb, struct genl_info *info); + +#endif /* #ifdef __KERNEL__ */ +#endif /* #ifndef __WIMAX_INTERNAL_H__ */ diff --git a/include/linux/wimax/debug.h b/include/linux/wimax/debug.h deleted file mode 100644 index cdae052bcdcd..000000000000 --- a/include/linux/wimax/debug.h +++ /dev/null @@ -1,491 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Linux WiMAX - * Collection of tools to manage debug operations. - * - * Copyright (C) 2005-2007 Intel Corporation - * Inaky Perez-Gonzalez - * - * Don't #include this file directly, read on! - * - * EXECUTING DEBUGGING ACTIONS OR NOT - * - * The main thing this framework provides is decission power to take a - * debug action (like printing a message) if the current debug level - * allows it. - * - * The decission power is at two levels: at compile-time (what does - * not make it is compiled out) and at run-time. The run-time - * selection is done per-submodule (as they are declared by the user - * of the framework). - * - * A call to d_test(L) (L being the target debug level) returns true - * if the action should be taken because the current debug levels - * allow it (both compile and run time). - * - * It follows that a call to d_test() that can be determined to be - * always false at compile time will get the code depending on it - * compiled out by optimization. - * - * DEBUG LEVELS - * - * It is up to the caller to define how much a debugging level is. - * - * Convention sets 0 as "no debug" (so an action marked as debug level 0 - * will always be taken). The increasing debug levels are used for - * increased verbosity. - * - * USAGE - * - * Group the code in modules and submodules inside each module [which - * in most cases maps to Linux modules and .c files that compose - * those]. - * - * For each module, there is: - * - * - a MODULENAME (single word, legal C identifier) - * - * - a debug-levels.h header file that declares the list of - * submodules and that is included by all .c files that use - * the debugging tools. The file name can be anything. - * - * - some (optional) .c code to manipulate the runtime debug levels - * through debugfs. - * - * The debug-levels.h file would look like: - * - * #ifndef __debug_levels__h__ - * #define __debug_levels__h__ - * - * #define D_MODULENAME modulename - * #define D_MASTER 10 - * - * #include - * - * enum d_module { - * D_SUBMODULE_DECLARE(submodule_1), - * D_SUBMODULE_DECLARE(submodule_2), - * ... - * D_SUBMODULE_DECLARE(submodule_N) - * }; - * - * #endif - * - * D_MASTER is the maximum compile-time debug level; any debug actions - * above this will be out. D_MODULENAME is the module name (legal C - * identifier), which has to be unique for each module (to avoid - * namespace collisions during linkage). Note those #defines need to - * be done before #including debug.h - * - * We declare N different submodules whose debug level can be - * independently controlled during runtime. - * - * In a .c file of the module (and only in one of them), define the - * following code: - * - * struct d_level D_LEVEL[] = { - * D_SUBMODULE_DEFINE(submodule_1), - * D_SUBMODULE_DEFINE(submodule_2), - * ... - * D_SUBMODULE_DEFINE(submodule_N), - * }; - * size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); - * - * Externs for d_level_MODULENAME and d_level_size_MODULENAME are used - * and declared in this file using the D_LEVEL and D_LEVEL_SIZE macros - * #defined also in this file. - * - * To manipulate from user space the levels, create a debugfs dentry - * and then register each submodule with: - * - * d_level_register_debugfs("PREFIX_", submodule_X, parent); - * - * Where PREFIX_ is a name of your chosing. This will create debugfs - * file with a single numeric value that can be use to tweak it. To - * remove the entires, just use debugfs_remove_recursive() on 'parent'. - * - * NOTE: remember that even if this will show attached to some - * particular instance of a device, the settings are *global*. - * - * On each submodule (for example, .c files), the debug infrastructure - * should be included like this: - * - * #define D_SUBMODULE submodule_x // matches one in debug-levels.h - * #include "debug-levels.h" - * - * after #including all your include files. - * - * Now you can use the d_*() macros below [d_test(), d_fnstart(), - * d_fnend(), d_printf(), d_dump()]. - * - * If their debug level is greater than D_MASTER, they will be - * compiled out. - * - * If their debug level is lower or equal than D_MASTER but greater - * than the current debug level of their submodule, they'll be - * ignored. - * - * Otherwise, the action will be performed. - */ -#ifndef __debug__h__ -#define __debug__h__ - -#include -#include - -struct device; - -/* Backend stuff */ - -/* - * Debug backend: generate a message header from a 'struct device' - * - * @head: buffer where to place the header - * @head_size: length of @head - * @dev: pointer to device used to generate a header from. If NULL, - * an empty ("") header is generated. - */ -static inline -void __d_head(char *head, size_t head_size, - struct device *dev) -{ - if (dev == NULL) - head[0] = 0; - else if ((unsigned long)dev < 4096) { - printk(KERN_ERR "E: Corrupt dev %p\n", dev); - WARN_ON(1); - } else - snprintf(head, head_size, "%s %s: ", - dev_driver_string(dev), dev_name(dev)); -} - - -/* - * Debug backend: log some message if debugging is enabled - * - * @l: intended debug level - * @tag: tag to prefix the message with - * @dev: 'struct device' associated to this message - * @f: printf-like format and arguments - * - * Note this is optimized out if it doesn't pass the compile-time - * check; however, it is *always* compiled. This is useful to make - * sure the printf-like formats and variables are always checked and - * they don't get bit rot if you have all the debugging disabled. - */ -#define _d_printf(l, tag, dev, f, a...) \ -do { \ - char head[64]; \ - if (!d_test(l)) \ - break; \ - __d_head(head, sizeof(head), dev); \ - printk(KERN_ERR "%s%s%s: " f, head, __func__, tag, ##a); \ -} while (0) - - -/* - * CPP syntactic sugar to generate A_B like symbol names when one of - * the arguments is a preprocessor #define. - */ -#define __D_PASTE__(varname, modulename) varname##_##modulename -#define __D_PASTE(varname, modulename) (__D_PASTE__(varname, modulename)) -#define _D_SUBMODULE_INDEX(_name) (D_SUBMODULE_DECLARE(_name)) - - -/* - * Store a submodule's runtime debug level and name - */ -struct d_level { - u8 level; - const char *name; -}; - - -/* - * List of available submodules and their debug levels - * - * We call them d_level_MODULENAME and d_level_size_MODULENAME; the - * macros D_LEVEL and D_LEVEL_SIZE contain the name already for - * convenience. - * - * This array and the size are defined on some .c file that is part of - * the current module. - */ -#define D_LEVEL __D_PASTE(d_level, D_MODULENAME) -#define D_LEVEL_SIZE __D_PASTE(d_level_size, D_MODULENAME) - -extern struct d_level D_LEVEL[]; -extern size_t D_LEVEL_SIZE; - - -/* - * Frontend stuff - * - * - * Stuff you need to declare prior to using the actual "debug" actions - * (defined below). - */ - -#ifndef D_MODULENAME -#error D_MODULENAME is not defined in your debug-levels.h file -/** - * D_MODULE - Name of the current module - * - * #define in your module's debug-levels.h, making sure it is - * unique. This has to be a legal C identifier. - */ -#define D_MODULENAME undefined_modulename -#endif - - -#ifndef D_MASTER -#warning D_MASTER not defined, but debug.h included! [see docs] -/** - * D_MASTER - Compile time maximum debug level - * - * #define in your debug-levels.h file to the maximum debug level the - * runtime code will be allowed to have. This allows you to provide a - * main knob. - * - * Anything above that level will be optimized out of the compile. - * - * Defaults to zero (no debug code compiled in). - * - * Maximum one definition per module (at the debug-levels.h file). - */ -#define D_MASTER 0 -#endif - -#ifndef D_SUBMODULE -#error D_SUBMODULE not defined, but debug.h included! [see docs] -/** - * D_SUBMODULE - Name of the current submodule - * - * #define in your submodule .c file before #including debug-levels.h - * to the name of the current submodule as previously declared and - * defined with D_SUBMODULE_DECLARE() (in your module's - * debug-levels.h) and D_SUBMODULE_DEFINE(). - * - * This is used to provide runtime-control over the debug levels. - * - * Maximum one per .c file! Can be shared among different .c files - * (meaning they belong to the same submodule categorization). - */ -#define D_SUBMODULE undefined_module -#endif - - -/** - * D_SUBMODULE_DECLARE - Declare a submodule for runtime debug level control - * - * @_name: name of the submodule, restricted to the chars that make up a - * valid C identifier ([a-zA-Z0-9_]). - * - * Declare in the module's debug-levels.h header file as: - * - * enum d_module { - * D_SUBMODULE_DECLARE(submodule_1), - * D_SUBMODULE_DECLARE(submodule_2), - * D_SUBMODULE_DECLARE(submodule_3), - * }; - * - * Some corresponding .c file needs to have a matching - * D_SUBMODULE_DEFINE(). - */ -#define D_SUBMODULE_DECLARE(_name) __D_SUBMODULE_##_name - - -/** - * D_SUBMODULE_DEFINE - Define a submodule for runtime debug level control - * - * @_name: name of the submodule, restricted to the chars that make up a - * valid C identifier ([a-zA-Z0-9_]). - * - * Use once per module (in some .c file) as: - * - * static - * struct d_level d_level_SUBMODULENAME[] = { - * D_SUBMODULE_DEFINE(submodule_1), - * D_SUBMODULE_DEFINE(submodule_2), - * D_SUBMODULE_DEFINE(submodule_3), - * }; - * size_t d_level_size_SUBDMODULENAME = ARRAY_SIZE(d_level_SUBDMODULENAME); - * - * Matching D_SUBMODULE_DECLARE()s have to be present in a - * debug-levels.h header file. - */ -#define D_SUBMODULE_DEFINE(_name) \ -[__D_SUBMODULE_##_name] = { \ - .level = 0, \ - .name = #_name \ -} - - - -/* The actual "debug" operations */ - - -/** - * d_test - Returns true if debugging should be enabled - * - * @l: intended debug level (unsigned) - * - * If the master debug switch is enabled and the current settings are - * higher or equal to the requested level, then debugging - * output/actions should be enabled. - * - * NOTE: - * - * This needs to be coded so that it can be evaluated in compile - * time; this is why the ugly BUG_ON() is placed in there, so the - * D_MASTER evaluation compiles all out if it is compile-time false. - */ -#define d_test(l) \ -({ \ - unsigned __l = l; /* type enforcer */ \ - (D_MASTER) >= __l \ - && ({ \ - BUG_ON(_D_SUBMODULE_INDEX(D_SUBMODULE) >= D_LEVEL_SIZE);\ - D_LEVEL[_D_SUBMODULE_INDEX(D_SUBMODULE)].level >= __l; \ - }); \ -}) - - -/** - * d_fnstart - log message at function start if debugging enabled - * - * @l: intended debug level - * @_dev: 'struct device' pointer, NULL if none (for context) - * @f: printf-like format and arguments - */ -#define d_fnstart(l, _dev, f, a...) _d_printf(l, " FNSTART", _dev, f, ## a) - - -/** - * d_fnend - log message at function end if debugging enabled - * - * @l: intended debug level - * @_dev: 'struct device' pointer, NULL if none (for context) - * @f: printf-like format and arguments - */ -#define d_fnend(l, _dev, f, a...) _d_printf(l, " FNEND", _dev, f, ## a) - - -/** - * d_printf - log message if debugging enabled - * - * @l: intended debug level - * @_dev: 'struct device' pointer, NULL if none (for context) - * @f: printf-like format and arguments - */ -#define d_printf(l, _dev, f, a...) _d_printf(l, "", _dev, f, ## a) - - -/** - * d_dump - log buffer hex dump if debugging enabled - * - * @l: intended debug level - * @_dev: 'struct device' pointer, NULL if none (for context) - * @f: printf-like format and arguments - */ -#define d_dump(l, dev, ptr, size) \ -do { \ - char head[64]; \ - if (!d_test(l)) \ - break; \ - __d_head(head, sizeof(head), dev); \ - print_hex_dump(KERN_ERR, head, 0, 16, 1, \ - ((void *) ptr), (size), 0); \ -} while (0) - - -/** - * Export a submodule's debug level over debugfs as PREFIXSUBMODULE - * - * @prefix: string to prefix the name with - * @submodule: name of submodule (not a string, just the name) - * @dentry: debugfs parent dentry - * - * For removing, just use debugfs_remove_recursive() on the parent. - */ -#define d_level_register_debugfs(prefix, name, parent) \ -({ \ - debugfs_create_u8( \ - prefix #name, 0600, parent, \ - &(D_LEVEL[__D_SUBMODULE_ ## name].level)); \ -}) - - -static inline -void d_submodule_set(struct d_level *d_level, size_t d_level_size, - const char *submodule, u8 level, const char *tag) -{ - struct d_level *itr, *top; - int index = -1; - - for (itr = d_level, top = itr + d_level_size; itr < top; itr++) { - index++; - if (itr->name == NULL) { - printk(KERN_ERR "%s: itr->name NULL?? (%p, #%d)\n", - tag, itr, index); - continue; - } - if (!strcmp(itr->name, submodule)) { - itr->level = level; - return; - } - } - printk(KERN_ERR "%s: unknown submodule %s\n", tag, submodule); -} - - -/** - * d_parse_params - Parse a string with debug parameters from the - * command line - * - * @d_level: level structure (D_LEVEL) - * @d_level_size: number of items in the level structure - * (D_LEVEL_SIZE). - * @_params: string with the parameters; this is a space (not tab!) - * separated list of NAME:VALUE, where value is the debug level - * and NAME is the name of the submodule. - * @tag: string for error messages (example: MODULE.ARGNAME). - */ -static inline -void d_parse_params(struct d_level *d_level, size_t d_level_size, - const char *_params, const char *tag) -{ - char submodule[130], *params, *params_orig, *token, *colon; - unsigned level, tokens; - - if (_params == NULL) - return; - params_orig = kstrdup(_params, GFP_KERNEL); - params = params_orig; - while (1) { - token = strsep(¶ms, " "); - if (token == NULL) - break; - if (*token == '\0') /* eat joint spaces */ - continue; - /* kernel's sscanf %s eats until whitespace, so we - * replace : by \n so it doesn't get eaten later by - * strsep */ - colon = strchr(token, ':'); - if (colon != NULL) - *colon = '\n'; - tokens = sscanf(token, "%s\n%u", submodule, &level); - if (colon != NULL) - *colon = ':'; /* set back, for error messages */ - if (tokens == 2) - d_submodule_set(d_level, d_level_size, - submodule, level, tag); - else - printk(KERN_ERR "%s: can't parse '%s' as a " - "SUBMODULE:LEVEL (%d tokens)\n", - tag, token, tokens); - } - kfree(params_orig); -} - -#endif /* #ifndef __debug__h__ */ diff --git a/include/net/wimax.h b/include/net/wimax.h deleted file mode 100644 index f6e31d2f47aa..000000000000 --- a/include/net/wimax.h +++ /dev/null @@ -1,503 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Linux WiMAX - * Kernel space API for accessing WiMAX devices - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - * - * The WiMAX stack provides an API for controlling and managing the - * system's WiMAX devices. This API affects the control plane; the - * data plane is accessed via the network stack (netdev). - * - * Parts of the WiMAX stack API and notifications are exported to - * user space via Generic Netlink. In user space, libwimax (part of - * the wimax-tools package) provides a shim layer for accessing those - * calls. - * - * The API is standarized for all WiMAX devices and different drivers - * implement the backend support for it. However, device-specific - * messaging pipes are provided that can be used to issue commands and - * receive notifications in free form. - * - * Currently the messaging pipes are the only means of control as it - * is not known (due to the lack of more devices in the market) what - * will be a good abstraction layer. Expect this to change as more - * devices show in the market. This API is designed to be growable in - * order to address this problem. - * - * USAGE - * - * Embed a `struct wimax_dev` at the beginning of the device's - * private structure, initialize and register it. For details, see - * `struct wimax_dev`s documentation. - * - * Once this is done, wimax-tools's libwimaxll can be used to - * communicate with the driver from user space. You user space - * application does not have to forcibily use libwimaxll and can talk - * the generic netlink protocol directly if desired. - * - * Remember this is a very low level API that will to provide all of - * WiMAX features. Other daemons and services running in user space - * are the expected clients of it. They offer a higher level API that - * applications should use (an example of this is the Intel's WiMAX - * Network Service for the i2400m). - * - * DESIGN - * - * Although not set on final stone, this very basic interface is - * mostly completed. Remember this is meant to grow as new common - * operations are decided upon. New operations will be added to the - * interface, intent being on keeping backwards compatibility as much - * as possible. - * - * This layer implements a set of calls to control a WiMAX device, - * exposing a frontend to the rest of the kernel and user space (via - * generic netlink) and a backend implementation in the driver through - * function pointers. - * - * WiMAX devices have a state, and a kernel-only API allows the - * drivers to manipulate that state. State transitions are atomic, and - * only some of them are allowed (see `enum wimax_st`). - * - * Most API calls will set the state automatically; in most cases - * drivers have to only report state changes due to external - * conditions. - * - * All API operations are 'atomic', serialized through a mutex in the - * `struct wimax_dev`. - * - * EXPORTING TO USER SPACE THROUGH GENERIC NETLINK - * - * The API is exported to user space using generic netlink (other - * methods can be added as needed). - * - * There is a Generic Netlink Family named "WiMAX", where interfaces - * supporting the WiMAX interface receive commands and broadcast their - * signals over a multicast group named "msg". - * - * Mapping to the source/destination interface is done by an interface - * index attribute. - * - * For user-to-kernel traffic (commands) we use a function call - * marshalling mechanism, where a message X with attributes A, B, C - * sent from user space to kernel space means executing the WiMAX API - * call wimax_X(A, B, C), sending the results back as a message. - * - * Kernel-to-user (notifications or signals) communication is sent - * over multicast groups. This allows to have multiple applications - * monitoring them. - * - * Each command/signal gets assigned it's own attribute policy. This - * way the validator will verify that all the attributes in there are - * only the ones that should be for each command/signal. Thing of an - * attribute mapping to a type+argumentname for each command/signal. - * - * If we had a single policy for *all* commands/signals, after running - * the validator we'd have to check "does this attribute belong in - * here"? for each one. It can be done manually, but it's just easier - * to have the validator do that job with multiple policies. As well, - * it makes it easier to later expand each command/signal signature - * without affecting others and keeping the namespace more or less - * sane. Not that it is too complicated, but it makes it even easier. - * - * No state information is maintained in the kernel for each user - * space connection (the connection is stateless). - * - * TESTING FOR THE INTERFACE AND VERSIONING - * - * If network interface X is a WiMAX device, there will be a Generic - * Netlink family named "WiMAX X" and the device will present a - * "wimax" directory in it's network sysfs directory - * (/sys/class/net/DEVICE/wimax) [used by HAL]. - * - * The inexistence of any of these means the device does not support - * this WiMAX API. - * - * By querying the generic netlink controller, versioning information - * and the multicast groups available can be found. Applications using - * the interface can either rely on that or use the generic netlink - * controller to figure out which generic netlink commands/signals are - * supported. - * - * NOTE: this versioning is a last resort to avoid hard - * incompatibilities. It is the intention of the design of this - * stack not to introduce backward incompatible changes. - * - * The version code has to fit in one byte (restrictions imposed by - * generic netlink); we use `version / 10` for the major version and - * `version % 10` for the minor. This gives 9 minors for each major - * and 25 majors. - * - * The version change protocol is as follow: - * - * - Major versions: needs to be increased if an existing message/API - * call is changed or removed. Doesn't need to be changed if a new - * message is added. - * - * - Minor version: needs to be increased if new messages/API calls are - * being added or some other consideration that doesn't impact the - * user-kernel interface too much (like some kind of bug fix) and - * that is kind of left up in the air to common sense. - * - * User space code should not try to work if the major version it was - * compiled for differs from what the kernel offers. As well, if the - * minor version of the kernel interface is lower than the one user - * space is expecting (the one it was compiled for), the kernel - * might be missing API calls; user space shall be ready to handle - * said condition. Use the generic netlink controller operations to - * find which ones are supported and which not. - * - * libwimaxll:wimaxll_open() takes care of checking versions. - * - * THE OPERATIONS: - * - * Each operation is defined in its on file (drivers/net/wimax/op-*.c) - * for clarity. The parts needed for an operation are: - * - * - a function pointer in `struct wimax_dev`: optional, as the - * operation might be implemented by the stack and not by the - * driver. - * - * All function pointers are named wimax_dev->op_*(), and drivers - * must implement them except where noted otherwise. - * - * - When exported to user space, a `struct nla_policy` to define the - * attributes of the generic netlink command and a `struct genl_ops` - * to define the operation. - * - * All the declarations for the operation codes (WIMAX_GNL_OP_) - * and generic netlink attributes (WIMAX_GNL__*) are declared in - * include/linux/wimax.h; this file is intended to be cloned by user - * space to gain access to those declarations. - * - * A few caveats to remember: - * - * - Need to define attribute numbers starting in 1; otherwise it - * fails. - * - * - the `struct genl_family` requires a maximum attribute id; when - * defining the `struct nla_policy` for each message, it has to have - * an array size of WIMAX_GNL_ATTR_MAX+1. - * - * The op_*() function pointers will not be called if the wimax_dev is - * in a state <= %WIMAX_ST_UNINITIALIZED. The exception is: - * - * - op_reset: can be called at any time after wimax_dev_add() has - * been called. - * - * THE PIPE INTERFACE: - * - * This interface is kept intentionally simple. The driver can send - * and receive free-form messages to/from user space through a - * pipe. See drivers/net/wimax/op-msg.c for details. - * - * The kernel-to-user messages are sent with - * wimax_msg(). user-to-kernel messages are delivered via - * wimax_dev->op_msg_from_user(). - * - * RFKILL: - * - * RFKILL support is built into the wimax_dev layer; the driver just - * needs to call wimax_report_rfkill_{hw,sw}() to inform of changes in - * the hardware or software RF kill switches. When the stack wants to - * turn the radio off, it will call wimax_dev->op_rfkill_sw_toggle(), - * which the driver implements. - * - * User space can set the software RF Kill switch by calling - * wimax_rfkill(). - * - * The code for now only supports devices that don't require polling; - * If the device needs to be polled, create a self-rearming delayed - * work struct for polling or look into adding polled support to the - * WiMAX stack. - * - * When initializing the hardware (_probe), after calling - * wimax_dev_add(), query the device for it's RF Kill switches status - * and feed it back to the WiMAX stack using - * wimax_report_rfkill_{hw,sw}(). If any switch is missing, always - * report it as ON. - * - * NOTE: the wimax stack uses an inverted terminology to that of the - * RFKILL subsystem: - * - * - ON: radio is ON, RFKILL is DISABLED or OFF. - * - OFF: radio is OFF, RFKILL is ENABLED or ON. - * - * MISCELLANEOUS OPS: - * - * wimax_reset() can be used to reset the device to power on state; by - * default it issues a warm reset that maintains the same device - * node. If that is not possible, it falls back to a cold reset - * (device reconnect). The driver implements the backend to this - * through wimax_dev->op_reset(). - */ - -#ifndef __NET__WIMAX_H__ -#define __NET__WIMAX_H__ - -#include -#include -#include - -struct net_device; -struct genl_info; -struct wimax_dev; - -/** - * struct wimax_dev - Generic WiMAX device - * - * @net_dev: [fill] Pointer to the &struct net_device this WiMAX - * device implements. - * - * @op_msg_from_user: [fill] Driver-specific operation to - * handle a raw message from user space to the driver. The - * driver can send messages to user space using with - * wimax_msg_to_user(). - * - * @op_rfkill_sw_toggle: [fill] Driver-specific operation to act on - * userspace (or any other agent) requesting the WiMAX device to - * change the RF Kill software switch (WIMAX_RF_ON or - * WIMAX_RF_OFF). - * If such hardware support is not present, it is assumed the - * radio cannot be switched off and it is always on (and the stack - * will error out when trying to switch it off). In such case, - * this function pointer can be left as NULL. - * - * @op_reset: [fill] Driver specific operation to reset the - * device. - * This operation should always attempt first a warm reset that - * does not disconnect the device from the bus and return 0. - * If that fails, it should resort to some sort of cold or bus - * reset (even if it implies a bus disconnection and device - * disappearance). In that case, -ENODEV should be returned to - * indicate the device is gone. - * This operation has to be synchronous, and return only when the - * reset is complete. In case of having had to resort to bus/cold - * reset implying a device disconnection, the call is allowed to - * return immediately. - * NOTE: wimax_dev->mutex is NOT locked when this op is being - * called; however, wimax_dev->mutex_reset IS locked to ensure - * serialization of calls to wimax_reset(). - * See wimax_reset()'s documentation. - * - * @name: [fill] A way to identify this device. We need to register a - * name with many subsystems (rfkill, workqueue creation, etc). - * We can't use the network device name as that - * might change and in some instances we don't know it yet (until - * we don't call register_netdev()). So we generate an unique one - * using the driver name and device bus id, place it here and use - * it across the board. Recommended naming: - * DRIVERNAME-BUSNAME:BUSID (dev->bus->name, dev->bus_id). - * - * @id_table_node: [private] link to the list of wimax devices kept by - * id-table.c. Protected by it's own spinlock. - * - * @mutex: [private] Serializes all concurrent access and execution of - * operations. - * - * @mutex_reset: [private] Serializes reset operations. Needs to be a - * different mutex because as part of the reset operation, the - * driver has to call back into the stack to do things such as - * state change, that require wimax_dev->mutex. - * - * @state: [private] Current state of the WiMAX device. - * - * @rfkill: [private] integration into the RF-Kill infrastructure. - * - * @rf_sw: [private] State of the software radio switch (OFF/ON) - * - * @rf_hw: [private] State of the hardware radio switch (OFF/ON) - * - * @debugfs_dentry: [private] Used to hook up a debugfs entry. This - * shows up in the debugfs root as wimax\:DEVICENAME. - * - * Description: - * This structure defines a common interface to access all WiMAX - * devices from different vendors and provides a common API as well as - * a free-form device-specific messaging channel. - * - * Usage: - * 1. Embed a &struct wimax_dev at *the beginning* the network - * device structure so that netdev_priv() points to it. - * - * 2. memset() it to zero - * - * 3. Initialize with wimax_dev_init(). This will leave the WiMAX - * device in the %__WIMAX_ST_NULL state. - * - * 4. Fill all the fields marked with [fill]; once called - * wimax_dev_add(), those fields CANNOT be modified. - * - * 5. Call wimax_dev_add() *after* registering the network - * device. This will leave the WiMAX device in the %WIMAX_ST_DOWN - * state. - * Protect the driver's net_device->open() against succeeding if - * the wimax device state is lower than %WIMAX_ST_DOWN. - * - * 6. Select when the device is going to be turned on/initialized; - * for example, it could be initialized on 'ifconfig up' (when the - * netdev op 'open()' is called on the driver). - * - * When the device is initialized (at `ifconfig up` time, or right - * after calling wimax_dev_add() from _probe(), make sure the - * following steps are taken - * - * a. Move the device to %WIMAX_ST_UNINITIALIZED. This is needed so - * some API calls that shouldn't work until the device is ready - * can be blocked. - * - * b. Initialize the device. Make sure to turn the SW radio switch - * off and move the device to state %WIMAX_ST_RADIO_OFF when - * done. When just initialized, a device should be left in RADIO - * OFF state until user space devices to turn it on. - * - * c. Query the device for the state of the hardware rfkill switch - * and call wimax_rfkill_report_hw() and wimax_rfkill_report_sw() - * as needed. See below. - * - * wimax_dev_rm() undoes before unregistering the network device. Once - * wimax_dev_add() is called, the driver can get called on the - * wimax_dev->op_* function pointers - * - * CONCURRENCY: - * - * The stack provides a mutex for each device that will disallow API - * calls happening concurrently; thus, op calls into the driver - * through the wimax_dev->op*() function pointers will always be - * serialized and *never* concurrent. - * - * For locking, take wimax_dev->mutex is taken; (most) operations in - * the API have to check for wimax_dev_is_ready() to return 0 before - * continuing (this is done internally). - * - * REFERENCE COUNTING: - * - * The WiMAX device is reference counted by the associated network - * device. The only operation that can be used to reference the device - * is wimax_dev_get_by_genl_info(), and the reference it acquires has - * to be released with dev_put(wimax_dev->net_dev). - * - * RFKILL: - * - * At startup, both HW and SW radio switchess are assumed to be off. - * - * At initialization time [after calling wimax_dev_add()], have the - * driver query the device for the status of the software and hardware - * RF kill switches and call wimax_report_rfkill_hw() and - * wimax_rfkill_report_sw() to indicate their state. If any is - * missing, just call it to indicate it is ON (radio always on). - * - * Whenever the driver detects a change in the state of the RF kill - * switches, it should call wimax_report_rfkill_hw() or - * wimax_report_rfkill_sw() to report it to the stack. - */ -struct wimax_dev { - struct net_device *net_dev; - struct list_head id_table_node; - struct mutex mutex; /* Protects all members and API calls */ - struct mutex mutex_reset; - enum wimax_st state; - - int (*op_msg_from_user)(struct wimax_dev *wimax_dev, - const char *, - const void *, size_t, - const struct genl_info *info); - int (*op_rfkill_sw_toggle)(struct wimax_dev *wimax_dev, - enum wimax_rf_state); - int (*op_reset)(struct wimax_dev *wimax_dev); - - struct rfkill *rfkill; - unsigned int rf_hw; - unsigned int rf_sw; - char name[32]; - - struct dentry *debugfs_dentry; -}; - - - -/* - * WiMAX stack public API for device drivers - * ----------------------------------------- - * - * These functions are not exported to user space. - */ -void wimax_dev_init(struct wimax_dev *); -int wimax_dev_add(struct wimax_dev *, struct net_device *); -void wimax_dev_rm(struct wimax_dev *); - -static inline -struct wimax_dev *net_dev_to_wimax(struct net_device *net_dev) -{ - return netdev_priv(net_dev); -} - -static inline -struct device *wimax_dev_to_dev(struct wimax_dev *wimax_dev) -{ - return wimax_dev->net_dev->dev.parent; -} - -void wimax_state_change(struct wimax_dev *, enum wimax_st); -enum wimax_st wimax_state_get(struct wimax_dev *); - -/* - * Radio Switch state reporting. - * - * enum wimax_rf_state is declared in linux/wimax.h so the exports - * to user space can use it. - */ -void wimax_report_rfkill_hw(struct wimax_dev *, enum wimax_rf_state); -void wimax_report_rfkill_sw(struct wimax_dev *, enum wimax_rf_state); - - -/* - * Free-form messaging to/from user space - * - * Sending a message: - * - * wimax_msg(wimax_dev, pipe_name, buf, buf_size, GFP_KERNEL); - * - * Broken up: - * - * skb = wimax_msg_alloc(wimax_dev, pipe_name, buf_size, GFP_KERNEL); - * ...fill up skb... - * wimax_msg_send(wimax_dev, pipe_name, skb); - * - * Be sure not to modify skb->data in the middle (ie: don't use - * skb_push()/skb_pull()/skb_reserve() on the skb). - * - * "pipe_name" is any string, that can be interpreted as the name of - * the pipe or recipient; the interpretation of it is driver - * specific, so the recipient can multiplex it as wished. It can be - * NULL, it won't be used - an example is using a "diagnostics" tag to - * send diagnostics information that a device-specific diagnostics - * tool would be interested in. - */ -struct sk_buff *wimax_msg_alloc(struct wimax_dev *, const char *, const void *, - size_t, gfp_t); -int wimax_msg_send(struct wimax_dev *, struct sk_buff *); -int wimax_msg(struct wimax_dev *, const char *, const void *, size_t, gfp_t); - -const void *wimax_msg_data_len(struct sk_buff *, size_t *); -const void *wimax_msg_data(struct sk_buff *); -ssize_t wimax_msg_len(struct sk_buff *); - - -/* - * WiMAX stack user space API - * -------------------------- - * - * This API is what gets exported to user space for general - * operations. As well, they can be called from within the kernel, - * (with a properly referenced `struct wimax_dev`). - * - * Properly referenced means: the 'struct net_device' that embeds the - * device's control structure and (as such) the 'struct wimax_dev' is - * referenced by the caller. - */ -int wimax_rfkill(struct wimax_dev *, enum wimax_rf_state); -int wimax_reset(struct wimax_dev *); - -#endif /* #ifndef __NET__WIMAX_H__ */ diff --git a/include/uapi/linux/wimax.h b/include/uapi/linux/wimax.h deleted file mode 100644 index 9f6b77af2f6d..000000000000 --- a/include/uapi/linux/wimax.h +++ /dev/null @@ -1,239 +0,0 @@ -/* - * Linux WiMax - * API for user space - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Inaky Perez-Gonzalez - * - Initial implementation - * - * - * This file declares the user/kernel protocol that is spoken over - * Generic Netlink, as well as any type declaration that is to be used - * by kernel and user space. - * - * It is intended for user space to clone it verbatim to use it as a - * primary reference for definitions. - * - * Stuff intended for kernel usage as well as full protocol and stack - * documentation is rooted in include/net/wimax.h. - */ - -#ifndef __LINUX__WIMAX_H__ -#define __LINUX__WIMAX_H__ - -#include - -enum { - /** - * Version of the interface (unsigned decimal, MMm, max 25.5) - * M - Major: change if removing or modifying an existing call. - * m - minor: change when adding a new call - */ - WIMAX_GNL_VERSION = 01, - /* Generic NetLink attributes */ - WIMAX_GNL_ATTR_INVALID = 0x00, - WIMAX_GNL_ATTR_MAX = 10, -}; - - -/* - * Generic NetLink operations - * - * Most of these map to an API call; _OP_ stands for operation, _RP_ - * for reply and _RE_ for report (aka: signal). - */ -enum { - WIMAX_GNL_OP_MSG_FROM_USER, /* User to kernel message */ - WIMAX_GNL_OP_MSG_TO_USER, /* Kernel to user message */ - WIMAX_GNL_OP_RFKILL, /* Run wimax_rfkill() */ - WIMAX_GNL_OP_RESET, /* Run wimax_rfkill() */ - WIMAX_GNL_RE_STATE_CHANGE, /* Report: status change */ - WIMAX_GNL_OP_STATE_GET, /* Request for current state */ -}; - - -/* Message from user / to user */ -enum { - WIMAX_GNL_MSG_IFIDX = 1, - WIMAX_GNL_MSG_PIPE_NAME, - WIMAX_GNL_MSG_DATA, -}; - - -/* - * wimax_rfkill() - * - * The state of the radio (ON/OFF) is mapped to the rfkill subsystem's - * switch state (DISABLED/ENABLED). - */ -enum wimax_rf_state { - WIMAX_RF_OFF = 0, /* Radio is off, rfkill on/enabled */ - WIMAX_RF_ON = 1, /* Radio is on, rfkill off/disabled */ - WIMAX_RF_QUERY = 2, -}; - -/* Attributes */ -enum { - WIMAX_GNL_RFKILL_IFIDX = 1, - WIMAX_GNL_RFKILL_STATE, -}; - - -/* Attributes for wimax_reset() */ -enum { - WIMAX_GNL_RESET_IFIDX = 1, -}; - -/* Attributes for wimax_state_get() */ -enum { - WIMAX_GNL_STGET_IFIDX = 1, -}; - -/* - * Attributes for the Report State Change - * - * For now we just have the old and new states; new attributes might - * be added later on. - */ -enum { - WIMAX_GNL_STCH_IFIDX = 1, - WIMAX_GNL_STCH_STATE_OLD, - WIMAX_GNL_STCH_STATE_NEW, -}; - - -/** - * enum wimax_st - The different states of a WiMAX device - * @__WIMAX_ST_NULL: The device structure has been allocated and zeroed, - * but still wimax_dev_add() hasn't been called. There is no state. - * - * @WIMAX_ST_DOWN: The device has been registered with the WiMAX and - * networking stacks, but it is not initialized (normally that is - * done with 'ifconfig DEV up' [or equivalent], which can upload - * firmware and enable communications with the device). - * In this state, the device is powered down and using as less - * power as possible. - * This state is the default after a call to wimax_dev_add(). It - * is ok to have drivers move directly to %WIMAX_ST_UNINITIALIZED - * or %WIMAX_ST_RADIO_OFF in _probe() after the call to - * wimax_dev_add(). - * It is recommended that the driver leaves this state when - * calling 'ifconfig DEV up' and enters it back on 'ifconfig DEV - * down'. - * - * @__WIMAX_ST_QUIESCING: The device is being torn down, so no API - * operations are allowed to proceed except the ones needed to - * complete the device clean up process. - * - * @WIMAX_ST_UNINITIALIZED: [optional] Communication with the device - * is setup, but the device still requires some configuration - * before being operational. - * Some WiMAX API calls might work. - * - * @WIMAX_ST_RADIO_OFF: The device is fully up; radio is off (wether - * by hardware or software switches). - * It is recommended to always leave the device in this state - * after initialization. - * - * @WIMAX_ST_READY: The device is fully up and radio is on. - * - * @WIMAX_ST_SCANNING: [optional] The device has been instructed to - * scan. In this state, the device cannot be actively connected to - * a network. - * - * @WIMAX_ST_CONNECTING: The device is connecting to a network. This - * state exists because in some devices, the connect process can - * include a number of negotiations between user space, kernel - * space and the device. User space needs to know what the device - * is doing. If the connect sequence in a device is atomic and - * fast, the device can transition directly to CONNECTED - * - * @WIMAX_ST_CONNECTED: The device is connected to a network. - * - * @__WIMAX_ST_INVALID: This is an invalid state used to mark the - * maximum numeric value of states. - * - * Description: - * - * Transitions from one state to another one are atomic and can only - * be caused in kernel space with wimax_state_change(). To read the - * state, use wimax_state_get(). - * - * States starting with __ are internal and shall not be used or - * referred to by drivers or userspace. They look ugly, but that's the - * point -- if any use is made non-internal to the stack, it is easier - * to catch on review. - * - * All API operations [with well defined exceptions] will take the - * device mutex before starting and then check the state. If the state - * is %__WIMAX_ST_NULL, %WIMAX_ST_DOWN, %WIMAX_ST_UNINITIALIZED or - * %__WIMAX_ST_QUIESCING, it will drop the lock and quit with - * -%EINVAL, -%ENOMEDIUM, -%ENOTCONN or -%ESHUTDOWN. - * - * The order of the definitions is important, so we can do numerical - * comparisons (eg: < %WIMAX_ST_RADIO_OFF means the device is not ready - * to operate). - */ -/* - * The allowed state transitions are described in the table below - * (states in rows can go to states in columns where there is an X): - * - * UNINI RADIO READY SCAN CONNEC CONNEC - * NULL DOWN QUIESCING TIALIZED OFF NING TING TED - * NULL - x - * DOWN - x x x - * QUIESCING x - - * UNINITIALIZED x - x - * RADIO_OFF x - x - * READY x x - x x x - * SCANNING x x x - x x - * CONNECTING x x x x - x - * CONNECTED x x x - - * - * This table not available in kernel-doc because the formatting messes it up. - */ - enum wimax_st { - __WIMAX_ST_NULL = 0, - WIMAX_ST_DOWN, - __WIMAX_ST_QUIESCING, - WIMAX_ST_UNINITIALIZED, - WIMAX_ST_RADIO_OFF, - WIMAX_ST_READY, - WIMAX_ST_SCANNING, - WIMAX_ST_CONNECTING, - WIMAX_ST_CONNECTED, - __WIMAX_ST_INVALID /* Always keep last */ -}; - - -#endif /* #ifndef __LINUX__WIMAX_H__ */ diff --git a/include/uapi/linux/wimax/i2400m.h b/include/uapi/linux/wimax/i2400m.h deleted file mode 100644 index fd198bc24a3c..000000000000 --- a/include/uapi/linux/wimax/i2400m.h +++ /dev/null @@ -1,572 +0,0 @@ -/* - * Intel Wireless WiMax Connection 2400m - * Host-Device protocol interface definitions - * - * - * Copyright (C) 2007-2008 Intel Corporation. All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * - * Intel Corporation - * Inaky Perez-Gonzalez - * - Initial implementation - * - * - * This header defines the data structures and constants used to - * communicate with the device. - * - * BOOTMODE/BOOTROM/FIRMWARE UPLOAD PROTOCOL - * - * The firmware upload protocol is quite simple and only requires a - * handful of commands. See drivers/net/wimax/i2400m/fw.c for more - * details. - * - * The BCF data structure is for the firmware file header. - * - * - * THE DATA / CONTROL PROTOCOL - * - * This is the normal protocol spoken with the device once the - * firmware is uploaded. It transports data payloads and control - * messages back and forth. - * - * It consists 'messages' that pack one or more payloads each. The - * format is described in detail in drivers/net/wimax/i2400m/rx.c and - * tx.c. - * - * - * THE L3L4 PROTOCOL - * - * The term L3L4 refers to Layer 3 (the device), Layer 4 (the - * driver/host software). - * - * This is the control protocol used by the host to control the i2400m - * device (scan, connect, disconnect...). This is sent to / received - * as control frames. These frames consist of a header and zero or - * more TLVs with information. We call each control frame a "message". - * - * Each message is composed of: - * - * HEADER - * [TLV0 + PAYLOAD0] - * [TLV1 + PAYLOAD1] - * [...] - * [TLVN + PAYLOADN] - * - * The HEADER is defined by 'struct i2400m_l3l4_hdr'. The payloads are - * defined by a TLV structure (Type Length Value) which is a 'header' - * (struct i2400m_tlv_hdr) and then the payload. - * - * All integers are represented as Little Endian. - * - * - REQUESTS AND EVENTS - * - * The requests can be clasified as follows: - * - * COMMAND: implies a request from the host to the device requesting - * an action being performed. The device will reply with a - * message (with the same type as the command), status and - * no (TLV) payload. Execution of a command might cause - * events (of different type) to be sent later on as - * device's state changes. - * - * GET/SET: similar to COMMAND, but will not cause other - * EVENTs. The reply, in the case of GET, will contain - * TLVs with the requested information. - * - * EVENT: asynchronous messages sent from the device, maybe as a - * consequence of previous COMMANDs but disassociated from - * them. - * - * Only one request might be pending at the same time (ie: don't - * parallelize nor post another GET request before the previous - * COMMAND has been acknowledged with it's corresponding reply by the - * device). - * - * The different requests and their formats are described below: - * - * I2400M_MT_* Message types - * I2400M_MS_* Message status (for replies, events) - * i2400m_tlv_* TLVs - * - * data types are named 'struct i2400m_msg_OPNAME', OPNAME matching the - * operation. - */ - -#ifndef __LINUX__WIMAX__I2400M_H__ -#define __LINUX__WIMAX__I2400M_H__ - -#include -#include - -/* - * Host Device Interface (HDI) common to all busses - */ - -/* Boot-mode (firmware upload mode) commands */ - -/* Header for the firmware file */ -struct i2400m_bcf_hdr { - __le32 module_type; - __le32 header_len; - __le32 header_version; - __le32 module_id; - __le32 module_vendor; - __le32 date; /* BCD YYYMMDD */ - __le32 size; /* in dwords */ - __le32 key_size; /* in dwords */ - __le32 modulus_size; /* in dwords */ - __le32 exponent_size; /* in dwords */ - __u8 reserved[88]; -} __attribute__ ((packed)); - -/* Boot mode opcodes */ -enum i2400m_brh_opcode { - I2400M_BRH_READ = 1, - I2400M_BRH_WRITE = 2, - I2400M_BRH_JUMP = 3, - I2400M_BRH_SIGNED_JUMP = 8, - I2400M_BRH_HASH_PAYLOAD_ONLY = 9, -}; - -/* Boot mode command masks and stuff */ -enum i2400m_brh { - I2400M_BRH_SIGNATURE = 0xcbbc0000, - I2400M_BRH_SIGNATURE_MASK = 0xffff0000, - I2400M_BRH_SIGNATURE_SHIFT = 16, - I2400M_BRH_OPCODE_MASK = 0x0000000f, - I2400M_BRH_RESPONSE_MASK = 0x000000f0, - I2400M_BRH_RESPONSE_SHIFT = 4, - I2400M_BRH_DIRECT_ACCESS = 0x00000400, - I2400M_BRH_RESPONSE_REQUIRED = 0x00000200, - I2400M_BRH_USE_CHECKSUM = 0x00000100, -}; - - -/** - * i2400m_bootrom_header - Header for a boot-mode command - * - * @cmd: the above command descriptor - * @target_addr: where on the device memory should the action be performed. - * @data_size: for read/write, amount of data to be read/written - * @block_checksum: checksum value (if applicable) - * @payload: the beginning of data attached to this header - */ -struct i2400m_bootrom_header { - __le32 command; /* Compose with enum i2400_brh */ - __le32 target_addr; - __le32 data_size; - __le32 block_checksum; - char payload[0]; -} __attribute__ ((packed)); - - -/* - * Data / control protocol - */ - -/* Packet types for the host-device interface */ -enum i2400m_pt { - I2400M_PT_DATA = 0, - I2400M_PT_CTRL, - I2400M_PT_TRACE, /* For device debug */ - I2400M_PT_RESET_WARM, /* device reset */ - I2400M_PT_RESET_COLD, /* USB[transport] reset, like reconnect */ - I2400M_PT_EDATA, /* Extended RX data */ - I2400M_PT_ILLEGAL -}; - - -/* - * Payload for a data packet - * - * This is prefixed to each and every outgoing DATA type. - */ -struct i2400m_pl_data_hdr { - __le32 reserved; -} __attribute__((packed)); - - -/* - * Payload for an extended data packet - * - * New in fw v1.4 - * - * @reorder: if this payload has to be reorder or not (and how) - * @cs: the type of data in the packet, as defined per (802.16e - * T11.13.19.1). Currently only 2 (IPv4 packet) supported. - * - * This is prefixed to each and every INCOMING DATA packet. - */ -struct i2400m_pl_edata_hdr { - __le32 reorder; /* bits defined in i2400m_ro */ - __u8 cs; - __u8 reserved[11]; -} __attribute__((packed)); - -enum i2400m_cs { - I2400M_CS_IPV4_0 = 0, - I2400M_CS_IPV4 = 2, -}; - -enum i2400m_ro { - I2400M_RO_NEEDED = 0x01, - I2400M_RO_TYPE = 0x03, - I2400M_RO_TYPE_SHIFT = 1, - I2400M_RO_CIN = 0x0f, - I2400M_RO_CIN_SHIFT = 4, - I2400M_RO_FBN = 0x07ff, - I2400M_RO_FBN_SHIFT = 8, - I2400M_RO_SN = 0x07ff, - I2400M_RO_SN_SHIFT = 21, -}; - -enum i2400m_ro_type { - I2400M_RO_TYPE_RESET = 0, - I2400M_RO_TYPE_PACKET, - I2400M_RO_TYPE_WS, - I2400M_RO_TYPE_PACKET_WS, -}; - - -/* Misc constants */ -enum { - I2400M_PL_ALIGN = 16, /* Payload data size alignment */ - I2400M_PL_SIZE_MAX = 0x3EFF, - I2400M_MAX_PLS_IN_MSG = 60, - /* protocol barkers: sync sequences; for notifications they - * are sent in groups of four. */ - I2400M_H2D_PREVIEW_BARKER = 0xcafe900d, - I2400M_COLD_RESET_BARKER = 0xc01dc01d, - I2400M_WARM_RESET_BARKER = 0x50f750f7, - I2400M_NBOOT_BARKER = 0xdeadbeef, - I2400M_SBOOT_BARKER = 0x0ff1c1a1, - I2400M_SBOOT_BARKER_6050 = 0x80000001, - I2400M_ACK_BARKER = 0xfeedbabe, - I2400M_D2H_MSG_BARKER = 0xbeefbabe, -}; - - -/* - * Hardware payload descriptor - * - * Bitfields encoded in a struct to enforce typing semantics. - * - * Look in rx.c and tx.c for a full description of the format. - */ -struct i2400m_pld { - __le32 val; -} __attribute__ ((packed)); - -#define I2400M_PLD_SIZE_MASK 0x00003fff -#define I2400M_PLD_TYPE_SHIFT 16 -#define I2400M_PLD_TYPE_MASK 0x000f0000 - -/* - * Header for a TX message or RX message - * - * @barker: preamble - * @size: used for management of the FIFO queue buffer; before - * sending, this is converted to be a real preamble. This - * indicates the real size of the TX message that starts at this - * point. If the highest bit is set, then this message is to be - * skipped. - * @sequence: sequence number of this message - * @offset: offset where the message itself starts -- see the comments - * in the file header about message header and payload descriptor - * alignment. - * @num_pls: number of payloads in this message - * @padding: amount of padding bytes at the end of the message to make - * it be of block-size aligned - * - * Look in rx.c and tx.c for a full description of the format. - */ -struct i2400m_msg_hdr { - union { - __le32 barker; - __u32 size; /* same size type as barker!! */ - }; - union { - __le32 sequence; - __u32 offset; /* same size type as barker!! */ - }; - __le16 num_pls; - __le16 rsv1; - __le16 padding; - __le16 rsv2; - struct i2400m_pld pld[0]; -} __attribute__ ((packed)); - - - -/* - * L3/L4 control protocol - */ - -enum { - /* Interface version */ - I2400M_L3L4_VERSION = 0x0100, -}; - -/* Message types */ -enum i2400m_mt { - I2400M_MT_RESERVED = 0x0000, - I2400M_MT_INVALID = 0xffff, - I2400M_MT_REPORT_MASK = 0x8000, - - I2400M_MT_GET_SCAN_RESULT = 0x4202, - I2400M_MT_SET_SCAN_PARAM = 0x4402, - I2400M_MT_CMD_RF_CONTROL = 0x4602, - I2400M_MT_CMD_SCAN = 0x4603, - I2400M_MT_CMD_CONNECT = 0x4604, - I2400M_MT_CMD_DISCONNECT = 0x4605, - I2400M_MT_CMD_EXIT_IDLE = 0x4606, - I2400M_MT_GET_LM_VERSION = 0x5201, - I2400M_MT_GET_DEVICE_INFO = 0x5202, - I2400M_MT_GET_LINK_STATUS = 0x5203, - I2400M_MT_GET_STATISTICS = 0x5204, - I2400M_MT_GET_STATE = 0x5205, - I2400M_MT_GET_MEDIA_STATUS = 0x5206, - I2400M_MT_SET_INIT_CONFIG = 0x5404, - I2400M_MT_CMD_INIT = 0x5601, - I2400M_MT_CMD_TERMINATE = 0x5602, - I2400M_MT_CMD_MODE_OF_OP = 0x5603, - I2400M_MT_CMD_RESET_DEVICE = 0x5604, - I2400M_MT_CMD_MONITOR_CONTROL = 0x5605, - I2400M_MT_CMD_ENTER_POWERSAVE = 0x5606, - I2400M_MT_GET_TLS_OPERATION_RESULT = 0x6201, - I2400M_MT_SET_EAP_SUCCESS = 0x6402, - I2400M_MT_SET_EAP_FAIL = 0x6403, - I2400M_MT_SET_EAP_KEY = 0x6404, - I2400M_MT_CMD_SEND_EAP_RESPONSE = 0x6602, - I2400M_MT_REPORT_SCAN_RESULT = 0xc002, - I2400M_MT_REPORT_STATE = 0xd002, - I2400M_MT_REPORT_POWERSAVE_READY = 0xd005, - I2400M_MT_REPORT_EAP_REQUEST = 0xe002, - I2400M_MT_REPORT_EAP_RESTART = 0xe003, - I2400M_MT_REPORT_ALT_ACCEPT = 0xe004, - I2400M_MT_REPORT_KEY_REQUEST = 0xe005, -}; - - -/* - * Message Ack Status codes - * - * When a message is replied-to, this status is reported. - */ -enum i2400m_ms { - I2400M_MS_DONE_OK = 0, - I2400M_MS_DONE_IN_PROGRESS = 1, - I2400M_MS_INVALID_OP = 2, - I2400M_MS_BAD_STATE = 3, - I2400M_MS_ILLEGAL_VALUE = 4, - I2400M_MS_MISSING_PARAMS = 5, - I2400M_MS_VERSION_ERROR = 6, - I2400M_MS_ACCESSIBILITY_ERROR = 7, - I2400M_MS_BUSY = 8, - I2400M_MS_CORRUPTED_TLV = 9, - I2400M_MS_UNINITIALIZED = 10, - I2400M_MS_UNKNOWN_ERROR = 11, - I2400M_MS_PRODUCTION_ERROR = 12, - I2400M_MS_NO_RF = 13, - I2400M_MS_NOT_READY_FOR_POWERSAVE = 14, - I2400M_MS_THERMAL_CRITICAL = 15, - I2400M_MS_MAX -}; - - -/** - * i2400m_tlv - enumeration of the different types of TLVs - * - * TLVs stand for type-length-value and are the header for a payload - * composed of almost anything. Each payload has a type assigned - * and a length. - */ -enum i2400m_tlv { - I2400M_TLV_L4_MESSAGE_VERSIONS = 129, - I2400M_TLV_SYSTEM_STATE = 141, - I2400M_TLV_MEDIA_STATUS = 161, - I2400M_TLV_RF_OPERATION = 162, - I2400M_TLV_RF_STATUS = 163, - I2400M_TLV_DEVICE_RESET_TYPE = 132, - I2400M_TLV_CONFIG_IDLE_PARAMETERS = 601, - I2400M_TLV_CONFIG_IDLE_TIMEOUT = 611, - I2400M_TLV_CONFIG_D2H_DATA_FORMAT = 614, - I2400M_TLV_CONFIG_DL_HOST_REORDER = 615, -}; - - -struct i2400m_tlv_hdr { - __le16 type; - __le16 length; /* payload's */ - __u8 pl[0]; -} __attribute__((packed)); - - -struct i2400m_l3l4_hdr { - __le16 type; - __le16 length; /* payload's */ - __le16 version; - __le16 resv1; - __le16 status; - __le16 resv2; - struct i2400m_tlv_hdr pl[0]; -} __attribute__((packed)); - - -/** - * i2400m_system_state - different states of the device - */ -enum i2400m_system_state { - I2400M_SS_UNINITIALIZED = 1, - I2400M_SS_INIT, - I2400M_SS_READY, - I2400M_SS_SCAN, - I2400M_SS_STANDBY, - I2400M_SS_CONNECTING, - I2400M_SS_WIMAX_CONNECTED, - I2400M_SS_DATA_PATH_CONNECTED, - I2400M_SS_IDLE, - I2400M_SS_DISCONNECTING, - I2400M_SS_OUT_OF_ZONE, - I2400M_SS_SLEEPACTIVE, - I2400M_SS_PRODUCTION, - I2400M_SS_CONFIG, - I2400M_SS_RF_OFF, - I2400M_SS_RF_SHUTDOWN, - I2400M_SS_DEVICE_DISCONNECT, - I2400M_SS_MAX, -}; - - -/** - * i2400m_tlv_system_state - report on the state of the system - * - * @state: see enum i2400m_system_state - */ -struct i2400m_tlv_system_state { - struct i2400m_tlv_hdr hdr; - __le32 state; -} __attribute__((packed)); - - -struct i2400m_tlv_l4_message_versions { - struct i2400m_tlv_hdr hdr; - __le16 major; - __le16 minor; - __le16 branch; - __le16 reserved; -} __attribute__((packed)); - - -struct i2400m_tlv_detailed_device_info { - struct i2400m_tlv_hdr hdr; - __u8 reserved1[400]; - __u8 mac_address[ETH_ALEN]; - __u8 reserved2[2]; -} __attribute__((packed)); - - -enum i2400m_rf_switch_status { - I2400M_RF_SWITCH_ON = 1, - I2400M_RF_SWITCH_OFF = 2, -}; - -struct i2400m_tlv_rf_switches_status { - struct i2400m_tlv_hdr hdr; - __u8 sw_rf_switch; /* 1 ON, 2 OFF */ - __u8 hw_rf_switch; /* 1 ON, 2 OFF */ - __u8 reserved[2]; -} __attribute__((packed)); - - -enum { - i2400m_rf_operation_on = 1, - i2400m_rf_operation_off = 2 -}; - -struct i2400m_tlv_rf_operation { - struct i2400m_tlv_hdr hdr; - __le32 status; /* 1 ON, 2 OFF */ -} __attribute__((packed)); - - -enum i2400m_tlv_reset_type { - I2400M_RESET_TYPE_COLD = 1, - I2400M_RESET_TYPE_WARM -}; - -struct i2400m_tlv_device_reset_type { - struct i2400m_tlv_hdr hdr; - __le32 reset_type; -} __attribute__((packed)); - - -struct i2400m_tlv_config_idle_parameters { - struct i2400m_tlv_hdr hdr; - __le32 idle_timeout; /* 100 to 300000 ms [5min], 100 increments - * 0 disabled */ - __le32 idle_paging_interval; /* frames */ -} __attribute__((packed)); - - -enum i2400m_media_status { - I2400M_MEDIA_STATUS_LINK_UP = 1, - I2400M_MEDIA_STATUS_LINK_DOWN, - I2400M_MEDIA_STATUS_LINK_RENEW, -}; - -struct i2400m_tlv_media_status { - struct i2400m_tlv_hdr hdr; - __le32 media_status; -} __attribute__((packed)); - - -/* New in v1.4 */ -struct i2400m_tlv_config_idle_timeout { - struct i2400m_tlv_hdr hdr; - __le32 timeout; /* 100 to 300000 ms [5min], 100 increments - * 0 disabled */ -} __attribute__((packed)); - -/* New in v1.4 -- for backward compat, will be removed */ -struct i2400m_tlv_config_d2h_data_format { - struct i2400m_tlv_hdr hdr; - __u8 format; /* 0 old format, 1 enhanced */ - __u8 reserved[3]; -} __attribute__((packed)); - -/* New in v1.4 */ -struct i2400m_tlv_config_dl_host_reorder { - struct i2400m_tlv_hdr hdr; - __u8 reorder; /* 0 disabled, 1 enabled */ - __u8 reserved[3]; -} __attribute__((packed)); - - -#endif /* #ifndef __LINUX__WIMAX__I2400M_H__ */ diff --git a/net/Kconfig b/net/Kconfig index d6567162c1cf..f4c32d982af6 100644 --- a/net/Kconfig +++ b/net/Kconfig @@ -386,8 +386,6 @@ source "net/mac80211/Kconfig" endif # WIRELESS -source "net/wimax/Kconfig" - source "net/rfkill/Kconfig" source "net/9p/Kconfig" source "net/caif/Kconfig" diff --git a/net/Makefile b/net/Makefile index 5744bf1997fd..d96b0aa8f39f 100644 --- a/net/Makefile +++ b/net/Makefile @@ -66,7 +66,6 @@ obj-$(CONFIG_MAC802154) += mac802154/ ifeq ($(CONFIG_NET),y) obj-$(CONFIG_SYSCTL) += sysctl_net.o endif -obj-$(CONFIG_WIMAX) += wimax/ obj-$(CONFIG_DNS_RESOLVER) += dns_resolver/ obj-$(CONFIG_CEPH_LIB) += ceph/ obj-$(CONFIG_BATMAN_ADV) += batman-adv/ diff --git a/net/wimax/Kconfig b/net/wimax/Kconfig deleted file mode 100644 index d13762bc4abc..000000000000 --- a/net/wimax/Kconfig +++ /dev/null @@ -1,40 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only -# -# WiMAX LAN device configuration -# - -menuconfig WIMAX - tristate "WiMAX Wireless Broadband support" - depends on RFKILL || !RFKILL - help - - Select to configure support for devices that provide - wireless broadband connectivity using the WiMAX protocol - (IEEE 802.16). - - Please note that most of these devices require signing up - for a service plan with a provider. - - The different WiMAX drivers can be enabled in the menu entry - - Device Drivers > Network device support > WiMAX Wireless - Broadband devices - - If unsure, it is safe to select M (module). - -config WIMAX_DEBUG_LEVEL - int "WiMAX debug level" - depends on WIMAX - default 8 - help - - Select the maximum debug verbosity level to be compiled into - the WiMAX stack code. - - By default, debug messages are disabled at runtime and can - be selectively enabled for different parts of the code using - the sysfs debug-levels file. - - If set at zero, this will compile out all the debug code. - - It is recommended that it is left at 8. diff --git a/net/wimax/Makefile b/net/wimax/Makefile deleted file mode 100644 index c2a71ae487ac..000000000000 --- a/net/wimax/Makefile +++ /dev/null @@ -1,13 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0 - -obj-$(CONFIG_WIMAX) += wimax.o - -wimax-y := \ - id-table.o \ - op-msg.o \ - op-reset.o \ - op-rfkill.o \ - op-state-get.o \ - stack.o - -wimax-$(CONFIG_DEBUG_FS) += debugfs.o diff --git a/net/wimax/debug-levels.h b/net/wimax/debug-levels.h deleted file mode 100644 index ebc287cde336..000000000000 --- a/net/wimax/debug-levels.h +++ /dev/null @@ -1,29 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Linux WiMAX Stack - * Debug levels control file for the wimax module - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - */ -#ifndef __debug_levels__h__ -#define __debug_levels__h__ - -/* Maximum compile and run time debug level for all submodules */ -#define D_MODULENAME wimax -#define D_MASTER CONFIG_WIMAX_DEBUG_LEVEL - -#include - -/* List of all the enabled modules */ -enum d_module { - D_SUBMODULE_DECLARE(debugfs), - D_SUBMODULE_DECLARE(id_table), - D_SUBMODULE_DECLARE(op_msg), - D_SUBMODULE_DECLARE(op_reset), - D_SUBMODULE_DECLARE(op_rfkill), - D_SUBMODULE_DECLARE(op_state_get), - D_SUBMODULE_DECLARE(stack), -}; - -#endif /* #ifndef __debug_levels__h__ */ diff --git a/net/wimax/debugfs.c b/net/wimax/debugfs.c deleted file mode 100644 index 3c54bb6b925a..000000000000 --- a/net/wimax/debugfs.c +++ /dev/null @@ -1,38 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * Debugfs support - * - * Copyright (C) 2005-2006 Intel Corporation - * Inaky Perez-Gonzalez - */ -#include -#include -#include "wimax-internal.h" - -#define D_SUBMODULE debugfs -#include "debug-levels.h" - -void wimax_debugfs_add(struct wimax_dev *wimax_dev) -{ - struct net_device *net_dev = wimax_dev->net_dev; - struct dentry *dentry; - char buf[128]; - - snprintf(buf, sizeof(buf), "wimax:%s", net_dev->name); - dentry = debugfs_create_dir(buf, NULL); - wimax_dev->debugfs_dentry = dentry; - - d_level_register_debugfs("wimax_dl_", debugfs, dentry); - d_level_register_debugfs("wimax_dl_", id_table, dentry); - d_level_register_debugfs("wimax_dl_", op_msg, dentry); - d_level_register_debugfs("wimax_dl_", op_reset, dentry); - d_level_register_debugfs("wimax_dl_", op_rfkill, dentry); - d_level_register_debugfs("wimax_dl_", op_state_get, dentry); - d_level_register_debugfs("wimax_dl_", stack, dentry); -} - -void wimax_debugfs_rm(struct wimax_dev *wimax_dev) -{ - debugfs_remove_recursive(wimax_dev->debugfs_dentry); -} diff --git a/net/wimax/id-table.c b/net/wimax/id-table.c deleted file mode 100644 index 02eee37b7e31..000000000000 --- a/net/wimax/id-table.c +++ /dev/null @@ -1,130 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * Mappping of generic netlink family IDs to net devices - * - * Copyright (C) 2005-2006 Intel Corporation - * Inaky Perez-Gonzalez - * - * We assign a single generic netlink family ID to each device (to - * simplify lookup). - * - * We need a way to map family ID to a wimax_dev pointer. - * - * The idea is to use a very simple lookup. Using a netlink attribute - * with (for example) the interface name implies a heavier search over - * all the network devices; seemed kind of a waste given that we know - * we are looking for a WiMAX device and that most systems will have - * just a single WiMAX adapter. - * - * We put all the WiMAX devices in the system in a linked list and - * match the generic link family ID against the list. - * - * By using a linked list, the case of a single adapter in the system - * becomes (almost) no overhead, while still working for many more. If - * it ever goes beyond two, I'll be surprised. - */ -#include -#include -#include -#include -#include -#include "wimax-internal.h" - - -#define D_SUBMODULE id_table -#include "debug-levels.h" - - -static DEFINE_SPINLOCK(wimax_id_table_lock); -static struct list_head wimax_id_table = LIST_HEAD_INIT(wimax_id_table); - - -/* - * wimax_id_table_add - add a gennetlink familiy ID / wimax_dev mapping - * - * @wimax_dev: WiMAX device descriptor to associate to the Generic - * Netlink family ID. - * - * Look for an empty spot in the ID table; if none found, double the - * table's size and get the first spot. - */ -void wimax_id_table_add(struct wimax_dev *wimax_dev) -{ - d_fnstart(3, NULL, "(wimax_dev %p)\n", wimax_dev); - spin_lock(&wimax_id_table_lock); - list_add(&wimax_dev->id_table_node, &wimax_id_table); - spin_unlock(&wimax_id_table_lock); - d_fnend(3, NULL, "(wimax_dev %p)\n", wimax_dev); -} - - -/* - * wimax_get_netdev_by_info - lookup a wimax_dev from the gennetlink info - * - * The generic netlink family ID has been filled out in the - * nlmsghdr->nlmsg_type field, so we pull it from there, look it up in - * the mapping table and reference the wimax_dev. - * - * When done, the reference should be dropped with - * 'dev_put(wimax_dev->net_dev)'. - */ -struct wimax_dev *wimax_dev_get_by_genl_info( - struct genl_info *info, int ifindex) -{ - struct wimax_dev *wimax_dev = NULL; - - d_fnstart(3, NULL, "(info %p ifindex %d)\n", info, ifindex); - spin_lock(&wimax_id_table_lock); - list_for_each_entry(wimax_dev, &wimax_id_table, id_table_node) { - if (wimax_dev->net_dev->ifindex == ifindex) { - dev_hold(wimax_dev->net_dev); - goto found; - } - } - wimax_dev = NULL; - d_printf(1, NULL, "wimax: no devices found with ifindex %d\n", - ifindex); -found: - spin_unlock(&wimax_id_table_lock); - d_fnend(3, NULL, "(info %p ifindex %d) = %p\n", - info, ifindex, wimax_dev); - return wimax_dev; -} - - -/* - * wimax_id_table_rm - Remove a gennetlink familiy ID / wimax_dev mapping - * - * @id: family ID to remove from the table - */ -void wimax_id_table_rm(struct wimax_dev *wimax_dev) -{ - spin_lock(&wimax_id_table_lock); - list_del_init(&wimax_dev->id_table_node); - spin_unlock(&wimax_id_table_lock); -} - - -/* - * Release the gennetlink family id / mapping table - * - * On debug, verify that the table is empty upon removal. We want the - * code always compiled, to ensure it doesn't bit rot. It will be - * compiled out if CONFIG_BUG is disabled. - */ -void wimax_id_table_release(void) -{ - struct wimax_dev *wimax_dev; - -#ifndef CONFIG_BUG - return; -#endif - spin_lock(&wimax_id_table_lock); - list_for_each_entry(wimax_dev, &wimax_id_table, id_table_node) { - pr_err("BUG: %s wimax_dev %p ifindex %d not cleared\n", - __func__, wimax_dev, wimax_dev->net_dev->ifindex); - WARN_ON(1); - } - spin_unlock(&wimax_id_table_lock); -} diff --git a/net/wimax/op-msg.c b/net/wimax/op-msg.c deleted file mode 100644 index 6460b5785758..000000000000 --- a/net/wimax/op-msg.c +++ /dev/null @@ -1,391 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * Generic messaging interface between userspace and driver/device - * - * Copyright (C) 2007-2008 Intel Corporation - * Inaky Perez-Gonzalez - * - * This implements a direct communication channel between user space and - * the driver/device, by which free form messages can be sent back and - * forth. - * - * This is intended for device-specific features, vendor quirks, etc. - * - * See include/net/wimax.h - * - * GENERIC NETLINK ENCODING AND CAPACITY - * - * A destination "pipe name" is added to each message; it is up to the - * drivers to assign or use those names (if using them at all). - * - * Messages are encoded as a binary netlink attribute using nla_put() - * using type NLA_UNSPEC (as some versions of libnl still in - * deployment don't yet understand NLA_BINARY). - * - * The maximum capacity of this transport is PAGESIZE per message (so - * the actual payload will be bit smaller depending on the - * netlink/generic netlink attributes and headers). - * - * RECEPTION OF MESSAGES - * - * When a message is received from user space, it is passed verbatim - * to the driver calling wimax_dev->op_msg_from_user(). The return - * value from this function is passed back to user space as an ack - * over the generic netlink protocol. - * - * The stack doesn't do any processing or interpretation of these - * messages. - * - * SENDING MESSAGES - * - * Messages can be sent with wimax_msg(). - * - * If the message delivery needs to happen on a different context to - * that of its creation, wimax_msg_alloc() can be used to get a - * pointer to the message that can be delivered later on with - * wimax_msg_send(). - * - * ROADMAP - * - * wimax_gnl_doit_msg_from_user() Process a message from user space - * wimax_dev_get_by_genl_info() - * wimax_dev->op_msg_from_user() Delivery of message to the driver - * - * wimax_msg() Send a message to user space - * wimax_msg_alloc() - * wimax_msg_send() - */ -#include -#include -#include -#include -#include -#include -#include -#include "wimax-internal.h" - - -#define D_SUBMODULE op_msg -#include "debug-levels.h" - - -/** - * wimax_msg_alloc - Create a new skb for sending a message to userspace - * - * @wimax_dev: WiMAX device descriptor - * @pipe_name: "named pipe" the message will be sent to - * @msg: pointer to the message data to send - * @size: size of the message to send (in bytes), including the header. - * @gfp_flags: flags for memory allocation. - * - * Returns: %0 if ok, negative errno code on error - * - * Description: - * - * Allocates an skb that will contain the message to send to user - * space over the messaging pipe and initializes it, copying the - * payload. - * - * Once this call is done, you can deliver it with - * wimax_msg_send(). - * - * IMPORTANT: - * - * Don't use skb_push()/skb_pull()/skb_reserve() on the skb, as - * wimax_msg_send() depends on skb->data being placed at the - * beginning of the user message. - * - * Unlike other WiMAX stack calls, this call can be used way early, - * even before wimax_dev_add() is called, as long as the - * wimax_dev->net_dev pointer is set to point to a proper - * net_dev. This is so that drivers can use it early in case they need - * to send stuff around or communicate with user space. - */ -struct sk_buff *wimax_msg_alloc(struct wimax_dev *wimax_dev, - const char *pipe_name, - const void *msg, size_t size, - gfp_t gfp_flags) -{ - int result; - struct device *dev = wimax_dev_to_dev(wimax_dev); - size_t msg_size; - void *genl_msg; - struct sk_buff *skb; - - msg_size = nla_total_size(size) - + nla_total_size(sizeof(u32)) - + (pipe_name ? nla_total_size(strlen(pipe_name)) : 0); - result = -ENOMEM; - skb = genlmsg_new(msg_size, gfp_flags); - if (skb == NULL) - goto error_new; - genl_msg = genlmsg_put(skb, 0, 0, &wimax_gnl_family, - 0, WIMAX_GNL_OP_MSG_TO_USER); - if (genl_msg == NULL) { - dev_err(dev, "no memory to create generic netlink message\n"); - goto error_genlmsg_put; - } - result = nla_put_u32(skb, WIMAX_GNL_MSG_IFIDX, - wimax_dev->net_dev->ifindex); - if (result < 0) { - dev_err(dev, "no memory to add ifindex attribute\n"); - goto error_nla_put; - } - if (pipe_name) { - result = nla_put_string(skb, WIMAX_GNL_MSG_PIPE_NAME, - pipe_name); - if (result < 0) { - dev_err(dev, "no memory to add pipe_name attribute\n"); - goto error_nla_put; - } - } - result = nla_put(skb, WIMAX_GNL_MSG_DATA, size, msg); - if (result < 0) { - dev_err(dev, "no memory to add payload (msg %p size %zu) in " - "attribute: %d\n", msg, size, result); - goto error_nla_put; - } - genlmsg_end(skb, genl_msg); - return skb; - -error_nla_put: -error_genlmsg_put: -error_new: - nlmsg_free(skb); - return ERR_PTR(result); -} -EXPORT_SYMBOL_GPL(wimax_msg_alloc); - - -/** - * wimax_msg_data_len - Return a pointer and size of a message's payload - * - * @msg: Pointer to a message created with wimax_msg_alloc() - * @size: Pointer to where to store the message's size - * - * Returns the pointer to the message data. - */ -const void *wimax_msg_data_len(struct sk_buff *msg, size_t *size) -{ - struct nlmsghdr *nlh = (void *) msg->head; - struct nlattr *nla; - - nla = nlmsg_find_attr(nlh, sizeof(struct genlmsghdr), - WIMAX_GNL_MSG_DATA); - if (nla == NULL) { - pr_err("Cannot find attribute WIMAX_GNL_MSG_DATA\n"); - return NULL; - } - *size = nla_len(nla); - return nla_data(nla); -} -EXPORT_SYMBOL_GPL(wimax_msg_data_len); - - -/** - * wimax_msg_data - Return a pointer to a message's payload - * - * @msg: Pointer to a message created with wimax_msg_alloc() - */ -const void *wimax_msg_data(struct sk_buff *msg) -{ - struct nlmsghdr *nlh = (void *) msg->head; - struct nlattr *nla; - - nla = nlmsg_find_attr(nlh, sizeof(struct genlmsghdr), - WIMAX_GNL_MSG_DATA); - if (nla == NULL) { - pr_err("Cannot find attribute WIMAX_GNL_MSG_DATA\n"); - return NULL; - } - return nla_data(nla); -} -EXPORT_SYMBOL_GPL(wimax_msg_data); - - -/** - * wimax_msg_len - Return a message's payload length - * - * @msg: Pointer to a message created with wimax_msg_alloc() - */ -ssize_t wimax_msg_len(struct sk_buff *msg) -{ - struct nlmsghdr *nlh = (void *) msg->head; - struct nlattr *nla; - - nla = nlmsg_find_attr(nlh, sizeof(struct genlmsghdr), - WIMAX_GNL_MSG_DATA); - if (nla == NULL) { - pr_err("Cannot find attribute WIMAX_GNL_MSG_DATA\n"); - return -EINVAL; - } - return nla_len(nla); -} -EXPORT_SYMBOL_GPL(wimax_msg_len); - - -/** - * wimax_msg_send - Send a pre-allocated message to user space - * - * @wimax_dev: WiMAX device descriptor - * - * @skb: &struct sk_buff returned by wimax_msg_alloc(). Note the - * ownership of @skb is transferred to this function. - * - * Returns: 0 if ok, < 0 errno code on error - * - * Description: - * - * Sends a free-form message that was preallocated with - * wimax_msg_alloc() and filled up. - * - * Assumes that once you pass an skb to this function for sending, it - * owns it and will release it when done (on success). - * - * IMPORTANT: - * - * Don't use skb_push()/skb_pull()/skb_reserve() on the skb, as - * wimax_msg_send() depends on skb->data being placed at the - * beginning of the user message. - * - * Unlike other WiMAX stack calls, this call can be used way early, - * even before wimax_dev_add() is called, as long as the - * wimax_dev->net_dev pointer is set to point to a proper - * net_dev. This is so that drivers can use it early in case they need - * to send stuff around or communicate with user space. - */ -int wimax_msg_send(struct wimax_dev *wimax_dev, struct sk_buff *skb) -{ - struct device *dev = wimax_dev_to_dev(wimax_dev); - void *msg = skb->data; - size_t size = skb->len; - might_sleep(); - - d_printf(1, dev, "CTX: wimax msg, %zu bytes\n", size); - d_dump(2, dev, msg, size); - genlmsg_multicast(&wimax_gnl_family, skb, 0, 0, GFP_KERNEL); - d_printf(1, dev, "CTX: genl multicast done\n"); - return 0; -} -EXPORT_SYMBOL_GPL(wimax_msg_send); - - -/** - * wimax_msg - Send a message to user space - * - * @wimax_dev: WiMAX device descriptor (properly referenced) - * @pipe_name: "named pipe" the message will be sent to - * @buf: pointer to the message to send. - * @size: size of the buffer pointed to by @buf (in bytes). - * @gfp_flags: flags for memory allocation. - * - * Returns: %0 if ok, negative errno code on error. - * - * Description: - * - * Sends a free-form message to user space on the device @wimax_dev. - * - * NOTES: - * - * Once the @skb is given to this function, who will own it and will - * release it when done (unless it returns error). - */ -int wimax_msg(struct wimax_dev *wimax_dev, const char *pipe_name, - const void *buf, size_t size, gfp_t gfp_flags) -{ - int result = -ENOMEM; - struct sk_buff *skb; - - skb = wimax_msg_alloc(wimax_dev, pipe_name, buf, size, gfp_flags); - if (IS_ERR(skb)) - result = PTR_ERR(skb); - else - result = wimax_msg_send(wimax_dev, skb); - return result; -} -EXPORT_SYMBOL_GPL(wimax_msg); - -/* - * Relays a message from user space to the driver - * - * The skb is passed to the driver-specific function with the netlink - * and generic netlink headers already stripped. - * - * This call will block while handling/relaying the message. - */ -int wimax_gnl_doit_msg_from_user(struct sk_buff *skb, struct genl_info *info) -{ - int result, ifindex; - struct wimax_dev *wimax_dev; - struct device *dev; - struct nlmsghdr *nlh = info->nlhdr; - char *pipe_name; - void *msg_buf; - size_t msg_len; - - might_sleep(); - d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); - result = -ENODEV; - if (info->attrs[WIMAX_GNL_MSG_IFIDX] == NULL) { - pr_err("WIMAX_GNL_MSG_FROM_USER: can't find IFIDX attribute\n"); - goto error_no_wimax_dev; - } - ifindex = nla_get_u32(info->attrs[WIMAX_GNL_MSG_IFIDX]); - wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); - if (wimax_dev == NULL) - goto error_no_wimax_dev; - dev = wimax_dev_to_dev(wimax_dev); - - /* Unpack arguments */ - result = -EINVAL; - if (info->attrs[WIMAX_GNL_MSG_DATA] == NULL) { - dev_err(dev, "WIMAX_GNL_MSG_FROM_USER: can't find MSG_DATA " - "attribute\n"); - goto error_no_data; - } - msg_buf = nla_data(info->attrs[WIMAX_GNL_MSG_DATA]); - msg_len = nla_len(info->attrs[WIMAX_GNL_MSG_DATA]); - - if (info->attrs[WIMAX_GNL_MSG_PIPE_NAME] == NULL) - pipe_name = NULL; - else { - struct nlattr *attr = info->attrs[WIMAX_GNL_MSG_PIPE_NAME]; - size_t attr_len = nla_len(attr); - /* libnl-1.1 does not yet support NLA_NUL_STRING */ - result = -ENOMEM; - pipe_name = kstrndup(nla_data(attr), attr_len + 1, GFP_KERNEL); - if (pipe_name == NULL) - goto error_alloc; - pipe_name[attr_len] = 0; - } - mutex_lock(&wimax_dev->mutex); - result = wimax_dev_is_ready(wimax_dev); - if (result == -ENOMEDIUM) - result = 0; - if (result < 0) - goto error_not_ready; - result = -ENOSYS; - if (wimax_dev->op_msg_from_user == NULL) - goto error_noop; - - d_printf(1, dev, - "CRX: nlmsghdr len %u type %u flags 0x%04x seq 0x%x pid %u\n", - nlh->nlmsg_len, nlh->nlmsg_type, nlh->nlmsg_flags, - nlh->nlmsg_seq, nlh->nlmsg_pid); - d_printf(1, dev, "CRX: wimax message %zu bytes\n", msg_len); - d_dump(2, dev, msg_buf, msg_len); - - result = wimax_dev->op_msg_from_user(wimax_dev, pipe_name, - msg_buf, msg_len, info); -error_noop: -error_not_ready: - mutex_unlock(&wimax_dev->mutex); -error_alloc: - kfree(pipe_name); -error_no_data: - dev_put(wimax_dev->net_dev); -error_no_wimax_dev: - d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); - return result; -} diff --git a/net/wimax/op-reset.c b/net/wimax/op-reset.c deleted file mode 100644 index 9899b2e56721..000000000000 --- a/net/wimax/op-reset.c +++ /dev/null @@ -1,108 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * Implement and export a method for resetting a WiMAX device - * - * Copyright (C) 2008 Intel Corporation - * Inaky Perez-Gonzalez - * - * This implements a simple synchronous call to reset a WiMAX device. - * - * Resets aim at being warm, keeping the device handles active; - * however, when that fails, it falls back to a cold reset (that will - * disconnect and reconnect the device). - */ - -#include -#include -#include -#include -#include -#include "wimax-internal.h" - -#define D_SUBMODULE op_reset -#include "debug-levels.h" - - -/** - * wimax_reset - Reset a WiMAX device - * - * @wimax_dev: WiMAX device descriptor - * - * Returns: - * - * %0 if ok and a warm reset was done (the device still exists in - * the system). - * - * -%ENODEV if a cold/bus reset had to be done (device has - * disconnected and reconnected, so current handle is not valid - * any more). - * - * -%EINVAL if the device is not even registered. - * - * Any other negative error code shall be considered as - * non-recoverable. - * - * Description: - * - * Called when wanting to reset the device for any reason. Device is - * taken back to power on status. - * - * This call blocks; on successful return, the device has completed the - * reset process and is ready to operate. - */ -int wimax_reset(struct wimax_dev *wimax_dev) -{ - int result = -EINVAL; - struct device *dev = wimax_dev_to_dev(wimax_dev); - enum wimax_st state; - - might_sleep(); - d_fnstart(3, dev, "(wimax_dev %p)\n", wimax_dev); - mutex_lock(&wimax_dev->mutex); - dev_hold(wimax_dev->net_dev); - state = wimax_dev->state; - mutex_unlock(&wimax_dev->mutex); - - if (state >= WIMAX_ST_DOWN) { - mutex_lock(&wimax_dev->mutex_reset); - result = wimax_dev->op_reset(wimax_dev); - mutex_unlock(&wimax_dev->mutex_reset); - } - dev_put(wimax_dev->net_dev); - - d_fnend(3, dev, "(wimax_dev %p) = %d\n", wimax_dev, result); - return result; -} -EXPORT_SYMBOL(wimax_reset); - - -/* - * Exporting to user space over generic netlink - * - * Parse the reset command from user space, return error code. - * - * No attributes. - */ -int wimax_gnl_doit_reset(struct sk_buff *skb, struct genl_info *info) -{ - int result, ifindex; - struct wimax_dev *wimax_dev; - - d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); - result = -ENODEV; - if (info->attrs[WIMAX_GNL_RESET_IFIDX] == NULL) { - pr_err("WIMAX_GNL_OP_RFKILL: can't find IFIDX attribute\n"); - goto error_no_wimax_dev; - } - ifindex = nla_get_u32(info->attrs[WIMAX_GNL_RESET_IFIDX]); - wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); - if (wimax_dev == NULL) - goto error_no_wimax_dev; - /* Execute the operation and send the result back to user space */ - result = wimax_reset(wimax_dev); - dev_put(wimax_dev->net_dev); -error_no_wimax_dev: - d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); - return result; -} diff --git a/net/wimax/op-rfkill.c b/net/wimax/op-rfkill.c deleted file mode 100644 index 248d10b60b05..000000000000 --- a/net/wimax/op-rfkill.c +++ /dev/null @@ -1,431 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * RF-kill framework integration - * - * Copyright (C) 2008 Intel Corporation - * Inaky Perez-Gonzalez - * - * This integrates into the Linux Kernel rfkill susbystem so that the - * drivers just have to do the bare minimal work, which is providing a - * method to set the software RF-Kill switch and to report changes in - * the software and hardware switch status. - * - * A non-polled generic rfkill device is embedded into the WiMAX - * subsystem's representation of a device. - * - * FIXME: Need polled support? Let drivers provide a poll routine - * and hand it to rfkill ops then? - * - * All device drivers have to do is after wimax_dev_init(), call - * wimax_report_rfkill_hw() and wimax_report_rfkill_sw() to update - * initial state and then every time it changes. See wimax.h:struct - * wimax_dev for more information. - * - * ROADMAP - * - * wimax_gnl_doit_rfkill() User space calling wimax_rfkill() - * wimax_rfkill() Kernel calling wimax_rfkill() - * __wimax_rf_toggle_radio() - * - * wimax_rfkill_set_radio_block() RF-Kill subsystem calling - * __wimax_rf_toggle_radio() - * - * __wimax_rf_toggle_radio() - * wimax_dev->op_rfkill_sw_toggle() Driver backend - * __wimax_state_change() - * - * wimax_report_rfkill_sw() Driver reports state change - * __wimax_state_change() - * - * wimax_report_rfkill_hw() Driver reports state change - * __wimax_state_change() - * - * wimax_rfkill_add() Initialize/shutdown rfkill support - * wimax_rfkill_rm() [called by wimax_dev_add/rm()] - */ - -#include -#include -#include -#include -#include -#include -#include "wimax-internal.h" - -#define D_SUBMODULE op_rfkill -#include "debug-levels.h" - -/** - * wimax_report_rfkill_hw - Reports changes in the hardware RF switch - * - * @wimax_dev: WiMAX device descriptor - * - * @state: New state of the RF Kill switch. %WIMAX_RF_ON radio on, - * %WIMAX_RF_OFF radio off. - * - * When the device detects a change in the state of thehardware RF - * switch, it must call this function to let the WiMAX kernel stack - * know that the state has changed so it can be properly propagated. - * - * The WiMAX stack caches the state (the driver doesn't need to). As - * well, as the change is propagated it will come back as a request to - * change the software state to mirror the hardware state. - * - * If the device doesn't have a hardware kill switch, just report - * it on initialization as always on (%WIMAX_RF_ON, radio on). - */ -void wimax_report_rfkill_hw(struct wimax_dev *wimax_dev, - enum wimax_rf_state state) -{ - int result; - struct device *dev = wimax_dev_to_dev(wimax_dev); - enum wimax_st wimax_state; - - d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); - BUG_ON(state == WIMAX_RF_QUERY); - BUG_ON(state != WIMAX_RF_ON && state != WIMAX_RF_OFF); - - mutex_lock(&wimax_dev->mutex); - result = wimax_dev_is_ready(wimax_dev); - if (result < 0) - goto error_not_ready; - - if (state != wimax_dev->rf_hw) { - wimax_dev->rf_hw = state; - if (wimax_dev->rf_hw == WIMAX_RF_ON && - wimax_dev->rf_sw == WIMAX_RF_ON) - wimax_state = WIMAX_ST_READY; - else - wimax_state = WIMAX_ST_RADIO_OFF; - - result = rfkill_set_hw_state(wimax_dev->rfkill, - state == WIMAX_RF_OFF); - - __wimax_state_change(wimax_dev, wimax_state); - } -error_not_ready: - mutex_unlock(&wimax_dev->mutex); - d_fnend(3, dev, "(wimax_dev %p state %u) = void [%d]\n", - wimax_dev, state, result); -} -EXPORT_SYMBOL_GPL(wimax_report_rfkill_hw); - - -/** - * wimax_report_rfkill_sw - Reports changes in the software RF switch - * - * @wimax_dev: WiMAX device descriptor - * - * @state: New state of the RF kill switch. %WIMAX_RF_ON radio on, - * %WIMAX_RF_OFF radio off. - * - * Reports changes in the software RF switch state to the WiMAX stack. - * - * The main use is during initialization, so the driver can query the - * device for its current software radio kill switch state and feed it - * to the system. - * - * On the side, the device does not change the software state by - * itself. In practice, this can happen, as the device might decide to - * switch (in software) the radio off for different reasons. - */ -void wimax_report_rfkill_sw(struct wimax_dev *wimax_dev, - enum wimax_rf_state state) -{ - int result; - struct device *dev = wimax_dev_to_dev(wimax_dev); - enum wimax_st wimax_state; - - d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); - BUG_ON(state == WIMAX_RF_QUERY); - BUG_ON(state != WIMAX_RF_ON && state != WIMAX_RF_OFF); - - mutex_lock(&wimax_dev->mutex); - result = wimax_dev_is_ready(wimax_dev); - if (result < 0) - goto error_not_ready; - - if (state != wimax_dev->rf_sw) { - wimax_dev->rf_sw = state; - if (wimax_dev->rf_hw == WIMAX_RF_ON && - wimax_dev->rf_sw == WIMAX_RF_ON) - wimax_state = WIMAX_ST_READY; - else - wimax_state = WIMAX_ST_RADIO_OFF; - __wimax_state_change(wimax_dev, wimax_state); - rfkill_set_sw_state(wimax_dev->rfkill, state == WIMAX_RF_OFF); - } -error_not_ready: - mutex_unlock(&wimax_dev->mutex); - d_fnend(3, dev, "(wimax_dev %p state %u) = void [%d]\n", - wimax_dev, state, result); -} -EXPORT_SYMBOL_GPL(wimax_report_rfkill_sw); - - -/* - * Callback for the RF Kill toggle operation - * - * This function is called by: - * - * - The rfkill subsystem when the RF-Kill key is pressed in the - * hardware and the driver notifies through - * wimax_report_rfkill_hw(). The rfkill subsystem ends up calling back - * here so the software RF Kill switch state is changed to reflect - * the hardware switch state. - * - * - When the user sets the state through sysfs' rfkill/state file - * - * - When the user calls wimax_rfkill(). - * - * This call blocks! - * - * WARNING! When we call rfkill_unregister(), this will be called with - * state 0! - * - * WARNING: wimax_dev must be locked - */ -static -int __wimax_rf_toggle_radio(struct wimax_dev *wimax_dev, - enum wimax_rf_state state) -{ - int result = 0; - struct device *dev = wimax_dev_to_dev(wimax_dev); - enum wimax_st wimax_state; - - might_sleep(); - d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); - if (wimax_dev->rf_sw == state) - goto out_no_change; - if (wimax_dev->op_rfkill_sw_toggle != NULL) - result = wimax_dev->op_rfkill_sw_toggle(wimax_dev, state); - else if (state == WIMAX_RF_OFF) /* No op? can't turn off */ - result = -ENXIO; - else /* No op? can turn on */ - result = 0; /* should never happen tho */ - if (result >= 0) { - result = 0; - wimax_dev->rf_sw = state; - wimax_state = state == WIMAX_RF_ON ? - WIMAX_ST_READY : WIMAX_ST_RADIO_OFF; - __wimax_state_change(wimax_dev, wimax_state); - } -out_no_change: - d_fnend(3, dev, "(wimax_dev %p state %u) = %d\n", - wimax_dev, state, result); - return result; -} - - -/* - * Translate from rfkill state to wimax state - * - * NOTE: Special state handling rules here - * - * Just pretend the call didn't happen if we are in a state where - * we know for sure it cannot be handled (WIMAX_ST_DOWN or - * __WIMAX_ST_QUIESCING). rfkill() needs it to register and - * unregister, as it will run this path. - * - * NOTE: This call will block until the operation is completed. - */ -static int wimax_rfkill_set_radio_block(void *data, bool blocked) -{ - int result; - struct wimax_dev *wimax_dev = data; - struct device *dev = wimax_dev_to_dev(wimax_dev); - enum wimax_rf_state rf_state; - - d_fnstart(3, dev, "(wimax_dev %p blocked %u)\n", wimax_dev, blocked); - rf_state = WIMAX_RF_ON; - if (blocked) - rf_state = WIMAX_RF_OFF; - mutex_lock(&wimax_dev->mutex); - if (wimax_dev->state <= __WIMAX_ST_QUIESCING) - result = 0; - else - result = __wimax_rf_toggle_radio(wimax_dev, rf_state); - mutex_unlock(&wimax_dev->mutex); - d_fnend(3, dev, "(wimax_dev %p blocked %u) = %d\n", - wimax_dev, blocked, result); - return result; -} - -static const struct rfkill_ops wimax_rfkill_ops = { - .set_block = wimax_rfkill_set_radio_block, -}; - -/** - * wimax_rfkill - Set the software RF switch state for a WiMAX device - * - * @wimax_dev: WiMAX device descriptor - * - * @state: New RF state. - * - * Returns: - * - * >= 0 toggle state if ok, < 0 errno code on error. The toggle state - * is returned as a bitmap, bit 0 being the hardware RF state, bit 1 - * the software RF state. - * - * 0 means disabled (%WIMAX_RF_ON, radio on), 1 means enabled radio - * off (%WIMAX_RF_OFF). - * - * Description: - * - * Called by the user when he wants to request the WiMAX radio to be - * switched on (%WIMAX_RF_ON) or off (%WIMAX_RF_OFF). With - * %WIMAX_RF_QUERY, just the current state is returned. - * - * NOTE: - * - * This call will block until the operation is complete. - */ -int wimax_rfkill(struct wimax_dev *wimax_dev, enum wimax_rf_state state) -{ - int result; - struct device *dev = wimax_dev_to_dev(wimax_dev); - - d_fnstart(3, dev, "(wimax_dev %p state %u)\n", wimax_dev, state); - mutex_lock(&wimax_dev->mutex); - result = wimax_dev_is_ready(wimax_dev); - if (result < 0) { - /* While initializing, < 1.4.3 wimax-tools versions use - * this call to check if the device is a valid WiMAX - * device; so we allow it to proceed always, - * considering the radios are all off. */ - if (result == -ENOMEDIUM && state == WIMAX_RF_QUERY) - result = WIMAX_RF_OFF << 1 | WIMAX_RF_OFF; - goto error_not_ready; - } - switch (state) { - case WIMAX_RF_ON: - case WIMAX_RF_OFF: - result = __wimax_rf_toggle_radio(wimax_dev, state); - if (result < 0) - goto error; - rfkill_set_sw_state(wimax_dev->rfkill, state == WIMAX_RF_OFF); - break; - case WIMAX_RF_QUERY: - break; - default: - result = -EINVAL; - goto error; - } - result = wimax_dev->rf_sw << 1 | wimax_dev->rf_hw; -error: -error_not_ready: - mutex_unlock(&wimax_dev->mutex); - d_fnend(3, dev, "(wimax_dev %p state %u) = %d\n", - wimax_dev, state, result); - return result; -} -EXPORT_SYMBOL(wimax_rfkill); - - -/* - * Register a new WiMAX device's RF Kill support - * - * WARNING: wimax_dev->mutex must be unlocked - */ -int wimax_rfkill_add(struct wimax_dev *wimax_dev) -{ - int result; - struct rfkill *rfkill; - struct device *dev = wimax_dev_to_dev(wimax_dev); - - d_fnstart(3, dev, "(wimax_dev %p)\n", wimax_dev); - /* Initialize RF Kill */ - result = -ENOMEM; - rfkill = rfkill_alloc(wimax_dev->name, dev, RFKILL_TYPE_WIMAX, - &wimax_rfkill_ops, wimax_dev); - if (rfkill == NULL) - goto error_rfkill_allocate; - - d_printf(1, dev, "rfkill %p\n", rfkill); - - wimax_dev->rfkill = rfkill; - - rfkill_init_sw_state(rfkill, 1); - result = rfkill_register(wimax_dev->rfkill); - if (result < 0) - goto error_rfkill_register; - - /* If there is no SW toggle op, SW RFKill is always on */ - if (wimax_dev->op_rfkill_sw_toggle == NULL) - wimax_dev->rf_sw = WIMAX_RF_ON; - - d_fnend(3, dev, "(wimax_dev %p) = 0\n", wimax_dev); - return 0; - -error_rfkill_register: - rfkill_destroy(wimax_dev->rfkill); -error_rfkill_allocate: - d_fnend(3, dev, "(wimax_dev %p) = %d\n", wimax_dev, result); - return result; -} - - -/* - * Deregister a WiMAX device's RF Kill support - * - * Ick, we can't call rfkill_free() after rfkill_unregister()...oh - * well. - * - * WARNING: wimax_dev->mutex must be unlocked - */ -void wimax_rfkill_rm(struct wimax_dev *wimax_dev) -{ - struct device *dev = wimax_dev_to_dev(wimax_dev); - d_fnstart(3, dev, "(wimax_dev %p)\n", wimax_dev); - rfkill_unregister(wimax_dev->rfkill); - rfkill_destroy(wimax_dev->rfkill); - d_fnend(3, dev, "(wimax_dev %p)\n", wimax_dev); -} - - -/* - * Exporting to user space over generic netlink - * - * Parse the rfkill command from user space, return a combination - * value that describe the states of the different toggles. - * - * Only one attribute: the new state requested (on, off or no change, - * just query). - */ - -int wimax_gnl_doit_rfkill(struct sk_buff *skb, struct genl_info *info) -{ - int result, ifindex; - struct wimax_dev *wimax_dev; - struct device *dev; - enum wimax_rf_state new_state; - - d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); - result = -ENODEV; - if (info->attrs[WIMAX_GNL_RFKILL_IFIDX] == NULL) { - pr_err("WIMAX_GNL_OP_RFKILL: can't find IFIDX attribute\n"); - goto error_no_wimax_dev; - } - ifindex = nla_get_u32(info->attrs[WIMAX_GNL_RFKILL_IFIDX]); - wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); - if (wimax_dev == NULL) - goto error_no_wimax_dev; - dev = wimax_dev_to_dev(wimax_dev); - result = -EINVAL; - if (info->attrs[WIMAX_GNL_RFKILL_STATE] == NULL) { - dev_err(dev, "WIMAX_GNL_RFKILL: can't find RFKILL_STATE " - "attribute\n"); - goto error_no_pid; - } - new_state = nla_get_u32(info->attrs[WIMAX_GNL_RFKILL_STATE]); - - /* Execute the operation and send the result back to user space */ - result = wimax_rfkill(wimax_dev, new_state); -error_no_pid: - dev_put(wimax_dev->net_dev); -error_no_wimax_dev: - d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); - return result; -} diff --git a/net/wimax/op-state-get.c b/net/wimax/op-state-get.c deleted file mode 100644 index 5bc712de1563..000000000000 --- a/net/wimax/op-state-get.c +++ /dev/null @@ -1,52 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * Implement and export a method for getting a WiMAX device current state - * - * Copyright (C) 2009 Paulius Zaleckas - * - * Based on previous WiMAX core work by: - * Copyright (C) 2008 Intel Corporation - * Inaky Perez-Gonzalez - */ - -#include -#include -#include -#include -#include "wimax-internal.h" - -#define D_SUBMODULE op_state_get -#include "debug-levels.h" - - -/* - * Exporting to user space over generic netlink - * - * Parse the state get command from user space, return a combination - * value that describe the current state. - * - * No attributes. - */ -int wimax_gnl_doit_state_get(struct sk_buff *skb, struct genl_info *info) -{ - int result, ifindex; - struct wimax_dev *wimax_dev; - - d_fnstart(3, NULL, "(skb %p info %p)\n", skb, info); - result = -ENODEV; - if (info->attrs[WIMAX_GNL_STGET_IFIDX] == NULL) { - pr_err("WIMAX_GNL_OP_STATE_GET: can't find IFIDX attribute\n"); - goto error_no_wimax_dev; - } - ifindex = nla_get_u32(info->attrs[WIMAX_GNL_STGET_IFIDX]); - wimax_dev = wimax_dev_get_by_genl_info(info, ifindex); - if (wimax_dev == NULL) - goto error_no_wimax_dev; - /* Execute the operation and send the result back to user space */ - result = wimax_state_get(wimax_dev); - dev_put(wimax_dev->net_dev); -error_no_wimax_dev: - d_fnend(3, NULL, "(skb %p info %p) = %d\n", skb, info, result); - return result; -} diff --git a/net/wimax/stack.c b/net/wimax/stack.c deleted file mode 100644 index 3a62af3f80bf..000000000000 --- a/net/wimax/stack.c +++ /dev/null @@ -1,616 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Linux WiMAX - * Initialization, addition and removal of wimax devices - * - * Copyright (C) 2005-2006 Intel Corporation - * Inaky Perez-Gonzalez - * - * This implements: - * - * - basic life cycle of 'struct wimax_dev' [wimax_dev_*()]; on - * addition/registration initialize all subfields and allocate - * generic netlink resources for user space communication. On - * removal/unregistration, undo all that. - * - * - device state machine [wimax_state_change()] and support to send - * reports to user space when the state changes - * [wimax_gnl_re_state_change*()]. - * - * See include/net/wimax.h for rationales and design. - * - * ROADMAP - * - * [__]wimax_state_change() Called by drivers to update device's state - * wimax_gnl_re_state_change_alloc() - * wimax_gnl_re_state_change_send() - * - * wimax_dev_init() Init a device - * wimax_dev_add() Register - * wimax_rfkill_add() - * wimax_gnl_add() Register all the generic netlink resources. - * wimax_id_table_add() - * wimax_dev_rm() Unregister - * wimax_id_table_rm() - * wimax_gnl_rm() - * wimax_rfkill_rm() - */ -#include -#include -#include -#include -#include -#include -#include "wimax-internal.h" - - -#define D_SUBMODULE stack -#include "debug-levels.h" - -static char wimax_debug_params[128]; -module_param_string(debug, wimax_debug_params, sizeof(wimax_debug_params), - 0644); -MODULE_PARM_DESC(debug, - "String of space-separated NAME:VALUE pairs, where NAMEs " - "are the different debug submodules and VALUE are the " - "initial debug value to set."); - -/* - * Authoritative source for the RE_STATE_CHANGE attribute policy - * - * We don't really use it here, but /me likes to keep the definition - * close to where the data is generated. - */ -/* -static const struct nla_policy wimax_gnl_re_status_change[WIMAX_GNL_ATTR_MAX + 1] = { - [WIMAX_GNL_STCH_STATE_OLD] = { .type = NLA_U8 }, - [WIMAX_GNL_STCH_STATE_NEW] = { .type = NLA_U8 }, -}; -*/ - - -/* - * Allocate a Report State Change message - * - * @header: save it, you need it for _send() - * - * Creates and fills a basic state change message; different code - * paths can then add more attributes to the message as needed. - * - * Use wimax_gnl_re_state_change_send() to send the returned skb. - * - * Returns: skb with the genl message if ok, IS_ERR() ptr on error - * with an errno code. - */ -static -struct sk_buff *wimax_gnl_re_state_change_alloc( - struct wimax_dev *wimax_dev, - enum wimax_st new_state, enum wimax_st old_state, - void **header) -{ - int result; - struct device *dev = wimax_dev_to_dev(wimax_dev); - void *data; - struct sk_buff *report_skb; - - d_fnstart(3, dev, "(wimax_dev %p new_state %u old_state %u)\n", - wimax_dev, new_state, old_state); - result = -ENOMEM; - report_skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); - if (report_skb == NULL) { - dev_err(dev, "RE_STCH: can't create message\n"); - goto error_new; - } - /* FIXME: sending a group ID as the seq is wrong */ - data = genlmsg_put(report_skb, 0, wimax_gnl_family.mcgrp_offset, - &wimax_gnl_family, 0, WIMAX_GNL_RE_STATE_CHANGE); - if (data == NULL) { - dev_err(dev, "RE_STCH: can't put data into message\n"); - goto error_put; - } - *header = data; - - result = nla_put_u8(report_skb, WIMAX_GNL_STCH_STATE_OLD, old_state); - if (result < 0) { - dev_err(dev, "RE_STCH: Error adding OLD attr: %d\n", result); - goto error_put; - } - result = nla_put_u8(report_skb, WIMAX_GNL_STCH_STATE_NEW, new_state); - if (result < 0) { - dev_err(dev, "RE_STCH: Error adding NEW attr: %d\n", result); - goto error_put; - } - result = nla_put_u32(report_skb, WIMAX_GNL_STCH_IFIDX, - wimax_dev->net_dev->ifindex); - if (result < 0) { - dev_err(dev, "RE_STCH: Error adding IFINDEX attribute\n"); - goto error_put; - } - d_fnend(3, dev, "(wimax_dev %p new_state %u old_state %u) = %p\n", - wimax_dev, new_state, old_state, report_skb); - return report_skb; - -error_put: - nlmsg_free(report_skb); -error_new: - d_fnend(3, dev, "(wimax_dev %p new_state %u old_state %u) = %d\n", - wimax_dev, new_state, old_state, result); - return ERR_PTR(result); -} - - -/* - * Send a Report State Change message (as created with _alloc). - * - * @report_skb: as returned by wimax_gnl_re_state_change_alloc() - * @header: as returned by wimax_gnl_re_state_change_alloc() - * - * Returns: 0 if ok, < 0 errno code on error. - * - * If the message is NULL, pretend it didn't happen. - */ -static -int wimax_gnl_re_state_change_send( - struct wimax_dev *wimax_dev, struct sk_buff *report_skb, - void *header) -{ - int result = 0; - struct device *dev = wimax_dev_to_dev(wimax_dev); - d_fnstart(3, dev, "(wimax_dev %p report_skb %p)\n", - wimax_dev, report_skb); - if (report_skb == NULL) { - result = -ENOMEM; - goto out; - } - genlmsg_end(report_skb, header); - genlmsg_multicast(&wimax_gnl_family, report_skb, 0, 0, GFP_KERNEL); -out: - d_fnend(3, dev, "(wimax_dev %p report_skb %p) = %d\n", - wimax_dev, report_skb, result); - return result; -} - - -static -void __check_new_state(enum wimax_st old_state, enum wimax_st new_state, - unsigned int allowed_states_bm) -{ - if (WARN_ON(((1 << new_state) & allowed_states_bm) == 0)) { - pr_err("SW BUG! Forbidden state change %u -> %u\n", - old_state, new_state); - } -} - - -/* - * Set the current state of a WiMAX device [unlocking version of - * wimax_state_change(). - */ -void __wimax_state_change(struct wimax_dev *wimax_dev, enum wimax_st new_state) -{ - struct device *dev = wimax_dev_to_dev(wimax_dev); - enum wimax_st old_state = wimax_dev->state; - struct sk_buff *stch_skb; - void *header; - - d_fnstart(3, dev, "(wimax_dev %p new_state %u [old %u])\n", - wimax_dev, new_state, old_state); - - if (WARN_ON(new_state >= __WIMAX_ST_INVALID)) { - dev_err(dev, "SW BUG: requesting invalid state %u\n", - new_state); - goto out; - } - if (old_state == new_state) - goto out; - header = NULL; /* gcc complains? can't grok why */ - stch_skb = wimax_gnl_re_state_change_alloc( - wimax_dev, new_state, old_state, &header); - - /* Verify the state transition and do exit-from-state actions */ - switch (old_state) { - case __WIMAX_ST_NULL: - __check_new_state(old_state, new_state, - 1 << WIMAX_ST_DOWN); - break; - case WIMAX_ST_DOWN: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_UNINITIALIZED - | 1 << WIMAX_ST_RADIO_OFF); - break; - case __WIMAX_ST_QUIESCING: - __check_new_state(old_state, new_state, 1 << WIMAX_ST_DOWN); - break; - case WIMAX_ST_UNINITIALIZED: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_RADIO_OFF); - break; - case WIMAX_ST_RADIO_OFF: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_READY); - break; - case WIMAX_ST_READY: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_RADIO_OFF - | 1 << WIMAX_ST_SCANNING - | 1 << WIMAX_ST_CONNECTING - | 1 << WIMAX_ST_CONNECTED); - break; - case WIMAX_ST_SCANNING: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_RADIO_OFF - | 1 << WIMAX_ST_READY - | 1 << WIMAX_ST_CONNECTING - | 1 << WIMAX_ST_CONNECTED); - break; - case WIMAX_ST_CONNECTING: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_RADIO_OFF - | 1 << WIMAX_ST_READY - | 1 << WIMAX_ST_SCANNING - | 1 << WIMAX_ST_CONNECTED); - break; - case WIMAX_ST_CONNECTED: - __check_new_state(old_state, new_state, - 1 << __WIMAX_ST_QUIESCING - | 1 << WIMAX_ST_RADIO_OFF - | 1 << WIMAX_ST_READY); - netif_tx_disable(wimax_dev->net_dev); - netif_carrier_off(wimax_dev->net_dev); - break; - case __WIMAX_ST_INVALID: - default: - dev_err(dev, "SW BUG: wimax_dev %p is in unknown state %u\n", - wimax_dev, wimax_dev->state); - WARN_ON(1); - goto out; - } - - /* Execute the actions of entry to the new state */ - switch (new_state) { - case __WIMAX_ST_NULL: - dev_err(dev, "SW BUG: wimax_dev %p entering NULL state " - "from %u\n", wimax_dev, wimax_dev->state); - WARN_ON(1); /* Nobody can enter this state */ - break; - case WIMAX_ST_DOWN: - break; - case __WIMAX_ST_QUIESCING: - break; - case WIMAX_ST_UNINITIALIZED: - break; - case WIMAX_ST_RADIO_OFF: - break; - case WIMAX_ST_READY: - break; - case WIMAX_ST_SCANNING: - break; - case WIMAX_ST_CONNECTING: - break; - case WIMAX_ST_CONNECTED: - netif_carrier_on(wimax_dev->net_dev); - netif_wake_queue(wimax_dev->net_dev); - break; - case __WIMAX_ST_INVALID: - default: - BUG(); - } - __wimax_state_set(wimax_dev, new_state); - if (!IS_ERR(stch_skb)) - wimax_gnl_re_state_change_send(wimax_dev, stch_skb, header); -out: - d_fnend(3, dev, "(wimax_dev %p new_state %u [old %u]) = void\n", - wimax_dev, new_state, old_state); -} - - -/** - * wimax_state_change - Set the current state of a WiMAX device - * - * @wimax_dev: WiMAX device descriptor (properly referenced) - * @new_state: New state to switch to - * - * This implements the state changes for the wimax devices. It will - * - * - verify that the state transition is legal (for now it'll just - * print a warning if not) according to the table in - * linux/wimax.h's documentation for 'enum wimax_st'. - * - * - perform the actions needed for leaving the current state and - * whichever are needed for entering the new state. - * - * - issue a report to user space indicating the new state (and an - * optional payload with information about the new state). - * - * NOTE: @wimax_dev must be locked - */ -void wimax_state_change(struct wimax_dev *wimax_dev, enum wimax_st new_state) -{ - /* - * A driver cannot take the wimax_dev out of the - * __WIMAX_ST_NULL state unless by calling wimax_dev_add(). If - * the wimax_dev's state is still NULL, we ignore any request - * to change its state because it means it hasn't been yet - * registered. - * - * There is no need to complain about it, as routines that - * call this might be shared from different code paths that - * are called before or after wimax_dev_add() has done its - * job. - */ - mutex_lock(&wimax_dev->mutex); - if (wimax_dev->state > __WIMAX_ST_NULL) - __wimax_state_change(wimax_dev, new_state); - mutex_unlock(&wimax_dev->mutex); -} -EXPORT_SYMBOL_GPL(wimax_state_change); - - -/** - * wimax_state_get() - Return the current state of a WiMAX device - * - * @wimax_dev: WiMAX device descriptor - * - * Returns: Current state of the device according to its driver. - */ -enum wimax_st wimax_state_get(struct wimax_dev *wimax_dev) -{ - enum wimax_st state; - mutex_lock(&wimax_dev->mutex); - state = wimax_dev->state; - mutex_unlock(&wimax_dev->mutex); - return state; -} -EXPORT_SYMBOL_GPL(wimax_state_get); - - -/** - * wimax_dev_init - initialize a newly allocated instance - * - * @wimax_dev: WiMAX device descriptor to initialize. - * - * Initializes fields of a freshly allocated @wimax_dev instance. This - * function assumes that after allocation, the memory occupied by - * @wimax_dev was zeroed. - */ -void wimax_dev_init(struct wimax_dev *wimax_dev) -{ - INIT_LIST_HEAD(&wimax_dev->id_table_node); - __wimax_state_set(wimax_dev, __WIMAX_ST_NULL); - mutex_init(&wimax_dev->mutex); - mutex_init(&wimax_dev->mutex_reset); -} -EXPORT_SYMBOL_GPL(wimax_dev_init); - -/* - * There are multiple enums reusing the same values, adding - * others is only possible if they use a compatible policy. - */ -static const struct nla_policy wimax_gnl_policy[WIMAX_GNL_ATTR_MAX + 1] = { - /* - * WIMAX_GNL_RESET_IFIDX, WIMAX_GNL_RFKILL_IFIDX, - * WIMAX_GNL_STGET_IFIDX, WIMAX_GNL_MSG_IFIDX - */ - [1] = { .type = NLA_U32, }, - /* - * WIMAX_GNL_RFKILL_STATE, WIMAX_GNL_MSG_PIPE_NAME - */ - [2] = { .type = NLA_U32, }, /* enum wimax_rf_state */ - /* - * WIMAX_GNL_MSG_DATA - */ - [3] = { .type = NLA_UNSPEC, }, /* libnl doesn't grok BINARY yet */ -}; - -static const struct genl_small_ops wimax_gnl_ops[] = { - { - .cmd = WIMAX_GNL_OP_MSG_FROM_USER, - .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, - .doit = wimax_gnl_doit_msg_from_user, - }, - { - .cmd = WIMAX_GNL_OP_RESET, - .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, - .doit = wimax_gnl_doit_reset, - }, - { - .cmd = WIMAX_GNL_OP_RFKILL, - .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, - .doit = wimax_gnl_doit_rfkill, - }, - { - .cmd = WIMAX_GNL_OP_STATE_GET, - .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, - .doit = wimax_gnl_doit_state_get, - }, -}; - - -static -size_t wimax_addr_scnprint(char *addr_str, size_t addr_str_size, - unsigned char *addr, size_t addr_len) -{ - unsigned int cnt, total; - - for (total = cnt = 0; cnt < addr_len; cnt++) - total += scnprintf(addr_str + total, addr_str_size - total, - "%02x%c", addr[cnt], - cnt == addr_len - 1 ? '\0' : ':'); - return total; -} - - -/** - * wimax_dev_add - Register a new WiMAX device - * - * @wimax_dev: WiMAX device descriptor (as embedded in your @net_dev's - * priv data). You must have called wimax_dev_init() on it before. - * - * @net_dev: net device the @wimax_dev is associated with. The - * function expects SET_NETDEV_DEV() and register_netdev() were - * already called on it. - * - * Registers the new WiMAX device, sets up the user-kernel control - * interface (generic netlink) and common WiMAX infrastructure. - * - * Note that the parts that will allow interaction with user space are - * setup at the very end, when the rest is in place, as once that - * happens, the driver might get user space control requests via - * netlink or from debugfs that might translate into calls into - * wimax_dev->op_*(). - */ -int wimax_dev_add(struct wimax_dev *wimax_dev, struct net_device *net_dev) -{ - int result; - struct device *dev = net_dev->dev.parent; - char addr_str[32]; - - d_fnstart(3, dev, "(wimax_dev %p net_dev %p)\n", wimax_dev, net_dev); - - /* Do the RFKILL setup before locking, as RFKILL will call - * into our functions. - */ - wimax_dev->net_dev = net_dev; - result = wimax_rfkill_add(wimax_dev); - if (result < 0) - goto error_rfkill_add; - - /* Set up user-space interaction */ - mutex_lock(&wimax_dev->mutex); - wimax_id_table_add(wimax_dev); - wimax_debugfs_add(wimax_dev); - - __wimax_state_set(wimax_dev, WIMAX_ST_DOWN); - mutex_unlock(&wimax_dev->mutex); - - wimax_addr_scnprint(addr_str, sizeof(addr_str), - net_dev->dev_addr, net_dev->addr_len); - dev_err(dev, "WiMAX interface %s (%s) ready\n", - net_dev->name, addr_str); - d_fnend(3, dev, "(wimax_dev %p net_dev %p) = 0\n", wimax_dev, net_dev); - return 0; - -error_rfkill_add: - d_fnend(3, dev, "(wimax_dev %p net_dev %p) = %d\n", - wimax_dev, net_dev, result); - return result; -} -EXPORT_SYMBOL_GPL(wimax_dev_add); - - -/** - * wimax_dev_rm - Unregister an existing WiMAX device - * - * @wimax_dev: WiMAX device descriptor - * - * Unregisters a WiMAX device previously registered for use with - * wimax_add_rm(). - * - * IMPORTANT! Must call before calling unregister_netdev(). - * - * After this function returns, you will not get any more user space - * control requests (via netlink or debugfs) and thus to wimax_dev->ops. - * - * Reentrancy control is ensured by setting the state to - * %__WIMAX_ST_QUIESCING. rfkill operations coming through - * wimax_*rfkill*() will be stopped by the quiescing state; ops coming - * from the rfkill subsystem will be stopped by the support being - * removed by wimax_rfkill_rm(). - */ -void wimax_dev_rm(struct wimax_dev *wimax_dev) -{ - d_fnstart(3, NULL, "(wimax_dev %p)\n", wimax_dev); - - mutex_lock(&wimax_dev->mutex); - __wimax_state_change(wimax_dev, __WIMAX_ST_QUIESCING); - wimax_debugfs_rm(wimax_dev); - wimax_id_table_rm(wimax_dev); - __wimax_state_change(wimax_dev, WIMAX_ST_DOWN); - mutex_unlock(&wimax_dev->mutex); - wimax_rfkill_rm(wimax_dev); - d_fnend(3, NULL, "(wimax_dev %p) = void\n", wimax_dev); -} -EXPORT_SYMBOL_GPL(wimax_dev_rm); - - -/* Debug framework control of debug levels */ -struct d_level D_LEVEL[] = { - D_SUBMODULE_DEFINE(debugfs), - D_SUBMODULE_DEFINE(id_table), - D_SUBMODULE_DEFINE(op_msg), - D_SUBMODULE_DEFINE(op_reset), - D_SUBMODULE_DEFINE(op_rfkill), - D_SUBMODULE_DEFINE(op_state_get), - D_SUBMODULE_DEFINE(stack), -}; -size_t D_LEVEL_SIZE = ARRAY_SIZE(D_LEVEL); - - -static const struct genl_multicast_group wimax_gnl_mcgrps[] = { - { .name = "msg", }, -}; - -struct genl_family wimax_gnl_family __ro_after_init = { - .name = "WiMAX", - .version = WIMAX_GNL_VERSION, - .hdrsize = 0, - .maxattr = WIMAX_GNL_ATTR_MAX, - .policy = wimax_gnl_policy, - .module = THIS_MODULE, - .small_ops = wimax_gnl_ops, - .n_small_ops = ARRAY_SIZE(wimax_gnl_ops), - .mcgrps = wimax_gnl_mcgrps, - .n_mcgrps = ARRAY_SIZE(wimax_gnl_mcgrps), -}; - - - -/* Shutdown the wimax stack */ -static -int __init wimax_subsys_init(void) -{ - int result; - - d_fnstart(4, NULL, "()\n"); - d_parse_params(D_LEVEL, D_LEVEL_SIZE, wimax_debug_params, - "wimax.debug"); - - result = genl_register_family(&wimax_gnl_family); - if (unlikely(result < 0)) { - pr_err("cannot register generic netlink family: %d\n", result); - goto error_register_family; - } - - d_fnend(4, NULL, "() = 0\n"); - return 0; - -error_register_family: - d_fnend(4, NULL, "() = %d\n", result); - return result; - -} -module_init(wimax_subsys_init); - - -/* Shutdown the wimax stack */ -static -void __exit wimax_subsys_exit(void) -{ - wimax_id_table_release(); - genl_unregister_family(&wimax_gnl_family); -} -module_exit(wimax_subsys_exit); - -MODULE_AUTHOR("Intel Corporation "); -MODULE_DESCRIPTION("Linux WiMAX stack"); -MODULE_LICENSE("GPL"); diff --git a/net/wimax/wimax-internal.h b/net/wimax/wimax-internal.h deleted file mode 100644 index 40751207296c..000000000000 --- a/net/wimax/wimax-internal.h +++ /dev/null @@ -1,85 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * Linux WiMAX - * Internal API for kernel space WiMAX stack - * - * Copyright (C) 2007 Intel Corporation - * Inaky Perez-Gonzalez - * - * This header file is for declarations and definitions internal to - * the WiMAX stack. For public APIs and documentation, see - * include/net/wimax.h and include/linux/wimax.h. - */ - -#ifndef __WIMAX_INTERNAL_H__ -#define __WIMAX_INTERNAL_H__ -#ifdef __KERNEL__ - -#ifdef pr_fmt -#undef pr_fmt -#endif - -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt - -#include -#include - - -/* - * Decide if a (locked) device is ready for use - * - * Before using the device structure, it must be locked - * (wimax_dev->mutex). As well, most operations need to call this - * function to check if the state is the right one. - * - * An error value will be returned if the state is not the right - * one. In that case, the caller should not attempt to use the device - * and just unlock it. - */ -static inline __must_check -int wimax_dev_is_ready(struct wimax_dev *wimax_dev) -{ - if (wimax_dev->state == __WIMAX_ST_NULL) - return -EINVAL; /* Device is not even registered! */ - if (wimax_dev->state == WIMAX_ST_DOWN) - return -ENOMEDIUM; - if (wimax_dev->state == __WIMAX_ST_QUIESCING) - return -ESHUTDOWN; - return 0; -} - - -static inline -void __wimax_state_set(struct wimax_dev *wimax_dev, enum wimax_st state) -{ - wimax_dev->state = state; -} -void __wimax_state_change(struct wimax_dev *, enum wimax_st); - -#ifdef CONFIG_DEBUG_FS -void wimax_debugfs_add(struct wimax_dev *); -void wimax_debugfs_rm(struct wimax_dev *); -#else -static inline void wimax_debugfs_add(struct wimax_dev *wimax_dev) {} -static inline void wimax_debugfs_rm(struct wimax_dev *wimax_dev) {} -#endif - -void wimax_id_table_add(struct wimax_dev *); -struct wimax_dev *wimax_dev_get_by_genl_info(struct genl_info *, int); -void wimax_id_table_rm(struct wimax_dev *); -void wimax_id_table_release(void); - -int wimax_rfkill_add(struct wimax_dev *); -void wimax_rfkill_rm(struct wimax_dev *); - -/* generic netlink */ -extern struct genl_family wimax_gnl_family; - -/* ops */ -int wimax_gnl_doit_msg_from_user(struct sk_buff *skb, struct genl_info *info); -int wimax_gnl_doit_reset(struct sk_buff *skb, struct genl_info *info); -int wimax_gnl_doit_rfkill(struct sk_buff *skb, struct genl_info *info); -int wimax_gnl_doit_state_get(struct sk_buff *skb, struct genl_info *info); - -#endif /* #ifdef __KERNEL__ */ -#endif /* #ifndef __WIMAX_INTERNAL_H__ */ -- cgit v1.2.3 From b1e8eb11fb9cf666d8ae36bbcf533233a504c921 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Fri, 23 Oct 2020 18:33:08 +0200 Subject: mac80211: fix kernel-doc markups Some identifiers have different names between their prototypes and the kernel-doc markup. Others need to be fixed, as kernel-doc markups should use this format: identifier - description In the specific case of __sta_info_flush(), add a documentation for sta_info_flush(), as this one is the one used outside sta_info.c. Signed-off-by: Mauro Carvalho Chehab Reviewed-by: Johannes Berg Link: https://lore.kernel.org/r/978d35eef2dc76e21c81931804e4eaefbd6d635e.1603469755.git.mchehab+huawei@kernel.org Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 9 +++++---- include/net/mac80211.h | 7 ++++--- net/mac80211/sta_info.h | 9 ++++++++- 3 files changed, 17 insertions(+), 8 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 661edfc8722e..d5ab8d99739f 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1444,7 +1444,7 @@ int cfg80211_check_station_change(struct wiphy *wiphy, enum cfg80211_station_type statype); /** - * enum station_info_rate_flags - bitrate info flags + * enum rate_info_flags - bitrate info flags * * Used by the driver to indicate the specific rate transmission * type for 802.11n transmissions. @@ -1517,7 +1517,7 @@ struct rate_info { }; /** - * enum station_info_rate_flags - bitrate info flags + * enum bss_param_flags - bitrate info flags * * Used by the driver to indicate the specific rate transmission * type for 802.11n transmissions. @@ -6467,7 +6467,8 @@ void cfg80211_ibss_joined(struct net_device *dev, const u8 *bssid, struct ieee80211_channel *channel, gfp_t gfp); /** - * cfg80211_notify_new_candidate - notify cfg80211 of a new mesh peer candidate + * cfg80211_notify_new_peer_candidate - notify cfg80211 of a new mesh peer + * candidate * * @dev: network device * @macaddr: the MAC address of the new candidate @@ -7606,7 +7607,7 @@ u32 cfg80211_calculate_bitrate(struct rate_info *rate); void cfg80211_unregister_wdev(struct wireless_dev *wdev); /** - * struct cfg80211_ft_event - FT Information Elements + * struct cfg80211_ft_event_params - FT Information Elements * @ies: FT IEs * @ies_len: length of the FT IE in bytes * @target_ap: target AP's MAC address diff --git a/include/net/mac80211.h b/include/net/mac80211.h index e8e295dae744..dcdba96814a2 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -3311,7 +3311,7 @@ enum ieee80211_roc_type { }; /** - * enum ieee80211_reconfig_complete_type - reconfig type + * enum ieee80211_reconfig_type - reconfig type * * This enum is used by the reconfig_complete() callback to indicate what * reconfiguration type was completed. @@ -6334,7 +6334,8 @@ bool ieee80211_tx_prepare_skb(struct ieee80211_hw *hw, int band, struct ieee80211_sta **sta); /** - * Sanity-check and parse the radiotap header of injected frames + * ieee80211_parse_tx_radiotap - Sanity-check and parse the radiotap header + * of injected frames * @skb: packet injected by userspace * @dev: the &struct device of this 802.11 device */ @@ -6389,7 +6390,7 @@ int ieee80211_parse_p2p_noa(const struct ieee80211_p2p_noa_attr *attr, void ieee80211_update_p2p_noa(struct ieee80211_noa_data *data, u32 tsf); /** - * ieee80211_tdls_oper - request userspace to perform a TDLS operation + * ieee80211_tdls_oper_request - request userspace to perform a TDLS operation * @vif: virtual interface * @peer: the peer's destination address * @oper: the requested TDLS operation diff --git a/net/mac80211/sta_info.h b/net/mac80211/sta_info.h index 00ae81e9e1a1..7afd07636b81 100644 --- a/net/mac80211/sta_info.h +++ b/net/mac80211/sta_info.h @@ -785,7 +785,7 @@ int sta_info_init(struct ieee80211_local *local); void sta_info_stop(struct ieee80211_local *local); /** - * sta_info_flush - flush matching STA entries from the STA table + * __sta_info_flush - flush matching STA entries from the STA table * * Returns the number of removed STA entries. * @@ -794,6 +794,13 @@ void sta_info_stop(struct ieee80211_local *local); */ int __sta_info_flush(struct ieee80211_sub_if_data *sdata, bool vlans); +/** + * sta_info_flush - flush matching STA entries from the STA table + * + * Returns the number of removed STA entries. + * + * @sdata: sdata to remove all stations from + */ static inline int sta_info_flush(struct ieee80211_sub_if_data *sdata) { return __sta_info_flush(sdata, false); -- cgit v1.2.3 From 965ae44412f8c0c19945b3f62bc945ad0b15a8aa Mon Sep 17 00:00:00 2001 From: Xin Long Date: Thu, 29 Oct 2020 15:04:58 +0800 Subject: sctp: create udp4 sock and add its encap_rcv This patch is to add the functions to create/release udp4 sock, and set the sock's encap_rcv to process the incoming udp encap sctp packets. In sctp_udp_rcv(), as we can see, all we need to do is fix the transport header for sctp_rcv(), then it would implement the part of rfc6951#section-5.4: "When an encapsulated packet is received, the UDP header is removed. Then, the generic lookup is performed, as done by an SCTP stack whenever a packet is received, to find the association for the received SCTP packet" Note that these functions will be called in the last patch of this patchset when enabling this feature. v1->v2: - Add pr_err() when fails to create udp v4 sock. v2->v3: - Add 'select NET_UDP_TUNNEL' in sctp Kconfig. v3->v4: - No change. v4->v5: - Change to set udp_port to 0 by default. Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Signed-off-by: Jakub Kicinski --- include/net/netns/sctp.h | 5 +++++ include/net/sctp/constants.h | 2 ++ include/net/sctp/sctp.h | 2 ++ net/sctp/Kconfig | 1 + net/sctp/protocol.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 53 insertions(+) (limited to 'include/net') diff --git a/include/net/netns/sctp.h b/include/net/netns/sctp.h index d8d02e4188d1..8cc9aff24329 100644 --- a/include/net/netns/sctp.h +++ b/include/net/netns/sctp.h @@ -22,6 +22,11 @@ struct netns_sctp { */ struct sock *ctl_sock; + /* UDP tunneling listening sock. */ + struct sock *udp4_sock; + /* UDP tunneling listening port. */ + int udp_port; + /* This is the global local address list. * We actively maintain this complete list of addresses on * the system by catching address add/delete events. diff --git a/include/net/sctp/constants.h b/include/net/sctp/constants.h index 122d9e2d8dfd..14a0d22c9113 100644 --- a/include/net/sctp/constants.h +++ b/include/net/sctp/constants.h @@ -286,6 +286,8 @@ enum { SCTP_MAX_GABS = 16 }; * functions simpler to write. */ +#define SCTP_DEFAULT_UDP_PORT 9899 /* default UDP tunneling port */ + /* These are the values for pf exposure, UNUSED is to keep compatible with old * applications by default. */ diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h index 4fc747b778eb..bfd87a0aa2f1 100644 --- a/include/net/sctp/sctp.h +++ b/include/net/sctp/sctp.h @@ -84,6 +84,8 @@ int sctp_copy_local_addr_list(struct net *net, struct sctp_bind_addr *addr, struct sctp_pf *sctp_get_pf_specific(sa_family_t family); int sctp_register_pf(struct sctp_pf *, sa_family_t); void sctp_addr_wq_mgmt(struct net *, struct sctp_sockaddr_entry *, int); +int sctp_udp_sock_start(struct net *net); +void sctp_udp_sock_stop(struct net *net); /* * sctp/socket.c diff --git a/net/sctp/Kconfig b/net/sctp/Kconfig index 39d7fa9569f8..5da599ff84a9 100644 --- a/net/sctp/Kconfig +++ b/net/sctp/Kconfig @@ -11,6 +11,7 @@ menuconfig IP_SCTP select CRYPTO_HMAC select CRYPTO_SHA1 select LIBCRC32C + select NET_UDP_TUNNEL help Stream Control Transmission Protocol diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 25833238fe93..0f7933422d8e 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -44,6 +44,7 @@ #include #include #include +#include #define MAX_SCTP_PORT_HASH_ENTRIES (64 * 1024) @@ -840,6 +841,45 @@ static int sctp_ctl_sock_init(struct net *net) return 0; } +static int sctp_udp_rcv(struct sock *sk, struct sk_buff *skb) +{ + skb_set_transport_header(skb, sizeof(struct udphdr)); + sctp_rcv(skb); + return 0; +} + +int sctp_udp_sock_start(struct net *net) +{ + struct udp_tunnel_sock_cfg tuncfg = {NULL}; + struct udp_port_cfg udp_conf = {0}; + struct socket *sock; + int err; + + udp_conf.family = AF_INET; + udp_conf.local_ip.s_addr = htonl(INADDR_ANY); + udp_conf.local_udp_port = htons(net->sctp.udp_port); + err = udp_sock_create(net, &udp_conf, &sock); + if (err) { + pr_err("Failed to create the SCTP UDP tunneling v4 sock\n"); + return err; + } + + tuncfg.encap_type = 1; + tuncfg.encap_rcv = sctp_udp_rcv; + setup_udp_tunnel_sock(net, sock, &tuncfg); + net->sctp.udp4_sock = sock->sk; + + return 0; +} + +void sctp_udp_sock_stop(struct net *net) +{ + if (net->sctp.udp4_sock) { + udp_tunnel_sock_release(net->sctp.udp4_sock->sk_socket); + net->sctp.udp4_sock = NULL; + } +} + /* Register address family specific functions. */ int sctp_register_af(struct sctp_af *af) { @@ -1271,6 +1311,9 @@ static int __net_init sctp_defaults_init(struct net *net) /* Enable ECN by default. */ net->sctp.ecn_enable = 1; + /* Set UDP tunneling listening port to 0 by default */ + net->sctp.udp_port = 0; + /* Set SCOPE policy to enabled */ net->sctp.scope_policy = SCTP_SCOPE_POLICY_ENABLE; -- cgit v1.2.3 From 9d6ba260a0734d5e56243929a69d57ebbf0cb245 Mon Sep 17 00:00:00 2001 From: Xin Long Date: Thu, 29 Oct 2020 15:04:59 +0800 Subject: sctp: create udp6 sock and set its encap_rcv This patch is to add the udp6 sock part in sctp_udp_sock_start/stop(). udp_conf.use_udp6_rx_checksums is set to true, as: "The SCTP checksum MUST be computed for IPv4 and IPv6, and the UDP checksum SHOULD be computed for IPv4 and IPv6" says in rfc6951#section-5.3. v1->v2: - Add pr_err() when fails to create udp v6 sock. - Add #if IS_ENABLED(CONFIG_IPV6) not to create v6 sock when ipv6 is disabled. Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Signed-off-by: Jakub Kicinski --- include/net/netns/sctp.h | 1 + net/sctp/protocol.c | 26 ++++++++++++++++++++++++++ 2 files changed, 27 insertions(+) (limited to 'include/net') diff --git a/include/net/netns/sctp.h b/include/net/netns/sctp.h index 8cc9aff24329..247b401a0f52 100644 --- a/include/net/netns/sctp.h +++ b/include/net/netns/sctp.h @@ -24,6 +24,7 @@ struct netns_sctp { /* UDP tunneling listening sock. */ struct sock *udp4_sock; + struct sock *udp6_sock; /* UDP tunneling listening port. */ int udp_port; diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 0f7933422d8e..8410c9a3f145 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -869,6 +869,28 @@ int sctp_udp_sock_start(struct net *net) setup_udp_tunnel_sock(net, sock, &tuncfg); net->sctp.udp4_sock = sock->sk; +#if IS_ENABLED(CONFIG_IPV6) + memset(&udp_conf, 0, sizeof(udp_conf)); + + udp_conf.family = AF_INET6; + udp_conf.local_ip6 = in6addr_any; + udp_conf.local_udp_port = htons(net->sctp.udp_port); + udp_conf.use_udp6_rx_checksums = true; + udp_conf.ipv6_v6only = true; + err = udp_sock_create(net, &udp_conf, &sock); + if (err) { + pr_err("Failed to create the SCTP UDP tunneling v6 sock\n"); + udp_tunnel_sock_release(net->sctp.udp4_sock->sk_socket); + net->sctp.udp4_sock = NULL; + return err; + } + + tuncfg.encap_type = 1; + tuncfg.encap_rcv = sctp_udp_rcv; + setup_udp_tunnel_sock(net, sock, &tuncfg); + net->sctp.udp6_sock = sock->sk; +#endif + return 0; } @@ -878,6 +900,10 @@ void sctp_udp_sock_stop(struct net *net) udp_tunnel_sock_release(net->sctp.udp4_sock->sk_socket); net->sctp.udp4_sock = NULL; } + if (net->sctp.udp6_sock) { + udp_tunnel_sock_release(net->sctp.udp6_sock->sk_socket); + net->sctp.udp6_sock = NULL; + } } /* Register address family specific functions. */ -- cgit v1.2.3 From e8a3001c2120c760017da81a307308572a3cdbbc Mon Sep 17 00:00:00 2001 From: Xin Long Date: Thu, 29 Oct 2020 15:05:01 +0800 Subject: sctp: add encap_port for netns sock asoc and transport encap_port is added as per netns/sock/assoc/transport, and the latter one's encap_port inherits the former one's by default. The transport's encap_port value would mostly decide if one packet should go out with udp encapsulated or not. This patch also allows users to set netns' encap_port by sysctl. v1->v2: - Change to define encap_port as __be16 for sctp_sock, asoc and transport. v2->v3: - No change. v3->v4: - Add 'encap_port' entry in ip-sysctl.rst. v4->v5: - Improve the description of encap_port in ip-sysctl.rst. Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Signed-off-by: Jakub Kicinski --- Documentation/networking/ip-sysctl.rst | 16 ++++++++++++++++ include/net/netns/sctp.h | 2 ++ include/net/sctp/structs.h | 6 ++++++ net/sctp/associola.c | 4 ++++ net/sctp/protocol.c | 3 +++ net/sctp/socket.c | 1 + net/sctp/sysctl.c | 10 ++++++++++ 7 files changed, 42 insertions(+) (limited to 'include/net') diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index 25e6673a085a..dad3ba952989 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -2642,6 +2642,22 @@ addr_scope_policy - INTEGER Default: 1 +encap_port - INTEGER + The default remote UDP encapsulation port. + + This value is used to set the dest port of the UDP header for the + outgoing UDP-encapsulated SCTP packets by default. Users can also + change the value for each sock/asoc/transport by using setsockopt. + For further information, please refer to RFC6951. + + Note that when connecting to a remote server, the client should set + this to the port that the UDP tunneling sock on the peer server is + listening to and the local UDP tunneling sock on the client also + must be started. On the server, it would get the encap_port from + the incoming packet's source port. + + Default: 0 + ``/proc/sys/net/core/*`` ======================== diff --git a/include/net/netns/sctp.h b/include/net/netns/sctp.h index 247b401a0f52..a0f315effa94 100644 --- a/include/net/netns/sctp.h +++ b/include/net/netns/sctp.h @@ -27,6 +27,8 @@ struct netns_sctp { struct sock *udp6_sock; /* UDP tunneling listening port. */ int udp_port; + /* UDP tunneling remote encap port. */ + int encap_port; /* This is the global local address list. * We actively maintain this complete list of addresses on diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 0bdff38eb4bb..aa98e7e659ac 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -178,6 +178,8 @@ struct sctp_sock { */ __u32 hbinterval; + __be16 encap_port; + /* This is the max_retrans value for new associations. */ __u16 pathmaxrxt; @@ -877,6 +879,8 @@ struct sctp_transport { */ unsigned long last_time_ecne_reduced; + __be16 encap_port; + /* This is the max_retrans value for the transport and will * be initialized from the assocs value. This can be changed * using the SCTP_SET_PEER_ADDR_PARAMS socket option. @@ -1790,6 +1794,8 @@ struct sctp_association { */ unsigned long hbinterval; + __be16 encap_port; + /* This is the max_retrans value for new transports in the * association. */ diff --git a/net/sctp/associola.c b/net/sctp/associola.c index fdb69d46276d..336df4b36655 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -99,6 +99,8 @@ static struct sctp_association *sctp_association_init( */ asoc->hbinterval = msecs_to_jiffies(sp->hbinterval); + asoc->encap_port = sp->encap_port; + /* Initialize path max retrans value. */ asoc->pathmaxrxt = sp->pathmaxrxt; @@ -624,6 +626,8 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, */ peer->hbinterval = asoc->hbinterval; + peer->encap_port = asoc->encap_port; + /* Set the path max_retrans. */ peer->pathmaxrxt = asoc->pathmaxrxt; diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 4d12a0c869b8..89dfd313e113 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -1359,6 +1359,9 @@ static int __net_init sctp_defaults_init(struct net *net) /* Set UDP tunneling listening port to 0 by default */ net->sctp.udp_port = 0; + /* Set remote encap port to 0 by default */ + net->sctp.encap_port = 0; + /* Set SCOPE policy to enabled */ net->sctp.scope_policy = SCTP_SCOPE_POLICY_ENABLE; diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 53d0a4161df3..09b94cd7ca37 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4876,6 +4876,7 @@ static int sctp_init_sock(struct sock *sk) * be modified via SCTP_PEER_ADDR_PARAMS */ sp->hbinterval = net->sctp.hb_interval; + sp->encap_port = htons(net->sctp.encap_port); sp->pathmaxrxt = net->sctp.max_retrans_path; sp->pf_retrans = net->sctp.pf_retrans; sp->ps_retrans = net->sctp.ps_retrans; diff --git a/net/sctp/sysctl.c b/net/sctp/sysctl.c index c16c80963e55..ecc1b5e4c297 100644 --- a/net/sctp/sysctl.c +++ b/net/sctp/sysctl.c @@ -36,6 +36,7 @@ static int rto_alpha_max = 1000; static int rto_beta_max = 1000; static int pf_expose_max = SCTP_PF_EXPOSE_MAX; static int ps_retrans_max = SCTP_PS_RETRANS_MAX; +static int udp_port_max = 65535; static unsigned long max_autoclose_min = 0; static unsigned long max_autoclose_max = @@ -290,6 +291,15 @@ static struct ctl_table sctp_net_table[] = { .mode = 0644, .proc_handler = proc_dointvec, }, + { + .procname = "encap_port", + .data = &init_net.sctp.encap_port, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + .extra1 = SYSCTL_ZERO, + .extra2 = &udp_port_max, + }, { .procname = "addr_scope_policy", .data = &init_net.sctp.scope_policy, -- cgit v1.2.3 From a1dd2cf2f1aedabc2ca9bb4f90231a521c52d8eb Mon Sep 17 00:00:00 2001 From: Xin Long Date: Thu, 29 Oct 2020 15:05:03 +0800 Subject: sctp: allow changing transport encap_port by peer packets As rfc6951#section-5.4 says: "After finding the SCTP association (which includes checking the verification tag), the UDP source port MUST be stored as the encapsulation port for the destination address the SCTP packet is received from (see Section 5.1). When a non-encapsulated SCTP packet is received by the SCTP stack, the encapsulation of outgoing packets belonging to the same association and the corresponding destination address MUST be disabled." transport encap_port should be updated by a validated incoming packet's udp src port. We save the udp src port in sctp_input_cb->encap_port, and then update the transport in two places: 1. right after vtag is verified, which is required by RFC, and this allows the existent transports to be updated by the chunks that can only be processed on an asoc. 2. right before processing the 'init' where the transports are added, and this allows building a sctp over udp connection by client with the server not knowing the remote encap port. 3. when processing ootb_pkt and creating the temporary transport for the reply pkt. Note that sctp_input_cb->header is removed, as it's not used any more in sctp. v1->v2: - Change encap_port as __be16 for sctp_input_cb. Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Signed-off-by: Jakub Kicinski --- include/net/sctp/sm.h | 1 + include/net/sctp/structs.h | 7 +------ net/sctp/ipv6.c | 1 + net/sctp/protocol.c | 11 ++++++++++- net/sctp/sm_make_chunk.c | 1 + net/sctp/sm_statefuns.c | 2 ++ 6 files changed, 16 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/sm.h b/include/net/sctp/sm.h index 5c491a3bc27e..a499341472ad 100644 --- a/include/net/sctp/sm.h +++ b/include/net/sctp/sm.h @@ -380,6 +380,7 @@ sctp_vtag_verify(const struct sctp_chunk *chunk, if (ntohl(chunk->sctp_hdr->vtag) == asoc->c.my_vtag) return 1; + chunk->transport->encap_port = SCTP_INPUT_CB(chunk->skb)->encap_port; return 0; } diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index aa98e7e659ac..81464ae2b137 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -1120,14 +1120,9 @@ static inline void sctp_outq_cork(struct sctp_outq *q) * sctp_input_cb is currently used on rx and sock rx queue */ struct sctp_input_cb { - union { - struct inet_skb_parm h4; -#if IS_ENABLED(CONFIG_IPV6) - struct inet6_skb_parm h6; -#endif - } header; struct sctp_chunk *chunk; struct sctp_af *af; + __be16 encap_port; }; #define SCTP_INPUT_CB(__skb) ((struct sctp_input_cb *)&((__skb)->cb[0])) diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index 8a58f42d6d19..a064bf252b17 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -1053,6 +1053,7 @@ static struct inet_protosw sctpv6_stream_protosw = { static int sctp6_rcv(struct sk_buff *skb) { + memset(skb->cb, 0, sizeof(skb->cb)); return sctp_rcv(skb) ? -1 : 0; } diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 89dfd313e113..f3de8c03a15e 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -843,6 +843,9 @@ static int sctp_ctl_sock_init(struct net *net) static int sctp_udp_rcv(struct sock *sk, struct sk_buff *skb) { + memset(skb->cb, 0, sizeof(skb->cb)); + SCTP_INPUT_CB(skb)->encap_port = udp_hdr(skb)->source; + skb_set_transport_header(skb, sizeof(struct udphdr)); sctp_rcv(skb); return 0; @@ -1139,9 +1142,15 @@ static struct inet_protosw sctp_stream_protosw = { .flags = SCTP_PROTOSW_FLAG }; +static int sctp4_rcv(struct sk_buff *skb) +{ + memset(skb->cb, 0, sizeof(skb->cb)); + return sctp_rcv(skb); +} + /* Register with IP layer. */ static const struct net_protocol sctp_protocol = { - .handler = sctp_rcv, + .handler = sctp4_rcv, .err_handler = sctp_v4_err, .no_policy = 1, .netns_ok = 1, diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index 9a56ae2f3651..21d0ff1c6ab9 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -2321,6 +2321,7 @@ int sctp_process_init(struct sctp_association *asoc, struct sctp_chunk *chunk, * added as the primary transport. The source address seems to * be a better choice than any of the embedded addresses. */ + asoc->encap_port = SCTP_INPUT_CB(chunk->skb)->encap_port; if (!sctp_assoc_add_peer(asoc, peer_addr, gfp, SCTP_ACTIVE)) goto nomem; diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index c669f8bd1eab..8edab1533057 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -6268,6 +6268,8 @@ static struct sctp_packet *sctp_ootb_pkt_new( if (!transport) goto nomem; + transport->encap_port = SCTP_INPUT_CB(chunk->skb)->encap_port; + /* Cache a route for the transport with the chunk's destination as * the source address. */ -- cgit v1.2.3 From f1bfe8b5415171b5e70c2a47d399c91bd7c2752e Mon Sep 17 00:00:00 2001 From: Xin Long Date: Thu, 29 Oct 2020 15:05:04 +0800 Subject: sctp: add udphdr to overhead when udp_port is set sctp_mtu_payload() is for calculating the frag size before making chunks from a msg. So we should only add udphdr size to overhead when udp socks are listening, as only then sctp can handle the incoming sctp over udp packets and outgoing sctp over udp packets will be possible. Note that we can't do this according to transport->encap_port, as different transports may be set to different values, while the chunks were made before choosing the transport, we could not be able to meet all rfc6951#section-5.6 recommends. v1->v2: - Add udp_port for sctp_sock to avoid a potential race issue, it will be used in xmit path in the next patch. Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Signed-off-by: Jakub Kicinski --- include/net/sctp/sctp.h | 7 +++++-- include/net/sctp/structs.h | 1 + net/sctp/socket.c | 1 + 3 files changed, 7 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h index bfd87a0aa2f1..86f74f2fe6de 100644 --- a/include/net/sctp/sctp.h +++ b/include/net/sctp/sctp.h @@ -578,10 +578,13 @@ static inline __u32 sctp_mtu_payload(const struct sctp_sock *sp, { __u32 overhead = sizeof(struct sctphdr) + extra; - if (sp) + if (sp) { overhead += sp->pf->af->net_header_len; - else + if (sp->udp_port) + overhead += sizeof(struct udphdr); + } else { overhead += sizeof(struct ipv6hdr); + } if (WARN_ON_ONCE(mtu && mtu <= overhead)) mtu = overhead; diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 81464ae2b137..80f71499b543 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -178,6 +178,7 @@ struct sctp_sock { */ __u32 hbinterval; + __be16 udp_port; __be16 encap_port; /* This is the max_retrans value for new associations. */ diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 2a9ee9b3e46c..a710917c5ac7 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4928,6 +4928,7 @@ static int sctp_init_sock(struct sock *sk) * be modified via SCTP_PEER_ADDR_PARAMS */ sp->hbinterval = net->sctp.hb_interval; + sp->udp_port = htons(net->sctp.udp_port); sp->encap_port = htons(net->sctp.encap_port); sp->pathmaxrxt = net->sctp.max_retrans_path; sp->pf_retrans = net->sctp.pf_retrans; -- cgit v1.2.3 From e38d86b354f96e3ed6f5e6c7dbb442d7107fc0bd Mon Sep 17 00:00:00 2001 From: Xin Long Date: Thu, 29 Oct 2020 15:05:08 +0800 Subject: sctp: add the error cause for new encapsulation port restart This patch is to add the function to make the abort chunk with the error cause for new encapsulation port restart, defined on Section 4.4 in draft-tuexen-tsvwg-sctp-udp-encaps-cons-03. v1->v2: - no change. v2->v3: - no need to call htons() when setting nep.cur_port/new_port. Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Signed-off-by: Jakub Kicinski --- include/linux/sctp.h | 20 ++++++++++++++++++++ include/net/sctp/sm.h | 3 +++ net/sctp/sm_make_chunk.c | 20 ++++++++++++++++++++ 3 files changed, 43 insertions(+) (limited to 'include/net') diff --git a/include/linux/sctp.h b/include/linux/sctp.h index 76731230bbc5..bb1926589693 100644 --- a/include/linux/sctp.h +++ b/include/linux/sctp.h @@ -482,11 +482,13 @@ enum sctp_error { * 11 Restart of an association with new addresses * 12 User Initiated Abort * 13 Protocol Violation + * 14 Restart of an Association with New Encapsulation Port */ SCTP_ERROR_RESTART = cpu_to_be16(0x0b), SCTP_ERROR_USER_ABORT = cpu_to_be16(0x0c), SCTP_ERROR_PROTO_VIOLATION = cpu_to_be16(0x0d), + SCTP_ERROR_NEW_ENCAP_PORT = cpu_to_be16(0x0e), /* ADDIP Section 3.3 New Error Causes * @@ -793,4 +795,22 @@ enum { SCTP_FLOWLABEL_VAL_MASK = 0xfffff }; +/* UDP Encapsulation + * draft-tuexen-tsvwg-sctp-udp-encaps-cons-03.html#section-4-4 + * + * The error cause indicating an "Restart of an Association with + * New Encapsulation Port" + * + * 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Cause Code = 14 | Cause Length = 8 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Current Encapsulation Port | New Encapsulation Port | + * +-------------------------------+-------------------------------+ + */ +struct sctp_new_encap_port_hdr { + __be16 cur_port; + __be16 new_port; +}; + #endif /* __LINUX_SCTP_H__ */ diff --git a/include/net/sctp/sm.h b/include/net/sctp/sm.h index a499341472ad..fd223c94589a 100644 --- a/include/net/sctp/sm.h +++ b/include/net/sctp/sm.h @@ -221,6 +221,9 @@ struct sctp_chunk *sctp_make_violation_paramlen( struct sctp_chunk *sctp_make_violation_max_retrans( const struct sctp_association *asoc, const struct sctp_chunk *chunk); +struct sctp_chunk *sctp_make_new_encap_port( + const struct sctp_association *asoc, + const struct sctp_chunk *chunk); struct sctp_chunk *sctp_make_heartbeat(const struct sctp_association *asoc, const struct sctp_transport *transport); struct sctp_chunk *sctp_make_heartbeat_ack(const struct sctp_association *asoc, diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index 21d0ff1c6ab9..f77484df097b 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -1142,6 +1142,26 @@ nodata: return retval; } +struct sctp_chunk *sctp_make_new_encap_port(const struct sctp_association *asoc, + const struct sctp_chunk *chunk) +{ + struct sctp_new_encap_port_hdr nep; + struct sctp_chunk *retval; + + retval = sctp_make_abort(asoc, chunk, + sizeof(struct sctp_errhdr) + sizeof(nep)); + if (!retval) + goto nodata; + + sctp_init_cause(retval, SCTP_ERROR_NEW_ENCAP_PORT, sizeof(nep)); + nep.cur_port = SCTP_INPUT_CB(chunk->skb)->encap_port; + nep.new_port = chunk->transport->encap_port; + sctp_addto_chunk(retval, sizeof(nep), &nep); + +nodata: + return retval; +} + /* Make a HEARTBEAT chunk. */ struct sctp_chunk *sctp_make_heartbeat(const struct sctp_association *asoc, const struct sctp_transport *transport) -- cgit v1.2.3 From fa538f7cf05aab61cd91e01c160d4a09c81b8ffe Mon Sep 17 00:00:00 2001 From: "Jose M. Guisado Gomez" Date: Thu, 22 Oct 2020 21:43:51 +0200 Subject: netfilter: nf_reject: add reject skbuff creation helpers Adds reject skbuff creation helper functions to ipv4/6 nf_reject infrastructure. Use these functions for reject verdict in bridge family. Can be reused by all different families that support reject and will not inject the reject packet through ip local out. Signed-off-by: Jose M. Guisado Gomez Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/ipv4/nf_reject.h | 10 ++ include/net/netfilter/ipv6/nf_reject.h | 9 ++ net/bridge/netfilter/Kconfig | 2 +- net/bridge/netfilter/nft_reject_bridge.c | 195 +------------------------------ net/ipv4/netfilter/nf_reject_ipv4.c | 122 +++++++++++++++++++ net/ipv6/netfilter/nf_reject_ipv6.c | 134 +++++++++++++++++++++ 6 files changed, 280 insertions(+), 192 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/ipv4/nf_reject.h b/include/net/netfilter/ipv4/nf_reject.h index 40e0e0623f46..0d8ff84a2588 100644 --- a/include/net/netfilter/ipv4/nf_reject.h +++ b/include/net/netfilter/ipv4/nf_reject.h @@ -18,4 +18,14 @@ struct iphdr *nf_reject_iphdr_put(struct sk_buff *nskb, void nf_reject_ip_tcphdr_put(struct sk_buff *nskb, const struct sk_buff *oldskb, const struct tcphdr *oth); +struct sk_buff *nf_reject_skb_v4_unreach(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook, u8 code); +struct sk_buff *nf_reject_skb_v4_tcp_reset(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook); + + #endif /* _IPV4_NF_REJECT_H */ diff --git a/include/net/netfilter/ipv6/nf_reject.h b/include/net/netfilter/ipv6/nf_reject.h index 4a3ef9ebdf6f..edcf6d1cd316 100644 --- a/include/net/netfilter/ipv6/nf_reject.h +++ b/include/net/netfilter/ipv6/nf_reject.h @@ -20,4 +20,13 @@ void nf_reject_ip6_tcphdr_put(struct sk_buff *nskb, const struct sk_buff *oldskb, const struct tcphdr *oth, unsigned int otcplen); +struct sk_buff *nf_reject_skb_v6_tcp_reset(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook); +struct sk_buff *nf_reject_skb_v6_unreach(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook, u8 code); + #endif /* _IPV6_NF_REJECT_H */ diff --git a/net/bridge/netfilter/Kconfig b/net/bridge/netfilter/Kconfig index 5040fe43f4b4..e4d287afc2c9 100644 --- a/net/bridge/netfilter/Kconfig +++ b/net/bridge/netfilter/Kconfig @@ -17,7 +17,7 @@ config NFT_BRIDGE_META config NFT_BRIDGE_REJECT tristate "Netfilter nf_tables bridge reject support" - depends on NFT_REJECT && NFT_REJECT_IPV4 && NFT_REJECT_IPV6 + depends on NFT_REJECT help Add support to reject packets. diff --git a/net/bridge/netfilter/nft_reject_bridge.c b/net/bridge/netfilter/nft_reject_bridge.c index deae2c9a0f69..25ca7723c6c2 100644 --- a/net/bridge/netfilter/nft_reject_bridge.c +++ b/net/bridge/netfilter/nft_reject_bridge.c @@ -39,30 +39,6 @@ static void nft_reject_br_push_etherhdr(struct sk_buff *oldskb, } } -static int nft_bridge_iphdr_validate(struct sk_buff *skb) -{ - struct iphdr *iph; - u32 len; - - if (!pskb_may_pull(skb, sizeof(struct iphdr))) - return 0; - - iph = ip_hdr(skb); - if (iph->ihl < 5 || iph->version != 4) - return 0; - - len = ntohs(iph->tot_len); - if (skb->len < len) - return 0; - else if (len < (iph->ihl*4)) - return 0; - - if (!pskb_may_pull(skb, iph->ihl*4)) - return 0; - - return 1; -} - /* We cannot use oldskb->dev, it can be either bridge device (NF_BRIDGE INPUT) * or the bridge port (NF_BRIDGE PREROUTING). */ @@ -72,29 +48,11 @@ static void nft_reject_br_send_v4_tcp_reset(struct net *net, int hook) { struct sk_buff *nskb; - struct iphdr *niph; - const struct tcphdr *oth; - struct tcphdr _oth; - if (!nft_bridge_iphdr_validate(oldskb)) - return; - - oth = nf_reject_ip_tcphdr_get(oldskb, &_oth, hook); - if (!oth) - return; - - nskb = alloc_skb(sizeof(struct iphdr) + sizeof(struct tcphdr) + - LL_MAX_HEADER, GFP_ATOMIC); + nskb = nf_reject_skb_v4_tcp_reset(net, oldskb, dev, hook); if (!nskb) return; - skb_reserve(nskb, LL_MAX_HEADER); - niph = nf_reject_iphdr_put(nskb, oldskb, IPPROTO_TCP, - net->ipv4.sysctl_ip_default_ttl); - nf_reject_ip_tcphdr_put(nskb, oldskb, oth); - niph->tot_len = htons(nskb->len); - ip_send_check(niph); - nft_reject_br_push_etherhdr(oldskb, nskb); br_forward(br_port_get_rcu(dev), nskb, false, true); @@ -106,139 +64,32 @@ static void nft_reject_br_send_v4_unreach(struct net *net, int hook, u8 code) { struct sk_buff *nskb; - struct iphdr *niph; - struct icmphdr *icmph; - unsigned int len; - __wsum csum; - u8 proto; - - if (!nft_bridge_iphdr_validate(oldskb)) - return; - /* IP header checks: fragment. */ - if (ip_hdr(oldskb)->frag_off & htons(IP_OFFSET)) - return; - - /* RFC says return as much as we can without exceeding 576 bytes. */ - len = min_t(unsigned int, 536, oldskb->len); - - if (!pskb_may_pull(oldskb, len)) - return; - - if (pskb_trim_rcsum(oldskb, ntohs(ip_hdr(oldskb)->tot_len))) - return; - - proto = ip_hdr(oldskb)->protocol; - - if (!skb_csum_unnecessary(oldskb) && - nf_reject_verify_csum(proto) && - nf_ip_checksum(oldskb, hook, ip_hdrlen(oldskb), proto)) - return; - - nskb = alloc_skb(sizeof(struct iphdr) + sizeof(struct icmphdr) + - LL_MAX_HEADER + len, GFP_ATOMIC); + nskb = nf_reject_skb_v4_unreach(net, oldskb, dev, hook, code); if (!nskb) return; - skb_reserve(nskb, LL_MAX_HEADER); - niph = nf_reject_iphdr_put(nskb, oldskb, IPPROTO_ICMP, - net->ipv4.sysctl_ip_default_ttl); - - skb_reset_transport_header(nskb); - icmph = skb_put_zero(nskb, sizeof(struct icmphdr)); - icmph->type = ICMP_DEST_UNREACH; - icmph->code = code; - - skb_put_data(nskb, skb_network_header(oldskb), len); - - csum = csum_partial((void *)icmph, len + sizeof(struct icmphdr), 0); - icmph->checksum = csum_fold(csum); - - niph->tot_len = htons(nskb->len); - ip_send_check(niph); - nft_reject_br_push_etherhdr(oldskb, nskb); br_forward(br_port_get_rcu(dev), nskb, false, true); } -static int nft_bridge_ip6hdr_validate(struct sk_buff *skb) -{ - struct ipv6hdr *hdr; - u32 pkt_len; - - if (!pskb_may_pull(skb, sizeof(struct ipv6hdr))) - return 0; - - hdr = ipv6_hdr(skb); - if (hdr->version != 6) - return 0; - - pkt_len = ntohs(hdr->payload_len); - if (pkt_len + sizeof(struct ipv6hdr) > skb->len) - return 0; - - return 1; -} - static void nft_reject_br_send_v6_tcp_reset(struct net *net, struct sk_buff *oldskb, const struct net_device *dev, int hook) { struct sk_buff *nskb; - const struct tcphdr *oth; - struct tcphdr _oth; - unsigned int otcplen; - struct ipv6hdr *nip6h; - - if (!nft_bridge_ip6hdr_validate(oldskb)) - return; - oth = nf_reject_ip6_tcphdr_get(oldskb, &_oth, &otcplen, hook); - if (!oth) - return; - - nskb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(struct tcphdr) + - LL_MAX_HEADER, GFP_ATOMIC); + nskb = nf_reject_skb_v6_tcp_reset(net, oldskb, dev, hook); if (!nskb) return; - skb_reserve(nskb, LL_MAX_HEADER); - nip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_TCP, - net->ipv6.devconf_all->hop_limit); - nf_reject_ip6_tcphdr_put(nskb, oldskb, oth, otcplen); - nip6h->payload_len = htons(nskb->len - sizeof(struct ipv6hdr)); - nft_reject_br_push_etherhdr(oldskb, nskb); br_forward(br_port_get_rcu(dev), nskb, false, true); } -static bool reject6_br_csum_ok(struct sk_buff *skb, int hook) -{ - const struct ipv6hdr *ip6h = ipv6_hdr(skb); - int thoff; - __be16 fo; - u8 proto = ip6h->nexthdr; - - if (skb_csum_unnecessary(skb)) - return true; - - if (ip6h->payload_len && - pskb_trim_rcsum(skb, ntohs(ip6h->payload_len) + sizeof(*ip6h))) - return false; - - ip6h = ipv6_hdr(skb); - thoff = ipv6_skip_exthdr(skb, ((u8*)(ip6h+1) - skb->data), &proto, &fo); - if (thoff < 0 || thoff >= skb->len || (fo & htons(~0x7)) != 0) - return false; - - if (!nf_reject_verify_csum(proto)) - return true; - - return nf_ip6_checksum(skb, hook, thoff, proto) == 0; -} static void nft_reject_br_send_v6_unreach(struct net *net, struct sk_buff *oldskb, @@ -246,49 +97,11 @@ static void nft_reject_br_send_v6_unreach(struct net *net, int hook, u8 code) { struct sk_buff *nskb; - struct ipv6hdr *nip6h; - struct icmp6hdr *icmp6h; - unsigned int len; - - if (!nft_bridge_ip6hdr_validate(oldskb)) - return; - - /* Include "As much of invoking packet as possible without the ICMPv6 - * packet exceeding the minimum IPv6 MTU" in the ICMP payload. - */ - len = min_t(unsigned int, 1220, oldskb->len); - - if (!pskb_may_pull(oldskb, len)) - return; - if (!reject6_br_csum_ok(oldskb, hook)) - return; - - nskb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(struct icmp6hdr) + - LL_MAX_HEADER + len, GFP_ATOMIC); + nskb = nf_reject_skb_v6_unreach(net, oldskb, dev, hook, code); if (!nskb) return; - skb_reserve(nskb, LL_MAX_HEADER); - nip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_ICMPV6, - net->ipv6.devconf_all->hop_limit); - - skb_reset_transport_header(nskb); - icmp6h = skb_put_zero(nskb, sizeof(struct icmp6hdr)); - icmp6h->icmp6_type = ICMPV6_DEST_UNREACH; - icmp6h->icmp6_code = code; - - skb_put_data(nskb, skb_network_header(oldskb), len); - nip6h->payload_len = htons(nskb->len - sizeof(struct ipv6hdr)); - - icmp6h->icmp6_cksum = - csum_ipv6_magic(&nip6h->saddr, &nip6h->daddr, - nskb->len - sizeof(struct ipv6hdr), - IPPROTO_ICMPV6, - csum_partial(icmp6h, - nskb->len - sizeof(struct ipv6hdr), - 0)); - nft_reject_br_push_etherhdr(oldskb, nskb); br_forward(br_port_get_rcu(dev), nskb, false, true); diff --git a/net/ipv4/netfilter/nf_reject_ipv4.c b/net/ipv4/netfilter/nf_reject_ipv4.c index 9dcfa4e461b6..8ca99342879c 100644 --- a/net/ipv4/netfilter/nf_reject_ipv4.c +++ b/net/ipv4/netfilter/nf_reject_ipv4.c @@ -12,6 +12,128 @@ #include #include +static int nf_reject_iphdr_validate(struct sk_buff *skb) +{ + struct iphdr *iph; + u32 len; + + if (!pskb_may_pull(skb, sizeof(struct iphdr))) + return 0; + + iph = ip_hdr(skb); + if (iph->ihl < 5 || iph->version != 4) + return 0; + + len = ntohs(iph->tot_len); + if (skb->len < len) + return 0; + else if (len < (iph->ihl*4)) + return 0; + + if (!pskb_may_pull(skb, iph->ihl*4)) + return 0; + + return 1; +} + +struct sk_buff *nf_reject_skb_v4_tcp_reset(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook) +{ + const struct tcphdr *oth; + struct sk_buff *nskb; + struct iphdr *niph; + struct tcphdr _oth; + + if (!nf_reject_iphdr_validate(oldskb)) + return NULL; + + oth = nf_reject_ip_tcphdr_get(oldskb, &_oth, hook); + if (!oth) + return NULL; + + nskb = alloc_skb(sizeof(struct iphdr) + sizeof(struct tcphdr) + + LL_MAX_HEADER, GFP_ATOMIC); + if (!nskb) + return NULL; + + nskb->dev = (struct net_device *)dev; + + skb_reserve(nskb, LL_MAX_HEADER); + niph = nf_reject_iphdr_put(nskb, oldskb, IPPROTO_TCP, + net->ipv4.sysctl_ip_default_ttl); + nf_reject_ip_tcphdr_put(nskb, oldskb, oth); + niph->tot_len = htons(nskb->len); + ip_send_check(niph); + + return nskb; +} +EXPORT_SYMBOL_GPL(nf_reject_skb_v4_tcp_reset); + +struct sk_buff *nf_reject_skb_v4_unreach(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook, u8 code) +{ + struct sk_buff *nskb; + struct iphdr *niph; + struct icmphdr *icmph; + unsigned int len; + __wsum csum; + u8 proto; + + if (!nf_reject_iphdr_validate(oldskb)) + return NULL; + + /* IP header checks: fragment. */ + if (ip_hdr(oldskb)->frag_off & htons(IP_OFFSET)) + return NULL; + + /* RFC says return as much as we can without exceeding 576 bytes. */ + len = min_t(unsigned int, 536, oldskb->len); + + if (!pskb_may_pull(oldskb, len)) + return NULL; + + if (pskb_trim_rcsum(oldskb, ntohs(ip_hdr(oldskb)->tot_len))) + return NULL; + + proto = ip_hdr(oldskb)->protocol; + + if (!skb_csum_unnecessary(oldskb) && + nf_reject_verify_csum(proto) && + nf_ip_checksum(oldskb, hook, ip_hdrlen(oldskb), proto)) + return NULL; + + nskb = alloc_skb(sizeof(struct iphdr) + sizeof(struct icmphdr) + + LL_MAX_HEADER + len, GFP_ATOMIC); + if (!nskb) + return NULL; + + nskb->dev = (struct net_device *)dev; + + skb_reserve(nskb, LL_MAX_HEADER); + niph = nf_reject_iphdr_put(nskb, oldskb, IPPROTO_ICMP, + net->ipv4.sysctl_ip_default_ttl); + + skb_reset_transport_header(nskb); + icmph = skb_put_zero(nskb, sizeof(struct icmphdr)); + icmph->type = ICMP_DEST_UNREACH; + icmph->code = code; + + skb_put_data(nskb, skb_network_header(oldskb), len); + + csum = csum_partial((void *)icmph, len + sizeof(struct icmphdr), 0); + icmph->checksum = csum_fold(csum); + + niph->tot_len = htons(nskb->len); + ip_send_check(niph); + + return nskb; +} +EXPORT_SYMBOL_GPL(nf_reject_skb_v4_unreach); + const struct tcphdr *nf_reject_ip_tcphdr_get(struct sk_buff *oldskb, struct tcphdr *_oth, int hook) { diff --git a/net/ipv6/netfilter/nf_reject_ipv6.c b/net/ipv6/netfilter/nf_reject_ipv6.c index 4aef6baaa55e..8dcceda6c32a 100644 --- a/net/ipv6/netfilter/nf_reject_ipv6.c +++ b/net/ipv6/netfilter/nf_reject_ipv6.c @@ -12,6 +12,140 @@ #include #include +static bool nf_reject_v6_csum_ok(struct sk_buff *skb, int hook) +{ + const struct ipv6hdr *ip6h = ipv6_hdr(skb); + int thoff; + __be16 fo; + u8 proto = ip6h->nexthdr; + + if (skb_csum_unnecessary(skb)) + return true; + + if (ip6h->payload_len && + pskb_trim_rcsum(skb, ntohs(ip6h->payload_len) + sizeof(*ip6h))) + return false; + + ip6h = ipv6_hdr(skb); + thoff = ipv6_skip_exthdr(skb, ((u8*)(ip6h+1) - skb->data), &proto, &fo); + if (thoff < 0 || thoff >= skb->len || (fo & htons(~0x7)) != 0) + return false; + + if (!nf_reject_verify_csum(proto)) + return true; + + return nf_ip6_checksum(skb, hook, thoff, proto) == 0; +} + +static int nf_reject_ip6hdr_validate(struct sk_buff *skb) +{ + struct ipv6hdr *hdr; + u32 pkt_len; + + if (!pskb_may_pull(skb, sizeof(struct ipv6hdr))) + return 0; + + hdr = ipv6_hdr(skb); + if (hdr->version != 6) + return 0; + + pkt_len = ntohs(hdr->payload_len); + if (pkt_len + sizeof(struct ipv6hdr) > skb->len) + return 0; + + return 1; +} + +struct sk_buff *nf_reject_skb_v6_tcp_reset(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook) +{ + struct sk_buff *nskb; + const struct tcphdr *oth; + struct tcphdr _oth; + unsigned int otcplen; + struct ipv6hdr *nip6h; + + if (!nf_reject_ip6hdr_validate(oldskb)) + return NULL; + + oth = nf_reject_ip6_tcphdr_get(oldskb, &_oth, &otcplen, hook); + if (!oth) + return NULL; + + nskb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(struct tcphdr) + + LL_MAX_HEADER, GFP_ATOMIC); + if (!nskb) + return NULL; + + nskb->dev = (struct net_device *)dev; + + skb_reserve(nskb, LL_MAX_HEADER); + nip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_TCP, + net->ipv6.devconf_all->hop_limit); + nf_reject_ip6_tcphdr_put(nskb, oldskb, oth, otcplen); + nip6h->payload_len = htons(nskb->len - sizeof(struct ipv6hdr)); + + return nskb; +} +EXPORT_SYMBOL_GPL(nf_reject_skb_v6_tcp_reset); + +struct sk_buff *nf_reject_skb_v6_unreach(struct net *net, + struct sk_buff *oldskb, + const struct net_device *dev, + int hook, u8 code) +{ + struct sk_buff *nskb; + struct ipv6hdr *nip6h; + struct icmp6hdr *icmp6h; + unsigned int len; + + if (!nf_reject_ip6hdr_validate(oldskb)) + return NULL; + + /* Include "As much of invoking packet as possible without the ICMPv6 + * packet exceeding the minimum IPv6 MTU" in the ICMP payload. + */ + len = min_t(unsigned int, 1220, oldskb->len); + + if (!pskb_may_pull(oldskb, len)) + return NULL; + + if (!nf_reject_v6_csum_ok(oldskb, hook)) + return NULL; + + nskb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(struct icmp6hdr) + + LL_MAX_HEADER + len, GFP_ATOMIC); + if (!nskb) + return NULL; + + nskb->dev = (struct net_device *)dev; + + skb_reserve(nskb, LL_MAX_HEADER); + nip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_ICMPV6, + net->ipv6.devconf_all->hop_limit); + + skb_reset_transport_header(nskb); + icmp6h = skb_put_zero(nskb, sizeof(struct icmp6hdr)); + icmp6h->icmp6_type = ICMPV6_DEST_UNREACH; + icmp6h->icmp6_code = code; + + skb_put_data(nskb, skb_network_header(oldskb), len); + nip6h->payload_len = htons(nskb->len - sizeof(struct ipv6hdr)); + + icmp6h->icmp6_cksum = + csum_ipv6_magic(&nip6h->saddr, &nip6h->daddr, + nskb->len - sizeof(struct ipv6hdr), + IPPROTO_ICMPV6, + csum_partial(icmp6h, + nskb->len - sizeof(struct ipv6hdr), + 0)); + + return nskb; +} +EXPORT_SYMBOL_GPL(nf_reject_skb_v6_unreach); + const struct tcphdr *nf_reject_ip6_tcphdr_get(struct sk_buff *oldskb, struct tcphdr *otcph, unsigned int *otcplen, int hook) -- cgit v1.2.3 From d97f3bdf7a1c0346d3a272aa756d16633f0b8b3b Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 16 Sep 2020 11:45:24 -0700 Subject: net: Un-hide lockdep_sock_is_held() for !LOCKDEP Currently, variables used only within lockdep expressions are flagged as unused, requiring that these variables' declarations be decorated with either #ifdef or __maybe_unused. This results in ugly code. This commit therefore causes the lockdep_sock_is_held() function to be visible even when lockdep is not enabled, thus removing the need for these decorations. This approach further relies on dead-code elimination to remove any references to functions or variables that are not available in non-lockdep kernels. Signed-off-by: Jakub Kicinski Signed-off-by: Paul E. McKenney --- include/net/sock.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index a5c6ae78df77..198d5486fb09 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1566,13 +1566,11 @@ do { \ lockdep_init_map(&(sk)->sk_lock.dep_map, (name), (key), 0); \ } while (0) -#ifdef CONFIG_LOCKDEP static inline bool lockdep_sock_is_held(const struct sock *sk) { return lockdep_is_held(&sk->sk_lock) || lockdep_is_held(&sk->sk_lock.slock); } -#endif void lock_sock_nested(struct sock *sk, int subclass); -- cgit v1.2.3 From a72e9d5472055ca53faed106dc9a11c6b656e66d Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 16 Sep 2020 11:45:25 -0700 Subject: net: sched: Remove broken definitions and un-hide for !LOCKDEP Currently, variables used only within lockdep expressions are flagged as unused, requiring that these variables' declarations be decorated with either #ifdef or __maybe_unused. This results in ugly code. This commit therefore causes the full definitions of the lockdep_tcf_chain_is_locked() and lockdep_tcf_proto_is_locked() functions to be visible even when lockdep is not enabled, thus removing the need for the previous empty functions that were provided in non-lockdep kernels. This approach further relies on dead-code elimination to remove any references to functions or variables that are not available in non-lockdep kernels. Signed-off-by: Jakub Kicinski -- CC: jhs@mojatatu.com CC: xiyou.wangcong@gmail.com CC: jiri@resnulli.us Signed-off-by: Paul E. McKenney --- include/net/sch_generic.h | 12 ------------ 1 file changed, 12 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index d8fd8676fc72..749db62f6215 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -435,7 +435,6 @@ struct tcf_block { struct mutex proto_destroy_lock; /* Lock for proto_destroy hashtable. */ }; -#ifdef CONFIG_PROVE_LOCKING static inline bool lockdep_tcf_chain_is_locked(struct tcf_chain *chain) { return lockdep_is_held(&chain->filter_chain_lock); @@ -445,17 +444,6 @@ static inline bool lockdep_tcf_proto_is_locked(struct tcf_proto *tp) { return lockdep_is_held(&tp->lock); } -#else -static inline bool lockdep_tcf_chain_is_locked(struct tcf_block *chain) -{ - return true; -} - -static inline bool lockdep_tcf_proto_is_locked(struct tcf_proto *tp) -{ - return true; -} -#endif /* #ifdef CONFIG_PROVE_LOCKING */ #define tcf_chain_dereference(p, chain) \ rcu_dereference_protected(p, lockdep_tcf_chain_is_locked(chain)) -- cgit v1.2.3 From 7e901ee7b6ab0b7c1a5e29b8513af23709285a29 Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Fri, 30 Oct 2020 18:34:12 -0700 Subject: tcp: avoid slow start during fast recovery on new losses During TCP fast recovery, the congestion control in charge is by default the Proportional Rate Reduction (PRR) unless the congestion control module specified otherwise (e.g. BBR). Previously when tcp_packets_in_flight() is below snd_ssthresh PRR would slow start upon receiving an ACK that 1) cumulatively acknowledges retransmitted data and 2) does not detect further lost retransmission Such conditions indicate the repair is in good steady progress after the first round trip of recovery. Otherwise PRR adopts the packet conservation principle to send only the amount that was newly delivered (indicated by this ACK). This patch generalizes the previous design principle to include also the newly sent data beside retransmission: as long as the delivery is making good progress, both retransmission and new data should be accounted to make PRR more cautious in slow starting. Suggested-by: Matt Mathis Suggested-by: Neal Cardwell Signed-off-by: Yuchung Cheng Signed-off-by: Neal Cardwell Signed-off-by: Eric Dumazet Link: https://lore.kernel.org/r/20201031013412.1973112-1-ycheng@google.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 2 +- net/ipv4/tcp_input.c | 9 ++++----- net/ipv4/tcp_recovery.c | 3 ++- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index d4ef5bf94168..4aba0f069b05 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -386,7 +386,7 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, int tcp_child_process(struct sock *parent, struct sock *child, struct sk_buff *skb); void tcp_enter_loss(struct sock *sk); -void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int flag); +void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int newly_lost, int flag); void tcp_clear_retrans(struct tcp_sock *tp); void tcp_update_metrics(struct sock *sk); void tcp_init_metrics(struct sock *sk); diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 389d1b340248..fb3a7750f623 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -2546,7 +2546,7 @@ static bool tcp_try_undo_loss(struct sock *sk, bool frto_undo) * 1) If the packets in flight is larger than ssthresh, PRR spreads the * cwnd reductions across a full RTT. * 2) Otherwise PRR uses packet conservation to send as much as delivered. - * But when the retransmits are acked without further losses, PRR + * But when SND_UNA is acked without further losses, * slow starts cwnd up to ssthresh to speed up the recovery. */ static void tcp_init_cwnd_reduction(struct sock *sk) @@ -2563,7 +2563,7 @@ static void tcp_init_cwnd_reduction(struct sock *sk) tcp_ecn_queue_cwr(tp); } -void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int flag) +void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int newly_lost, int flag) { struct tcp_sock *tp = tcp_sk(sk); int sndcnt = 0; @@ -2577,8 +2577,7 @@ void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int flag) u64 dividend = (u64)tp->snd_ssthresh * tp->prr_delivered + tp->prior_cwnd - 1; sndcnt = div_u64(dividend, tp->prior_cwnd) - tp->prr_out; - } else if ((flag & (FLAG_RETRANS_DATA_ACKED | FLAG_LOST_RETRANS)) == - FLAG_RETRANS_DATA_ACKED) { + } else if (flag & FLAG_SND_UNA_ADVANCED && !newly_lost) { sndcnt = min_t(int, delta, max_t(int, tp->prr_delivered - tp->prr_out, newly_acked_sacked) + 1); @@ -3419,7 +3418,7 @@ static void tcp_cong_control(struct sock *sk, u32 ack, u32 acked_sacked, if (tcp_in_cwnd_reduction(sk)) { /* Reduce cwnd if state mandates */ - tcp_cwnd_reduction(sk, acked_sacked, flag); + tcp_cwnd_reduction(sk, acked_sacked, rs->losses, flag); } else if (tcp_may_raise_cwnd(sk, flag)) { /* Advance cwnd if state allows */ tcp_cong_avoid(sk, ack, acked_sacked); diff --git a/net/ipv4/tcp_recovery.c b/net/ipv4/tcp_recovery.c index f65a3ddd0d58..177307a3081f 100644 --- a/net/ipv4/tcp_recovery.c +++ b/net/ipv4/tcp_recovery.c @@ -153,6 +153,7 @@ void tcp_rack_reo_timeout(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); u32 timeout, prior_inflight; + u32 lost = tp->lost; prior_inflight = tcp_packets_in_flight(tp); tcp_rack_detect_loss(sk, &timeout); @@ -160,7 +161,7 @@ void tcp_rack_reo_timeout(struct sock *sk) if (inet_csk(sk)->icsk_ca_state != TCP_CA_Recovery) { tcp_enter_recovery(sk, false); if (!inet_csk(sk)->icsk_ca_ops->cong_control) - tcp_cwnd_reduction(sk, 1, 0); + tcp_cwnd_reduction(sk, 1, tp->lost - lost, 0); } tcp_xmit_retransmit_queue(sk); } -- cgit v1.2.3 From 987a64850996db22bbcf2c1d0a051446a343fa2c Mon Sep 17 00:00:00 2001 From: Greg Kurz Date: Wed, 23 Sep 2020 22:11:44 +0800 Subject: fs/9p: track open fids This patch adds accounting of open fids in a list hanging off the i_private field of the corresponding inode. This allows faster lookups compared to searching the full 9p client list. The lookup code is modified accordingly. Link: http://lkml.kernel.org/r/20200923141146.90046-3-jianyong.wu@arm.com Signed-off-by: Greg Kurz Signed-off-by: Jianyong Wu Signed-off-by: Dominique Martinet --- fs/9p/fid.c | 32 +++++++++++++++++++++++--------- fs/9p/fid.h | 1 + fs/9p/vfs_dir.c | 3 +++ fs/9p/vfs_file.c | 1 + fs/9p/vfs_inode.c | 6 +++++- fs/9p/vfs_inode_dotl.c | 1 + include/net/9p/client.h | 1 + 7 files changed, 35 insertions(+), 10 deletions(-) (limited to 'include/net') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 3304984c0fad..d11dd430590d 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -39,7 +39,7 @@ void v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) } /** - * v9fs_fid_find_inode - search for a fid off of the client list + * v9fs_fid_find_inode - search for an open fid off of the inode list * @inode: return a fid pointing to a specific inode * @uid: return a fid belonging to the specified user * @@ -47,24 +47,38 @@ void v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) static struct p9_fid *v9fs_fid_find_inode(struct inode *inode, kuid_t uid) { - struct p9_client *clnt = v9fs_inode2v9ses(inode)->clnt; - struct p9_fid *fid, *fidptr, *ret = NULL; - unsigned long flags; + struct hlist_head *h; + struct p9_fid *fid, *ret = NULL; p9_debug(P9_DEBUG_VFS, " inode: %p\n", inode); - spin_lock_irqsave(&clnt->lock, flags); - list_for_each_entry_safe(fid, fidptr, &clnt->fidlist, flist) { - if (uid_eq(fid->uid, uid) && - (inode->i_ino == v9fs_qid2ino(&fid->qid))) { + spin_lock(&inode->i_lock); + h = (struct hlist_head *)&inode->i_private; + hlist_for_each_entry(fid, h, ilist) { + if (uid_eq(fid->uid, uid)) { ret = fid; break; } } - spin_unlock_irqrestore(&clnt->lock, flags); + spin_unlock(&inode->i_lock); return ret; } +/** + * v9fs_open_fid_add - add an open fid to an inode + * @dentry: inode that the fid is being added to + * @fid: fid to add + * + */ + +void v9fs_open_fid_add(struct inode *inode, struct p9_fid *fid) +{ + spin_lock(&inode->i_lock); + hlist_add_head(&fid->ilist, (struct hlist_head *)&inode->i_private); + spin_unlock(&inode->i_lock); +} + + /** * v9fs_fid_find - retrieve a fid that belongs to the specified uid * @dentry: dentry to look for fid in diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 928b1093f511..dfa11df02818 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -15,6 +15,7 @@ static inline struct p9_fid *v9fs_parent_fid(struct dentry *dentry) } void v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid); struct p9_fid *v9fs_writeback_fid(struct dentry *dentry); +void v9fs_open_fid_add(struct inode *inode, struct p9_fid *fid); static inline struct p9_fid *clone_fid(struct p9_fid *fid) { return IS_ERR(fid) ? fid : p9_client_walk(fid, 0, NULL, 1); diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index 674d22bf4f6f..d82d8a346f86 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -210,6 +210,9 @@ int v9fs_dir_release(struct inode *inode, struct file *filp) fid = filp->private_data; p9_debug(P9_DEBUG_VFS, "inode: %p filp: %p fid: %d\n", inode, filp, fid ? fid->fid : -1); + spin_lock(&inode->i_lock); + hlist_del(&fid->ilist); + spin_unlock(&inode->i_lock); if (fid) p9_client_clunk(fid); return 0; diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index b177fd3b1eb3..b0ef225cecd0 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -96,6 +96,7 @@ int v9fs_file_open(struct inode *inode, struct file *file) mutex_unlock(&v9inode->v_mutex); if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) v9fs_cache_inode_set_cookie(inode, file); + v9fs_open_fid_add(inode, fid); return 0; out_error: p9_client_clunk(file->private_data); diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 31c2fddabb82..6b243ffcbcf0 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -256,6 +256,7 @@ int v9fs_init_inode(struct v9fs_session_info *v9ses, inode->i_rdev = rdev; inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode); inode->i_mapping->a_ops = &v9fs_addr_operations; + inode->i_private = NULL; switch (mode & S_IFMT) { case S_IFIFO: @@ -796,6 +797,7 @@ v9fs_vfs_atomic_open(struct inode *dir, struct dentry *dentry, struct v9fs_session_info *v9ses; struct p9_fid *fid, *inode_fid; struct dentry *res = NULL; + struct inode *inode; if (d_in_lookup(dentry)) { res = v9fs_vfs_lookup(dir, dentry, 0); @@ -824,7 +826,8 @@ v9fs_vfs_atomic_open(struct inode *dir, struct dentry *dentry, } v9fs_invalidate_inode_attr(dir); - v9inode = V9FS_I(d_inode(dentry)); + inode = d_inode(dentry); + v9inode = V9FS_I(inode); mutex_lock(&v9inode->v_mutex); if ((v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) && !v9inode->writeback_fid && @@ -852,6 +855,7 @@ v9fs_vfs_atomic_open(struct inode *dir, struct dentry *dentry, file->private_data = fid; if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) v9fs_cache_inode_set_cookie(d_inode(dentry), file); + v9fs_open_fid_add(inode, fid); file->f_mode |= FMODE_CREATED; out: diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 0028eccb665a..08f2e089fb0e 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -342,6 +342,7 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, file->private_data = ofid; if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) v9fs_cache_inode_set_cookie(inode, file); + v9fs_open_fid_add(inode, ofid); file->f_mode |= FMODE_CREATED; out: v9fs_put_acl(dacl, pacl); diff --git a/include/net/9p/client.h b/include/net/9p/client.h index dd5b5bd781a4..ce7882da8e86 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -152,6 +152,7 @@ struct p9_fid { void *rdir; struct hlist_node dlist; /* list of all fids attached to a dentry */ + struct hlist_node ilist; }; /** -- cgit v1.2.3 From 0992d67bc2bcd723d57c6ee7883bda4524450179 Mon Sep 17 00:00:00 2001 From: Guillaume Nault Date: Sat, 31 Oct 2020 01:07:25 +0100 Subject: mpls: drop skb's dst in mpls_forward() Commit 394de110a733 ("net: Added pointer check for dst->ops->neigh_lookup in dst_neigh_lookup_skb") added a test in dst_neigh_lookup_skb() to avoid a NULL pointer dereference. The root cause was the MPLS forwarding code, which doesn't call skb_dst_drop() on incoming packets. That is, if the packet is received from a collect_md device, it has a metadata_dst attached to it that doesn't implement any dst_ops function. To align the MPLS behaviour with IPv4 and IPv6, let's drop the dst in mpls_forward(). This way, dst_neigh_lookup_skb() doesn't need to test ->neigh_lookup any more. Let's keep a WARN condition though, to document the precondition and to ease detection of such problems in the future. Signed-off-by: Guillaume Nault Link: https://lore.kernel.org/r/f8c2784c13faa54469a2aac339470b1049ca6b63.1604102750.git.gnault@redhat.com Signed-off-by: Jakub Kicinski --- include/net/dst.h | 12 +++++------- net/mpls/af_mpls.c | 2 ++ 2 files changed, 7 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/dst.h b/include/net/dst.h index 8ea8812b0b41..10f0a8399867 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -400,14 +400,12 @@ static inline struct neighbour *dst_neigh_lookup(const struct dst_entry *dst, co static inline struct neighbour *dst_neigh_lookup_skb(const struct dst_entry *dst, struct sk_buff *skb) { - struct neighbour *n = NULL; + struct neighbour *n; - /* The packets from tunnel devices (eg bareudp) may have only - * metadata in the dst pointer of skb. Hence a pointer check of - * neigh_lookup is needed. - */ - if (dst->ops->neigh_lookup) - n = dst->ops->neigh_lookup(dst, skb, NULL); + if (WARN_ON_ONCE(!dst->ops->neigh_lookup)) + return NULL; + + n = dst->ops->neigh_lookup(dst, skb, NULL); return IS_ERR(n) ? NULL : n; } diff --git a/net/mpls/af_mpls.c b/net/mpls/af_mpls.c index f2868a8a50c3..47bab701555f 100644 --- a/net/mpls/af_mpls.c +++ b/net/mpls/af_mpls.c @@ -377,6 +377,8 @@ static int mpls_forward(struct sk_buff *skb, struct net_device *dev, if (!pskb_may_pull(skb, sizeof(*hdr))) goto err; + skb_dst_drop(skb); + /* Read and decode the label */ hdr = mpls_hdr(skb); dec = mpls_entry_decode(hdr); -- cgit v1.2.3 From 5a369ca64364b49caff424d2f988901bc7658b6d Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Tue, 3 Nov 2020 11:05:05 -0800 Subject: tcp: propagate MPTCP skb extensions on xmit splits When the TCP stack splits a packet on the write queue, the tail half currently lose the associated skb extensions, and will not carry the DSM on the wire. The above does not cause functional problems and is allowed by the RFC, but interact badly with GRO and RX coalescing, as possible candidates for aggregation will carry different TCP options. This change tries to improve the MPTCP behavior, propagating the skb extensions on split. Additionally, we must prevent the MPTCP stack from updating the mapping after the split occur: that will both violate the RFC and fool the reader. Signed-off-by: Paolo Abeni Signed-off-by: Mat Martineau Signed-off-by: Jakub Kicinski --- include/net/mptcp.h | 21 ++++++++++++++++++++- net/ipv4/tcp_output.c | 3 +++ net/mptcp/protocol.c | 7 +++++-- 3 files changed, 28 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/mptcp.h b/include/net/mptcp.h index 753ba7e755d6..6e706d838e4e 100644 --- a/include/net/mptcp.h +++ b/include/net/mptcp.h @@ -29,7 +29,8 @@ struct mptcp_ext { use_ack:1, ack64:1, mpc_map:1, - __unused:2; + frozen:1, + __unused:1; /* one byte hole */ }; @@ -106,6 +107,19 @@ static inline void mptcp_skb_ext_move(struct sk_buff *to, from->active_extensions = 0; } +static inline void mptcp_skb_ext_copy(struct sk_buff *to, + struct sk_buff *from) +{ + struct mptcp_ext *from_ext; + + from_ext = skb_ext_find(from, SKB_EXT_MPTCP); + if (!from_ext) + return; + + from_ext->frozen = 1; + skb_ext_copy(to, from); +} + static inline bool mptcp_ext_matches(const struct mptcp_ext *to_ext, const struct mptcp_ext *from_ext) { @@ -193,6 +207,11 @@ static inline void mptcp_skb_ext_move(struct sk_buff *to, { } +static inline void mptcp_skb_ext_copy(struct sk_buff *to, + struct sk_buff *from) +{ +} + static inline bool mptcp_skb_can_collapse(const struct sk_buff *to, const struct sk_buff *from) { diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index bf48cd73e967..9abf5a0358d5 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -1569,6 +1569,7 @@ int tcp_fragment(struct sock *sk, enum tcp_queue tcp_queue, if (!buff) return -ENOMEM; /* We'll just try again later. */ skb_copy_decrypted(buff, skb); + mptcp_skb_ext_copy(buff, skb); sk_wmem_queued_add(sk, buff->truesize); sk_mem_charge(sk, buff->truesize); @@ -2123,6 +2124,7 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, if (unlikely(!buff)) return -ENOMEM; skb_copy_decrypted(buff, skb); + mptcp_skb_ext_copy(buff, skb); sk_wmem_queued_add(sk, buff->truesize); sk_mem_charge(sk, buff->truesize); @@ -2393,6 +2395,7 @@ static int tcp_mtu_probe(struct sock *sk) skb = tcp_send_head(sk); skb_copy_decrypted(nskb, skb); + mptcp_skb_ext_copy(nskb, skb); TCP_SKB_CB(nskb)->seq = TCP_SKB_CB(skb)->seq; TCP_SKB_CB(nskb)->end_seq = TCP_SKB_CB(skb)->seq + probe_size; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index f5bacfc55006..25d12183d1ca 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -771,8 +771,11 @@ static bool mptcp_skb_can_collapse_to(u64 write_seq, if (!tcp_skb_can_collapse_to(skb)) return false; - /* can collapse only if MPTCP level sequence is in order */ - return mpext && mpext->data_seq + mpext->data_len == write_seq; + /* can collapse only if MPTCP level sequence is in order and this + * mapping has not been xmitted yet + */ + return mpext && mpext->data_seq + mpext->data_len == write_seq && + !mpext->frozen; } static bool mptcp_frag_can_collapse_to(const struct mptcp_sock *msk, -- cgit v1.2.3 From 01ef09caad66f2e6b2a9d1b92b1db5619fa196f1 Mon Sep 17 00:00:00 2001 From: Kurt Kanzenbach Date: Tue, 3 Nov 2020 08:10:54 +0100 Subject: net: dsa: Add tag handling for Hirschmann Hellcreek switches The Hirschmann Hellcreek TSN switches have a special tagging protocol for frames exchanged between the CPU port and the master interface. The format is a one byte trailer indicating the destination or origin port. It's quite similar to the Micrel KSZ tagging. That's why the implementation is based on that code. Signed-off-by: Kurt Kanzenbach Reviewed-by: Vladimir Oltean Signed-off-by: Jakub Kicinski --- include/net/dsa.h | 2 ++ net/dsa/Kconfig | 6 +++++ net/dsa/Makefile | 1 + net/dsa/tag_hellcreek.c | 66 +++++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 75 insertions(+) create mode 100644 net/dsa/tag_hellcreek.c (limited to 'include/net') diff --git a/include/net/dsa.h b/include/net/dsa.h index 35429a140dfa..04e93bafb7bd 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -45,6 +45,7 @@ struct phylink_link_state; #define DSA_TAG_PROTO_OCELOT_VALUE 15 #define DSA_TAG_PROTO_AR9331_VALUE 16 #define DSA_TAG_PROTO_RTL4_A_VALUE 17 +#define DSA_TAG_PROTO_HELLCREEK_VALUE 18 enum dsa_tag_protocol { DSA_TAG_PROTO_NONE = DSA_TAG_PROTO_NONE_VALUE, @@ -65,6 +66,7 @@ enum dsa_tag_protocol { DSA_TAG_PROTO_OCELOT = DSA_TAG_PROTO_OCELOT_VALUE, DSA_TAG_PROTO_AR9331 = DSA_TAG_PROTO_AR9331_VALUE, DSA_TAG_PROTO_RTL4_A = DSA_TAG_PROTO_RTL4_A_VALUE, + DSA_TAG_PROTO_HELLCREEK = DSA_TAG_PROTO_HELLCREEK_VALUE, }; struct packet_type; diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig index 1f9b9b11008c..d975614f7dd6 100644 --- a/net/dsa/Kconfig +++ b/net/dsa/Kconfig @@ -56,6 +56,12 @@ config NET_DSA_TAG_BRCM_PREPEND Broadcom switches which places the tag before the Ethernet header (prepended). +config NET_DSA_TAG_HELLCREEK + tristate "Tag driver for Hirschmann Hellcreek TSN switches" + help + Say Y or M if you want to enable support for tagging frames + for the Hirschmann Hellcreek TSN switches. + config NET_DSA_TAG_GSWIP tristate "Tag driver for Lantiq / Intel GSWIP switches" help diff --git a/net/dsa/Makefile b/net/dsa/Makefile index 4f47b2025ff5..e25d5457964a 100644 --- a/net/dsa/Makefile +++ b/net/dsa/Makefile @@ -10,6 +10,7 @@ obj-$(CONFIG_NET_DSA_TAG_BRCM_COMMON) += tag_brcm.o obj-$(CONFIG_NET_DSA_TAG_DSA) += tag_dsa.o obj-$(CONFIG_NET_DSA_TAG_EDSA) += tag_edsa.o obj-$(CONFIG_NET_DSA_TAG_GSWIP) += tag_gswip.o +obj-$(CONFIG_NET_DSA_TAG_HELLCREEK) += tag_hellcreek.o obj-$(CONFIG_NET_DSA_TAG_KSZ) += tag_ksz.o obj-$(CONFIG_NET_DSA_TAG_RTL4_A) += tag_rtl4_a.o obj-$(CONFIG_NET_DSA_TAG_LAN9303) += tag_lan9303.o diff --git a/net/dsa/tag_hellcreek.c b/net/dsa/tag_hellcreek.c new file mode 100644 index 000000000000..2061de06eafb --- /dev/null +++ b/net/dsa/tag_hellcreek.c @@ -0,0 +1,66 @@ +// SPDX-License-Identifier: (GPL-2.0 OR MIT) +/* + * net/dsa/tag_hellcreek.c - Hirschmann Hellcreek switch tag format handling + * + * Copyright (C) 2019,2020 Linutronix GmbH + * Author Kurt Kanzenbach + * + * Based on tag_ksz.c. + */ + +#include +#include +#include +#include + +#include "dsa_priv.h" + +#define HELLCREEK_TAG_LEN 1 + +static struct sk_buff *hellcreek_xmit(struct sk_buff *skb, + struct net_device *dev) +{ + struct dsa_port *dp = dsa_slave_to_port(dev); + u8 *tag; + + /* Tag encoding */ + tag = skb_put(skb, HELLCREEK_TAG_LEN); + *tag = BIT(dp->index); + + return skb; +} + +static struct sk_buff *hellcreek_rcv(struct sk_buff *skb, + struct net_device *dev, + struct packet_type *pt) +{ + /* Tag decoding */ + u8 *tag = skb_tail_pointer(skb) - HELLCREEK_TAG_LEN; + unsigned int port = tag[0] & 0x03; + + skb->dev = dsa_master_find_slave(dev, 0, port); + if (!skb->dev) { + netdev_warn(dev, "Failed to get source port: %d\n", port); + return NULL; + } + + pskb_trim_rcsum(skb, skb->len - HELLCREEK_TAG_LEN); + + skb->offload_fwd_mark = true; + + return skb; +} + +static const struct dsa_device_ops hellcreek_netdev_ops = { + .name = "hellcreek", + .proto = DSA_TAG_PROTO_HELLCREEK, + .xmit = hellcreek_xmit, + .rcv = hellcreek_rcv, + .overhead = HELLCREEK_TAG_LEN, + .tail_tag = true, +}; + +MODULE_LICENSE("Dual MIT/GPL"); +MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_HELLCREEK); + +module_dsa_tag_driver(hellcreek_netdev_ops); -- cgit v1.2.3 From e358bef7c392b6a29381b9b8fbf646a88d5968f3 Mon Sep 17 00:00:00 2001 From: Vladimir Oltean Date: Tue, 3 Nov 2020 08:10:55 +0100 Subject: net: dsa: Give drivers the chance to veto certain upper devices Some switches rely on unique pvids to ensure port separation in standalone mode, because they don't have a port forwarding matrix configurable in hardware. So, setups like a group of 2 uppers with the same VLAN, swp0.100 and swp1.100, will cause traffic tagged with VLAN 100 to be autonomously forwarded between these switch ports, in spite of there being no bridge between swp0 and swp1. These drivers need to prevent this from happening. They need to have VLAN filtering enabled in standalone mode (so they'll drop frames tagged with unknown VLANs) and they can only accept an 8021q upper on a port as long as it isn't installed on any other port too. So give them the chance to veto bad user requests. Signed-off-by: Vladimir Oltean [Kurt: Pass info instead of ptr] Signed-off-by: Kurt Kanzenbach Reviewed-by: Florian Fainelli Signed-off-by: Jakub Kicinski --- include/net/dsa.h | 6 ++++++ net/dsa/slave.c | 12 ++++++++++++ 2 files changed, 18 insertions(+) (limited to 'include/net') diff --git a/include/net/dsa.h b/include/net/dsa.h index 04e93bafb7bd..4e60d2610f20 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -536,6 +536,12 @@ struct dsa_switch_ops { void (*get_regs)(struct dsa_switch *ds, int port, struct ethtool_regs *regs, void *p); + /* + * Upper device tracking. + */ + int (*port_prechangeupper)(struct dsa_switch *ds, int port, + struct netdev_notifier_changeupper_info *info); + /* * Bridge integration */ diff --git a/net/dsa/slave.c b/net/dsa/slave.c index c6806eef906f..59c80052e950 100644 --- a/net/dsa/slave.c +++ b/net/dsa/slave.c @@ -2032,10 +2032,22 @@ static int dsa_slave_netdevice_event(struct notifier_block *nb, switch (event) { case NETDEV_PRECHANGEUPPER: { struct netdev_notifier_changeupper_info *info = ptr; + struct dsa_switch *ds; + struct dsa_port *dp; + int err; if (!dsa_slave_dev_check(dev)) return dsa_prevent_bridging_8021q_upper(dev, ptr); + dp = dsa_slave_to_port(dev); + ds = dp->ds; + + if (ds->ops->port_prechangeupper) { + err = ds->ops->port_prechangeupper(ds, dp->index, info); + if (err) + return notifier_from_errno(err); + } + if (is_vlan_dev(info->upper_dev)) return dsa_slave_check_8021q_upper(dev, ptr); break; -- cgit v1.2.3 From 0356010d825eb18d9157408e8ca4ec4613d61398 Mon Sep 17 00:00:00 2001 From: Xin Long Date: Wed, 4 Nov 2020 14:55:32 +0800 Subject: sctp: bring inet(6)_skb_parm back to sctp_input_cb inet(6)_skb_parm was removed from sctp_input_cb by Commit a1dd2cf2f1ae ("sctp: allow changing transport encap_port by peer packets"), as it thought sctp_input_cb->header is not used any more in SCTP. syzbot reported a crash: [ ] BUG: KASAN: use-after-free in decode_session6+0xe7c/0x1580 [ ] [ ] Call Trace: [ ] [ ] dump_stack+0x107/0x163 [ ] kasan_report.cold+0x1f/0x37 [ ] decode_session6+0xe7c/0x1580 [ ] __xfrm_policy_check+0x2fa/0x2850 [ ] sctp_rcv+0x12b0/0x2e30 [ ] sctp6_rcv+0x22/0x40 [ ] ip6_protocol_deliver_rcu+0x2e8/0x1680 [ ] ip6_input_finish+0x7f/0x160 [ ] ip6_input+0x9c/0xd0 [ ] ipv6_rcv+0x28e/0x3c0 It was caused by sctp_input_cb->header/IP6CB(skb) still used in sctp rx path decode_session6() but some members overwritten by sctp6_rcv(). This patch is to fix it by bring inet(6)_skb_parm back to sctp_input_cb and not overwriting it in sctp4/6_rcv() and sctp_udp_rcv(). Reported-by: syzbot+5be8aebb1b7dfa90ef31@syzkaller.appspotmail.com Fixes: a1dd2cf2f1ae ("sctp: allow changing transport encap_port by peer packets") Signed-off-by: Xin Long Acked-by: Marcelo Ricardo Leitner Link: https://lore.kernel.org/r/136c1a7a419341487c504be6d1996928d9d16e02.1604472932.git.lucien.xin@gmail.com Signed-off-by: Jakub Kicinski --- include/net/sctp/structs.h | 6 ++++++ net/sctp/ipv6.c | 2 +- net/sctp/protocol.c | 3 +-- 3 files changed, 8 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 80f71499b543..1aa585216f34 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -1121,6 +1121,12 @@ static inline void sctp_outq_cork(struct sctp_outq *q) * sctp_input_cb is currently used on rx and sock rx queue */ struct sctp_input_cb { + union { + struct inet_skb_parm h4; +#if IS_ENABLED(CONFIG_IPV6) + struct inet6_skb_parm h6; +#endif + } header; struct sctp_chunk *chunk; struct sctp_af *af; __be16 encap_port; diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index 814754d7cab5..c3e89c776e66 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -1074,7 +1074,7 @@ static struct inet_protosw sctpv6_stream_protosw = { static int sctp6_rcv(struct sk_buff *skb) { - memset(skb->cb, 0, sizeof(skb->cb)); + SCTP_INPUT_CB(skb)->encap_port = 0; return sctp_rcv(skb) ? -1 : 0; } diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 41f287a13b54..6f2bbfeec3a4 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -843,7 +843,6 @@ static int sctp_ctl_sock_init(struct net *net) static int sctp_udp_rcv(struct sock *sk, struct sk_buff *skb) { - memset(skb->cb, 0, sizeof(skb->cb)); SCTP_INPUT_CB(skb)->encap_port = udp_hdr(skb)->source; skb_set_transport_header(skb, sizeof(struct udphdr)); @@ -1163,7 +1162,7 @@ static struct inet_protosw sctp_stream_protosw = { static int sctp4_rcv(struct sk_buff *skb) { - memset(skb->cb, 0, sizeof(skb->cb)); + SCTP_INPUT_CB(skb)->encap_port = 0; return sctp_rcv(skb); } -- cgit v1.2.3 From 9f0ffa418483938d25a15f6ad3891389f333bc59 Mon Sep 17 00:00:00 2001 From: Rohan Dutta Date: Tue, 27 Oct 2020 12:09:10 +0200 Subject: cfg80211: Add support to configure SAE PWE value to drivers Add support to configure SAE PWE preference from userspace to drivers in both AP and STA modes. This is needed for cases where the driver takes care of Authentication frame processing (SME in the driver) so that correct enforcement of the acceptable PWE derivation mechanism can be performed. The userspace applications can pass the sae_pwe value using the NL80211_ATTR_SAE_PWE attribute in the NL80211_CMD_CONNECT and NL80211_CMD_START_AP commands to the driver. This allows selection between the hunting-and-pecking loop and hash-to-element options for PWE derivation. For backwards compatibility, this new attribute is optional and if not included, the driver is notified of the value being unspecified. Signed-off-by: Rohan Dutta Signed-off-by: Jouni Malinen Link: https://lore.kernel.org/r/20201027100910.22283-1-jouni@codeaurora.org Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 9 +++++++++ include/uapi/linux/nl80211.h | 26 ++++++++++++++++++++++++++ net/wireless/nl80211.c | 9 +++++++++ 3 files changed, 44 insertions(+) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 661edfc8722e..0ba8d1fa6eb9 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1008,6 +1008,14 @@ struct survey_info { * @sae_pwd: password for SAE authentication (for devices supporting SAE * offload) * @sae_pwd_len: length of SAE password (for devices supporting SAE offload) + * @sae_pwe: The mechanisms allowed for SAE PWE derivation + * NL80211_SAE_PWE_UNSPECIFIED: Not-specified, used to indicate userspace + * did not specify any preference. The driver should follow its + * internal policy in such a scenario. + * NL80211_SAE_PWE_HUNT_AND_PECK: Allow hunting-and-pecking loop only + * NL80211_SAE_PWE_HASH_TO_ELEMENT: Allow hash-to-element only + * NL80211_SAE_PWE_BOTH: Allow either hunting-and-pecking loop + * or hash-to-element */ struct cfg80211_crypto_settings { u32 wpa_versions; @@ -1026,6 +1034,7 @@ struct cfg80211_crypto_settings { const u8 *psk; const u8 *sae_pwd; u8 sae_pwd_len; + enum nl80211_sae_pwe_mechanism sae_pwe; }; /** diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index 47700a2b9af9..2d733effcdaf 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -2527,6 +2527,11 @@ enum nl80211_commands { * override mask. Used with NL80211_ATTR_S1G_CAPABILITY in * NL80211_CMD_ASSOCIATE or NL80211_CMD_CONNECT. * + * @NL80211_ATTR_SAE_PWE: Indicates the mechanism(s) allowed for SAE PWE + * derivation in WPA3-Personal networks which are using SAE authentication. + * This is a u8 attribute that encapsulates one of the values from + * &enum nl80211_sae_pwe_mechanism. + * * @NUM_NL80211_ATTR: total number of nl80211_attrs available * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use @@ -3016,6 +3021,8 @@ enum nl80211_attrs { NL80211_ATTR_S1G_CAPABILITY, NL80211_ATTR_S1G_CAPABILITY_MASK, + NL80211_ATTR_SAE_PWE, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, @@ -7124,4 +7131,23 @@ enum nl80211_unsol_bcast_probe_resp_attributes { NL80211_UNSOL_BCAST_PROBE_RESP_ATTR_MAX = __NL80211_UNSOL_BCAST_PROBE_RESP_ATTR_LAST - 1 }; + +/** + * enum nl80211_sae_pwe_mechanism - The mechanism(s) allowed for SAE PWE + * derivation. Applicable only when WPA3-Personal SAE authentication is + * used. + * + * @NL80211_SAE_PWE_UNSPECIFIED: not specified, used internally to indicate that + * attribute is not present from userspace. + * @NL80211_SAE_PWE_HUNT_AND_PECK: hunting-and-pecking loop only + * @NL80211_SAE_PWE_HASH_TO_ELEMENT: hash-to-element only + * @NL80211_SAE_PWE_BOTH: both hunting-and-pecking loop and hash-to-element + * can be used. + */ +enum nl80211_sae_pwe_mechanism { + NL80211_SAE_PWE_UNSPECIFIED, + NL80211_SAE_PWE_HUNT_AND_PECK, + NL80211_SAE_PWE_HASH_TO_ELEMENT, + NL80211_SAE_PWE_BOTH, +}; #endif /* __LINUX_NL80211_H */ diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 554796a6c6fe..0928ecbe5bd6 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -715,6 +715,9 @@ static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { NLA_POLICY_EXACT_LEN(IEEE80211_S1G_CAPABILITY_LEN), [NL80211_ATTR_S1G_CAPABILITY_MASK] = NLA_POLICY_EXACT_LEN(IEEE80211_S1G_CAPABILITY_LEN), + [NL80211_ATTR_SAE_PWE] = + NLA_POLICY_RANGE(NLA_U8, NL80211_SAE_PWE_HUNT_AND_PECK, + NL80211_SAE_PWE_BOTH), }; /* policy for the key attributes */ @@ -9731,6 +9734,12 @@ static int nl80211_crypto_settings(struct cfg80211_registered_device *rdev, nla_len(info->attrs[NL80211_ATTR_SAE_PASSWORD]); } + if (info->attrs[NL80211_ATTR_SAE_PWE]) + settings->sae_pwe = + nla_get_u8(info->attrs[NL80211_ATTR_SAE_PWE]); + else + settings->sae_pwe = NL80211_SAE_PWE_UNSPECIFIED; + return 0; } -- cgit v1.2.3 From 70debba3ab7d1009e97310268339cee1d5c7d949 Mon Sep 17 00:00:00 2001 From: Pradeep Kumar Chitrapu Date: Tue, 20 Oct 2020 11:31:08 -0700 Subject: mac80211: save HE oper info in BSS config for mesh Currently he_support is set only for AP mode. Storing this information for mesh BSS as well helps driver to determine HE support. Also save HE operation element params in BSS conf so that drivers can access this for any configurations instead of having to parse the beacon to fetch that info. Signed-off-by: Pradeep Kumar Chitrapu Link: https://lore.kernel.org/r/20201020183111.25458-2-pradeepc@codeaurora.org Signed-off-by: Johannes Berg --- include/net/mac80211.h | 3 ++- net/mac80211/mesh.c | 30 ++++++++++++++++++++++++++++++ 2 files changed, 32 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index e8e295dae744..ee72ea5ec861 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -621,7 +621,8 @@ struct ieee80211_fils_discovery { * nontransmitted BSSIDs * @profile_periodicity: the least number of beacon frames need to be received * in order to discover all the nontransmitted BSSIDs in the set. - * @he_oper: HE operation information of the AP we are connected to + * @he_oper: HE operation information of the BSS (AP/Mesh) or of the AP we are + * connected to (STA) * @he_obss_pd: OBSS Packet Detection parameters. * @he_bss_color: BSS coloring settings, if BSS supports HE * @fils_discovery: FILS discovery configuration diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index ce5825d6f1d1..97095b7c9c64 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -667,6 +667,35 @@ void ieee80211_mesh_root_setup(struct ieee80211_if_mesh *ifmsh) } } +static void +ieee80211_mesh_update_bss_params(struct ieee80211_sub_if_data *sdata, + u8 *ie, u8 ie_len) +{ + struct ieee80211_supported_band *sband; + const u8 *cap; + const struct ieee80211_he_operation *he_oper = NULL; + + sband = ieee80211_get_sband(sdata); + if (!sband) + return; + + if (!ieee80211_get_he_iftype_cap(sband, NL80211_IFTYPE_MESH_POINT) || + sdata->vif.bss_conf.chandef.width == NL80211_CHAN_WIDTH_20_NOHT || + sdata->vif.bss_conf.chandef.width == NL80211_CHAN_WIDTH_5 || + sdata->vif.bss_conf.chandef.width == NL80211_CHAN_WIDTH_10) + return; + + sdata->vif.bss_conf.he_support = true; + + cap = cfg80211_find_ext_ie(WLAN_EID_EXT_HE_OPERATION, ie, ie_len); + if (cap && cap[1] >= ieee80211_he_oper_size(&cap[3])) + he_oper = (void *)(cap + 3); + + if (he_oper) + sdata->vif.bss_conf.he_oper.params = + __le32_to_cpu(he_oper->he_oper_params); +} + /** * ieee80211_fill_mesh_addresses - fill addresses of a locally originated mesh frame * @hdr: 802.11 frame header @@ -943,6 +972,7 @@ ieee80211_mesh_build_beacon(struct ieee80211_if_mesh *ifmsh) bcn->tail_len = skb->len; memcpy(bcn->tail, skb->data, bcn->tail_len); + ieee80211_mesh_update_bss_params(sdata, bcn->tail, bcn->tail_len); bcn->meshconf = (struct ieee80211_meshconf_ie *) (bcn->tail + ifmsh->meshconf_offset); -- cgit v1.2.3 From 30df81301c63643fe1c3b9f05a57059c35a6a953 Mon Sep 17 00:00:00 2001 From: Mathy Vanhoef Date: Wed, 4 Nov 2020 10:18:19 +0400 Subject: mac80211: add radiotap flag to assure frames are not reordered Add a new radiotap flag to indicate injected frames must not be reordered relative to other frames that also have this flag set, independent of priority field values in the transmitted frame. Parse this radiotap flag and define and set a corresponding Tx control flag. Note that this flag has recently been standardized as part of an update to radiotap. Signed-off-by: Mathy Vanhoef Link: https://lore.kernel.org/r/20201104061823.197407-2-Mathy.Vanhoef@kuleuven.be Signed-off-by: Johannes Berg --- include/net/ieee80211_radiotap.h | 1 + include/net/mac80211.h | 4 ++++ net/mac80211/tx.c | 3 +++ 3 files changed, 8 insertions(+) (limited to 'include/net') diff --git a/include/net/ieee80211_radiotap.h b/include/net/ieee80211_radiotap.h index 19c00d100096..c0854933e24f 100644 --- a/include/net/ieee80211_radiotap.h +++ b/include/net/ieee80211_radiotap.h @@ -118,6 +118,7 @@ enum ieee80211_radiotap_tx_flags { IEEE80211_RADIOTAP_F_TX_RTS = 0x0004, IEEE80211_RADIOTAP_F_TX_NOACK = 0x0008, IEEE80211_RADIOTAP_F_TX_NOSEQNO = 0x0010, + IEEE80211_RADIOTAP_F_TX_ORDER = 0x0020, }; /* for IEEE80211_RADIOTAP_MCS "have" flags */ diff --git a/include/net/mac80211.h b/include/net/mac80211.h index ee72ea5ec861..4805af149ef5 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -857,6 +857,9 @@ enum mac80211_tx_info_flags { * it can be sent out. * @IEEE80211_TX_CTRL_NO_SEQNO: Do not overwrite the sequence number that * has already been assigned to this frame. + * @IEEE80211_TX_CTRL_DONT_REORDER: This frame should not be reordered + * relative to other frames that have this flag set, independent + * of their QoS TID or other priority field values. * * These flags are used in tx_info->control.flags. */ @@ -869,6 +872,7 @@ enum mac80211_tx_control_flags { IEEE80211_TX_CTRL_SKIP_MPATH_LOOKUP = BIT(5), IEEE80211_TX_INTCFL_NEED_TXPROCESSING = BIT(6), IEEE80211_TX_CTRL_NO_SEQNO = BIT(7), + IEEE80211_TX_CTRL_DONT_REORDER = BIT(8), }; /* diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 8ba10a48ded4..d4e1a2720807 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -2102,6 +2102,9 @@ bool ieee80211_parse_tx_radiotap(struct sk_buff *skb, info->flags |= IEEE80211_TX_CTL_NO_ACK; if (txflags & IEEE80211_RADIOTAP_F_TX_NOSEQNO) info->control.flags |= IEEE80211_TX_CTRL_NO_SEQNO; + if (txflags & IEEE80211_RADIOTAP_F_TX_ORDER) + info->control.flags |= + IEEE80211_TX_CTRL_DONT_REORDER; break; case IEEE80211_RADIOTAP_RATE: -- cgit v1.2.3 From 1c9cac65cecd4f98d8713ec0c737c38a26e5b5c2 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Wed, 4 Nov 2020 15:30:23 +0200 Subject: nexthop: Add nexthop notification data structures Add data structures that will be used for nexthop replace and delete notifications in the previously introduced nexthop notification chain. New data structures are added instead of passing the existing nexthop code structures directly for several reasons. First, the existing structures encode a lot of bookkeeping information which is irrelevant for listeners of the notification chain. Second, the existing structures can be changed without worrying about introducing regressions in listeners since they are not accessed directly by them. Third, listeners of the notification chain do not need to each parse the relatively complex nexthop code structures. They are passing the required information in a simplified way. Note that a single 'has_encap' bit is added instead of the actual encapsulation information since current listeners do not support such nexthops. Changes since RFC: * s/is_encap/has_encap/ Signed-off-by: Ido Schimmel Reviewed-by: David Ahern Signed-off-by: Jakub Kicinski --- include/net/nexthop.h | 35 +++++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) (limited to 'include/net') diff --git a/include/net/nexthop.h b/include/net/nexthop.h index 2fd76a9b6dc8..4a17b040b502 100644 --- a/include/net/nexthop.h +++ b/include/net/nexthop.h @@ -108,6 +108,41 @@ enum nexthop_event_type { NEXTHOP_EVENT_DEL }; +struct nh_notifier_single_info { + struct net_device *dev; + u8 gw_family; + union { + __be32 ipv4; + struct in6_addr ipv6; + }; + u8 is_reject:1, + is_fdb:1, + has_encap:1; +}; + +struct nh_notifier_grp_entry_info { + u8 weight; + u32 id; + struct nh_notifier_single_info nh; +}; + +struct nh_notifier_grp_info { + u16 num_nh; + bool is_fdb; + struct nh_notifier_grp_entry_info nh_entries[]; +}; + +struct nh_notifier_info { + struct net *net; + struct netlink_ext_ack *extack; + u32 id; + bool is_grp; + union { + struct nh_notifier_single_info *nh; + struct nh_notifier_grp_info *nh_grp; + }; +}; + int register_nexthop_notifier(struct net *net, struct notifier_block *nb); int unregister_nexthop_notifier(struct net *net, struct notifier_block *nb); -- cgit v1.2.3 From e95f2592f633a334b175003f13f42d3c217dc657 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Wed, 4 Nov 2020 15:30:28 +0200 Subject: nexthop: Allow setting "offload" and "trap" indications on nexthops Add a function that can be called by device drivers to set "offload" or "trap" indication on nexthops following nexthop notifications. Changes since RFC: * s/nexthop_hw_flags_set/nexthop_set_hw_flags/ Signed-off-by: Ido Schimmel Signed-off-by: Jakub Kicinski --- include/net/nexthop.h | 1 + net/ipv4/nexthop.c | 21 +++++++++++++++++++++ 2 files changed, 22 insertions(+) (limited to 'include/net') diff --git a/include/net/nexthop.h b/include/net/nexthop.h index 4a17b040b502..aa7ac12c35e2 100644 --- a/include/net/nexthop.h +++ b/include/net/nexthop.h @@ -145,6 +145,7 @@ struct nh_notifier_info { int register_nexthop_notifier(struct net *net, struct notifier_block *nb); int unregister_nexthop_notifier(struct net *net, struct notifier_block *nb); +void nexthop_set_hw_flags(struct net *net, u32 id, bool offload, bool trap); /* caller is holding rcu or rtnl; no reference taken to nexthop */ struct nexthop *nexthop_find_by_id(struct net *net, u32 id); diff --git a/net/ipv4/nexthop.c b/net/ipv4/nexthop.c index 1d66f2439063..d1a1600aee18 100644 --- a/net/ipv4/nexthop.c +++ b/net/ipv4/nexthop.c @@ -2081,6 +2081,27 @@ int unregister_nexthop_notifier(struct net *net, struct notifier_block *nb) } EXPORT_SYMBOL(unregister_nexthop_notifier); +void nexthop_set_hw_flags(struct net *net, u32 id, bool offload, bool trap) +{ + struct nexthop *nexthop; + + rcu_read_lock(); + + nexthop = nexthop_find_by_id(net, id); + if (!nexthop) + goto out; + + nexthop->nh_flags &= ~(RTNH_F_OFFLOAD | RTNH_F_TRAP); + if (offload) + nexthop->nh_flags |= RTNH_F_OFFLOAD; + if (trap) + nexthop->nh_flags |= RTNH_F_TRAP; + +out: + rcu_read_unlock(); +} +EXPORT_SYMBOL(nexthop_set_hw_flags); + static void __net_exit nexthop_net_exit(struct net *net) { rtnl_lock(); -- cgit v1.2.3 From 732d167bf5f53a8c1e8c53cf7dbffe2a13f63752 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Wed, 4 Nov 2020 15:30:29 +0200 Subject: nexthop: Emit a notification when a nexthop is added Emit a notification in the nexthop notification chain when a new nexthop is added (not replaced). The nexthop can either be a new group or a single nexthop. The notification is sent after the nexthop is inserted into the red-black tree, as listeners might need to callback into the nexthop code with the nexthop ID in order to mark the nexthop as offloaded. A 'REPLACE' notification is emitted instead of 'ADD' as the distinction between the two is not important for in-kernel listeners. In case the listener is not familiar with the encoded nexthop ID, it can simply treat it as a new one. This is also consistent with the route offload API. Changes since RFC: * Reword commit message Signed-off-by: Ido Schimmel Signed-off-by: Jakub Kicinski --- include/net/nexthop.h | 3 ++- net/ipv4/nexthop.c | 6 +++++- 2 files changed, 7 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/nexthop.h b/include/net/nexthop.h index aa7ac12c35e2..9c85199b826e 100644 --- a/include/net/nexthop.h +++ b/include/net/nexthop.h @@ -105,7 +105,8 @@ struct nexthop { }; enum nexthop_event_type { - NEXTHOP_EVENT_DEL + NEXTHOP_EVENT_DEL, + NEXTHOP_EVENT_REPLACE, }; struct nh_notifier_single_info { diff --git a/net/ipv4/nexthop.c b/net/ipv4/nexthop.c index d1a1600aee18..4e9d0395f959 100644 --- a/net/ipv4/nexthop.c +++ b/net/ipv4/nexthop.c @@ -1278,7 +1278,11 @@ static int insert_nexthop(struct net *net, struct nexthop *new_nh, rb_link_node_rcu(&new_nh->rb_node, parent, pp); rb_insert_color(&new_nh->rb_node, root); - rc = 0; + + rc = call_nexthop_notifiers(net, NEXTHOP_EVENT_REPLACE, new_nh, extack); + if (rc) + rb_erase(&new_nh->rb_node, &net->nexthop.rb_root); + out: if (!rc) { nh_base_seq_inc(net); -- cgit v1.2.3 From ce7e9c8a080b6daca646a7740d916ed547503a12 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Wed, 4 Nov 2020 15:30:34 +0200 Subject: nexthop: Pass extack to register_nexthop_notifier() This will be used by the next patch which extends the function to replay all the existing nexthops to the notifier block being registered. Device drivers will be able to pass extack to the function since it is passed to them upon reload from devlink. Signed-off-by: Ido Schimmel Reviewed-by: David Ahern Signed-off-by: Jakub Kicinski --- drivers/net/vxlan.c | 3 ++- include/net/nexthop.h | 3 ++- net/ipv4/nexthop.c | 3 ++- 3 files changed, 6 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c index b9db20b4ebfd..183e708dc6cb 100644 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -4711,7 +4711,8 @@ static __net_init int vxlan_init_net(struct net *net) for (h = 0; h < PORT_HASH_SIZE; ++h) INIT_HLIST_HEAD(&vn->sock_list[h]); - return register_nexthop_notifier(net, &vn->nexthop_notifier_block); + return register_nexthop_notifier(net, &vn->nexthop_notifier_block, + NULL); } static void vxlan_destroy_tunnels(struct net *net, struct list_head *head) diff --git a/include/net/nexthop.h b/include/net/nexthop.h index 9c85199b826e..226930d66b63 100644 --- a/include/net/nexthop.h +++ b/include/net/nexthop.h @@ -144,7 +144,8 @@ struct nh_notifier_info { }; }; -int register_nexthop_notifier(struct net *net, struct notifier_block *nb); +int register_nexthop_notifier(struct net *net, struct notifier_block *nb, + struct netlink_ext_ack *extack); int unregister_nexthop_notifier(struct net *net, struct notifier_block *nb); void nexthop_set_hw_flags(struct net *net, u32 id, bool offload, bool trap); diff --git a/net/ipv4/nexthop.c b/net/ipv4/nexthop.c index d5a2c28bea49..9a235d5c5670 100644 --- a/net/ipv4/nexthop.c +++ b/net/ipv4/nexthop.c @@ -2118,7 +2118,8 @@ static struct notifier_block nh_netdev_notifier = { .notifier_call = nh_netdev_event, }; -int register_nexthop_notifier(struct net *net, struct notifier_block *nb) +int register_nexthop_notifier(struct net *net, struct notifier_block *nb, + struct netlink_ext_ack *extack) { return blocking_notifier_chain_register(&net->nexthop.notifier_chain, nb); -- cgit v1.2.3 From 6e1978a9a941d342a8b2e6432f52e07d4af86e21 Mon Sep 17 00:00:00 2001 From: Allen Pais Date: Tue, 3 Nov 2020 14:48:21 +0530 Subject: net: sched: convert tasklets to use new tasklet_setup() API In preparation for unconditionally passing the struct tasklet_struct pointer to all tasklet callbacks, switch to using the new tasklet_setup() and from_tasklet() to pass the tasklet pointer explicitly. Signed-off-by: Romain Perier Signed-off-by: Allen Pais Signed-off-by: Jakub Kicinski --- include/net/pkt_sched.h | 5 +++++ net/sched/sch_atm.c | 8 ++++---- 2 files changed, 9 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 4ed32e6b0201..15b1b30f454e 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -24,6 +24,11 @@ static inline void *qdisc_priv(struct Qdisc *q) return &q->privdata; } +static inline struct Qdisc *qdisc_from_priv(void *priv) +{ + return container_of(priv, struct Qdisc, privdata); +} + /* Timer resolution MUST BE < 10% of min_schedulable_packet_size/bandwidth diff --git a/net/sched/sch_atm.c b/net/sched/sch_atm.c index 1c281cc81f57..007bd2d9f1ff 100644 --- a/net/sched/sch_atm.c +++ b/net/sched/sch_atm.c @@ -466,10 +466,10 @@ drop: __maybe_unused * non-ATM interfaces. */ -static void sch_atm_dequeue(unsigned long data) +static void sch_atm_dequeue(struct tasklet_struct *t) { - struct Qdisc *sch = (struct Qdisc *)data; - struct atm_qdisc_data *p = qdisc_priv(sch); + struct atm_qdisc_data *p = from_tasklet(p, t, task); + struct Qdisc *sch = qdisc_from_priv(p); struct atm_flow_data *flow; struct sk_buff *skb; @@ -563,7 +563,7 @@ static int atm_tc_init(struct Qdisc *sch, struct nlattr *opt, if (err) return err; - tasklet_init(&p->task, sch_atm_dequeue, (unsigned long)sch); + tasklet_setup(&p->task, sch_atm_dequeue); return 0; } -- cgit v1.2.3 From 682036b2b9fbcf98e333bad717968ad6650a0d94 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Sat, 7 Nov 2020 21:55:11 +0100 Subject: net: remove ip_tunnel_get_stats64 After having migrated all users remove ip_tunnel_get_stats64(). Signed-off-by: Heiner Kallweit Signed-off-by: Jakub Kicinski --- include/net/ip_tunnels.h | 2 -- net/ipv4/ip_tunnel_core.c | 9 --------- 2 files changed, 11 deletions(-) (limited to 'include/net') diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h index 02ccd32542d0..1b7905eb793e 100644 --- a/include/net/ip_tunnels.h +++ b/include/net/ip_tunnels.h @@ -274,8 +274,6 @@ int ip_tunnel_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd); int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict); int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu); -void ip_tunnel_get_stats64(struct net_device *dev, - struct rtnl_link_stats64 *tot); struct ip_tunnel *ip_tunnel_lookup(struct ip_tunnel_net *itn, int link, __be16 flags, __be32 remote, __be32 local, diff --git a/net/ipv4/ip_tunnel_core.c b/net/ipv4/ip_tunnel_core.c index 25f1caf5abf9..923a9fa2ecfc 100644 --- a/net/ipv4/ip_tunnel_core.c +++ b/net/ipv4/ip_tunnel_core.c @@ -429,15 +429,6 @@ int skb_tunnel_check_pmtu(struct sk_buff *skb, struct dst_entry *encap_dst, } EXPORT_SYMBOL(skb_tunnel_check_pmtu); -/* Often modified stats are per cpu, other are shared (netdev->stats) */ -void ip_tunnel_get_stats64(struct net_device *dev, - struct rtnl_link_stats64 *tot) -{ - netdev_stats_to_stats64(tot, &dev->stats); - dev_fetch_sw_netstats(tot, dev->tstats); -} -EXPORT_SYMBOL_GPL(ip_tunnel_get_stats64); - static const struct nla_policy ip_tun_policy[LWTUNNEL_IP_MAX + 1] = { [LWTUNNEL_IP_UNSPEC] = { .strict_start_type = LWTUNNEL_IP_OPTS }, [LWTUNNEL_IP_ID] = { .type = NLA_U64 }, -- cgit v1.2.3 From d6bb2d1e86903d3fbf895752ac2c0c2465534579 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 9 Nov 2020 15:13:48 -0800 Subject: inet: constify inet_sdif() argument inet_sdif() does not modify the skb. This will permit propagating the const qualifier in udp{4|6}_lib_lookup_skb() functions. Signed-off-by: Eric Dumazet Acked-by: Alexander Lobakin Signed-off-by: Jakub Kicinski --- include/net/ip.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/ip.h b/include/net/ip.h index 2d6b985d11cc..e20874059f82 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -99,7 +99,7 @@ static inline void ipcm_init_sk(struct ipcm_cookie *ipcm, #define PKTINFO_SKB_CB(skb) ((struct in_pktinfo *)((skb)->cb)) /* return enslaved device index if relevant */ -static inline int inet_sdif(struct sk_buff *skb) +static inline int inet_sdif(const struct sk_buff *skb) { #if IS_ENABLED(CONFIG_NET_L3_MASTER_DEV) if (skb && ipv4_l3mdev_skb(IPCB(skb)->flags)) -- cgit v1.2.3 From 7b58e63e744cbcdeafe0a52423014fd9c9f7e346 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 9 Nov 2020 15:13:49 -0800 Subject: inet: udp{4|6}_lib_lookup_skb() skb argument is const The skb is needed only to fetch the keys for the lookup. Both functions are used from GRO stack, we do not want accidental modification of the skb. Signed-off-by: Eric Dumazet Acked-by: Alexander Lobakin Signed-off-by: Jakub Kicinski --- include/net/udp.h | 6 +++--- net/ipv4/udp.c | 2 +- net/ipv6/udp.c | 2 +- 3 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/include/net/udp.h b/include/net/udp.h index 295d52a73598..877832bed471 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -164,7 +164,7 @@ static inline void udp_csum_pull_header(struct sk_buff *skb) UDP_SKB_CB(skb)->cscov -= sizeof(struct udphdr); } -typedef struct sock *(*udp_lookup_t)(struct sk_buff *skb, __be16 sport, +typedef struct sock *(*udp_lookup_t)(const struct sk_buff *skb, __be16 sport, __be16 dport); INDIRECT_CALLABLE_DECLARE(struct sk_buff *udp4_gro_receive(struct list_head *, @@ -313,7 +313,7 @@ struct sock *udp4_lib_lookup(struct net *net, __be32 saddr, __be16 sport, struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr, __be16 sport, __be32 daddr, __be16 dport, int dif, int sdif, struct udp_table *tbl, struct sk_buff *skb); -struct sock *udp4_lib_lookup_skb(struct sk_buff *skb, +struct sock *udp4_lib_lookup_skb(const struct sk_buff *skb, __be16 sport, __be16 dport); struct sock *udp6_lib_lookup(struct net *net, const struct in6_addr *saddr, __be16 sport, @@ -324,7 +324,7 @@ struct sock *__udp6_lib_lookup(struct net *net, const struct in6_addr *daddr, __be16 dport, int dif, int sdif, struct udp_table *tbl, struct sk_buff *skb); -struct sock *udp6_lib_lookup_skb(struct sk_buff *skb, +struct sock *udp6_lib_lookup_skb(const struct sk_buff *skb, __be16 sport, __be16 dport); /* UDP uses skb->dev_scratch to cache as much information as possible and avoid diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 0d5cd6905749..c732f5acf720 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -541,7 +541,7 @@ static inline struct sock *__udp4_lib_lookup_skb(struct sk_buff *skb, inet_sdif(skb), udptable, skb); } -struct sock *udp4_lib_lookup_skb(struct sk_buff *skb, +struct sock *udp4_lib_lookup_skb(const struct sk_buff *skb, __be16 sport, __be16 dport) { const struct iphdr *iph = ip_hdr(skb); diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 559611bef0e6..e152f8000db4 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -276,7 +276,7 @@ static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb, inet6_sdif(skb), udptable, skb); } -struct sock *udp6_lib_lookup_skb(struct sk_buff *skb, +struct sock *udp6_lib_lookup_skb(const struct sk_buff *skb, __be16 sport, __be16 dport) { const struct ipv6hdr *iph = ipv6_hdr(skb); -- cgit v1.2.3 From e7e0517c1004991908bc7f20b4c9a7b678277358 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 9 Nov 2020 10:57:47 +0100 Subject: cfg80211: remove WDS code Remove all the code that was there to configure WDS interfaces, now that there's no way to reach it anymore. Link: https://lore.kernel.org/r/20201109105103.8f5b98e4068d.I5f5129041649ef2862b69683574bb3344743727b@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 5 ----- include/uapi/linux/nl80211.h | 3 ++- net/wireless/chan.c | 6 +++--- net/wireless/core.c | 8 +------ net/wireless/nl80211.c | 36 ++----------------------------- net/wireless/rdev-ops.h | 10 --------- net/wireless/trace.h | 5 ----- net/wireless/util.c | 5 ++--- net/wireless/wext-compat.c | 51 -------------------------------------------- 9 files changed, 10 insertions(+), 119 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 0ba8d1fa6eb9..4ff804a8bc1d 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -3745,8 +3745,6 @@ struct mgmt_frame_regs { * @get_tx_power: store the current TX power into the dbm variable; * return 0 if successful * - * @set_wds_peer: set the WDS peer for a WDS interface - * * @rfkill_poll: polls the hw rfkill line, use cfg80211 reporting * functions to adjust rfkill hw state * @@ -4067,9 +4065,6 @@ struct cfg80211_ops { int (*get_tx_power)(struct wiphy *wiphy, struct wireless_dev *wdev, int *dbm); - int (*set_wds_peer)(struct wiphy *wiphy, struct net_device *dev, - const u8 *addr); - void (*rfkill_poll)(struct wiphy *wiphy); #ifdef CONFIG_NL80211_TESTMODE diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index e1e5b3d4dd81..3e0d4a038ab6 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -757,7 +757,8 @@ * of any other interfaces, and other interfaces will again take * precedence when they are used. * - * @NL80211_CMD_SET_WDS_PEER: Set the MAC address of the peer on a WDS interface. + * @NL80211_CMD_SET_WDS_PEER: Set the MAC address of the peer on a WDS interface + * (no longer supported). * * @NL80211_CMD_SET_MULTICAST_TO_UNICAST: Configure if this AP should perform * multicast to unicast conversion. When enabled, all multicast packets diff --git a/net/wireless/chan.c b/net/wireless/chan.c index 22d1779ab2b1..e4030f1fbc60 100644 --- a/net/wireless/chan.c +++ b/net/wireless/chan.c @@ -530,10 +530,10 @@ int cfg80211_chandef_dfs_required(struct wiphy *wiphy, case NL80211_IFTYPE_P2P_CLIENT: case NL80211_IFTYPE_MONITOR: case NL80211_IFTYPE_AP_VLAN: - case NL80211_IFTYPE_WDS: case NL80211_IFTYPE_P2P_DEVICE: case NL80211_IFTYPE_NAN: break; + case NL80211_IFTYPE_WDS: case NL80211_IFTYPE_UNSPECIFIED: case NUM_NL80211_IFTYPES: WARN_ON(1); @@ -677,12 +677,12 @@ bool cfg80211_beaconing_iface_active(struct wireless_dev *wdev) case NL80211_IFTYPE_P2P_CLIENT: case NL80211_IFTYPE_MONITOR: case NL80211_IFTYPE_AP_VLAN: - case NL80211_IFTYPE_WDS: case NL80211_IFTYPE_P2P_DEVICE: /* Can NAN type be considered as beaconing interface? */ case NL80211_IFTYPE_NAN: break; case NL80211_IFTYPE_UNSPECIFIED: + case NL80211_IFTYPE_WDS: case NUM_NL80211_IFTYPES: WARN_ON(1); } @@ -1324,12 +1324,12 @@ cfg80211_get_chan_state(struct wireless_dev *wdev, break; case NL80211_IFTYPE_MONITOR: case NL80211_IFTYPE_AP_VLAN: - case NL80211_IFTYPE_WDS: case NL80211_IFTYPE_P2P_DEVICE: case NL80211_IFTYPE_NAN: /* these interface types don't really have a channel */ return; case NL80211_IFTYPE_UNSPECIFIED: + case NL80211_IFTYPE_WDS: case NUM_NL80211_IFTYPES: WARN_ON(1); } diff --git a/net/wireless/core.c b/net/wireless/core.c index 9f23923e8d29..f6b744e91ff4 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -631,10 +631,8 @@ static int wiphy_verify_combinations(struct wiphy *wiphy) return -EINVAL; } -#ifndef CONFIG_WIRELESS_WDS if (WARN_ON(all_iftypes & BIT(NL80211_IFTYPE_WDS))) return -EINVAL; -#endif /* You can't even choose that many! */ if (WARN_ON(cnt < c->max_interfaces)) @@ -675,10 +673,8 @@ int wiphy_register(struct wiphy *wiphy) !(wiphy->nan_supported_bands & BIT(NL80211_BAND_2GHZ))))) return -EINVAL; -#ifndef CONFIG_WIRELESS_WDS if (WARN_ON(wiphy->interface_modes & BIT(NL80211_IFTYPE_WDS))) return -EINVAL; -#endif if (WARN_ON(wiphy->pmsr_capa && !wiphy->pmsr_capa->ftm.supported)) return -EINVAL; @@ -1202,9 +1198,6 @@ void __cfg80211_leave(struct cfg80211_registered_device *rdev, case NL80211_IFTYPE_OCB: __cfg80211_leave_ocb(rdev, dev); break; - case NL80211_IFTYPE_WDS: - /* must be handled by mac80211/driver, has no APIs */ - break; case NL80211_IFTYPE_P2P_DEVICE: case NL80211_IFTYPE_NAN: /* cannot happen, has no netdev */ @@ -1214,6 +1207,7 @@ void __cfg80211_leave(struct cfg80211_registered_device *rdev, /* nothing to do */ break; case NL80211_IFTYPE_UNSPECIFIED: + case NL80211_IFTYPE_WDS: case NUM_NL80211_IFTYPES: /* invalid */ break; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index aad37e7c7f91..b76bdc8417c4 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -1885,7 +1885,6 @@ static int nl80211_add_commands_unsplit(struct cfg80211_registered_device *rdev, if (nla_put_u32(msg, i, NL80211_CMD_SET_CHANNEL)) goto nla_put_failure; } - CMD(set_wds_peer, SET_WDS_PEER); if (rdev->wiphy.flags & WIPHY_FLAG_SUPPORTS_TDLS) { CMD(tdls_mgmt, TDLS_MGMT); CMD(tdls_oper, TDLS_OPER); @@ -2863,8 +2862,8 @@ static int parse_txq_params(struct nlattr *tb[], static bool nl80211_can_set_dev_channel(struct wireless_dev *wdev) { /* - * You can only set the channel explicitly for WDS interfaces, - * all others have their channel managed via their respective + * You can only set the channel explicitly for some interfaces, + * most have their channel managed via their respective * "establish a connection" command (connect, join, ...) * * For AP/GO and mesh mode, the channel can be set with the @@ -3069,29 +3068,6 @@ static int nl80211_set_channel(struct sk_buff *skb, struct genl_info *info) return __nl80211_set_channel(rdev, netdev, info); } -static int nl80211_set_wds_peer(struct sk_buff *skb, struct genl_info *info) -{ - struct cfg80211_registered_device *rdev = info->user_ptr[0]; - struct net_device *dev = info->user_ptr[1]; - struct wireless_dev *wdev = dev->ieee80211_ptr; - const u8 *bssid; - - if (!info->attrs[NL80211_ATTR_MAC]) - return -EINVAL; - - if (netif_running(dev)) - return -EBUSY; - - if (!rdev->ops->set_wds_peer) - return -EOPNOTSUPP; - - if (wdev->iftype != NL80211_IFTYPE_WDS) - return -EOPNOTSUPP; - - bssid = nla_data(info->attrs[NL80211_ATTR_MAC]); - return rdev_set_wds_peer(rdev, dev, bssid); -} - static int nl80211_set_wiphy(struct sk_buff *skb, struct genl_info *info) { struct cfg80211_registered_device *rdev; @@ -15173,14 +15149,6 @@ static const struct genl_small_ops nl80211_small_ops[] = { .internal_flags = NL80211_FLAG_NEED_NETDEV | NL80211_FLAG_NEED_RTNL, }, - { - .cmd = NL80211_CMD_SET_WDS_PEER, - .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .doit = nl80211_set_wds_peer, - .flags = GENL_UNS_ADMIN_PERM, - .internal_flags = NL80211_FLAG_NEED_NETDEV | - NL80211_FLAG_NEED_RTNL, - }, { .cmd = NL80211_CMD_JOIN_MESH, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, diff --git a/net/wireless/rdev-ops.h b/net/wireless/rdev-ops.h index 950d57494168..5e2f349c92a8 100644 --- a/net/wireless/rdev-ops.h +++ b/net/wireless/rdev-ops.h @@ -582,16 +582,6 @@ static inline int rdev_get_tx_power(struct cfg80211_registered_device *rdev, return ret; } -static inline int rdev_set_wds_peer(struct cfg80211_registered_device *rdev, - struct net_device *dev, const u8 *addr) -{ - int ret; - trace_rdev_set_wds_peer(&rdev->wiphy, dev, addr); - ret = rdev->ops->set_wds_peer(&rdev->wiphy, dev, addr); - trace_rdev_return_int(&rdev->wiphy, ret); - return ret; -} - static inline int rdev_set_multicast_to_unicast(struct cfg80211_registered_device *rdev, struct net_device *dev, diff --git a/net/wireless/trace.h b/net/wireless/trace.h index 6e218a0acd4e..817c6fef13be 100644 --- a/net/wireless/trace.h +++ b/net/wireless/trace.h @@ -838,11 +838,6 @@ DEFINE_EVENT(wiphy_netdev_mac_evt, rdev_del_mpath, TP_ARGS(wiphy, netdev, mac) ); -DEFINE_EVENT(wiphy_netdev_mac_evt, rdev_set_wds_peer, - TP_PROTO(struct wiphy *wiphy, struct net_device *netdev, const u8 *mac), - TP_ARGS(wiphy, netdev, mac) -); - TRACE_EVENT(rdev_dump_station, TP_PROTO(struct wiphy *wiphy, struct net_device *netdev, int _idx, u8 *mac), diff --git a/net/wireless/util.c b/net/wireless/util.c index 5b6c80ae564a..5af88037f1fb 100644 --- a/net/wireless/util.c +++ b/net/wireless/util.c @@ -550,8 +550,7 @@ int ieee80211_data_to_8023_exthdr(struct sk_buff *skb, struct ethhdr *ehdr, return -1; break; case cpu_to_le16(IEEE80211_FCTL_TODS | IEEE80211_FCTL_FROMDS): - if (unlikely(iftype != NL80211_IFTYPE_WDS && - iftype != NL80211_IFTYPE_MESH_POINT && + if (unlikely(iftype != NL80211_IFTYPE_MESH_POINT && iftype != NL80211_IFTYPE_AP_VLAN && iftype != NL80211_IFTYPE_STATION)) return -1; @@ -1051,7 +1050,6 @@ int cfg80211_change_iface(struct cfg80211_registered_device *rdev, case NL80211_IFTYPE_P2P_GO: case NL80211_IFTYPE_AP: case NL80211_IFTYPE_AP_VLAN: - case NL80211_IFTYPE_WDS: case NL80211_IFTYPE_MESH_POINT: /* bridging OK */ break; @@ -1063,6 +1061,7 @@ int cfg80211_change_iface(struct cfg80211_registered_device *rdev, /* not happening */ break; case NL80211_IFTYPE_P2P_DEVICE: + case NL80211_IFTYPE_WDS: case NL80211_IFTYPE_NAN: WARN_ON(1); break; diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c index 78f2927ead7f..b84a345b2653 100644 --- a/net/wireless/wext-compat.c +++ b/net/wireless/wext-compat.c @@ -49,9 +49,6 @@ int cfg80211_wext_siwmode(struct net_device *dev, struct iw_request_info *info, case IW_MODE_ADHOC: type = NL80211_IFTYPE_ADHOC; break; - case IW_MODE_REPEAT: - type = NL80211_IFTYPE_WDS; - break; case IW_MODE_MONITOR: type = NL80211_IFTYPE_MONITOR; break; @@ -1150,50 +1147,6 @@ static int cfg80211_wext_giwpower(struct net_device *dev, return 0; } -static int cfg80211_wds_wext_siwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *addr, char *extra) -{ - struct wireless_dev *wdev = dev->ieee80211_ptr; - struct cfg80211_registered_device *rdev = wiphy_to_rdev(wdev->wiphy); - int err; - - if (WARN_ON(wdev->iftype != NL80211_IFTYPE_WDS)) - return -EINVAL; - - if (addr->sa_family != ARPHRD_ETHER) - return -EINVAL; - - if (netif_running(dev)) - return -EBUSY; - - if (!rdev->ops->set_wds_peer) - return -EOPNOTSUPP; - - err = rdev_set_wds_peer(rdev, dev, (u8 *)&addr->sa_data); - if (err) - return err; - - memcpy(&wdev->wext.bssid, (u8 *) &addr->sa_data, ETH_ALEN); - - return 0; -} - -static int cfg80211_wds_wext_giwap(struct net_device *dev, - struct iw_request_info *info, - struct sockaddr *addr, char *extra) -{ - struct wireless_dev *wdev = dev->ieee80211_ptr; - - if (WARN_ON(wdev->iftype != NL80211_IFTYPE_WDS)) - return -EINVAL; - - addr->sa_family = ARPHRD_ETHER; - memcpy(&addr->sa_data, wdev->wext.bssid, ETH_ALEN); - - return 0; -} - static int cfg80211_wext_siwrate(struct net_device *dev, struct iw_request_info *info, struct iw_param *rate, char *extra) @@ -1371,8 +1324,6 @@ static int cfg80211_wext_siwap(struct net_device *dev, return cfg80211_ibss_wext_siwap(dev, info, ap_addr, extra); case NL80211_IFTYPE_STATION: return cfg80211_mgd_wext_siwap(dev, info, ap_addr, extra); - case NL80211_IFTYPE_WDS: - return cfg80211_wds_wext_siwap(dev, info, ap_addr, extra); default: return -EOPNOTSUPP; } @@ -1389,8 +1340,6 @@ static int cfg80211_wext_giwap(struct net_device *dev, return cfg80211_ibss_wext_giwap(dev, info, ap_addr, extra); case NL80211_IFTYPE_STATION: return cfg80211_mgd_wext_giwap(dev, info, ap_addr, extra); - case NL80211_IFTYPE_WDS: - return cfg80211_wds_wext_giwap(dev, info, ap_addr, extra); default: return -EOPNOTSUPP; } -- cgit v1.2.3 From da1e9dd3a11cda85b58dafe64f091734934b2f6c Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 10 Nov 2020 09:49:11 +0100 Subject: nl80211: fix kernel-doc warning in the new SAE attribute Format the items as a definition list, to avoid the warning from the rst parsing. Fixes: 9f0ffa418483 ("cfg80211: Add support to configure SAE PWE value to drivers") Link: https://lore.kernel.org/r/20201110094911.bb020e863aa0.I960caf90e2a8cc23f6bf9245d77524df6a4d8f37@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 23 +++++++++++++++-------- 1 file changed, 15 insertions(+), 8 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 4ff804a8bc1d..9678d32dae83 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1008,14 +1008,21 @@ struct survey_info { * @sae_pwd: password for SAE authentication (for devices supporting SAE * offload) * @sae_pwd_len: length of SAE password (for devices supporting SAE offload) - * @sae_pwe: The mechanisms allowed for SAE PWE derivation - * NL80211_SAE_PWE_UNSPECIFIED: Not-specified, used to indicate userspace - * did not specify any preference. The driver should follow its - * internal policy in such a scenario. - * NL80211_SAE_PWE_HUNT_AND_PECK: Allow hunting-and-pecking loop only - * NL80211_SAE_PWE_HASH_TO_ELEMENT: Allow hash-to-element only - * NL80211_SAE_PWE_BOTH: Allow either hunting-and-pecking loop - * or hash-to-element + * @sae_pwe: The mechanisms allowed for SAE PWE derivation: + * + * NL80211_SAE_PWE_UNSPECIFIED + * Not-specified, used to indicate userspace did not specify any + * preference. The driver should follow its internal policy in + * such a scenario. + * + * NL80211_SAE_PWE_HUNT_AND_PECK + * Allow hunting-and-pecking loop only + * + * NL80211_SAE_PWE_HASH_TO_ELEMENT + * Allow hash-to-element only + * + * NL80211_SAE_PWE_BOTH + * Allow either hunting-and-pecking loop or hash-to-element */ struct cfg80211_crypto_settings { u32 wpa_versions; -- cgit v1.2.3 From 8e4597c627fb48f361e2a5b012202cb1b6cbcd5e Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 12 Nov 2020 13:13:13 -0800 Subject: bpf: Allow using bpf_sk_storage in FENTRY/FEXIT/RAW_TP This patch enables the FENTRY/FEXIT/RAW_TP tracing program to use the bpf_sk_storage_(get|delete) helper, so those tracing programs can access the sk's bpf_local_storage and the later selftest will show some examples. The bpf_sk_storage is currently used in bpf-tcp-cc, tc, cg sockops...etc which is running either in softirq or task context. This patch adds bpf_sk_storage_get_tracing_proto and bpf_sk_storage_delete_tracing_proto. They will check in runtime that the helpers can only be called when serving softirq or running in a task context. That should enable most common tracing use cases on sk. During the load time, the new tracing_allowed() function will ensure the tracing prog using the bpf_sk_storage_(get|delete) helper is not tracing any bpf_sk_storage*() function itself. The sk is passed as "void *" when calling into bpf_local_storage. This patch only allows tracing a kernel function. Signed-off-by: Martin KaFai Lau Signed-off-by: Alexei Starovoitov Acked-by: Song Liu Link: https://lore.kernel.org/bpf/20201112211313.2587383-1-kafai@fb.com --- include/net/bpf_sk_storage.h | 2 ++ kernel/trace/bpf_trace.c | 5 +++ net/core/bpf_sk_storage.c | 74 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 81 insertions(+) (limited to 'include/net') diff --git a/include/net/bpf_sk_storage.h b/include/net/bpf_sk_storage.h index 3c516dd07caf..0e85713f56df 100644 --- a/include/net/bpf_sk_storage.h +++ b/include/net/bpf_sk_storage.h @@ -20,6 +20,8 @@ void bpf_sk_storage_free(struct sock *sk); extern const struct bpf_func_proto bpf_sk_storage_get_proto; extern const struct bpf_func_proto bpf_sk_storage_delete_proto; +extern const struct bpf_func_proto bpf_sk_storage_get_tracing_proto; +extern const struct bpf_func_proto bpf_sk_storage_delete_tracing_proto; struct bpf_local_storage_elem; struct bpf_sk_storage_diag; diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index e4515b0f62a8..cfce60ad1cb5 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include @@ -1735,6 +1736,10 @@ tracing_prog_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_skc_to_tcp_request_sock_proto; case BPF_FUNC_skc_to_udp6_sock: return &bpf_skc_to_udp6_sock_proto; + case BPF_FUNC_sk_storage_get: + return &bpf_sk_storage_get_tracing_proto; + case BPF_FUNC_sk_storage_delete: + return &bpf_sk_storage_delete_tracing_proto; #endif case BPF_FUNC_seq_printf: return prog->expected_attach_type == BPF_TRACE_ITER ? diff --git a/net/core/bpf_sk_storage.c b/net/core/bpf_sk_storage.c index fd416678f236..359908a7d3c1 100644 --- a/net/core/bpf_sk_storage.c +++ b/net/core/bpf_sk_storage.c @@ -6,6 +6,7 @@ #include #include #include +#include #include #include #include @@ -378,6 +379,79 @@ const struct bpf_func_proto bpf_sk_storage_delete_proto = { .arg2_type = ARG_PTR_TO_BTF_ID_SOCK_COMMON, }; +static bool bpf_sk_storage_tracing_allowed(const struct bpf_prog *prog) +{ + const struct btf *btf_vmlinux; + const struct btf_type *t; + const char *tname; + u32 btf_id; + + if (prog->aux->dst_prog) + return false; + + /* Ensure the tracing program is not tracing + * any bpf_sk_storage*() function and also + * use the bpf_sk_storage_(get|delete) helper. + */ + switch (prog->expected_attach_type) { + case BPF_TRACE_RAW_TP: + /* bpf_sk_storage has no trace point */ + return true; + case BPF_TRACE_FENTRY: + case BPF_TRACE_FEXIT: + btf_vmlinux = bpf_get_btf_vmlinux(); + btf_id = prog->aux->attach_btf_id; + t = btf_type_by_id(btf_vmlinux, btf_id); + tname = btf_name_by_offset(btf_vmlinux, t->name_off); + return !!strncmp(tname, "bpf_sk_storage", + strlen("bpf_sk_storage")); + default: + return false; + } + + return false; +} + +BPF_CALL_4(bpf_sk_storage_get_tracing, struct bpf_map *, map, struct sock *, sk, + void *, value, u64, flags) +{ + if (!in_serving_softirq() && !in_task()) + return (unsigned long)NULL; + + return (unsigned long)____bpf_sk_storage_get(map, sk, value, flags); +} + +BPF_CALL_2(bpf_sk_storage_delete_tracing, struct bpf_map *, map, + struct sock *, sk) +{ + if (!in_serving_softirq() && !in_task()) + return -EPERM; + + return ____bpf_sk_storage_delete(map, sk); +} + +const struct bpf_func_proto bpf_sk_storage_get_tracing_proto = { + .func = bpf_sk_storage_get_tracing, + .gpl_only = false, + .ret_type = RET_PTR_TO_MAP_VALUE_OR_NULL, + .arg1_type = ARG_CONST_MAP_PTR, + .arg2_type = ARG_PTR_TO_BTF_ID, + .arg2_btf_id = &btf_sock_ids[BTF_SOCK_TYPE_SOCK_COMMON], + .arg3_type = ARG_PTR_TO_MAP_VALUE_OR_NULL, + .arg4_type = ARG_ANYTHING, + .allowed = bpf_sk_storage_tracing_allowed, +}; + +const struct bpf_func_proto bpf_sk_storage_delete_tracing_proto = { + .func = bpf_sk_storage_delete_tracing, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_CONST_MAP_PTR, + .arg2_type = ARG_PTR_TO_BTF_ID, + .arg2_btf_id = &btf_sock_ids[BTF_SOCK_TYPE_SOCK_COMMON], + .allowed = bpf_sk_storage_tracing_allowed, +}; + struct bpf_sk_storage_diag { u32 nr_maps; struct bpf_map *maps[]; -- cgit v1.2.3 From 8cf8821e15cd553339a5b48ee555a0439c2b2742 Mon Sep 17 00:00:00 2001 From: Jeff Dike Date: Thu, 12 Nov 2020 20:58:15 -0500 Subject: net: Exempt multicast addresses from five-second neighbor lifetime Commit 58956317c8de ("neighbor: Improve garbage collection") guarantees neighbour table entries a five-second lifetime. Processes which make heavy use of multicast can fill the neighour table with multicast addresses in five seconds. At that point, neighbour entries can't be GC-ed because they aren't five seconds old yet, the kernel log starts to fill up with "neighbor table overflow!" messages, and sends start to fail. This patch allows multicast addresses to be thrown out before they've lived out their five seconds. This makes room for non-multicast addresses and makes messages to all addresses more reliable in these circumstances. Fixes: 58956317c8de ("neighbor: Improve garbage collection") Signed-off-by: Jeff Dike Reviewed-by: David Ahern Link: https://lore.kernel.org/r/20201113015815.31397-1-jdike@akamai.com Signed-off-by: Jakub Kicinski --- include/net/neighbour.h | 1 + net/core/neighbour.c | 2 ++ net/ipv4/arp.c | 6 ++++++ net/ipv6/ndisc.c | 7 +++++++ 4 files changed, 16 insertions(+) (limited to 'include/net') diff --git a/include/net/neighbour.h b/include/net/neighbour.h index 81ee17594c32..22ced1381ede 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -204,6 +204,7 @@ struct neigh_table { int (*pconstructor)(struct pneigh_entry *); void (*pdestructor)(struct pneigh_entry *); void (*proxy_redo)(struct sk_buff *skb); + int (*is_multicast)(const void *pkey); bool (*allow_add)(const struct net_device *dev, struct netlink_ext_ack *extack); char *id; diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 8e39e28b0a8d..9500d28a43b0 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -235,6 +235,8 @@ static int neigh_forced_gc(struct neigh_table *tbl) write_lock(&n->lock); if ((n->nud_state == NUD_FAILED) || + (tbl->is_multicast && + tbl->is_multicast(n->primary_key)) || time_after(tref, n->updated)) remove = true; write_unlock(&n->lock); diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index 687971d83b4e..922dd73e5740 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -125,6 +125,7 @@ static int arp_constructor(struct neighbour *neigh); static void arp_solicit(struct neighbour *neigh, struct sk_buff *skb); static void arp_error_report(struct neighbour *neigh, struct sk_buff *skb); static void parp_redo(struct sk_buff *skb); +static int arp_is_multicast(const void *pkey); static const struct neigh_ops arp_generic_ops = { .family = AF_INET, @@ -156,6 +157,7 @@ struct neigh_table arp_tbl = { .key_eq = arp_key_eq, .constructor = arp_constructor, .proxy_redo = parp_redo, + .is_multicast = arp_is_multicast, .id = "arp_cache", .parms = { .tbl = &arp_tbl, @@ -928,6 +930,10 @@ static void parp_redo(struct sk_buff *skb) arp_process(dev_net(skb->dev), NULL, skb); } +static int arp_is_multicast(const void *pkey) +{ + return ipv4_is_multicast(*((__be32 *)pkey)); +} /* * Receive an arp request from the device layer. diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index 27f29b957ee7..76717478f173 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -81,6 +81,7 @@ static void ndisc_error_report(struct neighbour *neigh, struct sk_buff *skb); static int pndisc_constructor(struct pneigh_entry *n); static void pndisc_destructor(struct pneigh_entry *n); static void pndisc_redo(struct sk_buff *skb); +static int ndisc_is_multicast(const void *pkey); static const struct neigh_ops ndisc_generic_ops = { .family = AF_INET6, @@ -115,6 +116,7 @@ struct neigh_table nd_tbl = { .pconstructor = pndisc_constructor, .pdestructor = pndisc_destructor, .proxy_redo = pndisc_redo, + .is_multicast = ndisc_is_multicast, .allow_add = ndisc_allow_add, .id = "ndisc_cache", .parms = { @@ -1706,6 +1708,11 @@ static void pndisc_redo(struct sk_buff *skb) kfree_skb(skb); } +static int ndisc_is_multicast(const void *pkey) +{ + return ipv6_addr_is_multicast((struct in6_addr *)pkey); +} + static bool ndisc_suppress_frag_ndisc(struct sk_buff *skb) { struct inet6_dev *idev = __in6_dev_get(skb->dev); -- cgit v1.2.3 From 9c2e14b48119b39446031d29d994044ae958d8fc Mon Sep 17 00:00:00 2001 From: Yi-Hung Wei Date: Tue, 10 Nov 2020 16:16:40 -0800 Subject: ip_tunnels: Set tunnel option flag when tunnel metadata is present Currently, we may set the tunnel option flag when the size of metadata is zero. For example, we set TUNNEL_GENEVE_OPT in the receive function no matter the geneve option is present or not. As this may result in issues on the tunnel flags consumers, this patch fixes the issue. Related discussion: * https://lore.kernel.org/netdev/1604448694-19351-1-git-send-email-yihung.wei@gmail.com/T/#u Fixes: 256c87c17c53 ("net: check tunnel option type in tunnel flags") Signed-off-by: Yi-Hung Wei Link: https://lore.kernel.org/r/1605053800-74072-1-git-send-email-yihung.wei@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/geneve.c | 3 +-- include/net/ip_tunnels.h | 7 ++++--- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c index d07008a818df..1426bfc009bc 100644 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@ -224,8 +224,7 @@ static void geneve_rx(struct geneve_dev *geneve, struct geneve_sock *gs, if (ip_tunnel_collect_metadata() || gs->collect_md) { __be16 flags; - flags = TUNNEL_KEY | TUNNEL_GENEVE_OPT | - (gnvh->oam ? TUNNEL_OAM : 0) | + flags = TUNNEL_KEY | (gnvh->oam ? TUNNEL_OAM : 0) | (gnvh->critical ? TUNNEL_CRIT_OPT : 0); tun_dst = udp_tun_rx_dst(skb, geneve_get_sk_family(gs), flags, diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h index 02ccd32542d0..61620677b034 100644 --- a/include/net/ip_tunnels.h +++ b/include/net/ip_tunnels.h @@ -478,9 +478,11 @@ static inline void ip_tunnel_info_opts_set(struct ip_tunnel_info *info, const void *from, int len, __be16 flags) { - memcpy(ip_tunnel_info_opts(info), from, len); info->options_len = len; - info->key.tun_flags |= flags; + if (len > 0) { + memcpy(ip_tunnel_info_opts(info), from, len); + info->key.tun_flags |= flags; + } } static inline struct ip_tunnel_info *lwt_tun_info(struct lwtunnel_state *lwtstate) @@ -526,7 +528,6 @@ static inline void ip_tunnel_info_opts_set(struct ip_tunnel_info *info, __be16 flags) { info->options_len = 0; - info->key.tun_flags |= flags; } #endif /* CONFIG_INET */ -- cgit v1.2.3 From 8965398713d831f6b893805880c249e62e9059ae Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Fri, 13 Nov 2020 12:48:28 +0100 Subject: net: xdp: Introduce bulking for xdp tx return path XDP bulk APIs introduce a defer/flush mechanism to return pages belonging to the same xdp_mem_allocator object (identified via the mem.id field) in bulk to optimize I-cache and D-cache since xdp_return_frame is usually run inside the driver NAPI tx completion loop. The bulk queue size is set to 16 to be aligned to how XDP_REDIRECT bulking works. The bulk is flushed when it is full or when mem.id changes. xdp_frame_bulk is usually stored/allocated on the function call-stack to avoid locking penalties. Current implementation considers only page_pool memory model. Suggested-by: Jesper Dangaard Brouer Co-developed-by: Jesper Dangaard Brouer Signed-off-by: Jesper Dangaard Brouer Signed-off-by: Lorenzo Bianconi Signed-off-by: Daniel Borkmann Acked-by: John Fastabend Acked-by: Ilias Apalodimas Link: https://lore.kernel.org/bpf/e190c03eac71b20c8407ae0fc2c399eda7835f49.1605267335.git.lorenzo@kernel.org --- include/net/xdp.h | 17 +++++++++++++++- net/core/xdp.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 75 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/xdp.h b/include/net/xdp.h index 3814fb631d52..7d48b2ae217a 100644 --- a/include/net/xdp.h +++ b/include/net/xdp.h @@ -104,6 +104,18 @@ struct xdp_frame { struct net_device *dev_rx; /* used by cpumap */ }; +#define XDP_BULK_QUEUE_SIZE 16 +struct xdp_frame_bulk { + int count; + void *xa; + void *q[XDP_BULK_QUEUE_SIZE]; +}; + +static __always_inline void xdp_frame_bulk_init(struct xdp_frame_bulk *bq) +{ + /* bq->count will be zero'ed when bq->xa gets updated */ + bq->xa = NULL; +} static inline struct skb_shared_info * xdp_get_shared_info_from_frame(struct xdp_frame *frame) @@ -194,6 +206,9 @@ struct xdp_frame *xdp_convert_buff_to_frame(struct xdp_buff *xdp) void xdp_return_frame(struct xdp_frame *xdpf); void xdp_return_frame_rx_napi(struct xdp_frame *xdpf); void xdp_return_buff(struct xdp_buff *xdp); +void xdp_flush_frame_bulk(struct xdp_frame_bulk *bq); +void xdp_return_frame_bulk(struct xdp_frame *xdpf, + struct xdp_frame_bulk *bq); /* When sending xdp_frame into the network stack, then there is no * return point callback, which is needed to release e.g. DMA-mapping @@ -245,6 +260,6 @@ bool xdp_attachment_flags_ok(struct xdp_attachment_info *info, void xdp_attachment_setup(struct xdp_attachment_info *info, struct netdev_bpf *bpf); -#define DEV_MAP_BULK_SIZE 16 +#define DEV_MAP_BULK_SIZE XDP_BULK_QUEUE_SIZE #endif /* __LINUX_NET_XDP_H__ */ diff --git a/net/core/xdp.c b/net/core/xdp.c index 48aba933a5a8..bbaee7fdd44f 100644 --- a/net/core/xdp.c +++ b/net/core/xdp.c @@ -380,6 +380,65 @@ void xdp_return_frame_rx_napi(struct xdp_frame *xdpf) } EXPORT_SYMBOL_GPL(xdp_return_frame_rx_napi); +/* XDP bulk APIs introduce a defer/flush mechanism to return + * pages belonging to the same xdp_mem_allocator object + * (identified via the mem.id field) in bulk to optimize + * I-cache and D-cache. + * The bulk queue size is set to 16 to be aligned to how + * XDP_REDIRECT bulking works. The bulk is flushed when + * it is full or when mem.id changes. + * xdp_frame_bulk is usually stored/allocated on the function + * call-stack to avoid locking penalties. + */ +void xdp_flush_frame_bulk(struct xdp_frame_bulk *bq) +{ + struct xdp_mem_allocator *xa = bq->xa; + int i; + + if (unlikely(!xa)) + return; + + for (i = 0; i < bq->count; i++) { + struct page *page = virt_to_head_page(bq->q[i]); + + page_pool_put_full_page(xa->page_pool, page, false); + } + /* bq->xa is not cleared to save lookup, if mem.id same in next bulk */ + bq->count = 0; +} +EXPORT_SYMBOL_GPL(xdp_flush_frame_bulk); + +/* Must be called with rcu_read_lock held */ +void xdp_return_frame_bulk(struct xdp_frame *xdpf, + struct xdp_frame_bulk *bq) +{ + struct xdp_mem_info *mem = &xdpf->mem; + struct xdp_mem_allocator *xa; + + if (mem->type != MEM_TYPE_PAGE_POOL) { + __xdp_return(xdpf->data, &xdpf->mem, false); + return; + } + + xa = bq->xa; + if (unlikely(!xa)) { + xa = rhashtable_lookup(mem_id_ht, &mem->id, mem_id_rht_params); + bq->count = 0; + bq->xa = xa; + } + + if (bq->count == XDP_BULK_QUEUE_SIZE) + xdp_flush_frame_bulk(bq); + + if (unlikely(mem->id != xa->mem.id)) { + xdp_flush_frame_bulk(bq); + bq->xa = rhashtable_lookup(mem_id_ht, &mem->id, mem_id_rht_params); + } + + bq->q[bq->count++] = xdpf->data; +} +EXPORT_SYMBOL_GPL(xdp_return_frame_bulk); + void xdp_return_buff(struct xdp_buff *xdp) { __xdp_return(xdp->data, &xdp->rxq->mem, true); -- cgit v1.2.3 From 7886244736a4dbb49987f330772842130493e050 Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Fri, 13 Nov 2020 12:48:29 +0100 Subject: net: page_pool: Add bulk support for ptr_ring Introduce the capability to batch page_pool ptr_ring refill since it is usually run inside the driver NAPI tx completion loop. Suggested-by: Jesper Dangaard Brouer Co-developed-by: Jesper Dangaard Brouer Signed-off-by: Jesper Dangaard Brouer Signed-off-by: Lorenzo Bianconi Signed-off-by: Daniel Borkmann Acked-by: John Fastabend Acked-by: Ilias Apalodimas Link: https://lore.kernel.org/bpf/08dd249c9522c001313f520796faa777c4089e1c.1605267335.git.lorenzo@kernel.org --- include/net/page_pool.h | 26 ++++++++++++++++++ net/core/page_pool.c | 70 ++++++++++++++++++++++++++++++++++++++++++------- net/core/xdp.c | 9 ++----- 3 files changed, 88 insertions(+), 17 deletions(-) (limited to 'include/net') diff --git a/include/net/page_pool.h b/include/net/page_pool.h index 81d7773f96cd..b5b195305346 100644 --- a/include/net/page_pool.h +++ b/include/net/page_pool.h @@ -152,6 +152,8 @@ struct page_pool *page_pool_create(const struct page_pool_params *params); void page_pool_destroy(struct page_pool *pool); void page_pool_use_xdp_mem(struct page_pool *pool, void (*disconnect)(void *)); void page_pool_release_page(struct page_pool *pool, struct page *page); +void page_pool_put_page_bulk(struct page_pool *pool, void **data, + int count); #else static inline void page_pool_destroy(struct page_pool *pool) { @@ -165,6 +167,11 @@ static inline void page_pool_release_page(struct page_pool *pool, struct page *page) { } + +static inline void page_pool_put_page_bulk(struct page_pool *pool, void **data, + int count) +{ +} #endif void page_pool_put_page(struct page_pool *pool, struct page *page, @@ -215,4 +222,23 @@ static inline void page_pool_nid_changed(struct page_pool *pool, int new_nid) if (unlikely(pool->p.nid != new_nid)) page_pool_update_nid(pool, new_nid); } + +static inline void page_pool_ring_lock(struct page_pool *pool) + __acquires(&pool->ring.producer_lock) +{ + if (in_serving_softirq()) + spin_lock(&pool->ring.producer_lock); + else + spin_lock_bh(&pool->ring.producer_lock); +} + +static inline void page_pool_ring_unlock(struct page_pool *pool) + __releases(&pool->ring.producer_lock) +{ + if (in_serving_softirq()) + spin_unlock(&pool->ring.producer_lock); + else + spin_unlock_bh(&pool->ring.producer_lock); +} + #endif /* _NET_PAGE_POOL_H */ diff --git a/net/core/page_pool.c b/net/core/page_pool.c index ef98372facf6..f3c690b8c8e3 100644 --- a/net/core/page_pool.c +++ b/net/core/page_pool.c @@ -11,6 +11,8 @@ #include #include +#include + #include #include #include @@ -362,8 +364,9 @@ static bool pool_page_reusable(struct page_pool *pool, struct page *page) * If the page refcnt != 1, then the page will be returned to memory * subsystem. */ -void page_pool_put_page(struct page_pool *pool, struct page *page, - unsigned int dma_sync_size, bool allow_direct) +static __always_inline struct page * +__page_pool_put_page(struct page_pool *pool, struct page *page, + unsigned int dma_sync_size, bool allow_direct) { /* This allocator is optimized for the XDP mode that uses * one-frame-per-page, but have fallbacks that act like the @@ -379,15 +382,12 @@ void page_pool_put_page(struct page_pool *pool, struct page *page, page_pool_dma_sync_for_device(pool, page, dma_sync_size); - if (allow_direct && in_serving_softirq()) - if (page_pool_recycle_in_cache(page, pool)) - return; + if (allow_direct && in_serving_softirq() && + page_pool_recycle_in_cache(page, pool)) + return NULL; - if (!page_pool_recycle_in_ring(pool, page)) { - /* Cache full, fallback to free pages */ - page_pool_return_page(pool, page); - } - return; + /* Page found as candidate for recycling */ + return page; } /* Fallback/non-XDP mode: API user have elevated refcnt. * @@ -405,9 +405,59 @@ void page_pool_put_page(struct page_pool *pool, struct page *page, /* Do not replace this with page_pool_return_page() */ page_pool_release_page(pool, page); put_page(page); + + return NULL; +} + +void page_pool_put_page(struct page_pool *pool, struct page *page, + unsigned int dma_sync_size, bool allow_direct) +{ + page = __page_pool_put_page(pool, page, dma_sync_size, allow_direct); + if (page && !page_pool_recycle_in_ring(pool, page)) { + /* Cache full, fallback to free pages */ + page_pool_return_page(pool, page); + } } EXPORT_SYMBOL(page_pool_put_page); +/* Caller must not use data area after call, as this function overwrites it */ +void page_pool_put_page_bulk(struct page_pool *pool, void **data, + int count) +{ + int i, bulk_len = 0; + + for (i = 0; i < count; i++) { + struct page *page = virt_to_head_page(data[i]); + + page = __page_pool_put_page(pool, page, -1, false); + /* Approved for bulk recycling in ptr_ring cache */ + if (page) + data[bulk_len++] = page; + } + + if (unlikely(!bulk_len)) + return; + + /* Bulk producer into ptr_ring page_pool cache */ + page_pool_ring_lock(pool); + for (i = 0; i < bulk_len; i++) { + if (__ptr_ring_produce(&pool->ring, data[i])) + break; /* ring full */ + } + page_pool_ring_unlock(pool); + + /* Hopefully all pages was return into ptr_ring */ + if (likely(i == bulk_len)) + return; + + /* ptr_ring cache full, free remaining pages outside producer lock + * since put_page() with refcnt == 1 can be an expensive operation + */ + for (; i < bulk_len; i++) + page_pool_return_page(pool, data[i]); +} +EXPORT_SYMBOL(page_pool_put_page_bulk); + static void page_pool_empty_ring(struct page_pool *pool) { struct page *page; diff --git a/net/core/xdp.c b/net/core/xdp.c index bbaee7fdd44f..3d330ebda893 100644 --- a/net/core/xdp.c +++ b/net/core/xdp.c @@ -393,16 +393,11 @@ EXPORT_SYMBOL_GPL(xdp_return_frame_rx_napi); void xdp_flush_frame_bulk(struct xdp_frame_bulk *bq) { struct xdp_mem_allocator *xa = bq->xa; - int i; - if (unlikely(!xa)) + if (unlikely(!xa || !bq->count)) return; - for (i = 0; i < bq->count; i++) { - struct page *page = virt_to_head_page(bq->q[i]); - - page_pool_put_full_page(xa->page_pool, page, false); - } + page_pool_put_page_bulk(xa->page_pool, bq->q, bq->count); /* bq->xa is not cleared to save lookup, if mem.id same in next bulk */ bq->count = 0; } -- cgit v1.2.3 From d3cd4924e3853aa771c3acf6f3ea3fb2cccdfadd Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 13 Nov 2020 07:08:08 -0800 Subject: tcp: uninline tcp_stream_memory_free() Both IPv4 and IPv6 needs it via a function pointer. Following patch will avoid the indirect call. Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 13 +------------ net/ipv4/tcp_ipv4.c | 14 ++++++++++++++ 2 files changed, 15 insertions(+), 12 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 4aba0f069b05..d643ee4e4249 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1965,18 +1965,7 @@ static inline u32 tcp_notsent_lowat(const struct tcp_sock *tp) return tp->notsent_lowat ?: net->ipv4.sysctl_tcp_notsent_lowat; } -/* @wake is one when sk_stream_write_space() calls us. - * This sends EPOLLOUT only if notsent_bytes is half the limit. - * This mimics the strategy used in sock_def_write_space(). - */ -static inline bool tcp_stream_memory_free(const struct sock *sk, int wake) -{ - const struct tcp_sock *tp = tcp_sk(sk); - u32 notsent_bytes = READ_ONCE(tp->write_seq) - - READ_ONCE(tp->snd_nxt); - - return (notsent_bytes << wake) < tcp_notsent_lowat(tp); -} +bool tcp_stream_memory_free(const struct sock *sk, int wake); #ifdef CONFIG_PROC_FS int tcp4_proc_init(void); diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 7352c097ae48..c2d5132c523c 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2740,6 +2740,20 @@ void tcp4_proc_exit(void) } #endif /* CONFIG_PROC_FS */ +/* @wake is one when sk_stream_write_space() calls us. + * This sends EPOLLOUT only if notsent_bytes is half the limit. + * This mimics the strategy used in sock_def_write_space(). + */ +bool tcp_stream_memory_free(const struct sock *sk, int wake) +{ + const struct tcp_sock *tp = tcp_sk(sk); + u32 notsent_bytes = READ_ONCE(tp->write_seq) - + READ_ONCE(tp->snd_nxt); + + return (notsent_bytes << wake) < tcp_notsent_lowat(tp); +} +EXPORT_SYMBOL(tcp_stream_memory_free); + struct proto tcp_prot = { .name = "TCP", .owner = THIS_MODULE, -- cgit v1.2.3 From 1c5f2ced136a7196c41ca2c16b8c2646e9b042ce Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 13 Nov 2020 07:08:09 -0800 Subject: tcp: avoid indirect call to tcp_stream_memory_free() Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- include/net/sock.h | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index a5c6ae78df77..1d29aeae74fd 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -60,7 +60,7 @@ #include #include #include - +#include #include #include #include @@ -1264,13 +1264,17 @@ static inline void sk_refcnt_debug_release(const struct sock *sk) #define sk_refcnt_debug_release(sk) do { } while (0) #endif /* SOCK_REFCNT_DEBUG */ +INDIRECT_CALLABLE_DECLARE(bool tcp_stream_memory_free(const struct sock *sk, int wake)); + static inline bool __sk_stream_memory_free(const struct sock *sk, int wake) { if (READ_ONCE(sk->sk_wmem_queued) >= READ_ONCE(sk->sk_sndbuf)) return false; return sk->sk_prot->stream_memory_free ? - sk->sk_prot->stream_memory_free(sk, wake) : true; + INDIRECT_CALL_1(sk->sk_prot->stream_memory_free, + tcp_stream_memory_free, + sk, wake) : true; } static inline bool sk_stream_memory_free(const struct sock *sk) -- cgit v1.2.3 From 9ca718743ad8402958637bfc196d7b62371a1b9f Mon Sep 17 00:00:00 2001 From: Francis Laniel Date: Sun, 15 Nov 2020 18:08:05 +0100 Subject: Modify return value of nla_strlcpy to match that of strscpy. nla_strlcpy now returns -E2BIG if src was truncated when written to dst. It also returns this error value if dstsize is 0 or higher than INT_MAX. For example, if src is "foo\0" and dst is 3 bytes long, the result will be: 1. "foG" after memcpy (G means garbage). 2. "fo\0" after memset. 3. -E2BIG is returned because src was not completely written into dst. The callers of nla_strlcpy were modified to take into account this modification. Signed-off-by: Francis Laniel Reviewed-by: Kees Cook Signed-off-by: Jakub Kicinski --- include/net/netlink.h | 2 +- include/net/pkt_cls.h | 2 +- lib/nlattr.c | 39 +++++++++++++++++++++++++-------------- net/sched/act_api.c | 2 +- net/sched/cls_api.c | 2 +- net/sched/sch_api.c | 2 +- 6 files changed, 30 insertions(+), 19 deletions(-) (limited to 'include/net') diff --git a/include/net/netlink.h b/include/net/netlink.h index 7356f41d23ba..446ca182e13d 100644 --- a/include/net/netlink.h +++ b/include/net/netlink.h @@ -506,7 +506,7 @@ int __nla_parse(struct nlattr **tb, int maxtype, const struct nlattr *head, struct netlink_ext_ack *extack); int nla_policy_len(const struct nla_policy *, int); struct nlattr *nla_find(const struct nlattr *head, int len, int attrtype); -size_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize); +ssize_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize); char *nla_strdup(const struct nlattr *nla, gfp_t flags); int nla_memcpy(void *dest, const struct nlattr *src, int count); int nla_memcmp(const struct nlattr *nla, const void *data, size_t size); diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index d4d461236351..db9a828f4f4f 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -512,7 +512,7 @@ tcf_change_indev(struct net *net, struct nlattr *indev_tlv, char indev[IFNAMSIZ]; struct net_device *dev; - if (nla_strlcpy(indev, indev_tlv, IFNAMSIZ) >= IFNAMSIZ) { + if (nla_strlcpy(indev, indev_tlv, IFNAMSIZ) < 0) { NL_SET_ERR_MSG_ATTR(extack, indev_tlv, "Interface name too long"); return -EINVAL; diff --git a/lib/nlattr.c b/lib/nlattr.c index 07156e581997..447182543c03 100644 --- a/lib/nlattr.c +++ b/lib/nlattr.c @@ -710,33 +710,44 @@ EXPORT_SYMBOL(nla_find); /** * nla_strlcpy - Copy string attribute payload into a sized buffer - * @dst: where to copy the string to - * @nla: attribute to copy the string from - * @dstsize: size of destination buffer + * @dst: Where to copy the string to. + * @nla: Attribute to copy the string from. + * @dstsize: Size of destination buffer. * * Copies at most dstsize - 1 bytes into the destination buffer. - * The result is always a valid NUL-terminated string. Unlike - * strlcpy the destination buffer is always padded out. + * Unlike strlcpy the destination buffer is always padded out. * - * Returns the length of the source buffer. + * Return: + * * srclen - Returns @nla length (not including the trailing %NUL). + * * -E2BIG - If @dstsize is 0 or greater than U16_MAX or @nla length greater + * than @dstsize. */ -size_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize) +ssize_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize) { size_t srclen = nla_len(nla); char *src = nla_data(nla); + ssize_t ret; + size_t len; + + if (dstsize == 0 || WARN_ON_ONCE(dstsize > U16_MAX)) + return -E2BIG; if (srclen > 0 && src[srclen - 1] == '\0') srclen--; - if (dstsize > 0) { - size_t len = (srclen >= dstsize) ? dstsize - 1 : srclen; - - memcpy(dst, src, len); - /* Zero pad end of dst. */ - memset(dst + len, 0, dstsize - len); + if (srclen >= dstsize) { + len = dstsize - 1; + ret = -E2BIG; + } else { + len = srclen; + ret = len; } - return srclen; + memcpy(dst, src, len); + /* Zero pad end of dst. */ + memset(dst + len, 0, dstsize - len); + + return ret; } EXPORT_SYMBOL(nla_strlcpy); diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 60e1572ba606..fe540a89b16c 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -939,7 +939,7 @@ struct tc_action *tcf_action_init_1(struct net *net, struct tcf_proto *tp, NL_SET_ERR_MSG(extack, "TC action kind must be specified"); goto err_out; } - if (nla_strlcpy(act_name, kind, IFNAMSIZ) >= IFNAMSIZ) { + if (nla_strlcpy(act_name, kind, IFNAMSIZ) < 0) { NL_SET_ERR_MSG(extack, "TC action name too long"); goto err_out; } diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index ba0715ee9eac..c2e9661e20d3 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -223,7 +223,7 @@ static inline u32 tcf_auto_prio(struct tcf_proto *tp) static bool tcf_proto_check_kind(struct nlattr *kind, char *name) { if (kind) - return nla_strlcpy(name, kind, IFNAMSIZ) >= IFNAMSIZ; + return nla_strlcpy(name, kind, IFNAMSIZ) < 0; memset(name, 0, IFNAMSIZ); return false; } diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 2a76a2f5ed88..05449286d889 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -1170,7 +1170,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev, #ifdef CONFIG_MODULES if (ops == NULL && kind != NULL) { char name[IFNAMSIZ]; - if (nla_strlcpy(name, kind, IFNAMSIZ) < IFNAMSIZ) { + if (nla_strlcpy(name, kind, IFNAMSIZ) >= 0) { /* We dropped the RTNL semaphore in order to * perform the module load. So, even if we * succeeded in loading the module we have to -- cgit v1.2.3 From 872f690341948b502c93318f806d821c56772c42 Mon Sep 17 00:00:00 2001 From: Francis Laniel Date: Sun, 15 Nov 2020 18:08:06 +0100 Subject: treewide: rename nla_strlcpy to nla_strscpy. Calls to nla_strlcpy are now replaced by calls to nla_strscpy which is the new name of this function. Signed-off-by: Francis Laniel Reviewed-by: Kees Cook Signed-off-by: Jakub Kicinski --- drivers/infiniband/core/nldev.c | 10 +++++----- drivers/net/can/vxcan.c | 4 ++-- drivers/net/veth.c | 4 ++-- include/linux/genl_magic_struct.h | 2 +- include/net/netlink.h | 4 ++-- include/net/pkt_cls.h | 2 +- kernel/taskstats.c | 2 +- lib/nlattr.c | 6 +++--- net/core/fib_rules.c | 4 ++-- net/core/rtnetlink.c | 12 ++++++------ net/decnet/dn_dev.c | 2 +- net/ieee802154/nl-mac.c | 2 +- net/ipv4/devinet.c | 2 +- net/ipv4/fib_semantics.c | 2 +- net/ipv4/metrics.c | 2 +- net/netfilter/ipset/ip_set_hash_netiface.c | 4 ++-- net/netfilter/nf_tables_api.c | 6 +++--- net/netfilter/nfnetlink_acct.c | 2 +- net/netfilter/nfnetlink_cthelper.c | 4 ++-- net/netfilter/nft_ct.c | 2 +- net/netfilter/nft_log.c | 2 +- net/netlabel/netlabel_mgmt.c | 2 +- net/nfc/netlink.c | 2 +- net/sched/act_api.c | 2 +- net/sched/act_ipt.c | 2 +- net/sched/act_simple.c | 4 ++-- net/sched/cls_api.c | 2 +- net/sched/sch_api.c | 2 +- net/tipc/netlink_compat.c | 2 +- 29 files changed, 49 insertions(+), 49 deletions(-) (limited to 'include/net') diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c index 12d29d54a081..08366e254b1d 100644 --- a/drivers/infiniband/core/nldev.c +++ b/drivers/infiniband/core/nldev.c @@ -932,7 +932,7 @@ static int nldev_set_doit(struct sk_buff *skb, struct nlmsghdr *nlh, if (tb[RDMA_NLDEV_ATTR_DEV_NAME]) { char name[IB_DEVICE_NAME_MAX] = {}; - nla_strlcpy(name, tb[RDMA_NLDEV_ATTR_DEV_NAME], + nla_strscpy(name, tb[RDMA_NLDEV_ATTR_DEV_NAME], IB_DEVICE_NAME_MAX); if (strlen(name) == 0) { err = -EINVAL; @@ -1529,13 +1529,13 @@ static int nldev_newlink(struct sk_buff *skb, struct nlmsghdr *nlh, !tb[RDMA_NLDEV_ATTR_LINK_TYPE] || !tb[RDMA_NLDEV_ATTR_NDEV_NAME]) return -EINVAL; - nla_strlcpy(ibdev_name, tb[RDMA_NLDEV_ATTR_DEV_NAME], + nla_strscpy(ibdev_name, tb[RDMA_NLDEV_ATTR_DEV_NAME], sizeof(ibdev_name)); if (strchr(ibdev_name, '%') || strlen(ibdev_name) == 0) return -EINVAL; - nla_strlcpy(type, tb[RDMA_NLDEV_ATTR_LINK_TYPE], sizeof(type)); - nla_strlcpy(ndev_name, tb[RDMA_NLDEV_ATTR_NDEV_NAME], + nla_strscpy(type, tb[RDMA_NLDEV_ATTR_LINK_TYPE], sizeof(type)); + nla_strscpy(ndev_name, tb[RDMA_NLDEV_ATTR_NDEV_NAME], sizeof(ndev_name)); ndev = dev_get_by_name(sock_net(skb->sk), ndev_name); @@ -1602,7 +1602,7 @@ static int nldev_get_chardev(struct sk_buff *skb, struct nlmsghdr *nlh, if (err || !tb[RDMA_NLDEV_ATTR_CHARDEV_TYPE]) return -EINVAL; - nla_strlcpy(client_name, tb[RDMA_NLDEV_ATTR_CHARDEV_TYPE], + nla_strscpy(client_name, tb[RDMA_NLDEV_ATTR_CHARDEV_TYPE], sizeof(client_name)); if (tb[RDMA_NLDEV_ATTR_DEV_INDEX]) { diff --git a/drivers/net/can/vxcan.c b/drivers/net/can/vxcan.c index d6ba9426be4d..fa47bab510bb 100644 --- a/drivers/net/can/vxcan.c +++ b/drivers/net/can/vxcan.c @@ -186,7 +186,7 @@ static int vxcan_newlink(struct net *net, struct net_device *dev, } if (ifmp && tbp[IFLA_IFNAME]) { - nla_strlcpy(ifname, tbp[IFLA_IFNAME], IFNAMSIZ); + nla_strscpy(ifname, tbp[IFLA_IFNAME], IFNAMSIZ); name_assign_type = NET_NAME_USER; } else { snprintf(ifname, IFNAMSIZ, DRV_NAME "%%d"); @@ -223,7 +223,7 @@ static int vxcan_newlink(struct net *net, struct net_device *dev, /* register first device */ if (tb[IFLA_IFNAME]) - nla_strlcpy(dev->name, tb[IFLA_IFNAME], IFNAMSIZ); + nla_strscpy(dev->name, tb[IFLA_IFNAME], IFNAMSIZ); else snprintf(dev->name, IFNAMSIZ, DRV_NAME "%%d"); diff --git a/drivers/net/veth.c b/drivers/net/veth.c index 8c737668008a..359d3ab33c4d 100644 --- a/drivers/net/veth.c +++ b/drivers/net/veth.c @@ -1329,7 +1329,7 @@ static int veth_newlink(struct net *src_net, struct net_device *dev, } if (ifmp && tbp[IFLA_IFNAME]) { - nla_strlcpy(ifname, tbp[IFLA_IFNAME], IFNAMSIZ); + nla_strscpy(ifname, tbp[IFLA_IFNAME], IFNAMSIZ); name_assign_type = NET_NAME_USER; } else { snprintf(ifname, IFNAMSIZ, DRV_NAME "%%d"); @@ -1379,7 +1379,7 @@ static int veth_newlink(struct net *src_net, struct net_device *dev, eth_hw_addr_random(dev); if (tb[IFLA_IFNAME]) - nla_strlcpy(dev->name, tb[IFLA_IFNAME], IFNAMSIZ); + nla_strscpy(dev->name, tb[IFLA_IFNAME], IFNAMSIZ); else snprintf(dev->name, IFNAMSIZ, DRV_NAME "%%d"); diff --git a/include/linux/genl_magic_struct.h b/include/linux/genl_magic_struct.h index eeae59d3ceb7..35d21fddaf2d 100644 --- a/include/linux/genl_magic_struct.h +++ b/include/linux/genl_magic_struct.h @@ -89,7 +89,7 @@ static inline int nla_put_u64_0pad(struct sk_buff *skb, int attrtype, u64 value) nla_get_u64, nla_put_u64_0pad, false) #define __str_field(attr_nr, attr_flag, name, maxlen) \ __array(attr_nr, attr_flag, name, NLA_NUL_STRING, char, maxlen, \ - nla_strlcpy, nla_put, false) + nla_strscpy, nla_put, false) #define __bin_field(attr_nr, attr_flag, name, maxlen) \ __array(attr_nr, attr_flag, name, NLA_BINARY, char, maxlen, \ nla_memcpy, nla_put, false) diff --git a/include/net/netlink.h b/include/net/netlink.h index 446ca182e13d..1ceec518ab49 100644 --- a/include/net/netlink.h +++ b/include/net/netlink.h @@ -142,7 +142,7 @@ * Attribute Misc: * nla_memcpy(dest, nla, count) copy attribute into memory * nla_memcmp(nla, data, size) compare attribute with memory area - * nla_strlcpy(dst, nla, size) copy attribute to a sized string + * nla_strscpy(dst, nla, size) copy attribute to a sized string * nla_strcmp(nla, str) compare attribute with string * * Attribute Parsing: @@ -506,7 +506,7 @@ int __nla_parse(struct nlattr **tb, int maxtype, const struct nlattr *head, struct netlink_ext_ack *extack); int nla_policy_len(const struct nla_policy *, int); struct nlattr *nla_find(const struct nlattr *head, int len, int attrtype); -ssize_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize); +ssize_t nla_strscpy(char *dst, const struct nlattr *nla, size_t dstsize); char *nla_strdup(const struct nlattr *nla, gfp_t flags); int nla_memcpy(void *dest, const struct nlattr *src, int count); int nla_memcmp(const struct nlattr *nla, const void *data, size_t size); diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index db9a828f4f4f..133f9ad4d4f9 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -512,7 +512,7 @@ tcf_change_indev(struct net *net, struct nlattr *indev_tlv, char indev[IFNAMSIZ]; struct net_device *dev; - if (nla_strlcpy(indev, indev_tlv, IFNAMSIZ) < 0) { + if (nla_strscpy(indev, indev_tlv, IFNAMSIZ) < 0) { NL_SET_ERR_MSG_ATTR(extack, indev_tlv, "Interface name too long"); return -EINVAL; diff --git a/kernel/taskstats.c b/kernel/taskstats.c index a2802b6ff4bb..2b4898b4752e 100644 --- a/kernel/taskstats.c +++ b/kernel/taskstats.c @@ -346,7 +346,7 @@ static int parse(struct nlattr *na, struct cpumask *mask) data = kmalloc(len, GFP_KERNEL); if (!data) return -ENOMEM; - nla_strlcpy(data, na, len); + nla_strscpy(data, na, len); ret = cpulist_parse(data, mask); kfree(data); return ret; diff --git a/lib/nlattr.c b/lib/nlattr.c index 447182543c03..09aa181569e0 100644 --- a/lib/nlattr.c +++ b/lib/nlattr.c @@ -709,7 +709,7 @@ struct nlattr *nla_find(const struct nlattr *head, int len, int attrtype) EXPORT_SYMBOL(nla_find); /** - * nla_strlcpy - Copy string attribute payload into a sized buffer + * nla_strscpy - Copy string attribute payload into a sized buffer * @dst: Where to copy the string to. * @nla: Attribute to copy the string from. * @dstsize: Size of destination buffer. @@ -722,7 +722,7 @@ EXPORT_SYMBOL(nla_find); * * -E2BIG - If @dstsize is 0 or greater than U16_MAX or @nla length greater * than @dstsize. */ -ssize_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize) +ssize_t nla_strscpy(char *dst, const struct nlattr *nla, size_t dstsize) { size_t srclen = nla_len(nla); char *src = nla_data(nla); @@ -749,7 +749,7 @@ ssize_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize) return ret; } -EXPORT_SYMBOL(nla_strlcpy); +EXPORT_SYMBOL(nla_strscpy); /** * nla_strdup - Copy string attribute payload into a newly allocated buffer diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c index 7bcfb16854cb..cd80ffed6d26 100644 --- a/net/core/fib_rules.c +++ b/net/core/fib_rules.c @@ -563,7 +563,7 @@ static int fib_nl2rule(struct sk_buff *skb, struct nlmsghdr *nlh, struct net_device *dev; nlrule->iifindex = -1; - nla_strlcpy(nlrule->iifname, tb[FRA_IIFNAME], IFNAMSIZ); + nla_strscpy(nlrule->iifname, tb[FRA_IIFNAME], IFNAMSIZ); dev = __dev_get_by_name(net, nlrule->iifname); if (dev) nlrule->iifindex = dev->ifindex; @@ -573,7 +573,7 @@ static int fib_nl2rule(struct sk_buff *skb, struct nlmsghdr *nlh, struct net_device *dev; nlrule->oifindex = -1; - nla_strlcpy(nlrule->oifname, tb[FRA_OIFNAME], IFNAMSIZ); + nla_strscpy(nlrule->oifname, tb[FRA_OIFNAME], IFNAMSIZ); dev = __dev_get_by_name(net, nlrule->oifname); if (dev) nlrule->oifindex = dev->ifindex; diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 7d7223691783..60917ff4a00b 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -1939,7 +1939,7 @@ static const struct rtnl_link_ops *linkinfo_to_kind_ops(const struct nlattr *nla if (linfo[IFLA_INFO_KIND]) { char kind[MODULE_NAME_LEN]; - nla_strlcpy(kind, linfo[IFLA_INFO_KIND], sizeof(kind)); + nla_strscpy(kind, linfo[IFLA_INFO_KIND], sizeof(kind)); ops = rtnl_link_ops_get(kind); } @@ -2953,9 +2953,9 @@ static struct net_device *rtnl_dev_get(struct net *net, if (!ifname) { ifname = buffer; if (ifname_attr) - nla_strlcpy(ifname, ifname_attr, IFNAMSIZ); + nla_strscpy(ifname, ifname_attr, IFNAMSIZ); else if (altifname_attr) - nla_strlcpy(ifname, altifname_attr, ALTIFNAMSIZ); + nla_strscpy(ifname, altifname_attr, ALTIFNAMSIZ); else return NULL; } @@ -2983,7 +2983,7 @@ static int rtnl_setlink(struct sk_buff *skb, struct nlmsghdr *nlh, goto errout; if (tb[IFLA_IFNAME]) - nla_strlcpy(ifname, tb[IFLA_IFNAME], IFNAMSIZ); + nla_strscpy(ifname, tb[IFLA_IFNAME], IFNAMSIZ); else ifname[0] = '\0'; @@ -3264,7 +3264,7 @@ replay: return err; if (tb[IFLA_IFNAME]) - nla_strlcpy(ifname, tb[IFLA_IFNAME], IFNAMSIZ); + nla_strscpy(ifname, tb[IFLA_IFNAME], IFNAMSIZ); else ifname[0] = '\0'; @@ -3296,7 +3296,7 @@ replay: memset(linkinfo, 0, sizeof(linkinfo)); if (linkinfo[IFLA_INFO_KIND]) { - nla_strlcpy(kind, linkinfo[IFLA_INFO_KIND], sizeof(kind)); + nla_strscpy(kind, linkinfo[IFLA_INFO_KIND], sizeof(kind)); ops = rtnl_link_ops_get(kind); } else { kind[0] = '\0'; diff --git a/net/decnet/dn_dev.c b/net/decnet/dn_dev.c index 15d42353f1a3..d1c50a48614b 100644 --- a/net/decnet/dn_dev.c +++ b/net/decnet/dn_dev.c @@ -658,7 +658,7 @@ static int dn_nl_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, ifa->ifa_dev = dn_db; if (tb[IFA_LABEL]) - nla_strlcpy(ifa->ifa_label, tb[IFA_LABEL], IFNAMSIZ); + nla_strscpy(ifa->ifa_label, tb[IFA_LABEL], IFNAMSIZ); else memcpy(ifa->ifa_label, dev->name, IFNAMSIZ); diff --git a/net/ieee802154/nl-mac.c b/net/ieee802154/nl-mac.c index 6d091e419d3e..9c640d670ffe 100644 --- a/net/ieee802154/nl-mac.c +++ b/net/ieee802154/nl-mac.c @@ -149,7 +149,7 @@ static struct net_device *ieee802154_nl_get_dev(struct genl_info *info) if (info->attrs[IEEE802154_ATTR_DEV_NAME]) { char name[IFNAMSIZ + 1]; - nla_strlcpy(name, info->attrs[IEEE802154_ATTR_DEV_NAME], + nla_strscpy(name, info->attrs[IEEE802154_ATTR_DEV_NAME], sizeof(name)); dev = dev_get_by_name(&init_net, name); } else if (info->attrs[IEEE802154_ATTR_DEV_INDEX]) { diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index 43e04382c593..75f67994fc85 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -880,7 +880,7 @@ static struct in_ifaddr *rtm_to_ifaddr(struct net *net, struct nlmsghdr *nlh, ifa->ifa_broadcast = nla_get_in_addr(tb[IFA_BROADCAST]); if (tb[IFA_LABEL]) - nla_strlcpy(ifa->ifa_label, tb[IFA_LABEL], IFNAMSIZ); + nla_strscpy(ifa->ifa_label, tb[IFA_LABEL], IFNAMSIZ); else memcpy(ifa->ifa_label, dev->name, IFNAMSIZ); diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index 7612ff6111a7..b5400cec4f69 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -973,7 +973,7 @@ bool fib_metrics_match(struct fib_config *cfg, struct fib_info *fi) char tmp[TCP_CA_NAME_MAX]; bool ecn_ca = false; - nla_strlcpy(tmp, nla, sizeof(tmp)); + nla_strscpy(tmp, nla, sizeof(tmp)); val = tcp_ca_get_key_by_name(fi->fib_net, tmp, &ecn_ca); } else { if (nla_len(nla) != sizeof(u32)) diff --git a/net/ipv4/metrics.c b/net/ipv4/metrics.c index 3205d5f7c8c9..25ea6ac44db9 100644 --- a/net/ipv4/metrics.c +++ b/net/ipv4/metrics.c @@ -31,7 +31,7 @@ static int ip_metrics_convert(struct net *net, struct nlattr *fc_mx, if (type == RTAX_CC_ALGO) { char tmp[TCP_CA_NAME_MAX]; - nla_strlcpy(tmp, nla, sizeof(tmp)); + nla_strscpy(tmp, nla, sizeof(tmp)); val = tcp_ca_get_key_by_name(net, tmp, &ecn_ca); if (val == TCP_CA_UNSPEC) { NL_SET_ERR_MSG(extack, "Unknown tcp congestion algorithm"); diff --git a/net/netfilter/ipset/ip_set_hash_netiface.c b/net/netfilter/ipset/ip_set_hash_netiface.c index 3d74169b794c..ddd51c2e1cb3 100644 --- a/net/netfilter/ipset/ip_set_hash_netiface.c +++ b/net/netfilter/ipset/ip_set_hash_netiface.c @@ -226,7 +226,7 @@ hash_netiface4_uadt(struct ip_set *set, struct nlattr *tb[], if (e.cidr > HOST_MASK) return -IPSET_ERR_INVALID_CIDR; } - nla_strlcpy(e.iface, tb[IPSET_ATTR_IFACE], IFNAMSIZ); + nla_strscpy(e.iface, tb[IPSET_ATTR_IFACE], IFNAMSIZ); if (tb[IPSET_ATTR_CADT_FLAGS]) { u32 cadt_flags = ip_set_get_h32(tb[IPSET_ATTR_CADT_FLAGS]); @@ -443,7 +443,7 @@ hash_netiface6_uadt(struct ip_set *set, struct nlattr *tb[], ip6_netmask(&e.ip, e.cidr); - nla_strlcpy(e.iface, tb[IPSET_ATTR_IFACE], IFNAMSIZ); + nla_strscpy(e.iface, tb[IPSET_ATTR_IFACE], IFNAMSIZ); if (tb[IPSET_ATTR_CADT_FLAGS]) { u32 cadt_flags = ip_set_get_h32(tb[IPSET_ATTR_CADT_FLAGS]); diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index bd0e12bf5770..65aa98fc5eb6 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -1282,7 +1282,7 @@ static struct nft_chain *nft_chain_lookup(struct net *net, if (nla == NULL) return ERR_PTR(-EINVAL); - nla_strlcpy(search, nla, sizeof(search)); + nla_strscpy(search, nla, sizeof(search)); WARN_ON(!rcu_read_lock_held() && !lockdep_commit_lock_is_held(net)); @@ -1722,7 +1722,7 @@ static struct nft_hook *nft_netdev_hook_alloc(struct net *net, goto err_hook_alloc; } - nla_strlcpy(ifname, attr, IFNAMSIZ); + nla_strscpy(ifname, attr, IFNAMSIZ); dev = __dev_get_by_name(net, ifname); if (!dev) { err = -ENOENT; @@ -5735,7 +5735,7 @@ struct nft_object *nft_obj_lookup(const struct net *net, struct rhlist_head *tmp, *list; struct nft_object *obj; - nla_strlcpy(search, nla, sizeof(search)); + nla_strscpy(search, nla, sizeof(search)); k.name = search; WARN_ON_ONCE(!rcu_read_lock_held() && diff --git a/net/netfilter/nfnetlink_acct.c b/net/netfilter/nfnetlink_acct.c index 5bfec829c12f..5e511df8d709 100644 --- a/net/netfilter/nfnetlink_acct.c +++ b/net/netfilter/nfnetlink_acct.c @@ -112,7 +112,7 @@ static int nfnl_acct_new(struct net *net, struct sock *nfnl, nfacct->flags = flags; } - nla_strlcpy(nfacct->name, tb[NFACCT_NAME], NFACCT_NAME_MAX); + nla_strscpy(nfacct->name, tb[NFACCT_NAME], NFACCT_NAME_MAX); if (tb[NFACCT_BYTES]) { atomic64_set(&nfacct->bytes, diff --git a/net/netfilter/nfnetlink_cthelper.c b/net/netfilter/nfnetlink_cthelper.c index 5b0d0a77379c..0f94fce1d3ed 100644 --- a/net/netfilter/nfnetlink_cthelper.c +++ b/net/netfilter/nfnetlink_cthelper.c @@ -146,7 +146,7 @@ nfnl_cthelper_expect_policy(struct nf_conntrack_expect_policy *expect_policy, !tb[NFCTH_POLICY_EXPECT_TIMEOUT]) return -EINVAL; - nla_strlcpy(expect_policy->name, + nla_strscpy(expect_policy->name, tb[NFCTH_POLICY_NAME], NF_CT_HELPER_NAME_LEN); expect_policy->max_expected = ntohl(nla_get_be32(tb[NFCTH_POLICY_EXPECT_MAX])); @@ -233,7 +233,7 @@ nfnl_cthelper_create(const struct nlattr * const tb[], if (ret < 0) goto err1; - nla_strlcpy(helper->name, + nla_strscpy(helper->name, tb[NFCTH_NAME], NF_CT_HELPER_NAME_LEN); size = ntohl(nla_get_be32(tb[NFCTH_PRIV_DATA_LEN])); if (size > sizeof_field(struct nf_conn_help, data)) { diff --git a/net/netfilter/nft_ct.c b/net/netfilter/nft_ct.c index 322bd674963e..a8c4d442231c 100644 --- a/net/netfilter/nft_ct.c +++ b/net/netfilter/nft_ct.c @@ -990,7 +990,7 @@ static int nft_ct_helper_obj_init(const struct nft_ctx *ctx, if (!priv->l4proto) return -ENOENT; - nla_strlcpy(name, tb[NFTA_CT_HELPER_NAME], sizeof(name)); + nla_strscpy(name, tb[NFTA_CT_HELPER_NAME], sizeof(name)); if (tb[NFTA_CT_HELPER_L3PROTO]) family = ntohs(nla_get_be16(tb[NFTA_CT_HELPER_L3PROTO])); diff --git a/net/netfilter/nft_log.c b/net/netfilter/nft_log.c index 57899454a530..a06a46b039c5 100644 --- a/net/netfilter/nft_log.c +++ b/net/netfilter/nft_log.c @@ -152,7 +152,7 @@ static int nft_log_init(const struct nft_ctx *ctx, priv->prefix = kmalloc(nla_len(nla) + 1, GFP_KERNEL); if (priv->prefix == NULL) return -ENOMEM; - nla_strlcpy(priv->prefix, nla, nla_len(nla) + 1); + nla_strscpy(priv->prefix, nla, nla_len(nla) + 1); } else { priv->prefix = (char *)nft_log_null_prefix; } diff --git a/net/netlabel/netlabel_mgmt.c b/net/netlabel/netlabel_mgmt.c index eb1d66d20afb..df1b41ed73fd 100644 --- a/net/netlabel/netlabel_mgmt.c +++ b/net/netlabel/netlabel_mgmt.c @@ -95,7 +95,7 @@ static int netlbl_mgmt_add_common(struct genl_info *info, ret_val = -ENOMEM; goto add_free_entry; } - nla_strlcpy(entry->domain, + nla_strscpy(entry->domain, info->attrs[NLBL_MGMT_A_DOMAIN], tmp_size); } diff --git a/net/nfc/netlink.c b/net/nfc/netlink.c index 8709f3d4e7c4..573b38ad2f8e 100644 --- a/net/nfc/netlink.c +++ b/net/nfc/netlink.c @@ -1226,7 +1226,7 @@ static int nfc_genl_fw_download(struct sk_buff *skb, struct genl_info *info) if (!dev) return -ENODEV; - nla_strlcpy(firmware_name, info->attrs[NFC_ATTR_FIRMWARE_NAME], + nla_strscpy(firmware_name, info->attrs[NFC_ATTR_FIRMWARE_NAME], sizeof(firmware_name)); rc = nfc_fw_download(dev, firmware_name); diff --git a/net/sched/act_api.c b/net/sched/act_api.c index fe540a89b16c..fc23f46a315c 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -939,7 +939,7 @@ struct tc_action *tcf_action_init_1(struct net *net, struct tcf_proto *tp, NL_SET_ERR_MSG(extack, "TC action kind must be specified"); goto err_out; } - if (nla_strlcpy(act_name, kind, IFNAMSIZ) < 0) { + if (nla_strscpy(act_name, kind, IFNAMSIZ) < 0) { NL_SET_ERR_MSG(extack, "TC action name too long"); goto err_out; } diff --git a/net/sched/act_ipt.c b/net/sched/act_ipt.c index 8dc3bec0d325..ac7297f42355 100644 --- a/net/sched/act_ipt.c +++ b/net/sched/act_ipt.c @@ -166,7 +166,7 @@ static int __tcf_ipt_init(struct net *net, unsigned int id, struct nlattr *nla, if (unlikely(!tname)) goto err1; if (tb[TCA_IPT_TABLE] == NULL || - nla_strlcpy(tname, tb[TCA_IPT_TABLE], IFNAMSIZ) >= IFNAMSIZ) + nla_strscpy(tname, tb[TCA_IPT_TABLE], IFNAMSIZ) >= IFNAMSIZ) strcpy(tname, "mangle"); t = kmemdup(td, td->u.target_size, GFP_KERNEL); diff --git a/net/sched/act_simple.c b/net/sched/act_simple.c index a4f3d0f0daa9..726cc956d06f 100644 --- a/net/sched/act_simple.c +++ b/net/sched/act_simple.c @@ -52,7 +52,7 @@ static int alloc_defdata(struct tcf_defact *d, const struct nlattr *defdata) d->tcfd_defdata = kzalloc(SIMP_MAX_DATA, GFP_KERNEL); if (unlikely(!d->tcfd_defdata)) return -ENOMEM; - nla_strlcpy(d->tcfd_defdata, defdata, SIMP_MAX_DATA); + nla_strscpy(d->tcfd_defdata, defdata, SIMP_MAX_DATA); return 0; } @@ -71,7 +71,7 @@ static int reset_policy(struct tc_action *a, const struct nlattr *defdata, spin_lock_bh(&d->tcf_lock); goto_ch = tcf_action_set_ctrlact(a, p->action, goto_ch); memset(d->tcfd_defdata, 0, SIMP_MAX_DATA); - nla_strlcpy(d->tcfd_defdata, defdata, SIMP_MAX_DATA); + nla_strscpy(d->tcfd_defdata, defdata, SIMP_MAX_DATA); spin_unlock_bh(&d->tcf_lock); if (goto_ch) tcf_chain_put_by_act(goto_ch); diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index c2e9661e20d3..ff3e943febaa 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -223,7 +223,7 @@ static inline u32 tcf_auto_prio(struct tcf_proto *tp) static bool tcf_proto_check_kind(struct nlattr *kind, char *name) { if (kind) - return nla_strlcpy(name, kind, IFNAMSIZ) < 0; + return nla_strscpy(name, kind, IFNAMSIZ) < 0; memset(name, 0, IFNAMSIZ); return false; } diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 05449286d889..1a2d2471b078 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -1170,7 +1170,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev, #ifdef CONFIG_MODULES if (ops == NULL && kind != NULL) { char name[IFNAMSIZ]; - if (nla_strlcpy(name, kind, IFNAMSIZ) >= 0) { + if (nla_strscpy(name, kind, IFNAMSIZ) >= 0) { /* We dropped the RTNL semaphore in order to * perform the module load. So, even if we * succeeded in loading the module we have to diff --git a/net/tipc/netlink_compat.c b/net/tipc/netlink_compat.c index 5c6206c9d6a8..82f154989418 100644 --- a/net/tipc/netlink_compat.c +++ b/net/tipc/netlink_compat.c @@ -696,7 +696,7 @@ static int tipc_nl_compat_link_dump(struct tipc_nl_compat_msg *msg, link_info.dest = nla_get_flag(link[TIPC_NLA_LINK_DEST]); link_info.up = htonl(nla_get_flag(link[TIPC_NLA_LINK_UP])); - nla_strlcpy(link_info.str, link[TIPC_NLA_LINK_NAME], + nla_strscpy(link_info.str, link[TIPC_NLA_LINK_NAME], TIPC_MAX_LINK_NAME); return tipc_add_tlv(msg->rep, TIPC_TLV_LINK_INFO, -- cgit v1.2.3 From 9d9e937b1c8be97b424e3e11938e183fcde905c0 Mon Sep 17 00:00:00 2001 From: Georg Kohmann Date: Wed, 11 Nov 2020 12:50:25 +0100 Subject: ipv6/netfilter: Discard first fragment not including all headers Packets are processed even though the first fragment don't include all headers through the upper layer header. This breaks TAHI IPv6 Core Conformance Test v6LC.1.3.6. Referring to RFC8200 SECTION 4.5: "If the first fragment does not include all headers through an Upper-Layer header, then that fragment should be discarded and an ICMP Parameter Problem, Code 3, message should be sent to the source of the fragment, with the Pointer field set to zero." The fragment needs to be validated the same way it is done in commit 2efdaaaf883a ("IPv6: reply ICMP error if the first fragment don't include all headers") for ipv6. Wrap the validation into a common function, ipv6_frag_thdr_truncated() to check for truncation in the upper layer header. This validation does not fullfill all aspects of RFC 8200, section 4.5, but is at the moment sufficient to pass mentioned TAHI test. In netfilter, utilize the fragment offset returned by find_prev_fhdr() to let ipv6_frag_thdr_truncated() start it's traverse from the fragment header. Return 0 to drop the fragment in the netfilter. This is the same behaviour as used on other protocol errors in this function, e.g. when nf_ct_frag6_queue() returns -EPROTO. The Fragment will later be picked up by ipv6_frag_rcv() in reassembly.c. ipv6_frag_rcv() will then send an appropriate ICMP Parameter Problem message back to the source. References commit 2efdaaaf883a ("IPv6: reply ICMP error if the first fragment don't include all headers") Signed-off-by: Georg Kohmann Acked-by: Pablo Neira Ayuso Link: https://lore.kernel.org/r/20201111115025.28879-1-geokohma@cisco.com Signed-off-by: Jakub Kicinski --- include/net/ipv6.h | 2 ++ net/ipv6/netfilter/nf_conntrack_reasm.c | 9 ++++++ net/ipv6/reassembly.c | 55 +++++++++++++++++++++------------ 3 files changed, 46 insertions(+), 20 deletions(-) (limited to 'include/net') diff --git a/include/net/ipv6.h b/include/net/ipv6.h index bd1f396cc9c7..637cc6dd12b7 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -1064,6 +1064,8 @@ int ipv6_skip_exthdr(const struct sk_buff *, int start, u8 *nexthdrp, bool ipv6_ext_hdr(u8 nexthdr); +bool ipv6_frag_thdr_truncated(struct sk_buff *skb, int start, u8 *nexthdrp); + enum { IP6_FH_F_FRAG = (1 << 0), IP6_FH_F_AUTH = (1 << 1), diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index 054d287eb13d..b9cc0b330dbe 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -440,6 +440,7 @@ find_prev_fhdr(struct sk_buff *skb, u8 *prevhdrp, int *prevhoff, int *fhoff) int nf_ct_frag6_gather(struct net *net, struct sk_buff *skb, u32 user) { u16 savethdr = skb->transport_header; + u8 nexthdr = NEXTHDR_FRAGMENT; int fhoff, nhoff, ret; struct frag_hdr *fhdr; struct frag_queue *fq; @@ -455,6 +456,14 @@ int nf_ct_frag6_gather(struct net *net, struct sk_buff *skb, u32 user) if (find_prev_fhdr(skb, &prevhdr, &nhoff, &fhoff) < 0) return 0; + /* Discard the first fragment if it does not include all headers + * RFC 8200, Section 4.5 + */ + if (ipv6_frag_thdr_truncated(skb, fhoff, &nexthdr)) { + pr_debug("Drop incomplete fragment\n"); + return 0; + } + if (!pskb_may_pull(skb, fhoff + sizeof(*fhdr))) return -ENOMEM; diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c index c8cf1bbad74a..e3869bac9c88 100644 --- a/net/ipv6/reassembly.c +++ b/net/ipv6/reassembly.c @@ -318,15 +318,43 @@ out_fail: return -1; } +/* Check if the upper layer header is truncated in the first fragment. */ +bool ipv6_frag_thdr_truncated(struct sk_buff *skb, int start, u8 *nexthdrp) +{ + u8 nexthdr = *nexthdrp; + __be16 frag_off; + int offset; + + offset = ipv6_skip_exthdr(skb, start, &nexthdr, &frag_off); + if (offset < 0 || (frag_off & htons(IP6_OFFSET))) + return false; + switch (nexthdr) { + case NEXTHDR_TCP: + offset += sizeof(struct tcphdr); + break; + case NEXTHDR_UDP: + offset += sizeof(struct udphdr); + break; + case NEXTHDR_ICMP: + offset += sizeof(struct icmp6hdr); + break; + default: + offset += 1; + } + if (offset > skb->len) + return true; + return false; +} +EXPORT_SYMBOL(ipv6_frag_thdr_truncated); + static int ipv6_frag_rcv(struct sk_buff *skb) { struct frag_hdr *fhdr; struct frag_queue *fq; const struct ipv6hdr *hdr = ipv6_hdr(skb); struct net *net = dev_net(skb_dst(skb)->dev); - __be16 frag_off; - int iif, offset; u8 nexthdr; + int iif; if (IP6CB(skb)->flags & IP6SKB_FRAGMENTED) goto fail_hdr; @@ -362,24 +390,11 @@ static int ipv6_frag_rcv(struct sk_buff *skb) * the source of the fragment, with the Pointer field set to zero. */ nexthdr = hdr->nexthdr; - offset = ipv6_skip_exthdr(skb, skb_transport_offset(skb), &nexthdr, &frag_off); - if (offset >= 0) { - /* Check some common protocols' header */ - if (nexthdr == IPPROTO_TCP) - offset += sizeof(struct tcphdr); - else if (nexthdr == IPPROTO_UDP) - offset += sizeof(struct udphdr); - else if (nexthdr == IPPROTO_ICMPV6) - offset += sizeof(struct icmp6hdr); - else - offset += 1; - - if (!(frag_off & htons(IP6_OFFSET)) && offset > skb->len) { - __IP6_INC_STATS(net, __in6_dev_get_safely(skb->dev), - IPSTATS_MIB_INHDRERRORS); - icmpv6_param_prob(skb, ICMPV6_HDR_INCOMP, 0); - return -1; - } + if (ipv6_frag_thdr_truncated(skb, skb_transport_offset(skb), &nexthdr)) { + __IP6_INC_STATS(net, __in6_dev_get_safely(skb->dev), + IPSTATS_MIB_INHDRERRORS); + icmpv6_param_prob(skb, ICMPV6_HDR_INCOMP, 0); + return -1; } iif = skb->dev ? skb->dev->ifindex : 0; -- cgit v1.2.3 From b796d04bd014fd24e60ab4a6c604b258ac947825 Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Mon, 16 Nov 2020 10:48:02 +0100 Subject: tcp: factor out tcp_build_frag() Will be needed by the next patch, as MPTCP needs to handle directly the error/memory-allocation-needed path. No functional changes intended. Additionally let MPTCP code access the tcp_remove_empty_skb() helper. Signed-off-by: Paolo Abeni Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 3 ++ net/ipv4/tcp.c | 119 ++++++++++++++++++++++++++++++------------------------ 2 files changed, 70 insertions(+), 52 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index d643ee4e4249..8aee43ef0578 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -322,6 +322,7 @@ void tcp_shutdown(struct sock *sk, int how); int tcp_v4_early_demux(struct sk_buff *skb); int tcp_v4_rcv(struct sk_buff *skb); +void tcp_remove_empty_skb(struct sock *sk, struct sk_buff *skb); int tcp_v4_tw_remember_stamp(struct inet_timewait_sock *tw); int tcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t size); int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size); @@ -329,6 +330,8 @@ int tcp_sendpage(struct sock *sk, struct page *page, int offset, size_t size, int flags); int tcp_sendpage_locked(struct sock *sk, struct page *page, int offset, size_t size, int flags); +struct sk_buff *tcp_build_frag(struct sock *sk, int size_goal, int flags, + struct page *page, int offset, size_t *size); ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset, size_t size, int flags); int tcp_send_mss(struct sock *sk, int *size_goal, int flags); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index b2bc3d7fe9e8..a40981e347c0 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -954,7 +954,7 @@ int tcp_send_mss(struct sock *sk, int *size_goal, int flags) * importantly be able to generate EPOLLOUT for Edge Trigger epoll() * users. */ -static void tcp_remove_empty_skb(struct sock *sk, struct sk_buff *skb) +void tcp_remove_empty_skb(struct sock *sk, struct sk_buff *skb) { if (skb && !skb->len) { tcp_unlink_write_queue(skb, sk); @@ -964,6 +964,68 @@ static void tcp_remove_empty_skb(struct sock *sk, struct sk_buff *skb) } } +struct sk_buff *tcp_build_frag(struct sock *sk, int size_goal, int flags, + struct page *page, int offset, size_t *size) +{ + struct sk_buff *skb = tcp_write_queue_tail(sk); + struct tcp_sock *tp = tcp_sk(sk); + bool can_coalesce; + int copy, i; + + if (!skb || (copy = size_goal - skb->len) <= 0 || + !tcp_skb_can_collapse_to(skb)) { +new_segment: + if (!sk_stream_memory_free(sk)) + return NULL; + + skb = sk_stream_alloc_skb(sk, 0, sk->sk_allocation, + tcp_rtx_and_write_queues_empty(sk)); + if (!skb) + return NULL; + +#ifdef CONFIG_TLS_DEVICE + skb->decrypted = !!(flags & MSG_SENDPAGE_DECRYPTED); +#endif + skb_entail(sk, skb); + copy = size_goal; + } + + if (copy > *size) + copy = *size; + + i = skb_shinfo(skb)->nr_frags; + can_coalesce = skb_can_coalesce(skb, i, page, offset); + if (!can_coalesce && i >= sysctl_max_skb_frags) { + tcp_mark_push(tp, skb); + goto new_segment; + } + if (!sk_wmem_schedule(sk, copy)) + return NULL; + + if (can_coalesce) { + skb_frag_size_add(&skb_shinfo(skb)->frags[i - 1], copy); + } else { + get_page(page); + skb_fill_page_desc(skb, i, page, offset, copy); + } + + if (!(flags & MSG_NO_SHARED_FRAGS)) + skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; + + skb->len += copy; + skb->data_len += copy; + skb->truesize += copy; + sk_wmem_queued_add(sk, copy); + sk_mem_charge(sk, copy); + skb->ip_summed = CHECKSUM_PARTIAL; + WRITE_ONCE(tp->write_seq, tp->write_seq + copy); + TCP_SKB_CB(skb)->end_seq += copy; + tcp_skb_pcount_set(skb, 0); + + *size = copy; + return skb; +} + ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset, size_t size, int flags) { @@ -999,60 +1061,13 @@ ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset, goto out_err; while (size > 0) { - struct sk_buff *skb = tcp_write_queue_tail(sk); - int copy, i; - bool can_coalesce; - - if (!skb || (copy = size_goal - skb->len) <= 0 || - !tcp_skb_can_collapse_to(skb)) { -new_segment: - if (!sk_stream_memory_free(sk)) - goto wait_for_space; - - skb = sk_stream_alloc_skb(sk, 0, sk->sk_allocation, - tcp_rtx_and_write_queues_empty(sk)); - if (!skb) - goto wait_for_space; - -#ifdef CONFIG_TLS_DEVICE - skb->decrypted = !!(flags & MSG_SENDPAGE_DECRYPTED); -#endif - skb_entail(sk, skb); - copy = size_goal; - } + struct sk_buff *skb; + size_t copy = size; - if (copy > size) - copy = size; - - i = skb_shinfo(skb)->nr_frags; - can_coalesce = skb_can_coalesce(skb, i, page, offset); - if (!can_coalesce && i >= sysctl_max_skb_frags) { - tcp_mark_push(tp, skb); - goto new_segment; - } - if (!sk_wmem_schedule(sk, copy)) + skb = tcp_build_frag(sk, size_goal, flags, page, offset, ©); + if (!skb) goto wait_for_space; - if (can_coalesce) { - skb_frag_size_add(&skb_shinfo(skb)->frags[i - 1], copy); - } else { - get_page(page); - skb_fill_page_desc(skb, i, page, offset, copy); - } - - if (!(flags & MSG_NO_SHARED_FRAGS)) - skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG; - - skb->len += copy; - skb->data_len += copy; - skb->truesize += copy; - sk_wmem_queued_add(sk, copy); - sk_mem_charge(sk, copy); - skb->ip_summed = CHECKSUM_PARTIAL; - WRITE_ONCE(tp->write_seq, tp->write_seq + copy); - TCP_SKB_CB(skb)->end_seq += copy; - tcp_skb_pcount_set(skb, 0); - if (!copied) TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_PSH; -- cgit v1.2.3 From 77c3c95637526f1e4330cc9a4b2065f668c2c4fe Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Mon, 16 Nov 2020 10:48:04 +0100 Subject: tcp: factor out __tcp_close() helper unlocked version of protocol level close, will be used by MPTCP to allow decouple orphaning and subflow level close. Signed-off-by: Paolo Abeni Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 1 + net/ipv4/tcp.c | 9 +++++++-- 2 files changed, 8 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 8aee43ef0578..d5d22c411918 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -395,6 +395,7 @@ void tcp_update_metrics(struct sock *sk); void tcp_init_metrics(struct sock *sk); void tcp_metrics_init(void); bool tcp_peer_is_proven(struct request_sock *req, struct dst_entry *dst); +void __tcp_close(struct sock *sk, long timeout); void tcp_close(struct sock *sk, long timeout); void tcp_init_sock(struct sock *sk); void tcp_init_transfer(struct sock *sk, int bpf_op, struct sk_buff *skb); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index a40981e347c0..b285b338a019 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2420,13 +2420,12 @@ bool tcp_check_oom(struct sock *sk, int shift) return too_many_orphans || out_of_socket_memory; } -void tcp_close(struct sock *sk, long timeout) +void __tcp_close(struct sock *sk, long timeout) { struct sk_buff *skb; int data_was_unread = 0; int state; - lock_sock(sk); sk->sk_shutdown = SHUTDOWN_MASK; if (sk->sk_state == TCP_LISTEN) { @@ -2590,6 +2589,12 @@ adjudge_to_death: out: bh_unlock_sock(sk); local_bh_enable(); +} + +void tcp_close(struct sock *sk, long timeout) +{ + lock_sock(sk); + __tcp_close(sk, timeout); release_sock(sk); sock_put(sk); } -- cgit v1.2.3 From 9349eb3a9d2ae0151510dd98b6640dfaeebee9cc Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Mon, 16 Nov 2020 12:12:46 +0100 Subject: xsk: Introduce batched Tx descriptor interfaces Introduce batched descriptor interfaces in the xsk core code for the Tx path to be used in the driver to write a code path with higher performance. This interface will be used by the i40e driver in the next patch. Though other drivers would likely benefit from this new interface too. Note that batching is only implemented for the common case when there is only one socket bound to the same device and queue id. When this is not the case, we fall back to the old non-batched version of the function. Signed-off-by: Magnus Karlsson Signed-off-by: Daniel Borkmann Acked-by: John Fastabend Link: https://lore.kernel.org/bpf/1605525167-14450-5-git-send-email-magnus.karlsson@gmail.com --- include/net/xdp_sock_drv.h | 7 ++++ net/xdp/xsk.c | 57 +++++++++++++++++++++++++++++ net/xdp/xsk_queue.h | 89 +++++++++++++++++++++++++++++++++++++++------- 3 files changed, 140 insertions(+), 13 deletions(-) (limited to 'include/net') diff --git a/include/net/xdp_sock_drv.h b/include/net/xdp_sock_drv.h index 5b1ee8a9976d..4e295541e396 100644 --- a/include/net/xdp_sock_drv.h +++ b/include/net/xdp_sock_drv.h @@ -13,6 +13,7 @@ void xsk_tx_completed(struct xsk_buff_pool *pool, u32 nb_entries); bool xsk_tx_peek_desc(struct xsk_buff_pool *pool, struct xdp_desc *desc); +u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_desc *desc, u32 max); void xsk_tx_release(struct xsk_buff_pool *pool); struct xsk_buff_pool *xsk_get_pool_from_qid(struct net_device *dev, u16 queue_id); @@ -128,6 +129,12 @@ static inline bool xsk_tx_peek_desc(struct xsk_buff_pool *pool, return false; } +static inline u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_desc *desc, + u32 max) +{ + return 0; +} + static inline void xsk_tx_release(struct xsk_buff_pool *pool) { } diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index cfbec3989a76..b0141973f23e 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -332,6 +332,63 @@ out: } EXPORT_SYMBOL(xsk_tx_peek_desc); +static u32 xsk_tx_peek_release_fallback(struct xsk_buff_pool *pool, struct xdp_desc *descs, + u32 max_entries) +{ + u32 nb_pkts = 0; + + while (nb_pkts < max_entries && xsk_tx_peek_desc(pool, &descs[nb_pkts])) + nb_pkts++; + + xsk_tx_release(pool); + return nb_pkts; +} + +u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_desc *descs, + u32 max_entries) +{ + struct xdp_sock *xs; + u32 nb_pkts; + + rcu_read_lock(); + if (!list_is_singular(&pool->xsk_tx_list)) { + /* Fallback to the non-batched version */ + rcu_read_unlock(); + return xsk_tx_peek_release_fallback(pool, descs, max_entries); + } + + xs = list_first_or_null_rcu(&pool->xsk_tx_list, struct xdp_sock, tx_list); + if (!xs) { + nb_pkts = 0; + goto out; + } + + nb_pkts = xskq_cons_peek_desc_batch(xs->tx, descs, pool, max_entries); + if (!nb_pkts) { + xs->tx->queue_empty_descs++; + goto out; + } + + /* This is the backpressure mechanism for the Tx path. Try to + * reserve space in the completion queue for all packets, but + * if there are fewer slots available, just process that many + * packets. This avoids having to implement any buffering in + * the Tx path. + */ + nb_pkts = xskq_prod_reserve_addr_batch(pool->cq, descs, nb_pkts); + if (!nb_pkts) + goto out; + + xskq_cons_release_n(xs->tx, nb_pkts); + __xskq_cons_release(xs->tx); + xs->sk.sk_write_space(&xs->sk); + +out: + rcu_read_unlock(); + return nb_pkts; +} +EXPORT_SYMBOL(xsk_tx_peek_release_desc_batch); + static int xsk_wakeup(struct xdp_sock *xs, u8 flags) { struct net_device *dev = xs->dev; diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h index 74fac802cce1..b936c46b1e16 100644 --- a/net/xdp/xsk_queue.h +++ b/net/xdp/xsk_queue.h @@ -199,6 +199,30 @@ static inline bool xskq_cons_read_desc(struct xsk_queue *q, return false; } +static inline u32 xskq_cons_read_desc_batch(struct xsk_queue *q, + struct xdp_desc *descs, + struct xsk_buff_pool *pool, u32 max) +{ + u32 cached_cons = q->cached_cons, nb_entries = 0; + + while (cached_cons != q->cached_prod && nb_entries < max) { + struct xdp_rxtx_ring *ring = (struct xdp_rxtx_ring *)q->ring; + u32 idx = cached_cons & q->ring_mask; + + descs[nb_entries] = ring->desc[idx]; + if (unlikely(!xskq_cons_is_valid_desc(q, &descs[nb_entries], pool))) { + /* Skip the entry */ + cached_cons++; + continue; + } + + nb_entries++; + cached_cons++; + } + + return nb_entries; +} + /* Functions for consumers */ static inline void __xskq_cons_release(struct xsk_queue *q) @@ -220,17 +244,22 @@ static inline void xskq_cons_get_entries(struct xsk_queue *q) __xskq_cons_peek(q); } -static inline bool xskq_cons_has_entries(struct xsk_queue *q, u32 cnt) +static inline u32 xskq_cons_nb_entries(struct xsk_queue *q, u32 max) { u32 entries = q->cached_prod - q->cached_cons; - if (entries >= cnt) - return true; + if (entries >= max) + return max; __xskq_cons_peek(q); entries = q->cached_prod - q->cached_cons; - return entries >= cnt; + return entries >= max ? max : entries; +} + +static inline bool xskq_cons_has_entries(struct xsk_queue *q, u32 cnt) +{ + return xskq_cons_nb_entries(q, cnt) >= cnt ? true : false; } static inline bool xskq_cons_peek_addr_unchecked(struct xsk_queue *q, u64 *addr) @@ -249,16 +278,28 @@ static inline bool xskq_cons_peek_desc(struct xsk_queue *q, return xskq_cons_read_desc(q, desc, pool); } +static inline u32 xskq_cons_peek_desc_batch(struct xsk_queue *q, struct xdp_desc *descs, + struct xsk_buff_pool *pool, u32 max) +{ + u32 entries = xskq_cons_nb_entries(q, max); + + return xskq_cons_read_desc_batch(q, descs, pool, entries); +} + +/* To improve performance in the xskq_cons_release functions, only update local state here. + * Reflect this to global state when we get new entries from the ring in + * xskq_cons_get_entries() and whenever Rx or Tx processing are completed in the NAPI loop. + */ static inline void xskq_cons_release(struct xsk_queue *q) { - /* To improve performance, only update local state here. - * Reflect this to global state when we get new entries - * from the ring in xskq_cons_get_entries() and whenever - * Rx or Tx processing are completed in the NAPI loop. - */ q->cached_cons++; } +static inline void xskq_cons_release_n(struct xsk_queue *q, u32 cnt) +{ + q->cached_cons += cnt; +} + static inline bool xskq_cons_is_full(struct xsk_queue *q) { /* No barriers needed since data is not accessed */ @@ -268,18 +309,23 @@ static inline bool xskq_cons_is_full(struct xsk_queue *q) /* Functions for producers */ -static inline bool xskq_prod_is_full(struct xsk_queue *q) +static inline u32 xskq_prod_nb_free(struct xsk_queue *q, u32 max) { u32 free_entries = q->nentries - (q->cached_prod - q->cached_cons); - if (free_entries) - return false; + if (free_entries >= max) + return max; /* Refresh the local tail pointer */ q->cached_cons = READ_ONCE(q->ring->consumer); free_entries = q->nentries - (q->cached_prod - q->cached_cons); - return !free_entries; + return free_entries >= max ? max : free_entries; +} + +static inline bool xskq_prod_is_full(struct xsk_queue *q) +{ + return xskq_prod_nb_free(q, 1) ? false : true; } static inline int xskq_prod_reserve(struct xsk_queue *q) @@ -304,6 +350,23 @@ static inline int xskq_prod_reserve_addr(struct xsk_queue *q, u64 addr) return 0; } +static inline u32 xskq_prod_reserve_addr_batch(struct xsk_queue *q, struct xdp_desc *descs, + u32 max) +{ + struct xdp_umem_ring *ring = (struct xdp_umem_ring *)q->ring; + u32 nb_entries, i, cached_prod; + + nb_entries = xskq_prod_nb_free(q, max); + + /* A, matches D */ + cached_prod = q->cached_prod; + for (i = 0; i < nb_entries; i++) + ring->desc[cached_prod++ & q->ring_mask] = descs[i].addr; + q->cached_prod = cached_prod; + + return nb_entries; +} + static inline int xskq_prod_reserve_desc(struct xsk_queue *q, u64 addr, u32 len) { -- cgit v1.2.3 From 138559b9f99d3b6b1d5e75c78facc067a23871c6 Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Sun, 15 Nov 2020 15:14:48 +0200 Subject: net/tls: Fix wrong record sn in async mode of device resync In async_resync mode, we log the TCP seq of records until the async request is completed. Later, in case one of the logged seqs matches the resync request, we return it, together with its record serial number. Before this fix, we mistakenly returned the serial number of the current record instead. Fixes: ed9b7646b06a ("net/tls: Add asynchronous resync") Signed-off-by: Tariq Toukan Reviewed-by: Boris Pismenny Link: https://lore.kernel.org/r/20201115131448.2702-1-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- include/net/tls.h | 16 +++++++++++++++- net/tls/tls_device.c | 37 +++++++++++++++++++++++++++---------- 2 files changed, 42 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/tls.h b/include/net/tls.h index baf1e99d8193..cf1473099453 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -300,7 +300,8 @@ enum tls_offload_sync_type { #define TLS_DEVICE_RESYNC_ASYNC_LOGMAX 13 struct tls_offload_resync_async { atomic64_t req; - u32 loglen; + u16 loglen; + u16 rcd_delta; u32 log[TLS_DEVICE_RESYNC_ASYNC_LOGMAX]; }; @@ -471,6 +472,18 @@ static inline bool tls_bigint_increment(unsigned char *seq, int len) return (i == -1); } +static inline void tls_bigint_subtract(unsigned char *seq, int n) +{ + u64 rcd_sn; + __be64 *p; + + BUILD_BUG_ON(TLS_MAX_REC_SEQ_SIZE != 8); + + p = (__be64 *)seq; + rcd_sn = be64_to_cpu(*p); + *p = cpu_to_be64(rcd_sn - n); +} + static inline struct tls_context *tls_get_ctx(const struct sock *sk) { struct inet_connection_sock *icsk = inet_csk(sk); @@ -639,6 +652,7 @@ tls_offload_rx_resync_async_request_start(struct sock *sk, __be32 seq, u16 len) atomic64_set(&rx_ctx->resync_async->req, ((u64)ntohl(seq) << 32) | ((u64)len << 16) | RESYNC_REQ | RESYNC_REQ_ASYNC); rx_ctx->resync_async->loglen = 0; + rx_ctx->resync_async->rcd_delta = 0; } static inline void diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c index cec86229a6a0..54d3e161d198 100644 --- a/net/tls/tls_device.c +++ b/net/tls/tls_device.c @@ -694,36 +694,51 @@ static void tls_device_resync_rx(struct tls_context *tls_ctx, static bool tls_device_rx_resync_async(struct tls_offload_resync_async *resync_async, - s64 resync_req, u32 *seq) + s64 resync_req, u32 *seq, u16 *rcd_delta) { u32 is_async = resync_req & RESYNC_REQ_ASYNC; u32 req_seq = resync_req >> 32; u32 req_end = req_seq + ((resync_req >> 16) & 0xffff); + u16 i; + + *rcd_delta = 0; if (is_async) { + /* shouldn't get to wraparound: + * too long in async stage, something bad happened + */ + if (WARN_ON_ONCE(resync_async->rcd_delta == USHRT_MAX)) + return false; + /* asynchronous stage: log all headers seq such that * req_seq <= seq <= end_seq, and wait for real resync request */ - if (between(*seq, req_seq, req_end) && + if (before(*seq, req_seq)) + return false; + if (!after(*seq, req_end) && resync_async->loglen < TLS_DEVICE_RESYNC_ASYNC_LOGMAX) resync_async->log[resync_async->loglen++] = *seq; + resync_async->rcd_delta++; + return false; } /* synchronous stage: check against the logged entries and * proceed to check the next entries if no match was found */ - while (resync_async->loglen) { - if (req_seq == resync_async->log[resync_async->loglen - 1] && - atomic64_try_cmpxchg(&resync_async->req, - &resync_req, 0)) { - resync_async->loglen = 0; + for (i = 0; i < resync_async->loglen; i++) + if (req_seq == resync_async->log[i] && + atomic64_try_cmpxchg(&resync_async->req, &resync_req, 0)) { + *rcd_delta = resync_async->rcd_delta - i; *seq = req_seq; + resync_async->loglen = 0; + resync_async->rcd_delta = 0; return true; } - resync_async->loglen--; - } + + resync_async->loglen = 0; + resync_async->rcd_delta = 0; if (req_seq == *seq && atomic64_try_cmpxchg(&resync_async->req, @@ -741,6 +756,7 @@ void tls_device_rx_resync_new_rec(struct sock *sk, u32 rcd_len, u32 seq) u32 sock_data, is_req_pending; struct tls_prot_info *prot; s64 resync_req; + u16 rcd_delta; u32 req_seq; if (tls_ctx->rx_conf != TLS_HW) @@ -786,8 +802,9 @@ void tls_device_rx_resync_new_rec(struct sock *sk, u32 rcd_len, u32 seq) return; if (!tls_device_rx_resync_async(rx_ctx->resync_async, - resync_req, &seq)) + resync_req, &seq, &rcd_delta)) return; + tls_bigint_subtract(rcd_sn, rcd_delta); break; } -- cgit v1.2.3 From 6636b6dcc3db2258cd0585b8078c1c225c4b6dde Mon Sep 17 00:00:00 2001 From: Jianyong Wu Date: Wed, 23 Sep 2020 22:11:46 +0800 Subject: 9p: add refcount to p9_fid struct Fix race issue in fid contention. Eric's and Greg's patch offer a mechanism to fix open-unlink-f*syscall bug in 9p. But there is race issue in fid parallel accesses. As Greg's patch stores all of fids from opened files into according inode, so all the lookup fid ops can retrieve fid from inode preferentially. But there is no mechanism to handle the fid contention issue. For example, there are two threads get the same fid in the same time and one of them clunk the fid before the other thread ready to discard the fid. In this scenario, it will lead to some fatal problems, even kernel core dump. I introduce a mechanism to fix this race issue. A counter field introduced into p9_fid struct to store the reference counter to the fid. When a fid is allocated from the inode or dentry, the counter will increase, and will decrease at the end of its occupation. It is guaranteed that the fid won't be clunked before the reference counter go down to 0, then we can avoid the clunked fid to be used. tests: race issue test from the old test case: for file in {01..50}; do touch f.${file}; done seq 1 1000 | xargs -n 1 -P 50 -I{} cat f.* > /dev/null open-unlink-f*syscall test: I have tested for f*syscall include: ftruncate fstat fchown fchmod faccessat. Link: http://lkml.kernel.org/r/20200923141146.90046-5-jianyong.wu@arm.com Fixes: 478ba09edc1f ("fs/9p: search open fids first") Signed-off-by: Jianyong Wu Signed-off-by: Dominique Martinet --- fs/9p/fid.c | 22 ++++++++++++++++++---- fs/9p/fid.h | 10 +++++++++- fs/9p/vfs_dentry.c | 2 ++ fs/9p/vfs_dir.c | 9 +++++---- fs/9p/vfs_inode.c | 37 +++++++++++++++++++++++++++++-------- fs/9p/vfs_inode_dotl.c | 34 ++++++++++++++++++++++++++++------ fs/9p/vfs_super.c | 1 + fs/9p/xattr.c | 16 +++++++++++++--- include/net/9p/client.h | 6 ++++++ net/9p/client.c | 9 +++++---- 10 files changed, 116 insertions(+), 30 deletions(-) (limited to 'include/net') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 0b23b0fe6c51..89643dabcdae 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -28,6 +28,7 @@ static inline void __add_fid(struct dentry *dentry, struct p9_fid *fid) { + atomic_set(&fid->count, 1); hlist_add_head(&fid->dlist, (struct hlist_head *)&dentry->d_fsdata); } @@ -60,6 +61,8 @@ static struct p9_fid *v9fs_fid_find_inode(struct inode *inode, kuid_t uid) break; } } + if (ret && !IS_ERR(ret)) + atomic_inc(&ret->count); spin_unlock(&inode->i_lock); return ret; } @@ -74,6 +77,7 @@ static struct p9_fid *v9fs_fid_find_inode(struct inode *inode, kuid_t uid) void v9fs_open_fid_add(struct inode *inode, struct p9_fid *fid) { spin_lock(&inode->i_lock); + atomic_set(&fid->count, 1); hlist_add_head(&fid->ilist, (struct hlist_head *)&inode->i_private); spin_unlock(&inode->i_lock); } @@ -106,6 +110,7 @@ static struct p9_fid *v9fs_fid_find(struct dentry *dentry, kuid_t uid, int any) hlist_for_each_entry(fid, h, dlist) { if (any || uid_eq(fid->uid, uid)) { ret = fid; + atomic_inc(&ret->count); break; } } @@ -167,7 +172,10 @@ static struct p9_fid *v9fs_fid_lookup_with_uid(struct dentry *dentry, fid = v9fs_fid_find(ds, uid, any); if (fid) { /* Found the parent fid do a lookup with that */ - fid = p9_client_walk(fid, 1, &dentry->d_name.name, 1); + struct p9_fid *ofid = fid; + + fid = p9_client_walk(ofid, 1, &dentry->d_name.name, 1); + p9_client_clunk(ofid); goto fid_out; } up_read(&v9ses->rename_sem); @@ -192,8 +200,10 @@ static struct p9_fid *v9fs_fid_lookup_with_uid(struct dentry *dentry, v9fs_fid_add(dentry->d_sb->s_root, fid); } /* If we are root ourself just return that */ - if (dentry->d_sb->s_root == dentry) + if (dentry->d_sb->s_root == dentry) { + atomic_inc(&fid->count); return fid; + } /* * Do a multipath walk with attached root. * When walking parent we need to make sure we @@ -240,6 +250,7 @@ fid_out: fid = ERR_PTR(-ENOENT); } else { __add_fid(dentry, fid); + atomic_inc(&fid->count); spin_unlock(&dentry->d_lock); } } @@ -290,11 +301,14 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) { int err; - struct p9_fid *fid; + struct p9_fid *fid, *ofid; - fid = clone_fid(v9fs_fid_lookup_with_uid(dentry, GLOBAL_ROOT_UID, 0)); + ofid = v9fs_fid_lookup_with_uid(dentry, GLOBAL_ROOT_UID, 0); + if (ofid && !IS_ERR(ofid)) + fid = clone_fid(ofid); if (IS_ERR(fid)) goto error_out; + p9_client_clunk(ofid); /* * writeback fid will only be used to write back the * dirty pages. We always request for the open fid in read-write diff --git a/fs/9p/fid.h b/fs/9p/fid.h index dfa11df02818..1fed96546728 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -22,6 +22,14 @@ static inline struct p9_fid *clone_fid(struct p9_fid *fid) } static inline struct p9_fid *v9fs_fid_clone(struct dentry *dentry) { - return clone_fid(v9fs_fid_lookup(dentry)); + struct p9_fid *fid, *nfid; + + fid = v9fs_fid_lookup(dentry); + if (!fid || IS_ERR(fid)) + return fid; + + nfid = p9_client_walk(fid, 0, NULL, 1); + p9_client_clunk(fid); + return nfid; } #endif diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index 7d6f69aefd45..4b4292123b3d 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -85,6 +85,8 @@ static int v9fs_lookup_revalidate(struct dentry *dentry, unsigned int flags) retval = v9fs_refresh_inode_dotl(fid, inode); else retval = v9fs_refresh_inode(fid, inode); + p9_client_clunk(fid); + if (retval == -ENOENT) return 0; if (retval < 0) diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index d82d8a346f86..b6a5a0be444d 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -210,11 +210,12 @@ int v9fs_dir_release(struct inode *inode, struct file *filp) fid = filp->private_data; p9_debug(P9_DEBUG_VFS, "inode: %p filp: %p fid: %d\n", inode, filp, fid ? fid->fid : -1); - spin_lock(&inode->i_lock); - hlist_del(&fid->ilist); - spin_unlock(&inode->i_lock); - if (fid) + if (fid) { + spin_lock(&inode->i_lock); + hlist_del(&fid->ilist); + spin_unlock(&inode->i_lock); p9_client_clunk(fid); + } return 0; } diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 6b243ffcbcf0..4a937fac1acb 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -551,6 +551,7 @@ static int v9fs_remove(struct inode *dir, struct dentry *dentry, int flags) if (v9fs_proto_dotl(v9ses)) retval = p9_client_unlinkat(dfid, dentry->d_name.name, v9fs_at_to_dotl_flags(flags)); + p9_client_clunk(dfid); if (retval == -EOPNOTSUPP) { /* Try the one based on path */ v9fid = v9fs_fid_clone(dentry); @@ -595,14 +596,12 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, { int err; const unsigned char *name; - struct p9_fid *dfid, *ofid, *fid; + struct p9_fid *dfid, *ofid = NULL, *fid = NULL; struct inode *inode; p9_debug(P9_DEBUG_VFS, "name %pd\n", dentry); err = 0; - ofid = NULL; - fid = NULL; name = dentry->d_name.name; dfid = v9fs_parent_fid(dentry); if (IS_ERR(dfid)) { @@ -616,12 +615,14 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, if (IS_ERR(ofid)) { err = PTR_ERR(ofid); p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err); + p9_client_clunk(dfid); return ERR_PTR(err); } err = p9_client_fcreate(ofid, name, perm, mode, extension); if (err < 0) { p9_debug(P9_DEBUG_VFS, "p9_client_fcreate failed %d\n", err); + p9_client_clunk(dfid); goto error; } @@ -633,6 +634,7 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err); fid = NULL; + p9_client_clunk(dfid); goto error; } /* @@ -643,11 +645,13 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, err = PTR_ERR(inode); p9_debug(P9_DEBUG_VFS, "inode creation failed %d\n", err); + p9_client_clunk(dfid); goto error; } v9fs_fid_add(dentry, fid); d_instantiate(dentry, inode); } + p9_client_clunk(dfid); return ofid; error: if (ofid) @@ -760,6 +764,7 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, */ name = dentry->d_name.name; fid = p9_client_walk(dfid, 1, &name, 1); + p9_client_clunk(dfid); if (fid == ERR_PTR(-ENOENT)) inode = NULL; else if (IS_ERR(fid)) @@ -910,7 +915,7 @@ v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *old_inode; struct inode *new_inode; struct v9fs_session_info *v9ses; - struct p9_fid *oldfid; + struct p9_fid *oldfid, *dfid; struct p9_fid *olddirfid; struct p9_fid *newdirfid; struct p9_wstat wstat; @@ -927,13 +932,20 @@ v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (IS_ERR(oldfid)) return PTR_ERR(oldfid); - olddirfid = clone_fid(v9fs_parent_fid(old_dentry)); + dfid = v9fs_parent_fid(old_dentry); + olddirfid = clone_fid(dfid); + if (dfid && !IS_ERR(dfid)) + p9_client_clunk(dfid); + if (IS_ERR(olddirfid)) { retval = PTR_ERR(olddirfid); goto done; } - newdirfid = clone_fid(v9fs_parent_fid(new_dentry)); + dfid = v9fs_parent_fid(new_dentry); + newdirfid = clone_fid(dfid); + p9_client_clunk(dfid); + if (IS_ERR(newdirfid)) { retval = PTR_ERR(newdirfid); goto clunk_olddir; @@ -990,6 +1002,7 @@ clunk_olddir: p9_client_clunk(olddirfid); done: + p9_client_clunk(oldfid); return retval; } @@ -1022,6 +1035,7 @@ v9fs_vfs_getattr(const struct path *path, struct kstat *stat, return PTR_ERR(fid); st = p9_client_stat(fid); + p9_client_clunk(fid); if (IS_ERR(st)) return PTR_ERR(st); @@ -1042,7 +1056,7 @@ v9fs_vfs_getattr(const struct path *path, struct kstat *stat, static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) { - int retval; + int retval, use_dentry = 0; struct v9fs_session_info *v9ses; struct p9_fid *fid = NULL; struct p9_wstat wstat; @@ -1058,8 +1072,10 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) fid = iattr->ia_file->private_data; WARN_ON(!fid); } - if (!fid) + if (!fid) { fid = v9fs_fid_lookup(dentry); + use_dentry = 1; + } if(IS_ERR(fid)) return PTR_ERR(fid); @@ -1089,6 +1105,10 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) filemap_write_and_wait(d_inode(dentry)->i_mapping); retval = p9_client_wstat(fid, &wstat); + + if (use_dentry) + p9_client_clunk(fid); + if (retval < 0) return retval; @@ -1213,6 +1233,7 @@ static const char *v9fs_vfs_get_link(struct dentry *dentry, return ERR_PTR(-EBADF); st = p9_client_stat(fid); + p9_client_clunk(fid); if (IS_ERR(st)) return ERR_CAST(st); diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 08f2e089fb0e..823c2eb5f1bf 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -296,6 +296,7 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, /* instantiate inode and assign the unopened fid to the dentry */ fid = p9_client_walk(dfid, 1, &name, 1); + p9_client_clunk(dfid); if (IS_ERR(fid)) { err = PTR_ERR(fid); p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err); @@ -408,7 +409,6 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, err = p9_client_mkdir_dotl(dfid, name, mode, gid, &qid); if (err < 0) goto error; - fid = p9_client_walk(dfid, 1, &name, 1); if (IS_ERR(fid)) { err = PTR_ERR(fid); @@ -452,6 +452,7 @@ error: if (fid) p9_client_clunk(fid); v9fs_put_acl(dacl, pacl); + p9_client_clunk(dfid); return err; } @@ -479,6 +480,7 @@ v9fs_vfs_getattr_dotl(const struct path *path, struct kstat *stat, */ st = p9_client_getattr_dotl(fid, P9_STATS_ALL); + p9_client_clunk(fid); if (IS_ERR(st)) return PTR_ERR(st); @@ -540,7 +542,7 @@ static int v9fs_mapped_iattr_valid(int iattr_valid) int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) { - int retval; + int retval, use_dentry = 0; struct p9_fid *fid = NULL; struct p9_iattr_dotl p9attr; struct inode *inode = d_inode(dentry); @@ -565,8 +567,10 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) fid = iattr->ia_file->private_data; WARN_ON(!fid); } - if (!fid) + if (!fid) { fid = v9fs_fid_lookup(dentry); + use_dentry = 1; + } if (IS_ERR(fid)) return PTR_ERR(fid); @@ -575,8 +579,11 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) filemap_write_and_wait(inode->i_mapping); retval = p9_client_setattr(fid, &p9attr); - if (retval < 0) + if (retval < 0) { + if (use_dentry) + p9_client_clunk(fid); return retval; + } if ((iattr->ia_valid & ATTR_SIZE) && iattr->ia_size != i_size_read(inode)) @@ -588,9 +595,15 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) if (iattr->ia_valid & ATTR_MODE) { /* We also want to update ACL when we update mode bits */ retval = v9fs_acl_chmod(inode, fid); - if (retval < 0) + if (retval < 0) { + if (use_dentry) + p9_client_clunk(fid); return retval; + } } + if (use_dentry) + p9_client_clunk(fid); + return 0; } @@ -742,6 +755,7 @@ error: if (fid) p9_client_clunk(fid); + p9_client_clunk(dfid); return err; } @@ -770,11 +784,15 @@ v9fs_vfs_link_dotl(struct dentry *old_dentry, struct inode *dir, return PTR_ERR(dfid); oldfid = v9fs_fid_lookup(old_dentry); - if (IS_ERR(oldfid)) + if (IS_ERR(oldfid)) { + p9_client_clunk(dfid); return PTR_ERR(oldfid); + } err = p9_client_link(dfid, oldfid, dentry->d_name.name); + p9_client_clunk(dfid); + p9_client_clunk(oldfid); if (err < 0) { p9_debug(P9_DEBUG_VFS, "p9_client_link failed %d\n", err); return err; @@ -789,6 +807,7 @@ v9fs_vfs_link_dotl(struct dentry *old_dentry, struct inode *dir, return PTR_ERR(fid); v9fs_refresh_inode_dotl(fid, d_inode(old_dentry)); + p9_client_clunk(fid); } ihold(d_inode(old_dentry)); d_instantiate(dentry, d_inode(old_dentry)); @@ -887,6 +906,8 @@ error: if (fid) p9_client_clunk(fid); v9fs_put_acl(dacl, pacl); + p9_client_clunk(dfid); + return err; } @@ -915,6 +936,7 @@ v9fs_vfs_get_link_dotl(struct dentry *dentry, if (IS_ERR(fid)) return ERR_CAST(fid); retval = p9_client_readlink(fid, &target); + p9_client_clunk(fid); if (retval) return ERR_PTR(retval); set_delayed_call(done, kfree_link, target); diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 9a21269b7234..5fce6e30bc5a 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -268,6 +268,7 @@ static int v9fs_statfs(struct dentry *dentry, struct kstatfs *buf) } res = simple_statfs(dentry, buf); done: + p9_client_clunk(fid); return res; } diff --git a/fs/9p/xattr.c b/fs/9p/xattr.c index ac8ff8ca4c11..87217dd0433e 100644 --- a/fs/9p/xattr.c +++ b/fs/9p/xattr.c @@ -71,14 +71,17 @@ ssize_t v9fs_xattr_get(struct dentry *dentry, const char *name, void *buffer, size_t buffer_size) { struct p9_fid *fid; + int ret; p9_debug(P9_DEBUG_VFS, "name = %s value_len = %zu\n", name, buffer_size); fid = v9fs_fid_lookup(dentry); if (IS_ERR(fid)) return PTR_ERR(fid); + ret = v9fs_fid_xattr_get(fid, name, buffer, buffer_size); + p9_client_clunk(fid); - return v9fs_fid_xattr_get(fid, name, buffer, buffer_size); + return ret; } /* @@ -96,8 +99,15 @@ ssize_t v9fs_xattr_get(struct dentry *dentry, const char *name, int v9fs_xattr_set(struct dentry *dentry, const char *name, const void *value, size_t value_len, int flags) { - struct p9_fid *fid = v9fs_fid_lookup(dentry); - return v9fs_fid_xattr_set(fid, name, value, value_len, flags); + int ret; + struct p9_fid *fid; + + fid = v9fs_fid_lookup(dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + ret = v9fs_fid_xattr_set(fid, name, value, value_len, flags); + p9_client_clunk(fid); + return ret; } int v9fs_fid_xattr_set(struct p9_fid *fid, const char *name, diff --git a/include/net/9p/client.h b/include/net/9p/client.h index ce7882da8e86..58ed9bd306bd 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -140,10 +140,16 @@ struct p9_client { * * TODO: This needs lots of explanation. */ +enum fid_source { + FID_FROM_OTHER, + FID_FROM_INODE, + FID_FROM_DENTRY, +}; struct p9_fid { struct p9_client *clnt; u32 fid; + atomic_t count; int mode; struct p9_qid qid; u32 iounit; diff --git a/net/9p/client.c b/net/9p/client.c index 1a3f72bf45fc..a6c8a915e0d8 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -901,6 +901,7 @@ static struct p9_fid *p9_fid_create(struct p9_client *clnt) fid->clnt = clnt; fid->rdir = NULL; fid->fid = 0; + atomic_set(&fid->count, 1); idr_preload(GFP_KERNEL); spin_lock_irq(&clnt->lock); @@ -908,7 +909,6 @@ static struct p9_fid *p9_fid_create(struct p9_client *clnt) GFP_NOWAIT); spin_unlock_irq(&clnt->lock); idr_preload_end(); - if (!ret) return fid; @@ -1187,7 +1187,6 @@ struct p9_fid *p9_client_walk(struct p9_fid *oldfid, uint16_t nwname, p9_debug(P9_DEBUG_9P, ">>> TWALK fids %d,%d nwname %ud wname[0] %s\n", oldfid->fid, fid->fid, nwname, wnames ? wnames[0] : NULL); - req = p9_client_rpc(clnt, P9_TWALK, "ddT", oldfid->fid, fid->fid, nwname, wnames); if (IS_ERR(req)) { @@ -1461,12 +1460,14 @@ int p9_client_clunk(struct p9_fid *fid) struct p9_req_t *req; int retries = 0; - if (!fid) { - pr_warn("%s (%d): Trying to clunk with NULL fid\n", + if (!fid || IS_ERR(fid)) { + pr_warn("%s (%d): Trying to clunk with invalid fid\n", __func__, task_pid_nr(current)); dump_stack(); return 0; } + if (!atomic_dec_and_test(&fid->count)) + return 0; again: p9_debug(P9_DEBUG_9P, ">>> TCLUNK fid %d (try %d)\n", fid->fid, -- cgit v1.2.3 From ff5e72ebef41068789c93b0666cebde80cc8bd8c Mon Sep 17 00:00:00 2001 From: Dominique Martinet Date: Tue, 3 Nov 2020 09:35:57 +0100 Subject: 9p: apply review requests for fid refcounting Fix style issues in parent commit ("apply review requests for fid refcounting"), no functional change. Link: http://lkml.kernel.org/r/1605802012-31133-2-git-send-email-asmadeus@codewreck.org Fixes: 6636b6dcc3db ("9p: add refcount to p9_fid struct") Signed-off-by: Dominique Martinet --- fs/9p/fid.c | 10 ++++------ fs/9p/fid.h | 2 +- include/net/9p/client.h | 2 +- net/9p/client.c | 4 ++-- 4 files changed, 8 insertions(+), 10 deletions(-) (limited to 'include/net') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 89643dabcdae..50118ec72a92 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -28,7 +28,6 @@ static inline void __add_fid(struct dentry *dentry, struct p9_fid *fid) { - atomic_set(&fid->count, 1); hlist_add_head(&fid->dlist, (struct hlist_head *)&dentry->d_fsdata); } @@ -62,7 +61,7 @@ static struct p9_fid *v9fs_fid_find_inode(struct inode *inode, kuid_t uid) } } if (ret && !IS_ERR(ret)) - atomic_inc(&ret->count); + refcount_inc(&ret->count); spin_unlock(&inode->i_lock); return ret; } @@ -77,7 +76,6 @@ static struct p9_fid *v9fs_fid_find_inode(struct inode *inode, kuid_t uid) void v9fs_open_fid_add(struct inode *inode, struct p9_fid *fid) { spin_lock(&inode->i_lock); - atomic_set(&fid->count, 1); hlist_add_head(&fid->ilist, (struct hlist_head *)&inode->i_private); spin_unlock(&inode->i_lock); } @@ -110,7 +108,7 @@ static struct p9_fid *v9fs_fid_find(struct dentry *dentry, kuid_t uid, int any) hlist_for_each_entry(fid, h, dlist) { if (any || uid_eq(fid->uid, uid)) { ret = fid; - atomic_inc(&ret->count); + refcount_inc(&ret->count); break; } } @@ -201,7 +199,7 @@ static struct p9_fid *v9fs_fid_lookup_with_uid(struct dentry *dentry, } /* If we are root ourself just return that */ if (dentry->d_sb->s_root == dentry) { - atomic_inc(&fid->count); + refcount_inc(&fid->count); return fid; } /* @@ -250,7 +248,7 @@ fid_out: fid = ERR_PTR(-ENOENT); } else { __add_fid(dentry, fid); - atomic_inc(&fid->count); + refcount_inc(&fid->count); spin_unlock(&dentry->d_lock); } } diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 1fed96546728..f7f33509e169 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -28,7 +28,7 @@ static inline struct p9_fid *v9fs_fid_clone(struct dentry *dentry) if (!fid || IS_ERR(fid)) return fid; - nfid = p9_client_walk(fid, 0, NULL, 1); + nfid = clone_fid(fid); p9_client_clunk(fid); return nfid; } diff --git a/include/net/9p/client.h b/include/net/9p/client.h index 58ed9bd306bd..e1c308d8d288 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -149,7 +149,7 @@ enum fid_source { struct p9_fid { struct p9_client *clnt; u32 fid; - atomic_t count; + refcount_t count; int mode; struct p9_qid qid; u32 iounit; diff --git a/net/9p/client.c b/net/9p/client.c index a6c8a915e0d8..ba4910138c5b 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -901,7 +901,7 @@ static struct p9_fid *p9_fid_create(struct p9_client *clnt) fid->clnt = clnt; fid->rdir = NULL; fid->fid = 0; - atomic_set(&fid->count, 1); + refcount_set(&fid->count, 1); idr_preload(GFP_KERNEL); spin_lock_irq(&clnt->lock); @@ -1466,7 +1466,7 @@ int p9_client_clunk(struct p9_fid *fid) dump_stack(); return 0; } - if (!atomic_dec_and_test(&fid->count)) + if (!refcount_dec_and_test(&fid->count)) return 0; again: -- cgit v1.2.3 From 2d8f6481c17db9fa5238b277cdbc392084060b09 Mon Sep 17 00:00:00 2001 From: Georg Kohmann Date: Thu, 19 Nov 2020 10:58:33 +0100 Subject: ipv6: Remove dependency of ipv6_frag_thdr_truncated on ipv6 module IPV6=m NF_DEFRAG_IPV6=y ld: net/ipv6/netfilter/nf_conntrack_reasm.o: in function `nf_ct_frag6_gather': net/ipv6/netfilter/nf_conntrack_reasm.c:462: undefined reference to `ipv6_frag_thdr_truncated' Netfilter is depending on ipv6 symbol ipv6_frag_thdr_truncated. This dependency is forcing IPV6=y. Remove this dependency by moving ipv6_frag_thdr_truncated out of ipv6. This is the same solution as used with a similar issues: Referring to commit 70b095c843266 ("ipv6: remove dependency of nf_defrag_ipv6 on ipv6 module") Fixes: 9d9e937b1c8b ("ipv6/netfilter: Discard first fragment not including all headers") Reported-by: Randy Dunlap Reported-by: kernel test robot Signed-off-by: Georg Kohmann Acked-by: Pablo Neira Ayuso Acked-by: Randy Dunlap # build-tested Link: https://lore.kernel.org/r/20201119095833.8409-1-geokohma@cisco.com Signed-off-by: Jakub Kicinski --- include/net/ipv6.h | 2 -- include/net/ipv6_frag.h | 30 ++++++++++++++++++++++++++++++ net/ipv6/netfilter/nf_conntrack_reasm.c | 2 +- net/ipv6/reassembly.c | 31 +------------------------------ 4 files changed, 32 insertions(+), 33 deletions(-) (limited to 'include/net') diff --git a/include/net/ipv6.h b/include/net/ipv6.h index 637cc6dd12b7..bd1f396cc9c7 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -1064,8 +1064,6 @@ int ipv6_skip_exthdr(const struct sk_buff *, int start, u8 *nexthdrp, bool ipv6_ext_hdr(u8 nexthdr); -bool ipv6_frag_thdr_truncated(struct sk_buff *skb, int start, u8 *nexthdrp); - enum { IP6_FH_F_FRAG = (1 << 0), IP6_FH_F_AUTH = (1 << 1), diff --git a/include/net/ipv6_frag.h b/include/net/ipv6_frag.h index a21e8b1381a1..851029ecff13 100644 --- a/include/net/ipv6_frag.h +++ b/include/net/ipv6_frag.h @@ -108,5 +108,35 @@ out_rcu_unlock: rcu_read_unlock(); inet_frag_put(&fq->q); } + +/* Check if the upper layer header is truncated in the first fragment. */ +static inline bool +ipv6frag_thdr_truncated(struct sk_buff *skb, int start, u8 *nexthdrp) +{ + u8 nexthdr = *nexthdrp; + __be16 frag_off; + int offset; + + offset = ipv6_skip_exthdr(skb, start, &nexthdr, &frag_off); + if (offset < 0 || (frag_off & htons(IP6_OFFSET))) + return false; + switch (nexthdr) { + case NEXTHDR_TCP: + offset += sizeof(struct tcphdr); + break; + case NEXTHDR_UDP: + offset += sizeof(struct udphdr); + break; + case NEXTHDR_ICMP: + offset += sizeof(struct icmp6hdr); + break; + default: + offset += 1; + } + if (offset > skb->len) + return true; + return false; +} + #endif #endif diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index b9cc0b330dbe..c129ad334eb3 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -459,7 +459,7 @@ int nf_ct_frag6_gather(struct net *net, struct sk_buff *skb, u32 user) /* Discard the first fragment if it does not include all headers * RFC 8200, Section 4.5 */ - if (ipv6_frag_thdr_truncated(skb, fhoff, &nexthdr)) { + if (ipv6frag_thdr_truncated(skb, fhoff, &nexthdr)) { pr_debug("Drop incomplete fragment\n"); return 0; } diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c index e3869bac9c88..47a0dc46cbdb 100644 --- a/net/ipv6/reassembly.c +++ b/net/ipv6/reassembly.c @@ -318,35 +318,6 @@ out_fail: return -1; } -/* Check if the upper layer header is truncated in the first fragment. */ -bool ipv6_frag_thdr_truncated(struct sk_buff *skb, int start, u8 *nexthdrp) -{ - u8 nexthdr = *nexthdrp; - __be16 frag_off; - int offset; - - offset = ipv6_skip_exthdr(skb, start, &nexthdr, &frag_off); - if (offset < 0 || (frag_off & htons(IP6_OFFSET))) - return false; - switch (nexthdr) { - case NEXTHDR_TCP: - offset += sizeof(struct tcphdr); - break; - case NEXTHDR_UDP: - offset += sizeof(struct udphdr); - break; - case NEXTHDR_ICMP: - offset += sizeof(struct icmp6hdr); - break; - default: - offset += 1; - } - if (offset > skb->len) - return true; - return false; -} -EXPORT_SYMBOL(ipv6_frag_thdr_truncated); - static int ipv6_frag_rcv(struct sk_buff *skb) { struct frag_hdr *fhdr; @@ -390,7 +361,7 @@ static int ipv6_frag_rcv(struct sk_buff *skb) * the source of the fragment, with the Pointer field set to zero. */ nexthdr = hdr->nexthdr; - if (ipv6_frag_thdr_truncated(skb, skb_transport_offset(skb), &nexthdr)) { + if (ipv6frag_thdr_truncated(skb, skb_transport_offset(skb), &nexthdr)) { __IP6_INC_STATS(net, __in6_dev_get_safely(skb->dev), IPSTATS_MIB_INHDRERRORS); icmpv6_param_prob(skb, ICMPV6_HDR_INCOMP, 0); -- cgit v1.2.3 From b44cfd4f5b912454387a4bf735d42eb4e7078ca8 Mon Sep 17 00:00:00 2001 From: Jacob Keller Date: Wed, 18 Nov 2020 11:06:35 -0800 Subject: devlink: move request_firmware out of driver All drivers which implement the devlink flash update support, with the exception of netdevsim, use either request_firmware or request_firmware_direct to locate the firmware file. Rather than having each driver do this separately as part of its .flash_update implementation, perform the request_firmware within net/core/devlink.c Replace the file_name parameter in the struct devlink_flash_update_params with a pointer to the fw object. Use request_firmware rather than request_firmware_direct. Although most Linux distributions today do not have the fallback mechanism implemented, only about half the drivers used the _direct request, as compared to the generic request_firmware. In the event that a distribution does support the fallback mechanism, the devlink flash update ought to be able to use it to provide the firmware contents. For distributions which do not support the fallback userspace mechanism, there should be essentially no difference between request_firmware and request_firmware_direct. Signed-off-by: Jacob Keller Acked-by: Shannon Nelson Acked-by: Vasundhara Volam Reviewed-by: Jiri Pirko Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c | 2 +- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 33 ++++++++++++++-------- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h | 4 +-- drivers/net/ethernet/huawei/hinic/hinic_devlink.c | 12 +------- drivers/net/ethernet/intel/ice/ice_devlink.c | 14 +-------- drivers/net/ethernet/mellanox/mlx5/core/devlink.c | 11 +------- drivers/net/ethernet/mellanox/mlxsw/core.c | 11 +------- drivers/net/ethernet/netronome/nfp/nfp_devlink.c | 2 +- drivers/net/ethernet/netronome/nfp/nfp_main.c | 17 ++--------- drivers/net/ethernet/netronome/nfp/nfp_main.h | 2 +- .../net/ethernet/pensando/ionic/ionic_devlink.c | 2 +- .../net/ethernet/pensando/ionic/ionic_devlink.h | 2 +- drivers/net/ethernet/pensando/ionic/ionic_fw.c | 12 ++------ include/net/devlink.h | 7 +++-- net/core/devlink.c | 26 +++++++++++++---- 15 files changed, 62 insertions(+), 95 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c index 184b6d0513b2..4ebae8a236fd 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c @@ -32,7 +32,7 @@ bnxt_dl_flash_update(struct devlink *dl, devlink_flash_update_begin_notify(dl); devlink_flash_update_status_notify(dl, "Preparing to flash", NULL, 0, 0); - rc = bnxt_flash_package_from_file(bp->dev, params->file_name, 0); + rc = bnxt_flash_package_from_fw_obj(bp->dev, params->fw, 0); if (!rc) devlink_flash_update_status_notify(dl, "Flashing done", NULL, 0, 0); else diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index 1471c9a36238..7b444fcb6289 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -2419,13 +2419,12 @@ static int bnxt_flash_firmware_from_file(struct net_device *dev, return rc; } -int bnxt_flash_package_from_file(struct net_device *dev, const char *filename, - u32 install_type) +int bnxt_flash_package_from_fw_obj(struct net_device *dev, const struct firmware *fw, + u32 install_type) { struct bnxt *bp = netdev_priv(dev); struct hwrm_nvm_install_update_output *resp = bp->hwrm_cmd_resp_addr; struct hwrm_nvm_install_update_input install = {0}; - const struct firmware *fw; u32 item_len; int rc = 0; u16 index; @@ -2440,13 +2439,6 @@ int bnxt_flash_package_from_file(struct net_device *dev, const char *filename, return rc; } - rc = request_firmware(&fw, filename, &dev->dev); - if (rc != 0) { - netdev_err(dev, "PKG error %d requesting file: %s\n", - rc, filename); - return rc; - } - if (fw->size > item_len) { netdev_err(dev, "PKG insufficient update area in nvram: %lu\n", (unsigned long)fw->size); @@ -2478,7 +2470,6 @@ int bnxt_flash_package_from_file(struct net_device *dev, const char *filename, dma_handle); } } - release_firmware(fw); if (rc) goto err_exit; @@ -2517,6 +2508,26 @@ err_exit: return rc; } +static int bnxt_flash_package_from_file(struct net_device *dev, const char *filename, + u32 install_type) +{ + const struct firmware *fw; + int rc; + + rc = request_firmware(&fw, filename, &dev->dev); + if (rc != 0) { + netdev_err(dev, "PKG error %d requesting file: %s\n", + rc, filename); + return rc; + } + + rc = bnxt_flash_package_from_fw_obj(dev, fw, install_type); + + release_firmware(fw); + + return rc; +} + static int bnxt_flash_device(struct net_device *dev, struct ethtool_flash *flash) { diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h index fa6fbde52bea..0a57cb6a4a4b 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h @@ -94,8 +94,8 @@ u32 bnxt_fw_to_ethtool_speed(u16); u16 bnxt_get_fw_auto_link_speeds(u32); int bnxt_hwrm_nvm_get_dev_info(struct bnxt *bp, struct hwrm_nvm_get_dev_info_output *nvm_dev_info); -int bnxt_flash_package_from_file(struct net_device *dev, const char *filename, - u32 install_type); +int bnxt_flash_package_from_fw_obj(struct net_device *dev, const struct firmware *fw, + u32 install_type); void bnxt_ethtool_init(struct bnxt *bp); void bnxt_ethtool_free(struct bnxt *bp); diff --git a/drivers/net/ethernet/huawei/hinic/hinic_devlink.c b/drivers/net/ethernet/huawei/hinic/hinic_devlink.c index 2630d667f393..58d5646444b0 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_devlink.c +++ b/drivers/net/ethernet/huawei/hinic/hinic_devlink.c @@ -285,18 +285,8 @@ static int hinic_devlink_flash_update(struct devlink *devlink, struct netlink_ext_ack *extack) { struct hinic_devlink_priv *priv = devlink_priv(devlink); - const struct firmware *fw; - int err; - - err = request_firmware_direct(&fw, params->file_name, - &priv->hwdev->hwif->pdev->dev); - if (err) - return err; - - err = hinic_firmware_update(priv, fw, extack); - release_firmware(fw); - return err; + return hinic_firmware_update(priv, params->fw, extack); } static const struct devlink_ops hinic_devlink_ops = { diff --git a/drivers/net/ethernet/intel/ice/ice_devlink.c b/drivers/net/ethernet/intel/ice/ice_devlink.c index 511da59bd6f2..0036d3e7df0b 100644 --- a/drivers/net/ethernet/intel/ice/ice_devlink.c +++ b/drivers/net/ethernet/intel/ice/ice_devlink.c @@ -247,9 +247,7 @@ ice_devlink_flash_update(struct devlink *devlink, struct netlink_ext_ack *extack) { struct ice_pf *pf = devlink_priv(devlink); - struct device *dev = &pf->pdev->dev; struct ice_hw *hw = &pf->hw; - const struct firmware *fw; u8 preservation; int err; @@ -277,21 +275,11 @@ ice_devlink_flash_update(struct devlink *devlink, if (err) return err; - err = request_firmware(&fw, params->file_name, dev); - if (err) { - NL_SET_ERR_MSG_MOD(extack, "Unable to read file from disk"); - return err; - } - - dev_dbg(dev, "Beginning flash update with file '%s'\n", params->file_name); - devlink_flash_update_begin_notify(devlink); devlink_flash_update_status_notify(devlink, "Preparing to flash", NULL, 0, 0); - err = ice_flash_pldm_image(pf, fw, preservation, extack); + err = ice_flash_pldm_image(pf, params->fw, preservation, extack); devlink_flash_update_end_notify(devlink); - release_firmware(fw); - return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c index a28f95df2901..e2ed341648e4 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c @@ -13,17 +13,8 @@ static int mlx5_devlink_flash_update(struct devlink *devlink, struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); - const struct firmware *fw; - int err; - - err = request_firmware_direct(&fw, params->file_name, &dev->pdev->dev); - if (err) - return err; - - err = mlx5_firmware_flash(dev, fw, extack); - release_firmware(fw); - return err; + return mlx5_firmware_flash(dev, params->fw, extack); } static u8 mlx5_fw_ver_major(u32 version) diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c index 1a86535c4968..630109f139a0 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core.c +++ b/drivers/net/ethernet/mellanox/mlxsw/core.c @@ -1117,16 +1117,7 @@ static int mlxsw_core_fw_flash_update(struct mlxsw_core *mlxsw_core, struct devlink_flash_update_params *params, struct netlink_ext_ack *extack) { - const struct firmware *firmware; - int err; - - err = request_firmware_direct(&firmware, params->file_name, mlxsw_core->bus_info->dev); - if (err) - return err; - err = mlxsw_core_fw_flash(mlxsw_core, firmware, extack); - release_firmware(firmware); - - return err; + return mlxsw_core_fw_flash(mlxsw_core, params->fw, extack); } static int mlxsw_core_devlink_param_fw_load_policy_validate(struct devlink *devlink, u32 id, diff --git a/drivers/net/ethernet/netronome/nfp/nfp_devlink.c b/drivers/net/ethernet/netronome/nfp/nfp_devlink.c index 97d2b03208de..713ee3041d49 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_devlink.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_devlink.c @@ -333,7 +333,7 @@ nfp_devlink_flash_update(struct devlink *devlink, struct devlink_flash_update_params *params, struct netlink_ext_ack *extack) { - return nfp_flash_update_common(devlink_priv(devlink), params->file_name, extack); + return nfp_flash_update_common(devlink_priv(devlink), params->fw, extack); } const struct devlink_ops nfp_devlink_ops = { diff --git a/drivers/net/ethernet/netronome/nfp/nfp_main.c b/drivers/net/ethernet/netronome/nfp/nfp_main.c index e672614d2906..742a420152b3 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_main.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_main.c @@ -301,11 +301,10 @@ static int nfp_pcie_sriov_configure(struct pci_dev *pdev, int num_vfs) return nfp_pcie_sriov_enable(pdev, num_vfs); } -int nfp_flash_update_common(struct nfp_pf *pf, const char *path, +int nfp_flash_update_common(struct nfp_pf *pf, const struct firmware *fw, struct netlink_ext_ack *extack) { struct device *dev = &pf->pdev->dev; - const struct firmware *fw; struct nfp_nsp *nsp; int err; @@ -319,24 +318,12 @@ int nfp_flash_update_common(struct nfp_pf *pf, const char *path, return err; } - err = request_firmware_direct(&fw, path, dev); - if (err) { - NL_SET_ERR_MSG_MOD(extack, - "unable to read flash file from disk"); - goto exit_close_nsp; - } - - dev_info(dev, "Please be patient while writing flash image: %s\n", - path); - err = nfp_nsp_write_flash(nsp, fw); if (err < 0) - goto exit_release_fw; + goto exit_close_nsp; dev_info(dev, "Finished writing flash image\n"); err = 0; -exit_release_fw: - release_firmware(fw); exit_close_nsp: nfp_nsp_close(nsp); return err; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_main.h b/drivers/net/ethernet/netronome/nfp/nfp_main.h index fa6b13a05941..a7dede946a33 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_main.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_main.h @@ -166,7 +166,7 @@ nfp_pf_map_rtsym(struct nfp_pf *pf, const char *name, const char *sym_fmt, unsigned int min_size, struct nfp_cpp_area **area); int nfp_mbox_cmd(struct nfp_pf *pf, u32 cmd, void *in_data, u64 in_length, void *out_data, u64 out_length); -int nfp_flash_update_common(struct nfp_pf *pf, const char *path, +int nfp_flash_update_common(struct nfp_pf *pf, const struct firmware *fw, struct netlink_ext_ack *extack); enum nfp_dump_diag { diff --git a/drivers/net/ethernet/pensando/ionic/ionic_devlink.c b/drivers/net/ethernet/pensando/ionic/ionic_devlink.c index 51d64718ed9f..b41301a5b0df 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_devlink.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_devlink.c @@ -15,7 +15,7 @@ static int ionic_dl_flash_update(struct devlink *dl, { struct ionic *ionic = devlink_priv(dl); - return ionic_firmware_update(ionic->lif, params->file_name, extack); + return ionic_firmware_update(ionic->lif, params->fw, extack); } static int ionic_dl_info_get(struct devlink *dl, struct devlink_info_req *req, diff --git a/drivers/net/ethernet/pensando/ionic/ionic_devlink.h b/drivers/net/ethernet/pensando/ionic/ionic_devlink.h index 5c01a9e306d8..0a77e8e810c5 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_devlink.h +++ b/drivers/net/ethernet/pensando/ionic/ionic_devlink.h @@ -6,7 +6,7 @@ #include -int ionic_firmware_update(struct ionic_lif *lif, const char *fw_name, +int ionic_firmware_update(struct ionic_lif *lif, const struct firmware *fw, struct netlink_ext_ack *extack); struct ionic *ionic_devlink_alloc(struct device *dev); diff --git a/drivers/net/ethernet/pensando/ionic/ionic_fw.c b/drivers/net/ethernet/pensando/ionic/ionic_fw.c index d7bbf336c6f6..8922c316abe3 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_fw.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_fw.c @@ -91,7 +91,7 @@ static int ionic_fw_status_long_wait(struct ionic *ionic, return err; } -int ionic_firmware_update(struct ionic_lif *lif, const char *fw_name, +int ionic_firmware_update(struct ionic_lif *lif, const struct firmware *fw, struct netlink_ext_ack *extack) { struct ionic_dev *idev = &lif->ionic->idev; @@ -99,24 +99,17 @@ int ionic_firmware_update(struct ionic_lif *lif, const char *fw_name, struct ionic *ionic = lif->ionic; union ionic_dev_cmd_comp comp; u32 buf_sz, copy_sz, offset; - const struct firmware *fw; struct devlink *dl; int next_interval; int err = 0; u8 fw_slot; - netdev_info(netdev, "Installing firmware %s\n", fw_name); + netdev_info(netdev, "Installing firmware\n"); dl = priv_to_devlink(ionic); devlink_flash_update_begin_notify(dl); devlink_flash_update_status_notify(dl, "Preparing to flash", NULL, 0, 0); - err = request_firmware(&fw, fw_name, ionic->dev); - if (err) { - NL_SET_ERR_MSG_MOD(extack, "Unable to find firmware file"); - goto err_out; - } - buf_sz = sizeof(idev->dev_cmd_regs->data); netdev_dbg(netdev, @@ -200,7 +193,6 @@ err_out: devlink_flash_update_status_notify(dl, "Flash failed", NULL, 0, 0); else devlink_flash_update_status_notify(dl, "Flash done", NULL, 0, 0); - release_firmware(fw); devlink_flash_update_end_notify(dl); return err; } diff --git a/include/net/devlink.h b/include/net/devlink.h index b01bb9bca5a2..d1d125a33322 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -19,6 +19,7 @@ #include #include #include +#include #define DEVLINK_RELOAD_STATS_ARRAY_SIZE \ (__DEVLINK_RELOAD_LIMIT_MAX * __DEVLINK_RELOAD_ACTION_MAX) @@ -566,15 +567,15 @@ enum devlink_param_generic_id { /** * struct devlink_flash_update_params - Flash Update parameters - * @file_name: the name of the flash firmware file to update from + * @fw: pointer to the firmware data to update from * @component: the flash component to update * - * With the exception of file_name, drivers must opt-in to parameters by + * With the exception of fw, drivers must opt-in to parameters by * setting the appropriate bit in the supported_flash_update_params field in * their devlink_ops structure. */ struct devlink_flash_update_params { - const char *file_name; + const struct firmware *fw; const char *component; u32 overwrite_mask; }; diff --git a/net/core/devlink.c b/net/core/devlink.c index 4b0211590aac..7c5a5da8e338 100644 --- a/net/core/devlink.c +++ b/net/core/devlink.c @@ -3431,10 +3431,12 @@ EXPORT_SYMBOL_GPL(devlink_flash_update_timeout_notify); static int devlink_nl_cmd_flash_update(struct sk_buff *skb, struct genl_info *info) { - struct nlattr *nla_component, *nla_overwrite_mask; + struct nlattr *nla_component, *nla_overwrite_mask, *nla_file_name; struct devlink_flash_update_params params = {}; struct devlink *devlink = info->user_ptr[0]; + const char *file_name; u32 supported_params; + int ret; if (!devlink->ops->flash_update) return -EOPNOTSUPP; @@ -3444,8 +3446,6 @@ static int devlink_nl_cmd_flash_update(struct sk_buff *skb, supported_params = devlink->ops->supported_flash_update_params; - params.file_name = nla_data(info->attrs[DEVLINK_ATTR_FLASH_UPDATE_FILE_NAME]); - nla_component = info->attrs[DEVLINK_ATTR_FLASH_UPDATE_COMPONENT]; if (nla_component) { if (!(supported_params & DEVLINK_SUPPORT_FLASH_UPDATE_COMPONENT)) { @@ -3469,7 +3469,19 @@ static int devlink_nl_cmd_flash_update(struct sk_buff *skb, params.overwrite_mask = sections.value & sections.selector; } - return devlink->ops->flash_update(devlink, ¶ms, info->extack); + nla_file_name = info->attrs[DEVLINK_ATTR_FLASH_UPDATE_FILE_NAME]; + file_name = nla_data(nla_file_name); + ret = request_firmware(¶ms.fw, file_name, devlink->dev); + if (ret) { + NL_SET_ERR_MSG_ATTR(info->extack, nla_file_name, "failed to locate the requested firmware file"); + return ret; + } + + ret = devlink->ops->flash_update(devlink, ¶ms, info->extack); + + release_firmware(params.fw); + + return ret; } static const struct devlink_param devlink_param_generic[] = { @@ -10227,12 +10239,16 @@ int devlink_compat_flash_update(struct net_device *dev, const char *file_name) goto out; } - params.file_name = file_name; + ret = request_firmware(¶ms.fw, file_name, devlink->dev); + if (ret) + goto out; mutex_lock(&devlink->lock); ret = devlink->ops->flash_update(devlink, ¶ms, NULL); mutex_unlock(&devlink->lock); + release_firmware(params.fw); + out: rtnl_lock(); dev_put(dev); -- cgit v1.2.3 From 52cc5f3a166a33012ebca2cdefebf4c689110068 Mon Sep 17 00:00:00 2001 From: Jacob Keller Date: Wed, 18 Nov 2020 11:06:36 -0800 Subject: devlink: move flash end and begin to core devlink When performing a flash update via devlink, device drivers may inform user space of status updates via devlink_flash_update_(begin|end|timeout|status)_notify functions. It is expected that drivers do not send any status notifications unless they send a begin and end message. If a driver sends a status notification without sending the appropriate end notification upon finishing (regardless of success or failure), the current implementation of the devlink userspace program can get stuck endlessly waiting for the end notification that will never come. The current ice driver implementation may send such a status message without the appropriate end notification in rare cases. Fixing the ice driver is relatively simple: we just need to send the begin_notify at the start of the function and always send an end_notify no matter how the function exits. Rather than assuming driver authors will always get this right in the future, lets just fix the API so that it is not possible to get wrong. Make devlink_flash_update_begin_notify and devlink_flash_update_end_notify static, and call them in devlink.c core code. Always send the begin_notify just before calling the driver's flash_update routine. Always send the end_notify just after the routine returns regardless of success or failure. Doing this makes the status notification easier to use from the driver, as it no longer needs to worry about catching failures and cleaning up by calling devlink_flash_update_end_notify. It is now no longer possible to do the wrong thing in this regard. We also save a couple of lines of code in each driver. Signed-off-by: Jacob Keller Acked-by: Vasundhara Volam Reviewed-by: Jiri Pirko Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c | 2 -- drivers/net/ethernet/intel/ice/ice_devlink.c | 5 +---- drivers/net/ethernet/mellanox/mlxfw/mlxfw_fsm.c | 3 --- drivers/net/ethernet/pensando/ionic/ionic_fw.c | 2 -- drivers/net/netdevsim/dev.c | 2 -- include/net/devlink.h | 2 -- net/core/devlink.c | 10 ++++++---- 7 files changed, 7 insertions(+), 19 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c index 4ebae8a236fd..6b7b69ed62db 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c @@ -30,14 +30,12 @@ bnxt_dl_flash_update(struct devlink *dl, return -EPERM; } - devlink_flash_update_begin_notify(dl); devlink_flash_update_status_notify(dl, "Preparing to flash", NULL, 0, 0); rc = bnxt_flash_package_from_fw_obj(bp->dev, params->fw, 0); if (!rc) devlink_flash_update_status_notify(dl, "Flashing done", NULL, 0, 0); else devlink_flash_update_status_notify(dl, "Flashing failed", NULL, 0, 0); - devlink_flash_update_end_notify(dl); return rc; } diff --git a/drivers/net/ethernet/intel/ice/ice_devlink.c b/drivers/net/ethernet/intel/ice/ice_devlink.c index 0036d3e7df0b..29d6192b15f3 100644 --- a/drivers/net/ethernet/intel/ice/ice_devlink.c +++ b/drivers/net/ethernet/intel/ice/ice_devlink.c @@ -275,12 +275,9 @@ ice_devlink_flash_update(struct devlink *devlink, if (err) return err; - devlink_flash_update_begin_notify(devlink); devlink_flash_update_status_notify(devlink, "Preparing to flash", NULL, 0, 0); - err = ice_flash_pldm_image(pf, params->fw, preservation, extack); - devlink_flash_update_end_notify(devlink); - return err; + return ice_flash_pldm_image(pf, params->fw, preservation, extack); } static const struct devlink_ops ice_devlink_ops = { diff --git a/drivers/net/ethernet/mellanox/mlxfw/mlxfw_fsm.c b/drivers/net/ethernet/mellanox/mlxfw/mlxfw_fsm.c index bcd166911d44..46245e0b2462 100644 --- a/drivers/net/ethernet/mellanox/mlxfw/mlxfw_fsm.c +++ b/drivers/net/ethernet/mellanox/mlxfw/mlxfw_fsm.c @@ -368,7 +368,6 @@ int mlxfw_firmware_flash(struct mlxfw_dev *mlxfw_dev, } mlxfw_info(mlxfw_dev, "Initialize firmware flash process\n"); - devlink_flash_update_begin_notify(mlxfw_dev->devlink); mlxfw_status_notify(mlxfw_dev, "Initializing firmware flash process", NULL, 0, 0); err = mlxfw_dev->ops->fsm_lock(mlxfw_dev, &fwhandle); @@ -417,7 +416,6 @@ int mlxfw_firmware_flash(struct mlxfw_dev *mlxfw_dev, mlxfw_info(mlxfw_dev, "Firmware flash done\n"); mlxfw_status_notify(mlxfw_dev, "Firmware flash done", NULL, 0, 0); mlxfw_mfa2_file_fini(mfa2_file); - devlink_flash_update_end_notify(mlxfw_dev->devlink); return 0; err_state_wait_activate_to_locked: @@ -429,7 +427,6 @@ err_state_wait_idle_to_locked: mlxfw_dev->ops->fsm_release(mlxfw_dev, fwhandle); err_fsm_lock: mlxfw_mfa2_file_fini(mfa2_file); - devlink_flash_update_end_notify(mlxfw_dev->devlink); return err; } EXPORT_SYMBOL(mlxfw_firmware_flash); diff --git a/drivers/net/ethernet/pensando/ionic/ionic_fw.c b/drivers/net/ethernet/pensando/ionic/ionic_fw.c index 8922c316abe3..5f40324cd243 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_fw.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_fw.c @@ -107,7 +107,6 @@ int ionic_firmware_update(struct ionic_lif *lif, const struct firmware *fw, netdev_info(netdev, "Installing firmware\n"); dl = priv_to_devlink(ionic); - devlink_flash_update_begin_notify(dl); devlink_flash_update_status_notify(dl, "Preparing to flash", NULL, 0, 0); buf_sz = sizeof(idev->dev_cmd_regs->data); @@ -193,6 +192,5 @@ err_out: devlink_flash_update_status_notify(dl, "Flash failed", NULL, 0, 0); else devlink_flash_update_status_notify(dl, "Flash done", NULL, 0, 0); - devlink_flash_update_end_notify(dl); return err; } diff --git a/drivers/net/netdevsim/dev.c b/drivers/net/netdevsim/dev.c index fe48817b3985..816af1f55e2c 100644 --- a/drivers/net/netdevsim/dev.c +++ b/drivers/net/netdevsim/dev.c @@ -766,7 +766,6 @@ static int nsim_dev_flash_update(struct devlink *devlink, return -EOPNOTSUPP; if (nsim_dev->fw_update_status) { - devlink_flash_update_begin_notify(devlink); devlink_flash_update_status_notify(devlink, "Preparing to flash", params->component, 0, 0); @@ -790,7 +789,6 @@ static int nsim_dev_flash_update(struct devlink *devlink, params->component, 81); devlink_flash_update_status_notify(devlink, "Flashing done", params->component, 0, 0); - devlink_flash_update_end_notify(devlink); } return 0; diff --git a/include/net/devlink.h b/include/net/devlink.h index d1d125a33322..457c537d0ef2 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -1577,8 +1577,6 @@ void devlink_remote_reload_actions_performed(struct devlink *devlink, enum devlink_reload_limit limit, u32 actions_performed); -void devlink_flash_update_begin_notify(struct devlink *devlink); -void devlink_flash_update_end_notify(struct devlink *devlink); void devlink_flash_update_status_notify(struct devlink *devlink, const char *status_msg, const char *component, diff --git a/net/core/devlink.c b/net/core/devlink.c index 7c5a5da8e338..e6fb1fdedded 100644 --- a/net/core/devlink.c +++ b/net/core/devlink.c @@ -3372,7 +3372,7 @@ out_free_msg: nlmsg_free(msg); } -void devlink_flash_update_begin_notify(struct devlink *devlink) +static void devlink_flash_update_begin_notify(struct devlink *devlink) { struct devlink_flash_notify params = { 0 }; @@ -3380,9 +3380,8 @@ void devlink_flash_update_begin_notify(struct devlink *devlink) DEVLINK_CMD_FLASH_UPDATE, ¶ms); } -EXPORT_SYMBOL_GPL(devlink_flash_update_begin_notify); -void devlink_flash_update_end_notify(struct devlink *devlink) +static void devlink_flash_update_end_notify(struct devlink *devlink) { struct devlink_flash_notify params = { 0 }; @@ -3390,7 +3389,6 @@ void devlink_flash_update_end_notify(struct devlink *devlink) DEVLINK_CMD_FLASH_UPDATE_END, ¶ms); } -EXPORT_SYMBOL_GPL(devlink_flash_update_end_notify); void devlink_flash_update_status_notify(struct devlink *devlink, const char *status_msg, @@ -3477,7 +3475,9 @@ static int devlink_nl_cmd_flash_update(struct sk_buff *skb, return ret; } + devlink_flash_update_begin_notify(devlink); ret = devlink->ops->flash_update(devlink, ¶ms, info->extack); + devlink_flash_update_end_notify(devlink); release_firmware(params.fw); @@ -10244,7 +10244,9 @@ int devlink_compat_flash_update(struct net_device *dev, const char *file_name) goto out; mutex_lock(&devlink->lock); + devlink_flash_update_begin_notify(devlink); ret = devlink->ops->flash_update(devlink, ¶ms, NULL); + devlink_flash_update_end_notify(devlink); mutex_unlock(&devlink->lock); release_firmware(params.fw); -- cgit v1.2.3 From 537cf4e3cc2f6cc9088dcd6162de573f603adc29 Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Fri, 20 Nov 2020 12:53:39 +0100 Subject: xsk: Fix umem cleanup bug at socket destruct Fix a bug that is triggered when a partially setup socket is destroyed. For a fully setup socket, a socket that has been bound to a device, the cleanup of the umem is performed at the end of the buffer pool's cleanup work queue item. This has to be performed in a work queue, and not in RCU cleanup, as it is doing a vunmap that cannot execute in interrupt context. However, when a socket has only been partially set up so that a umem has been created but the buffer pool has not, the code erroneously directly calls the umem cleanup function instead of using a work queue, and this leads to a BUG_ON() in vunmap(). As there in this case is no buffer pool, we cannot use its work queue, so we need to introduce a work queue for the umem and schedule this for the cleanup. So in the case there is no pool, we are going to use the umem's own work queue to schedule the cleanup. But if there is a pool, the cleanup of the umem is still being performed by the pool's work queue, as it is important that the umem is cleaned up after the pool. Fixes: e5e1a4bc916d ("xsk: Fix possible memory leak at socket close") Reported-by: Marek Majtyka Signed-off-by: Magnus Karlsson Signed-off-by: Daniel Borkmann Tested-by: Marek Majtyka Link: https://lore.kernel.org/bpf/1605873219-21629-1-git-send-email-magnus.karlsson@gmail.com --- include/net/xdp_sock.h | 1 + net/xdp/xdp_umem.c | 19 ++++++++++++++++--- net/xdp/xdp_umem.h | 2 +- net/xdp/xsk.c | 2 +- net/xdp/xsk_buff_pool.c | 2 +- 5 files changed, 20 insertions(+), 6 deletions(-) (limited to 'include/net') diff --git a/include/net/xdp_sock.h b/include/net/xdp_sock.h index 1a9559c0cbdd..4f4e93bf814c 100644 --- a/include/net/xdp_sock.h +++ b/include/net/xdp_sock.h @@ -31,6 +31,7 @@ struct xdp_umem { struct page **pgs; int id; struct list_head xsk_dma_list; + struct work_struct work; }; struct xsk_map { diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c index 56d052bc65cb..56a28a686988 100644 --- a/net/xdp/xdp_umem.c +++ b/net/xdp/xdp_umem.c @@ -66,18 +66,31 @@ static void xdp_umem_release(struct xdp_umem *umem) kfree(umem); } +static void xdp_umem_release_deferred(struct work_struct *work) +{ + struct xdp_umem *umem = container_of(work, struct xdp_umem, work); + + xdp_umem_release(umem); +} + void xdp_get_umem(struct xdp_umem *umem) { refcount_inc(&umem->users); } -void xdp_put_umem(struct xdp_umem *umem) +void xdp_put_umem(struct xdp_umem *umem, bool defer_cleanup) { if (!umem) return; - if (refcount_dec_and_test(&umem->users)) - xdp_umem_release(umem); + if (refcount_dec_and_test(&umem->users)) { + if (defer_cleanup) { + INIT_WORK(&umem->work, xdp_umem_release_deferred); + schedule_work(&umem->work); + } else { + xdp_umem_release(umem); + } + } } static int xdp_umem_pin_pages(struct xdp_umem *umem, unsigned long address) diff --git a/net/xdp/xdp_umem.h b/net/xdp/xdp_umem.h index 181fdda2f2a8..aa9fe2780410 100644 --- a/net/xdp/xdp_umem.h +++ b/net/xdp/xdp_umem.h @@ -9,7 +9,7 @@ #include void xdp_get_umem(struct xdp_umem *umem); -void xdp_put_umem(struct xdp_umem *umem); +void xdp_put_umem(struct xdp_umem *umem, bool defer_cleanup); struct xdp_umem *xdp_umem_create(struct xdp_umem_reg *mr); #endif /* XDP_UMEM_H_ */ diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index cfbec3989a76..5a6cdf7b320d 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -1147,7 +1147,7 @@ static void xsk_destruct(struct sock *sk) return; if (!xp_put_pool(xs->pool)) - xdp_put_umem(xs->umem); + xdp_put_umem(xs->umem, !xs->pool); sk_refcnt_debug_dec(sk); } diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 8a3bf4e1318e..3c5a1423d922 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -242,7 +242,7 @@ static void xp_release_deferred(struct work_struct *work) pool->cq = NULL; } - xdp_put_umem(pool->umem); + xdp_put_umem(pool->umem, false); xp_destroy(pool); } -- cgit v1.2.3 From e2ef5203c817a60bfb591343ffd851b6537370ff Mon Sep 17 00:00:00 2001 From: Numan Siddique Date: Mon, 16 Nov 2020 18:31:26 +0530 Subject: net: openvswitch: Be liberal in tcp conntrack. There is no easy way to distinguish if a conntracked tcp packet is marked invalid because of tcp_in_window() check error or because it doesn't belong to an existing connection. With this patch, openvswitch sets liberal tcp flag for the established sessions so that out of window packets are not marked invalid. A helper function - nf_ct_set_tcp_be_liberal(nf_conn) is added which sets this flag for both the directions of the nf_conn. Suggested-by: Florian Westphal Signed-off-by: Numan Siddique Acked-by: Florian Westphal Link: https://lore.kernel.org/r/20201116130126.3065077-1-nusiddiq@redhat.com Signed-off-by: Jakub Kicinski --- include/net/netfilter/nf_conntrack_l4proto.h | 14 ++++++++++++++ net/netfilter/nf_conntrack_proto_tcp.c | 6 ------ net/openvswitch/conntrack.c | 8 ++++++++ 3 files changed, 22 insertions(+), 6 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h index 88186b95b3c2..9be7320b994f 100644 --- a/include/net/netfilter/nf_conntrack_l4proto.h +++ b/include/net/netfilter/nf_conntrack_l4proto.h @@ -203,6 +203,20 @@ static inline struct nf_icmp_net *nf_icmpv6_pernet(struct net *net) { return &net->ct.nf_ct_proto.icmpv6; } + +/* Caller must check nf_ct_protonum(ct) is IPPROTO_TCP before calling. */ +static inline void nf_ct_set_tcp_be_liberal(struct nf_conn *ct) +{ + ct->proto.tcp.seen[0].flags |= IP_CT_TCP_FLAG_BE_LIBERAL; + ct->proto.tcp.seen[1].flags |= IP_CT_TCP_FLAG_BE_LIBERAL; +} + +/* Caller must check nf_ct_protonum(ct) is IPPROTO_TCP before calling. */ +static inline bool nf_conntrack_tcp_established(const struct nf_conn *ct) +{ + return ct->proto.tcp.state == TCP_CONNTRACK_ESTABLISHED && + test_bit(IPS_ASSURED_BIT, &ct->status); +} #endif #ifdef CONFIG_NF_CT_PROTO_DCCP diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index c8fb2187ad4b..811c6c9b59e1 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -834,12 +834,6 @@ static noinline bool tcp_new(struct nf_conn *ct, const struct sk_buff *skb, return true; } -static bool nf_conntrack_tcp_established(const struct nf_conn *ct) -{ - return ct->proto.tcp.state == TCP_CONNTRACK_ESTABLISHED && - test_bit(IPS_ASSURED_BIT, &ct->status); -} - /* Returns verdict for packet, or -1 for invalid. */ int nf_conntrack_tcp_packet(struct nf_conn *ct, struct sk_buff *skb, diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index 4beb96139d77..6a88daab0190 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -1037,6 +1037,14 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, ovs_ct_helper(skb, info->family) != NF_ACCEPT) { return -EINVAL; } + + if (nf_ct_protonum(ct) == IPPROTO_TCP && + nf_ct_is_confirmed(ct) && nf_conntrack_tcp_established(ct)) { + /* Be liberal for tcp packets so that out-of-window + * packets are not marked invalid. + */ + nf_ct_set_tcp_be_liberal(ct); + } } return 0; -- cgit v1.2.3 From 12f4bd86225e348ef3a3c8d2bb42dc23ee0f0a4c Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Tue, 17 Nov 2020 19:43:49 +0100 Subject: net: add annotation for sock_{lock,unlock}_fast The static checker is fooled by the non-static locking scheme implemented by the mentioned helpers. Let's make its life easier adding some unconditional annotation so that the helpers are now interpreted as a plain spinlock from sparse. v1 -> v2: - add __releases() annotation to unlock_sock_fast() Signed-off-by: Paolo Abeni Link: https://lore.kernel.org/r/6ed7ae627d8271fb7f20e0a9c6750fbba1ac2635.1605634911.git.pabeni@redhat.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 10 +++++++--- net/core/sock.c | 3 ++- 2 files changed, 9 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 1d29aeae74fd..093b51719c69 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1595,7 +1595,8 @@ void release_sock(struct sock *sk); SINGLE_DEPTH_NESTING) #define bh_unlock_sock(__sk) spin_unlock(&((__sk)->sk_lock.slock)) -bool lock_sock_fast(struct sock *sk); +bool lock_sock_fast(struct sock *sk) __acquires(&sk->sk_lock.slock); + /** * unlock_sock_fast - complement of lock_sock_fast * @sk: socket @@ -1605,11 +1606,14 @@ bool lock_sock_fast(struct sock *sk); * If slow mode is on, we call regular release_sock() */ static inline void unlock_sock_fast(struct sock *sk, bool slow) + __releases(&sk->sk_lock.slock) { - if (slow) + if (slow) { release_sock(sk); - else + __release(&sk->sk_lock.slock); + } else { spin_unlock_bh(&sk->sk_lock.slock); + } } /* Used by processes to "lock" a socket state, so that diff --git a/net/core/sock.c b/net/core/sock.c index 727ea1cc633c..9badbe7bb4e4 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3078,7 +3078,7 @@ EXPORT_SYMBOL(release_sock); * * sk_lock.slock unlocked, owned = 1, BH enabled */ -bool lock_sock_fast(struct sock *sk) +bool lock_sock_fast(struct sock *sk) __acquires(&sk->sk_lock.slock) { might_sleep(); spin_lock_bh(&sk->sk_lock.slock); @@ -3096,6 +3096,7 @@ bool lock_sock_fast(struct sock *sk) * The sk_lock has mutex_lock() semantics here: */ mutex_acquire(&sk->sk_lock.dep_map, 0, 0, _RET_IP_); + __acquire(&sk->sk_lock.slock); local_bh_enable(); return true; } -- cgit v1.2.3 From fc9840fbef0c3740e0fc3640eb8628d70fcb2215 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 18 Nov 2020 11:44:38 -0800 Subject: net: stream: fix TCP references when INET is not enabled Fix build of net/core/stream.o when CONFIG_INET is not enabled. Fixes these build errors (sample): ld: net/core/stream.o: in function `sk_stream_write_space': (.text+0x27e): undefined reference to `tcp_stream_memory_free' ld: (.text+0x29c): undefined reference to `tcp_stream_memory_free' ld: (.text+0x2ab): undefined reference to `tcp_stream_memory_free' ld: net/core/stream.o: in function `sk_stream_wait_memory': (.text+0x5a1): undefined reference to `tcp_stream_memory_free' ld: (.text+0x5bf): undefined reference to `tcp_stream_memory_free' Fixes: 1c5f2ced136a ("tcp: avoid indirect call to tcp_stream_memory_free()") Signed-off-by: Randy Dunlap Reported-by: Randy Dunlap Link: https://lore.kernel.org/r/20201118194438.674-1-rdunlap@infradead.org Signed-off-by: Jakub Kicinski --- include/net/sock.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 093b51719c69..80469c2c448d 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1271,10 +1271,15 @@ static inline bool __sk_stream_memory_free(const struct sock *sk, int wake) if (READ_ONCE(sk->sk_wmem_queued) >= READ_ONCE(sk->sk_sndbuf)) return false; +#ifdef CONFIG_INET return sk->sk_prot->stream_memory_free ? INDIRECT_CALL_1(sk->sk_prot->stream_memory_free, tcp_stream_memory_free, sk, wake) : true; +#else + return sk->sk_prot->stream_memory_free ? + sk->sk_prot->stream_memory_free(sk, wake) : true; +#endif } static inline bool sk_stream_memory_free(const struct sock *sk) -- cgit v1.2.3 From fa3fe2b150316b294f2c662653501273ff25bba8 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 19 Nov 2020 11:46:02 -0800 Subject: mptcp: track window announced to peer OoO handling attempts to detect when packet is out-of-window by testing current ack sequence and remaining space vs. sequence number. This doesn't work reliably. Store the highest allowed sequence number that we've announced and use it to detect oow packets. Do this when mptcp options get written to the packet (wire format). For this to work we need to move the write_options call until after stack selected a new tcp window. Acked-by: Paolo Abeni Signed-off-by: Florian Westphal Signed-off-by: Mat Martineau Signed-off-by: Jakub Kicinski --- include/net/mptcp.h | 3 ++- net/ipv4/tcp_output.c | 11 +++++++---- net/mptcp/options.c | 22 +++++++++++++++++++++- net/mptcp/protocol.c | 12 +++++++----- net/mptcp/protocol.h | 1 + 5 files changed, 38 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/mptcp.h b/include/net/mptcp.h index 6e706d838e4e..b6cf07143a8a 100644 --- a/include/net/mptcp.h +++ b/include/net/mptcp.h @@ -88,7 +88,8 @@ bool mptcp_established_options(struct sock *sk, struct sk_buff *skb, struct mptcp_out_options *opts); void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb); -void mptcp_write_options(__be32 *ptr, struct mptcp_out_options *opts); +void mptcp_write_options(__be32 *ptr, const struct tcp_sock *tp, + struct mptcp_out_options *opts); /* move the skb extension owership, with the assumption that 'to' is * newly allocated diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 99905bc01d40..41880d3521ed 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -445,11 +445,12 @@ struct tcp_out_options { struct mptcp_out_options mptcp; }; -static void mptcp_options_write(__be32 *ptr, struct tcp_out_options *opts) +static void mptcp_options_write(__be32 *ptr, const struct tcp_sock *tp, + struct tcp_out_options *opts) { #if IS_ENABLED(CONFIG_MPTCP) if (unlikely(OPTION_MPTCP & opts->options)) - mptcp_write_options(ptr, &opts->mptcp); + mptcp_write_options(ptr, tp, &opts->mptcp); #endif } @@ -701,7 +702,7 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, smc_options_write(ptr, &options); - mptcp_options_write(ptr, opts); + mptcp_options_write(ptr, tp, opts); } static void smc_set_option(const struct tcp_sock *tp, @@ -1346,7 +1347,6 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, } } - tcp_options_write((__be32 *)(th + 1), tp, &opts); skb_shinfo(skb)->gso_type = sk->sk_gso_type; if (likely(!(tcb->tcp_flags & TCPHDR_SYN))) { th->window = htons(tcp_select_window(sk)); @@ -1357,6 +1357,9 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, */ th->window = htons(min(tp->rcv_wnd, 65535U)); } + + tcp_options_write((__be32 *)(th + 1), tp, &opts); + #ifdef CONFIG_TCP_MD5SIG /* Calculate the MD5 hash, as we have all we need now */ if (md5) { diff --git a/net/mptcp/options.c b/net/mptcp/options.c index d7afffcc87c1..248e3930c0cb 100644 --- a/net/mptcp/options.c +++ b/net/mptcp/options.c @@ -1010,7 +1010,24 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb) } } -void mptcp_write_options(__be32 *ptr, struct mptcp_out_options *opts) +static void mptcp_set_rwin(const struct tcp_sock *tp) +{ + const struct sock *ssk = (const struct sock *)tp; + const struct mptcp_subflow_context *subflow; + struct mptcp_sock *msk; + u64 ack_seq; + + subflow = mptcp_subflow_ctx(ssk); + msk = mptcp_sk(subflow->conn); + + ack_seq = READ_ONCE(msk->ack_seq) + tp->rcv_wnd; + + if (after64(ack_seq, READ_ONCE(msk->rcv_wnd_sent))) + WRITE_ONCE(msk->rcv_wnd_sent, ack_seq); +} + +void mptcp_write_options(__be32 *ptr, const struct tcp_sock *tp, + struct mptcp_out_options *opts) { if ((OPTION_MPTCP_MPC_SYN | OPTION_MPTCP_MPC_SYNACK | OPTION_MPTCP_MPC_ACK) & opts->suboptions) { @@ -1167,4 +1184,7 @@ mp_capable_done: TCPOPT_NOP << 8 | TCPOPT_NOP, ptr); } } + + if (tp) + mptcp_set_rwin(tp); } diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 8f303de7f500..e1606e072218 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -168,19 +168,19 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk, struct sk_buff *skb) struct rb_node **p, *parent; u64 seq, end_seq, max_seq; struct sk_buff *skb1; - int space; seq = MPTCP_SKB_CB(skb)->map_seq; end_seq = MPTCP_SKB_CB(skb)->end_seq; - space = tcp_space(sk); - max_seq = space > 0 ? space + msk->ack_seq : msk->ack_seq; + max_seq = READ_ONCE(msk->rcv_wnd_sent); pr_debug("msk=%p seq=%llx limit=%llx empty=%d", msk, seq, max_seq, RB_EMPTY_ROOT(&msk->out_of_order_queue)); - if (after64(seq, max_seq)) { + if (after64(end_seq, max_seq)) { /* out of window */ mptcp_drop(sk, skb); - pr_debug("oow by %ld", (unsigned long)seq - (unsigned long)max_seq); + pr_debug("oow by %lld, rcv_wnd_sent %llu\n", + (unsigned long long)end_seq - (unsigned long)max_seq, + (unsigned long long)msk->rcv_wnd_sent); MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_NODSSWINDOW); return; } @@ -2295,6 +2295,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk, mptcp_crypto_key_sha(msk->remote_key, NULL, &ack_seq); ack_seq++; WRITE_ONCE(msk->ack_seq, ack_seq); + WRITE_ONCE(msk->rcv_wnd_sent, ack_seq); } sock_reset_flag(nsk, SOCK_RCU_FREE); @@ -2587,6 +2588,7 @@ void mptcp_finish_connect(struct sock *ssk) WRITE_ONCE(msk->write_seq, subflow->idsn + 1); WRITE_ONCE(msk->snd_nxt, msk->write_seq); WRITE_ONCE(msk->ack_seq, ack_seq); + WRITE_ONCE(msk->rcv_wnd_sent, ack_seq); WRITE_ONCE(msk->can_ack, 1); atomic64_set(&msk->snd_una, msk->write_seq); diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 4d6dd4adaaaf..67f86818203f 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -216,6 +216,7 @@ struct mptcp_sock { u64 write_seq; u64 snd_nxt; u64 ack_seq; + u64 rcv_wnd_sent; u64 rcv_data_fin_seq; struct sock *last_snd; int snd_burst; -- cgit v1.2.3 From b9ad3e9f5a7a760ab068e33e1f18d240ba32ce92 Mon Sep 17 00:00:00 2001 From: Jamie Iles Date: Fri, 20 Nov 2020 14:28:27 +0000 Subject: bonding: wait for sysfs kobject destruction before freeing struct slave syzkaller found that with CONFIG_DEBUG_KOBJECT_RELEASE=y, releasing a struct slave device could result in the following splat: kobject: 'bonding_slave' (00000000cecdd4fe): kobject_release, parent 0000000074ceb2b2 (delayed 1000) bond0 (unregistering): (slave bond_slave_1): Releasing backup interface ------------[ cut here ]------------ ODEBUG: free active (active state 0) object type: timer_list hint: workqueue_select_cpu_near kernel/workqueue.c:1549 [inline] ODEBUG: free active (active state 0) object type: timer_list hint: delayed_work_timer_fn+0x0/0x98 kernel/workqueue.c:1600 WARNING: CPU: 1 PID: 842 at lib/debugobjects.c:485 debug_print_object+0x180/0x240 lib/debugobjects.c:485 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 842 Comm: kworker/u4:4 Tainted: G S 5.9.0-rc8+ #96 Hardware name: linux,dummy-virt (DT) Workqueue: netns cleanup_net Call trace: dump_backtrace+0x0/0x4d8 include/linux/bitmap.h:239 show_stack+0x34/0x48 arch/arm64/kernel/traps.c:142 __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x174/0x1f8 lib/dump_stack.c:118 panic+0x360/0x7a0 kernel/panic.c:231 __warn+0x244/0x2ec kernel/panic.c:600 report_bug+0x240/0x398 lib/bug.c:198 bug_handler+0x50/0xc0 arch/arm64/kernel/traps.c:974 call_break_hook+0x160/0x1d8 arch/arm64/kernel/debug-monitors.c:322 brk_handler+0x30/0xc0 arch/arm64/kernel/debug-monitors.c:329 do_debug_exception+0x184/0x340 arch/arm64/mm/fault.c:864 el1_dbg+0x48/0xb0 arch/arm64/kernel/entry-common.c:65 el1_sync_handler+0x170/0x1c8 arch/arm64/kernel/entry-common.c:93 el1_sync+0x80/0x100 arch/arm64/kernel/entry.S:594 debug_print_object+0x180/0x240 lib/debugobjects.c:485 __debug_check_no_obj_freed lib/debugobjects.c:967 [inline] debug_check_no_obj_freed+0x200/0x430 lib/debugobjects.c:998 slab_free_hook mm/slub.c:1536 [inline] slab_free_freelist_hook+0x190/0x210 mm/slub.c:1577 slab_free mm/slub.c:3138 [inline] kfree+0x13c/0x460 mm/slub.c:4119 bond_free_slave+0x8c/0xf8 drivers/net/bonding/bond_main.c:1492 __bond_release_one+0xe0c/0xec8 drivers/net/bonding/bond_main.c:2190 bond_slave_netdev_event drivers/net/bonding/bond_main.c:3309 [inline] bond_netdev_event+0x8f0/0xa70 drivers/net/bonding/bond_main.c:3420 notifier_call_chain+0xf0/0x200 kernel/notifier.c:83 __raw_notifier_call_chain kernel/notifier.c:361 [inline] raw_notifier_call_chain+0x44/0x58 kernel/notifier.c:368 call_netdevice_notifiers_info+0xbc/0x150 net/core/dev.c:2033 call_netdevice_notifiers_extack net/core/dev.c:2045 [inline] call_netdevice_notifiers net/core/dev.c:2059 [inline] rollback_registered_many+0x6a4/0xec0 net/core/dev.c:9347 unregister_netdevice_many.part.0+0x2c/0x1c0 net/core/dev.c:10509 unregister_netdevice_many net/core/dev.c:10508 [inline] default_device_exit_batch+0x294/0x338 net/core/dev.c:10992 ops_exit_list.isra.0+0xec/0x150 net/core/net_namespace.c:189 cleanup_net+0x44c/0x888 net/core/net_namespace.c:603 process_one_work+0x96c/0x18c0 kernel/workqueue.c:2269 worker_thread+0x3f0/0xc30 kernel/workqueue.c:2415 kthread+0x390/0x498 kernel/kthread.c:292 ret_from_fork+0x10/0x18 arch/arm64/kernel/entry.S:925 This is a potential use-after-free if the sysfs nodes are being accessed whilst removing the struct slave, so wait for the object destruction to complete before freeing the struct slave itself. Fixes: 07699f9a7c8d ("bonding: add sysfs /slave dir for bond slave devices.") Fixes: a068aab42258 ("bonding: Fix reference count leak in bond_sysfs_slave_add.") Cc: Qiushi Wu Cc: Jay Vosburgh Cc: Veaceslav Falico Cc: Andy Gospodarek Signed-off-by: Jamie Iles Reviewed-by: Greg Kroah-Hartman Link: https://lore.kernel.org/r/20201120142827.879226-1-jamie@nuviainc.com Signed-off-by: Jakub Kicinski --- drivers/net/bonding/bond_main.c | 61 ++++++++++++++++++++++++---------- drivers/net/bonding/bond_sysfs_slave.c | 18 +--------- include/net/bonding.h | 8 +++++ 3 files changed, 52 insertions(+), 35 deletions(-) (limited to 'include/net') diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 84ecbc6fa0ff..47afc5938c26 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1460,7 +1460,39 @@ static void bond_upper_dev_unlink(struct bonding *bond, struct slave *slave) slave->dev->flags &= ~IFF_SLAVE; } -static struct slave *bond_alloc_slave(struct bonding *bond) +static void slave_kobj_release(struct kobject *kobj) +{ + struct slave *slave = to_slave(kobj); + struct bonding *bond = bond_get_bond_by_slave(slave); + + cancel_delayed_work_sync(&slave->notify_work); + if (BOND_MODE(bond) == BOND_MODE_8023AD) + kfree(SLAVE_AD_INFO(slave)); + + kfree(slave); +} + +static struct kobj_type slave_ktype = { + .release = slave_kobj_release, +#ifdef CONFIG_SYSFS + .sysfs_ops = &slave_sysfs_ops, +#endif +}; + +static int bond_kobj_init(struct slave *slave) +{ + int err; + + err = kobject_init_and_add(&slave->kobj, &slave_ktype, + &(slave->dev->dev.kobj), "bonding_slave"); + if (err) + kobject_put(&slave->kobj); + + return err; +} + +static struct slave *bond_alloc_slave(struct bonding *bond, + struct net_device *slave_dev) { struct slave *slave = NULL; @@ -1468,11 +1500,17 @@ static struct slave *bond_alloc_slave(struct bonding *bond) if (!slave) return NULL; + slave->bond = bond; + slave->dev = slave_dev; + + if (bond_kobj_init(slave)) + return NULL; + if (BOND_MODE(bond) == BOND_MODE_8023AD) { SLAVE_AD_INFO(slave) = kzalloc(sizeof(struct ad_slave_info), GFP_KERNEL); if (!SLAVE_AD_INFO(slave)) { - kfree(slave); + kobject_put(&slave->kobj); return NULL; } } @@ -1481,17 +1519,6 @@ static struct slave *bond_alloc_slave(struct bonding *bond) return slave; } -static void bond_free_slave(struct slave *slave) -{ - struct bonding *bond = bond_get_bond_by_slave(slave); - - cancel_delayed_work_sync(&slave->notify_work); - if (BOND_MODE(bond) == BOND_MODE_8023AD) - kfree(SLAVE_AD_INFO(slave)); - - kfree(slave); -} - static void bond_fill_ifbond(struct bonding *bond, struct ifbond *info) { info->bond_mode = BOND_MODE(bond); @@ -1678,14 +1705,12 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev, goto err_undo_flags; } - new_slave = bond_alloc_slave(bond); + new_slave = bond_alloc_slave(bond, slave_dev); if (!new_slave) { res = -ENOMEM; goto err_undo_flags; } - new_slave->bond = bond; - new_slave->dev = slave_dev; /* Set the new_slave's queue_id to be zero. Queue ID mapping * is set via sysfs or module option if desired. */ @@ -2007,7 +2032,7 @@ err_restore_mtu: dev_set_mtu(slave_dev, new_slave->original_mtu); err_free: - bond_free_slave(new_slave); + kobject_put(&new_slave->kobj); err_undo_flags: /* Enslave of first slave has failed and we need to fix master's mac */ @@ -2187,7 +2212,7 @@ static int __bond_release_one(struct net_device *bond_dev, if (!netif_is_bond_master(slave_dev)) slave_dev->priv_flags &= ~IFF_BONDING; - bond_free_slave(slave); + kobject_put(&slave->kobj); return 0; } diff --git a/drivers/net/bonding/bond_sysfs_slave.c b/drivers/net/bonding/bond_sysfs_slave.c index 9b8346638f69..fd07561da034 100644 --- a/drivers/net/bonding/bond_sysfs_slave.c +++ b/drivers/net/bonding/bond_sysfs_slave.c @@ -121,7 +121,6 @@ static const struct slave_attribute *slave_attrs[] = { }; #define to_slave_attr(_at) container_of(_at, struct slave_attribute, attr) -#define to_slave(obj) container_of(obj, struct slave, kobj) static ssize_t slave_show(struct kobject *kobj, struct attribute *attr, char *buf) @@ -132,28 +131,15 @@ static ssize_t slave_show(struct kobject *kobj, return slave_attr->show(slave, buf); } -static const struct sysfs_ops slave_sysfs_ops = { +const struct sysfs_ops slave_sysfs_ops = { .show = slave_show, }; -static struct kobj_type slave_ktype = { -#ifdef CONFIG_SYSFS - .sysfs_ops = &slave_sysfs_ops, -#endif -}; - int bond_sysfs_slave_add(struct slave *slave) { const struct slave_attribute **a; int err; - err = kobject_init_and_add(&slave->kobj, &slave_ktype, - &(slave->dev->dev.kobj), "bonding_slave"); - if (err) { - kobject_put(&slave->kobj); - return err; - } - for (a = slave_attrs; *a; ++a) { err = sysfs_create_file(&slave->kobj, &((*a)->attr)); if (err) { @@ -171,6 +157,4 @@ void bond_sysfs_slave_del(struct slave *slave) for (a = slave_attrs; *a; ++a) sysfs_remove_file(&slave->kobj, &((*a)->attr)); - - kobject_put(&slave->kobj); } diff --git a/include/net/bonding.h b/include/net/bonding.h index 7d132cc1e584..d9d0ff3b0ad3 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -185,6 +185,11 @@ struct slave { struct rtnl_link_stats64 slave_stats; }; +static inline struct slave *to_slave(struct kobject *kobj) +{ + return container_of(kobj, struct slave, kobj); +} + struct bond_up_slave { unsigned int count; struct rcu_head rcu; @@ -750,6 +755,9 @@ extern struct bond_parm_tbl ad_select_tbl[]; /* exported from bond_netlink.c */ extern struct rtnl_link_ops bond_link_ops; +/* exported from bond_sysfs_slave.c */ +extern const struct sysfs_ops slave_sysfs_ops; + static inline netdev_tx_t bond_tx_drop(struct net_device *dev, struct sk_buff *skb) { atomic_long_inc(&dev->tx_dropped); -- cgit v1.2.3 From fc0d3b24bdb7a523e973e49648c45d240320ee95 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Sat, 21 Nov 2020 13:48:44 -0800 Subject: compat: always include linux/compat.h from net/compat.h MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We're about to do reshuffling in networking headers and eliminate some implicit includes. This results in: In file included from ../net/ipv4/netfilter/arp_tables.c:26: include/net/compat.h:60:40: error: unknown type name ‘compat_uptr_t’; did you mean ‘compat_ptr_ioctl’? struct sockaddr __user **save_addr, compat_uptr_t *ptr, ^~~~~~~~~~~~~ compat_ptr_ioctl include/net/compat.h:61:4: error: unknown type name ‘compat_size_t’; did you mean ‘compat_sigset_t’? compat_size_t *len); ^~~~~~~~~~~~~ compat_sigset_t Currently net/compat.h depends on linux/compat.h being included first. After the upcoming changes this would break the 32bit build. Reviewed-by: Arnd Bergmann Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20201121214844.1488283-1-kuba@kernel.org Signed-off-by: Jakub Kicinski --- include/net/compat.h | 10 ---------- 1 file changed, 10 deletions(-) (limited to 'include/net') diff --git a/include/net/compat.h b/include/net/compat.h index 745db0d605b6..84805bdc4435 100644 --- a/include/net/compat.h +++ b/include/net/compat.h @@ -5,8 +5,6 @@ struct sock; -#if defined(CONFIG_COMPAT) - #include struct compat_msghdr { @@ -48,14 +46,6 @@ struct compat_rtentry { unsigned short rt_irtt; /* Initial RTT */ }; -#else /* defined(CONFIG_COMPAT) */ -/* - * To avoid compiler warnings: - */ -#define compat_msghdr msghdr -#define compat_mmsghdr mmsghdr -#endif /* defined(CONFIG_COMPAT) */ - int __get_compat_msghdr(struct msghdr *kmsg, struct compat_msghdr __user *umsg, struct sockaddr __user **save_addr, compat_uptr_t *ptr, compat_size_t *len); -- cgit v1.2.3 From 3df98d79215ace13d1e91ddfc5a67a0f5acbd83f Mon Sep 17 00:00:00 2001 From: Paul Moore Date: Sun, 27 Sep 2020 22:38:26 -0400 Subject: lsm,selinux: pass flowi_common instead of flowi to the LSM hooks As pointed out by Herbert in a recent related patch, the LSM hooks do not have the necessary address family information to use the flowi struct safely. As none of the LSMs currently use any of the protocol specific flowi information, replace the flowi pointers with pointers to the address family independent flowi_common struct. Reported-by: Herbert Xu Acked-by: James Morris Signed-off-by: Paul Moore --- .../chelsio/inline_crypto/chtls/chtls_cm.c | 2 +- drivers/net/wireguard/socket.c | 4 ++-- include/linux/lsm_hook_defs.h | 4 ++-- include/linux/lsm_hooks.h | 2 +- include/linux/security.h | 23 +++++++++++++--------- include/net/flow.h | 10 ++++++++++ include/net/route.h | 6 +++--- net/dccp/ipv4.c | 2 +- net/dccp/ipv6.c | 6 +++--- net/ipv4/icmp.c | 4 ++-- net/ipv4/inet_connection_sock.c | 4 ++-- net/ipv4/ip_output.c | 2 +- net/ipv4/ping.c | 2 +- net/ipv4/raw.c | 2 +- net/ipv4/syncookies.c | 2 +- net/ipv4/udp.c | 2 +- net/ipv6/af_inet6.c | 2 +- net/ipv6/datagram.c | 2 +- net/ipv6/icmp.c | 6 +++--- net/ipv6/inet6_connection_sock.c | 4 ++-- net/ipv6/netfilter/nf_reject_ipv6.c | 2 +- net/ipv6/ping.c | 2 +- net/ipv6/raw.c | 2 +- net/ipv6/syncookies.c | 2 +- net/ipv6/tcp_ipv6.c | 4 ++-- net/ipv6/udp.c | 2 +- net/l2tp/l2tp_ip6.c | 2 +- net/netfilter/nf_synproxy_core.c | 2 +- net/xfrm/xfrm_state.c | 6 ++++-- security/security.c | 17 ++++++++-------- security/selinux/hooks.c | 4 ++-- security/selinux/include/xfrm.h | 2 +- security/selinux/xfrm.c | 13 ++++++------ 33 files changed, 85 insertions(+), 66 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c b/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c index ec4f79049a06..42e4e43cdd91 100644 --- a/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c +++ b/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c @@ -1148,7 +1148,7 @@ static struct sock *chtls_recv_sock(struct sock *lsk, fl6.daddr = ip6h->saddr; fl6.fl6_dport = inet_rsk(oreq)->ir_rmt_port; fl6.fl6_sport = htons(inet_rsk(oreq)->ir_num); - security_req_classify_flow(oreq, flowi6_to_flowi(&fl6)); + security_req_classify_flow(oreq, flowi6_to_flowi_common(&fl6)); dst = ip6_dst_lookup_flow(sock_net(lsk), lsk, &fl6, NULL); if (IS_ERR(dst)) goto free_sk; diff --git a/drivers/net/wireguard/socket.c b/drivers/net/wireguard/socket.c index c33e2c81635f..410b318e57fb 100644 --- a/drivers/net/wireguard/socket.c +++ b/drivers/net/wireguard/socket.c @@ -49,7 +49,7 @@ static int send4(struct wg_device *wg, struct sk_buff *skb, rt = dst_cache_get_ip4(cache, &fl.saddr); if (!rt) { - security_sk_classify_flow(sock, flowi4_to_flowi(&fl)); + security_sk_classify_flow(sock, flowi4_to_flowi_common(&fl)); if (unlikely(!inet_confirm_addr(sock_net(sock), NULL, 0, fl.saddr, RT_SCOPE_HOST))) { endpoint->src4.s_addr = 0; @@ -129,7 +129,7 @@ static int send6(struct wg_device *wg, struct sk_buff *skb, dst = dst_cache_get_ip6(cache, &fl.saddr); if (!dst) { - security_sk_classify_flow(sock, flowi6_to_flowi(&fl)); + security_sk_classify_flow(sock, flowi6_to_flowi_common(&fl)); if (unlikely(!ipv6_addr_any(&fl.saddr) && !ipv6_chk_addr(sock_net(sock), &fl.saddr, NULL, 0))) { endpoint->src6 = fl.saddr = in6addr_any; diff --git a/include/linux/lsm_hook_defs.h b/include/linux/lsm_hook_defs.h index 32a940117e7a..f70984c8318b 100644 --- a/include/linux/lsm_hook_defs.h +++ b/include/linux/lsm_hook_defs.h @@ -311,7 +311,7 @@ LSM_HOOK(int, 0, secmark_relabel_packet, u32 secid) LSM_HOOK(void, LSM_RET_VOID, secmark_refcount_inc, void) LSM_HOOK(void, LSM_RET_VOID, secmark_refcount_dec, void) LSM_HOOK(void, LSM_RET_VOID, req_classify_flow, const struct request_sock *req, - struct flowi *fl) + struct flowi_common *flic) LSM_HOOK(int, 0, tun_dev_alloc_security, void **security) LSM_HOOK(void, LSM_RET_VOID, tun_dev_free_security, void *security) LSM_HOOK(int, 0, tun_dev_create, void) @@ -351,7 +351,7 @@ LSM_HOOK(int, 0, xfrm_state_delete_security, struct xfrm_state *x) LSM_HOOK(int, 0, xfrm_policy_lookup, struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir) LSM_HOOK(int, 1, xfrm_state_pol_flow_match, struct xfrm_state *x, - struct xfrm_policy *xp, const struct flowi *fl) + struct xfrm_policy *xp, const struct flowi_common *flic) LSM_HOOK(int, 0, xfrm_decode_session, struct sk_buff *skb, u32 *secid, int ckall) #endif /* CONFIG_SECURITY_NETWORK_XFRM */ diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h index c503f7ab8afb..a19adef1f088 100644 --- a/include/linux/lsm_hooks.h +++ b/include/linux/lsm_hooks.h @@ -1105,7 +1105,7 @@ * @xfrm_state_pol_flow_match: * @x contains the state to match. * @xp contains the policy to check for a match. - * @fl contains the flow to check for a match. + * @flic contains the flowi_common struct to check for a match. * Return 1 if there is a match. * @xfrm_decode_session: * @skb points to skb to decode. diff --git a/include/linux/security.h b/include/linux/security.h index bc2725491560..469273e8b46d 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -167,7 +167,7 @@ struct sk_buff; struct sock; struct sockaddr; struct socket; -struct flowi; +struct flowi_common; struct dst_entry; struct xfrm_selector; struct xfrm_policy; @@ -1355,8 +1355,9 @@ int security_socket_getpeersec_dgram(struct socket *sock, struct sk_buff *skb, u int security_sk_alloc(struct sock *sk, int family, gfp_t priority); void security_sk_free(struct sock *sk); void security_sk_clone(const struct sock *sk, struct sock *newsk); -void security_sk_classify_flow(struct sock *sk, struct flowi *fl); -void security_req_classify_flow(const struct request_sock *req, struct flowi *fl); +void security_sk_classify_flow(struct sock *sk, struct flowi_common *flic); +void security_req_classify_flow(const struct request_sock *req, + struct flowi_common *flic); void security_sock_graft(struct sock*sk, struct socket *parent); int security_inet_conn_request(struct sock *sk, struct sk_buff *skb, struct request_sock *req); @@ -1507,11 +1508,13 @@ static inline void security_sk_clone(const struct sock *sk, struct sock *newsk) { } -static inline void security_sk_classify_flow(struct sock *sk, struct flowi *fl) +static inline void security_sk_classify_flow(struct sock *sk, + struct flowi_common *flic) { } -static inline void security_req_classify_flow(const struct request_sock *req, struct flowi *fl) +static inline void security_req_classify_flow(const struct request_sock *req, + struct flowi_common *flic) { } @@ -1638,9 +1641,9 @@ void security_xfrm_state_free(struct xfrm_state *x); int security_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir); int security_xfrm_state_pol_flow_match(struct xfrm_state *x, struct xfrm_policy *xp, - const struct flowi *fl); + const struct flowi_common *flic); int security_xfrm_decode_session(struct sk_buff *skb, u32 *secid); -void security_skb_classify_flow(struct sk_buff *skb, struct flowi *fl); +void security_skb_classify_flow(struct sk_buff *skb, struct flowi_common *flic); #else /* CONFIG_SECURITY_NETWORK_XFRM */ @@ -1692,7 +1695,8 @@ static inline int security_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_s } static inline int security_xfrm_state_pol_flow_match(struct xfrm_state *x, - struct xfrm_policy *xp, const struct flowi *fl) + struct xfrm_policy *xp, + const struct flowi_common *flic) { return 1; } @@ -1702,7 +1706,8 @@ static inline int security_xfrm_decode_session(struct sk_buff *skb, u32 *secid) return 0; } -static inline void security_skb_classify_flow(struct sk_buff *skb, struct flowi *fl) +static inline void security_skb_classify_flow(struct sk_buff *skb, + struct flowi_common *flic) { } diff --git a/include/net/flow.h b/include/net/flow.h index b2531df3f65f..39d0cedcddee 100644 --- a/include/net/flow.h +++ b/include/net/flow.h @@ -195,11 +195,21 @@ static inline struct flowi *flowi4_to_flowi(struct flowi4 *fl4) return container_of(fl4, struct flowi, u.ip4); } +static inline struct flowi_common *flowi4_to_flowi_common(struct flowi4 *fl4) +{ + return &(flowi4_to_flowi(fl4)->u.__fl_common); +} + static inline struct flowi *flowi6_to_flowi(struct flowi6 *fl6) { return container_of(fl6, struct flowi, u.ip6); } +static inline struct flowi_common *flowi6_to_flowi_common(struct flowi6 *fl6) +{ + return &(flowi6_to_flowi(fl6)->u.__fl_common); +} + static inline struct flowi *flowidn_to_flowi(struct flowidn *fldn) { return container_of(fldn, struct flowi, u.dn); diff --git a/include/net/route.h b/include/net/route.h index ff021cab657e..2e6c0e153e3a 100644 --- a/include/net/route.h +++ b/include/net/route.h @@ -165,7 +165,7 @@ static inline struct rtable *ip_route_output_ports(struct net *net, struct flowi sk ? inet_sk_flowi_flags(sk) : 0, daddr, saddr, dport, sport, sock_net_uid(net, sk)); if (sk) - security_sk_classify_flow(sk, flowi4_to_flowi(fl4)); + security_sk_classify_flow(sk, flowi4_to_flowi_common(fl4)); return ip_route_output_flow(net, fl4, sk); } @@ -322,7 +322,7 @@ static inline struct rtable *ip_route_connect(struct flowi4 *fl4, ip_rt_put(rt); flowi4_update_output(fl4, oif, tos, fl4->daddr, fl4->saddr); } - security_sk_classify_flow(sk, flowi4_to_flowi(fl4)); + security_sk_classify_flow(sk, flowi4_to_flowi_common(fl4)); return ip_route_output_flow(net, fl4, sk); } @@ -338,7 +338,7 @@ static inline struct rtable *ip_route_newports(struct flowi4 *fl4, struct rtable flowi4_update_output(fl4, sk->sk_bound_dev_if, RT_CONN_FLAGS(sk), fl4->daddr, fl4->saddr); - security_sk_classify_flow(sk, flowi4_to_flowi(fl4)); + security_sk_classify_flow(sk, flowi4_to_flowi_common(fl4)); return ip_route_output_flow(sock_net(sk), fl4, sk); } return rt; diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index bb3d70664dde..3d0b58d7bf3f 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -464,7 +464,7 @@ static struct dst_entry* dccp_v4_route_skb(struct net *net, struct sock *sk, .fl4_dport = dccp_hdr(skb)->dccph_sport, }; - security_skb_classify_flow(skb, flowi4_to_flowi(&fl4)); + security_skb_classify_flow(skb, flowi4_to_flowi_common(&fl4)); rt = ip_route_output_flow(net, &fl4, sk); if (IS_ERR(rt)) { IP_INC_STATS(net, IPSTATS_MIB_OUTNOROUTES); diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index ef4ab28cfde0..179dcf02a572 100644 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@ -203,7 +203,7 @@ static int dccp_v6_send_response(const struct sock *sk, struct request_sock *req fl6.flowi6_oif = ireq->ir_iif; fl6.fl6_dport = ireq->ir_rmt_port; fl6.fl6_sport = htons(ireq->ir_num); - security_req_classify_flow(req, flowi6_to_flowi(&fl6)); + security_req_classify_flow(req, flowi6_to_flowi_common(&fl6)); rcu_read_lock(); @@ -279,7 +279,7 @@ static void dccp_v6_ctl_send_reset(const struct sock *sk, struct sk_buff *rxskb) fl6.flowi6_oif = inet6_iif(rxskb); fl6.fl6_dport = dccp_hdr(skb)->dccph_dport; fl6.fl6_sport = dccp_hdr(skb)->dccph_sport; - security_skb_classify_flow(rxskb, flowi6_to_flowi(&fl6)); + security_skb_classify_flow(rxskb, flowi6_to_flowi_common(&fl6)); /* sk = NULL, but it is safe for now. RST socket required. */ dst = ip6_dst_lookup_flow(sock_net(ctl_sk), ctl_sk, &fl6, NULL); @@ -907,7 +907,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr, fl6.flowi6_oif = sk->sk_bound_dev_if; fl6.fl6_dport = usin->sin6_port; fl6.fl6_sport = inet->inet_sport; - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); opt = rcu_dereference_protected(np->opt, lockdep_sock_is_held(sk)); final_p = fl6_update_dst(&fl6, opt, &final); diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 005faea415a4..396b492c804f 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -447,7 +447,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb) fl4.flowi4_tos = RT_TOS(ip_hdr(skb)->tos); fl4.flowi4_proto = IPPROTO_ICMP; fl4.flowi4_oif = l3mdev_master_ifindex(skb->dev); - security_skb_classify_flow(skb, flowi4_to_flowi(&fl4)); + security_skb_classify_flow(skb, flowi4_to_flowi_common(&fl4)); rt = ip_route_output_key(net, &fl4); if (IS_ERR(rt)) goto out_unlock; @@ -503,7 +503,7 @@ static struct rtable *icmp_route_lookup(struct net *net, route_lookup_dev = icmp_get_route_lookup_dev(skb_in); fl4->flowi4_oif = l3mdev_master_ifindex(route_lookup_dev); - security_skb_classify_flow(skb_in, flowi4_to_flowi(fl4)); + security_skb_classify_flow(skb_in, flowi4_to_flowi_common(fl4)); rt = ip_route_output_key_hash(net, fl4, skb_in); if (IS_ERR(rt)) return rt; diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 4148f5f78f31..1b7801371f7d 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -602,7 +602,7 @@ struct dst_entry *inet_csk_route_req(const struct sock *sk, (opt && opt->opt.srr) ? opt->opt.faddr : ireq->ir_rmt_addr, ireq->ir_loc_addr, ireq->ir_rmt_port, htons(ireq->ir_num), sk->sk_uid); - security_req_classify_flow(req, flowi4_to_flowi(fl4)); + security_req_classify_flow(req, flowi4_to_flowi_common(fl4)); rt = ip_route_output_flow(net, fl4, sk); if (IS_ERR(rt)) goto no_route; @@ -640,7 +640,7 @@ struct dst_entry *inet_csk_route_child_sock(const struct sock *sk, (opt && opt->opt.srr) ? opt->opt.faddr : ireq->ir_rmt_addr, ireq->ir_loc_addr, ireq->ir_rmt_port, htons(ireq->ir_num), sk->sk_uid); - security_req_classify_flow(req, flowi4_to_flowi(fl4)); + security_req_classify_flow(req, flowi4_to_flowi_common(fl4)); rt = ip_route_output_flow(net, fl4, sk); if (IS_ERR(rt)) goto no_route; diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 879b76ae4435..89fff5f59eea 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -1700,7 +1700,7 @@ void ip_send_unicast_reply(struct sock *sk, struct sk_buff *skb, daddr, saddr, tcp_hdr(skb)->source, tcp_hdr(skb)->dest, arg->uid); - security_skb_classify_flow(skb, flowi4_to_flowi(&fl4)); + security_skb_classify_flow(skb, flowi4_to_flowi_common(&fl4)); rt = ip_route_output_key(net, &fl4); if (IS_ERR(rt)) return; diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index 248856b301c4..8b943f85fff9 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -778,7 +778,7 @@ static int ping_v4_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) fl4.fl4_icmp_type = user_icmph.type; fl4.fl4_icmp_code = user_icmph.code; - security_sk_classify_flow(sk, flowi4_to_flowi(&fl4)); + security_sk_classify_flow(sk, flowi4_to_flowi_common(&fl4)); rt = ip_route_output_flow(net, &fl4, sk); if (IS_ERR(rt)) { err = PTR_ERR(rt); diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 7d26e0f8bdae..50a73178d63a 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -640,7 +640,7 @@ static int raw_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) goto done; } - security_sk_classify_flow(sk, flowi4_to_flowi(&fl4)); + security_sk_classify_flow(sk, flowi4_to_flowi_common(&fl4)); rt = ip_route_output_flow(net, &fl4, sk); if (IS_ERR(rt)) { err = PTR_ERR(rt); diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c index 6ac473b47f30..36b3b9b72b75 100644 --- a/net/ipv4/syncookies.c +++ b/net/ipv4/syncookies.c @@ -418,7 +418,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb) inet_sk_flowi_flags(sk), opt->srr ? opt->faddr : ireq->ir_rmt_addr, ireq->ir_loc_addr, th->source, th->dest, sk->sk_uid); - security_req_classify_flow(req, flowi4_to_flowi(&fl4)); + security_req_classify_flow(req, flowi4_to_flowi_common(&fl4)); rt = ip_route_output_key(sock_net(sk), &fl4); if (IS_ERR(rt)) { reqsk_free(req); diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 09f0a23d1a01..594632ff094f 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -1197,7 +1197,7 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) faddr, saddr, dport, inet->inet_sport, sk->sk_uid); - security_sk_classify_flow(sk, flowi4_to_flowi(fl4)); + security_sk_classify_flow(sk, flowi4_to_flowi_common(fl4)); rt = ip_route_output_flow(net, fl4, sk); if (IS_ERR(rt)) { err = PTR_ERR(rt); diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index e648fbebb167..b8e491d3acf0 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -819,7 +819,7 @@ int inet6_sk_rebuild_header(struct sock *sk) fl6.fl6_dport = inet->inet_dport; fl6.fl6_sport = inet->inet_sport; fl6.flowi6_uid = sk->sk_uid; - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); rcu_read_lock(); final_p = fl6_update_dst(&fl6, rcu_dereference(np->opt), diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index cc8ad7ddecda..206f66310a88 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -60,7 +60,7 @@ static void ip6_datagram_flow_key_init(struct flowi6 *fl6, struct sock *sk) if (!fl6->flowi6_oif && ipv6_addr_is_multicast(&fl6->daddr)) fl6->flowi6_oif = np->mcast_oif; - security_sk_classify_flow(sk, flowi6_to_flowi(fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(fl6)); } int ip6_datagram_dst_update(struct sock *sk, bool fix_sk_saddr) diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c index ec448b71bf9a..91d9136b6e9c 100644 --- a/net/ipv6/icmp.c +++ b/net/ipv6/icmp.c @@ -567,7 +567,7 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, fl6.fl6_icmp_code = code; fl6.flowi6_uid = sock_net_uid(net, NULL); fl6.mp_hash = rt6_multipath_hash(net, &fl6, skb, NULL); - security_skb_classify_flow(skb, flowi6_to_flowi(&fl6)); + security_skb_classify_flow(skb, flowi6_to_flowi_common(&fl6)); np = inet6_sk(sk); @@ -749,7 +749,7 @@ static void icmpv6_echo_reply(struct sk_buff *skb) fl6.fl6_icmp_type = ICMPV6_ECHO_REPLY; fl6.flowi6_mark = mark; fl6.flowi6_uid = sock_net_uid(net, NULL); - security_skb_classify_flow(skb, flowi6_to_flowi(&fl6)); + security_skb_classify_flow(skb, flowi6_to_flowi_common(&fl6)); local_bh_disable(); sk = icmpv6_xmit_lock(net); @@ -1002,7 +1002,7 @@ void icmpv6_flow_init(struct sock *sk, struct flowi6 *fl6, fl6->fl6_icmp_type = type; fl6->fl6_icmp_code = 0; fl6->flowi6_oif = oif; - security_sk_classify_flow(sk, flowi6_to_flowi(fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(fl6)); } static void __net_exit icmpv6_sk_exit(struct net *net) diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c index e315526fa244..5a9f4d722f35 100644 --- a/net/ipv6/inet6_connection_sock.c +++ b/net/ipv6/inet6_connection_sock.c @@ -46,7 +46,7 @@ struct dst_entry *inet6_csk_route_req(const struct sock *sk, fl6->fl6_dport = ireq->ir_rmt_port; fl6->fl6_sport = htons(ireq->ir_num); fl6->flowi6_uid = sk->sk_uid; - security_req_classify_flow(req, flowi6_to_flowi(fl6)); + security_req_classify_flow(req, flowi6_to_flowi_common(fl6)); dst = ip6_dst_lookup_flow(sock_net(sk), sk, fl6, final_p); if (IS_ERR(dst)) @@ -95,7 +95,7 @@ static struct dst_entry *inet6_csk_route_socket(struct sock *sk, fl6->fl6_sport = inet->inet_sport; fl6->fl6_dport = inet->inet_dport; fl6->flowi6_uid = sk->sk_uid; - security_sk_classify_flow(sk, flowi6_to_flowi(fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(fl6)); rcu_read_lock(); final_p = fl6_update_dst(fl6, rcu_dereference(np->opt), &final); diff --git a/net/ipv6/netfilter/nf_reject_ipv6.c b/net/ipv6/netfilter/nf_reject_ipv6.c index 4aef6baaa55e..bf95513736c9 100644 --- a/net/ipv6/netfilter/nf_reject_ipv6.c +++ b/net/ipv6/netfilter/nf_reject_ipv6.c @@ -179,7 +179,7 @@ void nf_send_reset6(struct net *net, struct sk_buff *oldskb, int hook) fl6.flowi6_oif = l3mdev_master_ifindex(skb_dst(oldskb)->dev); fl6.flowi6_mark = IP6_REPLY_MARK(net, oldskb->mark); - security_skb_classify_flow(oldskb, flowi6_to_flowi(&fl6)); + security_skb_classify_flow(oldskb, flowi6_to_flowi_common(&fl6)); dst = ip6_route_output(net, NULL, &fl6); if (dst->error) { dst_release(dst); diff --git a/net/ipv6/ping.c b/net/ipv6/ping.c index 6caa062f68e7..6ac88fe24a8e 100644 --- a/net/ipv6/ping.c +++ b/net/ipv6/ping.c @@ -111,7 +111,7 @@ static int ping_v6_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) fl6.flowi6_uid = sk->sk_uid; fl6.fl6_icmp_type = user_icmph.icmp6_type; fl6.fl6_icmp_code = user_icmph.icmp6_code; - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); ipcm6_init_sk(&ipc6, np); ipc6.sockc.mark = sk->sk_mark; diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 6e4ab80a3b94..1f56d9aae589 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -915,7 +915,7 @@ static int rawv6_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) fl6.flowi6_oif = np->mcast_oif; else if (!fl6.flowi6_oif) fl6.flowi6_oif = np->ucast_oif; - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); if (hdrincl) fl6.flowi6_flags |= FLOWI_FLAG_KNOWN_NH; diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c index e796a64be308..0351e2ffc707 100644 --- a/net/ipv6/syncookies.c +++ b/net/ipv6/syncookies.c @@ -233,7 +233,7 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) fl6.fl6_dport = ireq->ir_rmt_port; fl6.fl6_sport = inet_sk(sk)->inet_sport; fl6.flowi6_uid = sk->sk_uid; - security_req_classify_flow(req, flowi6_to_flowi(&fl6)); + security_req_classify_flow(req, flowi6_to_flowi_common(&fl6)); dst = ip6_dst_lookup_flow(sock_net(sk), sk, &fl6, final_p); if (IS_ERR(dst)) diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 8db59f4e5f13..e5f75bbcc66f 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -278,7 +278,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, opt = rcu_dereference_protected(np->opt, lockdep_sock_is_held(sk)); final_p = fl6_update_dst(&fl6, opt, &final); - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); dst = ip6_dst_lookup_flow(sock_net(sk), sk, &fl6, final_p); if (IS_ERR(dst)) { @@ -954,7 +954,7 @@ static void tcp_v6_send_response(const struct sock *sk, struct sk_buff *skb, u32 fl6.fl6_dport = t1->dest; fl6.fl6_sport = t1->source; fl6.flowi6_uid = sock_net_uid(net, sk && sk_fullsock(sk) ? sk : NULL); - security_skb_classify_flow(skb, flowi6_to_flowi(&fl6)); + security_skb_classify_flow(skb, flowi6_to_flowi_common(&fl6)); /* Pass a socket to ip6_dst_lookup either it is for RST * Underlying function will use this to retrieve the network diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 29d9691359b9..724942afc035 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -1496,7 +1496,7 @@ do_udp_sendmsg: } else if (!fl6.flowi6_oif) fl6.flowi6_oif = np->ucast_oif; - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); if (ipc6.tclass < 0) ipc6.tclass = np->tclass; diff --git a/net/l2tp/l2tp_ip6.c b/net/l2tp/l2tp_ip6.c index e5e5036257b0..96f975777438 100644 --- a/net/l2tp/l2tp_ip6.c +++ b/net/l2tp/l2tp_ip6.c @@ -606,7 +606,7 @@ static int l2tp_ip6_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) else if (!fl6.flowi6_oif) fl6.flowi6_oif = np->ucast_oif; - security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); + security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); if (ipc6.tclass < 0) ipc6.tclass = np->tclass; diff --git a/net/netfilter/nf_synproxy_core.c b/net/netfilter/nf_synproxy_core.c index 9cca35d22927..fb64e8515f7d 100644 --- a/net/netfilter/nf_synproxy_core.c +++ b/net/netfilter/nf_synproxy_core.c @@ -849,7 +849,7 @@ synproxy_send_tcp_ipv6(struct net *net, fl6.fl6_sport = nth->source; fl6.fl6_dport = nth->dest; security_skb_classify_flow((struct sk_buff *)skb, - flowi6_to_flowi(&fl6)); + flowi6_to_flowi_common(&fl6)); err = nf_ip6_route(net, &dst, flowi6_to_flowi(&fl6), false); if (err) { goto free_nskb; diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index bbd4643d7e82..ac25b0c2d87e 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -1021,7 +1021,8 @@ static void xfrm_state_look_at(struct xfrm_policy *pol, struct xfrm_state *x, if ((x->sel.family && (x->sel.family != family || !xfrm_selector_match(&x->sel, fl, family))) || - !security_xfrm_state_pol_flow_match(x, pol, fl)) + !security_xfrm_state_pol_flow_match(x, pol, + &fl->u.__fl_common)) return; if (!*best || @@ -1036,7 +1037,8 @@ static void xfrm_state_look_at(struct xfrm_policy *pol, struct xfrm_state *x, if ((!x->sel.family || (x->sel.family == family && xfrm_selector_match(&x->sel, fl, family))) && - security_xfrm_state_pol_flow_match(x, pol, fl)) + security_xfrm_state_pol_flow_match(x, pol, + &fl->u.__fl_common)) *error = -ESRCH; } } diff --git a/security/security.c b/security/security.c index a28045dc9e7f..c08e0eec8b9e 100644 --- a/security/security.c +++ b/security/security.c @@ -2207,15 +2207,16 @@ void security_sk_clone(const struct sock *sk, struct sock *newsk) } EXPORT_SYMBOL(security_sk_clone); -void security_sk_classify_flow(struct sock *sk, struct flowi *fl) +void security_sk_classify_flow(struct sock *sk, struct flowi_common *flic) { - call_void_hook(sk_getsecid, sk, &fl->flowi_secid); + call_void_hook(sk_getsecid, sk, &flic->flowic_secid); } EXPORT_SYMBOL(security_sk_classify_flow); -void security_req_classify_flow(const struct request_sock *req, struct flowi *fl) +void security_req_classify_flow(const struct request_sock *req, + struct flowi_common *flic) { - call_void_hook(req_classify_flow, req, fl); + call_void_hook(req_classify_flow, req, flic); } EXPORT_SYMBOL(security_req_classify_flow); @@ -2407,7 +2408,7 @@ int security_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir) int security_xfrm_state_pol_flow_match(struct xfrm_state *x, struct xfrm_policy *xp, - const struct flowi *fl) + const struct flowi_common *flic) { struct security_hook_list *hp; int rc = LSM_RET_DEFAULT(xfrm_state_pol_flow_match); @@ -2423,7 +2424,7 @@ int security_xfrm_state_pol_flow_match(struct xfrm_state *x, */ hlist_for_each_entry(hp, &security_hook_heads.xfrm_state_pol_flow_match, list) { - rc = hp->hook.xfrm_state_pol_flow_match(x, xp, fl); + rc = hp->hook.xfrm_state_pol_flow_match(x, xp, flic); break; } return rc; @@ -2434,9 +2435,9 @@ int security_xfrm_decode_session(struct sk_buff *skb, u32 *secid) return call_int_hook(xfrm_decode_session, 0, skb, secid, 1); } -void security_skb_classify_flow(struct sk_buff *skb, struct flowi *fl) +void security_skb_classify_flow(struct sk_buff *skb, struct flowi_common *flic) { - int rc = call_int_hook(xfrm_decode_session, 0, skb, &fl->flowi_secid, + int rc = call_int_hook(xfrm_decode_session, 0, skb, &flic->flowic_secid, 0); BUG_ON(rc); diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index 943c2693cec7..a515abdf115b 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -5437,9 +5437,9 @@ static void selinux_secmark_refcount_dec(void) } static void selinux_req_classify_flow(const struct request_sock *req, - struct flowi *fl) + struct flowi_common *flic) { - fl->flowi_secid = req->secid; + flic->flowic_secid = req->secid; } static int selinux_tun_dev_alloc_security(void **security) diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h index a0b465316292..0a6f34a7a971 100644 --- a/security/selinux/include/xfrm.h +++ b/security/selinux/include/xfrm.h @@ -26,7 +26,7 @@ int selinux_xfrm_state_delete(struct xfrm_state *x); int selinux_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir); int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x, struct xfrm_policy *xp, - const struct flowi *fl); + const struct flowi_common *flic); #ifdef CONFIG_SECURITY_NETWORK_XFRM extern atomic_t selinux_xfrm_refcount; diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c index 7314196185d1..c367d36965d4 100644 --- a/security/selinux/xfrm.c +++ b/security/selinux/xfrm.c @@ -175,9 +175,10 @@ int selinux_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir) */ int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x, struct xfrm_policy *xp, - const struct flowi *fl) + const struct flowi_common *flic) { u32 state_sid; + u32 flic_sid; if (!xp->security) if (x->security) @@ -196,17 +197,17 @@ int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x, return 0; state_sid = x->security->ctx_sid; + flic_sid = flic->flowic_secid; - if (fl->flowi_secid != state_sid) + if (flic_sid != state_sid) return 0; /* We don't need a separate SA Vs. policy polmatch check since the SA * is now of the same label as the flow and a flow Vs. policy polmatch * check had already happened in selinux_xfrm_policy_lookup() above. */ - return (avc_has_perm(&selinux_state, - fl->flowi_secid, state_sid, - SECCLASS_ASSOCIATION, ASSOCIATION__SENDTO, - NULL) ? 0 : 1); + return (avc_has_perm(&selinux_state, flic_sid, state_sid, + SECCLASS_ASSOCIATION, ASSOCIATION__SENDTO, + NULL) ? 0 : 1); } static u32 selinux_xfrm_skb_sid_egress(struct sk_buff *skb) -- cgit v1.2.3 From 01770a166165738a6e05c3d911fb4609cc4eb416 Mon Sep 17 00:00:00 2001 From: Ricardo Dias Date: Fri, 20 Nov 2020 11:11:33 +0000 Subject: tcp: fix race condition when creating child sockets from syncookies When the TCP stack is in SYN flood mode, the server child socket is created from the SYN cookie received in a TCP packet with the ACK flag set. The child socket is created when the server receives the first TCP packet with a valid SYN cookie from the client. Usually, this packet corresponds to the final step of the TCP 3-way handshake, the ACK packet. But is also possible to receive a valid SYN cookie from the first TCP data packet sent by the client, and thus create a child socket from that SYN cookie. Since a client socket is ready to send data as soon as it receives the SYN+ACK packet from the server, the client can send the ACK packet (sent by the TCP stack code), and the first data packet (sent by the userspace program) almost at the same time, and thus the server will equally receive the two TCP packets with valid SYN cookies almost at the same instant. When such event happens, the TCP stack code has a race condition that occurs between the momement a lookup is done to the established connections hashtable to check for the existence of a connection for the same client, and the moment that the child socket is added to the established connections hashtable. As a consequence, this race condition can lead to a situation where we add two child sockets to the established connections hashtable and deliver two sockets to the userspace program to the same client. This patch fixes the race condition by checking if an existing child socket exists for the same client when we are adding the second child socket to the established connections socket. If an existing child socket exists, we drop the packet and discard the second child socket to the same client. Signed-off-by: Ricardo Dias Signed-off-by: Eric Dumazet Link: https://lore.kernel.org/r/20201120111133.GA67501@rdias-suse-pc.lan Signed-off-by: Jakub Kicinski --- include/net/inet_hashtables.h | 5 +-- net/dccp/ipv4.c | 2 +- net/dccp/ipv6.c | 2 +- net/ipv4/inet_connection_sock.c | 2 +- net/ipv4/inet_hashtables.c | 68 ++++++++++++++++++++++++++++++++++++----- net/ipv4/tcp_ipv4.c | 15 +++++++-- net/ipv6/tcp_ipv6.c | 13 +++++++- 7 files changed, 91 insertions(+), 16 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h index 92560974ea67..ca6a3ea9057e 100644 --- a/include/net/inet_hashtables.h +++ b/include/net/inet_hashtables.h @@ -247,8 +247,9 @@ void inet_hashinfo2_init(struct inet_hashinfo *h, const char *name, unsigned long high_limit); int inet_hashinfo2_init_mod(struct inet_hashinfo *h); -bool inet_ehash_insert(struct sock *sk, struct sock *osk); -bool inet_ehash_nolisten(struct sock *sk, struct sock *osk); +bool inet_ehash_insert(struct sock *sk, struct sock *osk, bool *found_dup_sk); +bool inet_ehash_nolisten(struct sock *sk, struct sock *osk, + bool *found_dup_sk); int __inet_hash(struct sock *sk, struct sock *osk); int inet_hash(struct sock *sk); void inet_unhash(struct sock *sk); diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index bb3d70664dde..b0b6e6a4784e 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -427,7 +427,7 @@ struct sock *dccp_v4_request_recv_sock(const struct sock *sk, if (__inet_inherit_port(sk, newsk) < 0) goto put_and_exit; - *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash)); + *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash), NULL); if (*own_req) ireq->ireq_opt = NULL; else diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index ef4ab28cfde0..78ee1b5acf1f 100644 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@ -533,7 +533,7 @@ static struct sock *dccp_v6_request_recv_sock(const struct sock *sk, dccp_done(newsk); goto out; } - *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash)); + *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash), NULL); /* Clone pktoptions received with SYN, if we own the req */ if (*own_req && ireq->pktopts) { newnp->pktoptions = skb_clone(ireq->pktopts, GFP_ATOMIC); diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 4148f5f78f31..f60869acbef0 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -787,7 +787,7 @@ static void reqsk_queue_hash_req(struct request_sock *req, timer_setup(&req->rsk_timer, reqsk_timer_handler, TIMER_PINNED); mod_timer(&req->rsk_timer, jiffies + timeout); - inet_ehash_insert(req_to_sk(req), NULL); + inet_ehash_insert(req_to_sk(req), NULL, NULL); /* before letting lookups find us, make sure all req fields * are committed to memory and refcnt initialized. */ diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c index 8cbe74313f38..45fb450b4522 100644 --- a/net/ipv4/inet_hashtables.c +++ b/net/ipv4/inet_hashtables.c @@ -20,6 +20,9 @@ #include #include #include +#if IS_ENABLED(CONFIG_IPV6) +#include +#endif #include #include #include @@ -508,10 +511,52 @@ static u32 inet_sk_port_offset(const struct sock *sk) inet->inet_dport); } -/* insert a socket into ehash, and eventually remove another one - * (The another one can be a SYN_RECV or TIMEWAIT +/* Searches for an exsiting socket in the ehash bucket list. + * Returns true if found, false otherwise. */ -bool inet_ehash_insert(struct sock *sk, struct sock *osk) +static bool inet_ehash_lookup_by_sk(struct sock *sk, + struct hlist_nulls_head *list) +{ + const __portpair ports = INET_COMBINED_PORTS(sk->sk_dport, sk->sk_num); + const int sdif = sk->sk_bound_dev_if; + const int dif = sk->sk_bound_dev_if; + const struct hlist_nulls_node *node; + struct net *net = sock_net(sk); + struct sock *esk; + + INET_ADDR_COOKIE(acookie, sk->sk_daddr, sk->sk_rcv_saddr); + + sk_nulls_for_each_rcu(esk, node, list) { + if (esk->sk_hash != sk->sk_hash) + continue; + if (sk->sk_family == AF_INET) { + if (unlikely(INET_MATCH(esk, net, acookie, + sk->sk_daddr, + sk->sk_rcv_saddr, + ports, dif, sdif))) { + return true; + } + } +#if IS_ENABLED(CONFIG_IPV6) + else if (sk->sk_family == AF_INET6) { + if (unlikely(INET6_MATCH(esk, net, + &sk->sk_v6_daddr, + &sk->sk_v6_rcv_saddr, + ports, dif, sdif))) { + return true; + } + } +#endif + } + return false; +} + +/* Insert a socket into ehash, and eventually remove another one + * (The another one can be a SYN_RECV or TIMEWAIT) + * If an existing socket already exists, socket sk is not inserted, + * and sets found_dup_sk parameter to true. + */ +bool inet_ehash_insert(struct sock *sk, struct sock *osk, bool *found_dup_sk) { struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo; struct hlist_nulls_head *list; @@ -530,16 +575,23 @@ bool inet_ehash_insert(struct sock *sk, struct sock *osk) if (osk) { WARN_ON_ONCE(sk->sk_hash != osk->sk_hash); ret = sk_nulls_del_node_init_rcu(osk); + } else if (found_dup_sk) { + *found_dup_sk = inet_ehash_lookup_by_sk(sk, list); + if (*found_dup_sk) + ret = false; } + if (ret) __sk_nulls_add_node_rcu(sk, list); + spin_unlock(lock); + return ret; } -bool inet_ehash_nolisten(struct sock *sk, struct sock *osk) +bool inet_ehash_nolisten(struct sock *sk, struct sock *osk, bool *found_dup_sk) { - bool ok = inet_ehash_insert(sk, osk); + bool ok = inet_ehash_insert(sk, osk, found_dup_sk); if (ok) { sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); @@ -583,7 +635,7 @@ int __inet_hash(struct sock *sk, struct sock *osk) int err = 0; if (sk->sk_state != TCP_LISTEN) { - inet_ehash_nolisten(sk, osk); + inet_ehash_nolisten(sk, osk, NULL); return 0; } WARN_ON(!sk_unhashed(sk)); @@ -679,7 +731,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row, tb = inet_csk(sk)->icsk_bind_hash; spin_lock_bh(&head->lock); if (sk_head(&tb->owners) == sk && !sk->sk_bind_node.next) { - inet_ehash_nolisten(sk, NULL); + inet_ehash_nolisten(sk, NULL, NULL); spin_unlock_bh(&head->lock); return 0; } @@ -758,7 +810,7 @@ ok: inet_bind_hash(sk, tb, port); if (sk_unhashed(sk)) { inet_sk(sk)->inet_sport = htons(port); - inet_ehash_nolisten(sk, (struct sock *)tw); + inet_ehash_nolisten(sk, (struct sock *)tw, NULL); } if (tw) inet_twsk_bind_unhash(tw, hinfo); diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 25d35a56ceba..c95fcdfeed42 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1499,6 +1499,7 @@ struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb, bool *own_req) { struct inet_request_sock *ireq; + bool found_dup_sk = false; struct inet_sock *newinet; struct tcp_sock *newtp; struct sock *newsk; @@ -1576,12 +1577,22 @@ struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb, if (__inet_inherit_port(sk, newsk) < 0) goto put_and_exit; - *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash)); + *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash), + &found_dup_sk); if (likely(*own_req)) { tcp_move_syn(newtp, req); ireq->ireq_opt = NULL; } else { - newinet->inet_opt = NULL; + if (!req_unhash && found_dup_sk) { + /* This code path should only be executed in the + * syncookie case only + */ + bh_unlock_sock(newsk); + sock_put(newsk); + newsk = NULL; + } else { + newinet->inet_opt = NULL; + } } return newsk; diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 3d49e8d0afee..d2502911b7fa 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1193,6 +1193,7 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * const struct ipv6_pinfo *np = tcp_inet6_sk(sk); struct ipv6_txoptions *opt; struct inet_sock *newinet; + bool found_dup_sk = false; struct tcp_sock *newtp; struct sock *newsk; #ifdef CONFIG_TCP_MD5SIG @@ -1368,7 +1369,8 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * tcp_done(newsk); goto out; } - *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash)); + *own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash), + &found_dup_sk); if (*own_req) { tcp_move_syn(newtp, req); @@ -1383,6 +1385,15 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * skb_set_owner_r(newnp->pktoptions, newsk); } } + } else { + if (!req_unhash && found_dup_sk) { + /* This code path should only be executed in the + * syncookie case only + */ + bh_unlock_sock(newsk); + sock_put(newsk); + newsk = NULL; + } } return newsk; -- cgit v1.2.3 From cc69837fcaf467426ca19e5790085c26146a2300 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 20 Nov 2020 14:50:52 -0800 Subject: net: don't include ethtool.h from netdevice.h linux/netdevice.h is included in very many places, touching any of its dependecies causes large incremental builds. Drop the linux/ethtool.h include, linux/netdevice.h just needs a forward declaration of struct ethtool_ops. Fix all the places which made use of this implicit include. Acked-by: Johannes Berg Acked-by: Shannon Nelson Reviewed-by: Jesse Brandeburg Link: https://lore.kernel.org/r/20201120225052.1427503-1-kuba@kernel.org Signed-off-by: Jakub Kicinski --- drivers/isdn/capi/capi.c | 1 + drivers/media/pci/ttpci/av7110_av.c | 1 + drivers/net/bonding/bond_procfs.c | 1 + drivers/net/can/usb/gs_usb.c | 1 + drivers/net/ethernet/amazon/ena/ena_ethtool.c | 1 + drivers/net/ethernet/aquantia/atlantic/aq_nic.h | 2 ++ drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + drivers/net/ethernet/broadcom/bnxt/bnxt_sriov.c | 1 + drivers/net/ethernet/cavium/liquidio/lio_ethtool.c | 1 + drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c | 1 + drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 1 + drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c | 1 + drivers/net/ethernet/google/gve/gve_ethtool.c | 1 + drivers/net/ethernet/hisilicon/hns3/hnae3.h | 1 + drivers/net/ethernet/huawei/hinic/hinic_port.h | 1 + drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c | 1 + drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h | 1 + drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 1 + drivers/net/ethernet/mellanox/mlxsw/core_env.h | 3 +++ drivers/net/ethernet/mellanox/mlxsw/spectrum.h | 1 + drivers/net/ethernet/mellanox/mlxsw/switchx2.c | 1 + drivers/net/ethernet/pensando/ionic/ionic_lif.c | 1 + drivers/net/ethernet/pensando/ionic/ionic_stats.c | 1 + drivers/net/ethernet/qualcomm/rmnet/rmnet_vnd.c | 1 + drivers/net/geneve.c | 1 + drivers/net/hyperv/netvsc_drv.c | 1 + drivers/net/hyperv/rndis_filter.c | 1 + drivers/net/ipvlan/ipvlan_main.c | 2 ++ drivers/net/nlmon.c | 1 + drivers/net/team/team.c | 1 + drivers/net/vrf.c | 1 + drivers/net/vsockmon.c | 1 + drivers/scsi/bnx2fc/bnx2fc_fcoe.c | 2 ++ drivers/scsi/fcoe/fcoe_transport.c | 1 + drivers/staging/fsl-dpaa2/ethsw/ethsw-ethtool.c | 2 ++ drivers/staging/wimax/i2400m/usb.c | 1 + include/linux/netdevice.h | 2 +- include/linux/qed/qed_if.h | 1 + include/net/cfg80211.h | 1 + include/rdma/ib_addr.h | 1 + include/rdma/ib_verbs.h | 1 + net/packet/af_packet.c | 1 + net/sched/sch_cbs.c | 1 + net/sched/sch_taprio.c | 1 + net/socket.c | 1 + 45 files changed, 51 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/drivers/isdn/capi/capi.c b/drivers/isdn/capi/capi.c index 85767f52fe3c..fdf87acccd06 100644 --- a/drivers/isdn/capi/capi.c +++ b/drivers/isdn/capi/capi.c @@ -11,6 +11,7 @@ #include #include +#include #include #include #include diff --git a/drivers/media/pci/ttpci/av7110_av.c b/drivers/media/pci/ttpci/av7110_av.c index ea9f7d0058a2..91f4866c7e59 100644 --- a/drivers/media/pci/ttpci/av7110_av.c +++ b/drivers/media/pci/ttpci/av7110_av.c @@ -11,6 +11,7 @@ * the project's page is at https://linuxtv.org */ +#include #include #include #include diff --git a/drivers/net/bonding/bond_procfs.c b/drivers/net/bonding/bond_procfs.c index fd5c9cbe45b1..56d34be5e797 100644 --- a/drivers/net/bonding/bond_procfs.c +++ b/drivers/net/bonding/bond_procfs.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #include +#include #include #include #include diff --git a/drivers/net/can/usb/gs_usb.c b/drivers/net/can/usb/gs_usb.c index b3431c3feba1..a0336e895d94 100644 --- a/drivers/net/can/usb/gs_usb.c +++ b/drivers/net/can/usb/gs_usb.c @@ -9,6 +9,7 @@ * Many thanks to all socketcan devs! */ +#include #include #include #include diff --git a/drivers/net/ethernet/amazon/ena/ena_ethtool.c b/drivers/net/ethernet/amazon/ena/ena_ethtool.c index 3b2cd28f962d..6cdd9efe8df3 100644 --- a/drivers/net/ethernet/amazon/ena/ena_ethtool.c +++ b/drivers/net/ethernet/amazon/ena/ena_ethtool.c @@ -3,6 +3,7 @@ * Copyright 2015-2020 Amazon.com, Inc. or its affiliates. All rights reserved. */ +#include #include #include "ena_netdev.h" diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.h b/drivers/net/ethernet/aquantia/atlantic/aq_nic.h index 926cca9a0c83..1a7148041e3d 100644 --- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.h +++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.h @@ -10,6 +10,8 @@ #ifndef AQ_NIC_H #define AQ_NIC_H +#include + #include "aq_common.h" #include "aq_rss.h" #include "aq_hw.h" diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h index 47b3c3127879..950ea26ae0d2 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h @@ -20,6 +20,7 @@ #define DRV_VER_MIN 10 #define DRV_VER_UPD 1 +#include #include #include #include diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_sriov.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_sriov.c index 23b80aa171dd..a217316228f4 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_sriov.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_sriov.c @@ -8,6 +8,7 @@ * the Free Software Foundation. */ +#include #include #include #include diff --git a/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c b/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c index 16eebfc52109..66f2c553370c 100644 --- a/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c +++ b/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c @@ -15,6 +15,7 @@ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or * NONINFRINGEMENT. See the GNU General Public License for more details. ***********************************************************************/ +#include #include #include #include diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c b/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c index c7bdac79299a..2f218fbfed06 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c @@ -5,6 +5,7 @@ /* ETHTOOL Support for VNIC_VF Device*/ +#include #include #include diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h index 27308600da15..8e681ce72d62 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h @@ -39,6 +39,7 @@ #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c b/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c index cd8f9a481d73..d546993bda09 100644 --- a/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c @@ -33,6 +33,7 @@ * SOFTWARE. */ +#include #include #include "t4vf_common.h" diff --git a/drivers/net/ethernet/google/gve/gve_ethtool.c b/drivers/net/ethernet/google/gve/gve_ethtool.c index 7b44769bd87c..2fb197fd3daf 100644 --- a/drivers/net/ethernet/google/gve/gve_ethtool.c +++ b/drivers/net/ethernet/google/gve/gve_ethtool.c @@ -4,6 +4,7 @@ * Copyright (C) 2015-2019 Google, Inc. */ +#include #include #include "gve.h" #include "gve_adminq.h" diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 5bae5e859c81..f6fac2418648 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/huawei/hinic/hinic_port.h b/drivers/net/ethernet/huawei/hinic/hinic_port.h index 9c3cbe45c9ec..c9ae3d4dc547 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_port.h +++ b/drivers/net/ethernet/huawei/hinic/hinic_port.h @@ -8,6 +8,7 @@ #define HINIC_PORT_H #include +#include #include #include diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c b/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c index 908fefaa6b85..66776ba7bfb6 100644 --- a/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c +++ b/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright(c) 2013 - 2019 Intel Corporation. */ +#include #include #include "fm10k.h" diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h index ceec649bdd13..103430400a8a 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h @@ -11,6 +11,7 @@ #ifndef OTX2_COMMON_H #define OTX2_COMMON_H +#include #include #include #include diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index 014ce8d3d97b..1c50d0f22199 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -36,6 +36,7 @@ #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_env.h b/drivers/net/ethernet/mellanox/mlxsw/core_env.h index 8e36a2634ef5..2b23f8a87862 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core_env.h +++ b/drivers/net/ethernet/mellanox/mlxsw/core_env.h @@ -4,6 +4,9 @@ #ifndef _MLXSW_CORE_ENV_H #define _MLXSW_CORE_ENV_H +struct ethtool_modinfo; +struct ethtool_eeprom; + int mlxsw_env_module_temp_thresholds_get(struct mlxsw_core *core, int module, int off, int *temp); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index 74b3959b36d4..642099fee380 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -4,6 +4,7 @@ #ifndef _MLXSW_SPECTRUM_H #define _MLXSW_SPECTRUM_H +#include #include #include #include diff --git a/drivers/net/ethernet/mellanox/mlxsw/switchx2.c b/drivers/net/ethernet/mellanox/mlxsw/switchx2.c index 5023d91269f4..40e2e79d4517 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/switchx2.c +++ b/drivers/net/ethernet/mellanox/mlxsw/switchx2.c @@ -6,6 +6,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/pensando/ionic/ionic_lif.c b/drivers/net/ethernet/pensando/ionic/ionic_lif.c index deabd813e3fe..0afec2fa572d 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_lif.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_lif.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright(c) 2017 - 2019 Pensando Systems, Inc */ +#include #include #include #include diff --git a/drivers/net/ethernet/pensando/ionic/ionic_stats.c b/drivers/net/ethernet/pensando/ionic/ionic_stats.c index ff20a2ac4c2f..6ae75b771a15 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_stats.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_stats.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright(c) 2017 - 2019 Pensando Systems, Inc */ +#include #include #include #include diff --git a/drivers/net/ethernet/qualcomm/rmnet/rmnet_vnd.c b/drivers/net/ethernet/qualcomm/rmnet/rmnet_vnd.c index d58b51d277f1..ca1535ebb6e7 100644 --- a/drivers/net/ethernet/qualcomm/rmnet/rmnet_vnd.c +++ b/drivers/net/ethernet/qualcomm/rmnet/rmnet_vnd.c @@ -5,6 +5,7 @@ */ #include +#include #include #include #include "rmnet_config.h" diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c index ef9b5ea9073b..5523f069b9a5 100644 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@ -7,6 +7,7 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt +#include #include #include #include diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c index 261e6e55a907..d17bbc75f5e7 100644 --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@ -10,6 +10,7 @@ #include #include +#include #include #include #include diff --git a/drivers/net/hyperv/rndis_filter.c b/drivers/net/hyperv/rndis_filter.c index b22e47bcfeca..2c2b55c32a7a 100644 --- a/drivers/net/hyperv/rndis_filter.c +++ b/drivers/net/hyperv/rndis_filter.c @@ -6,6 +6,7 @@ * Haiyang Zhang * Hank Janssen */ +#include #include #include #include diff --git a/drivers/net/ipvlan/ipvlan_main.c b/drivers/net/ipvlan/ipvlan_main.c index 60b7d93bb834..a707502a0c0f 100644 --- a/drivers/net/ipvlan/ipvlan_main.c +++ b/drivers/net/ipvlan/ipvlan_main.c @@ -2,6 +2,8 @@ /* Copyright (c) 2014 Mahesh Bandewar */ +#include + #include "ipvlan.h" static int ipvlan_set_port_mode(struct ipvl_port *port, u16 nval, diff --git a/drivers/net/nlmon.c b/drivers/net/nlmon.c index afb119f38325..5e19a6839dea 100644 --- a/drivers/net/nlmon.c +++ b/drivers/net/nlmon.c @@ -1,4 +1,5 @@ // SPDX-License-Identifier: GPL-2.0-only +#include #include #include #include diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c index b4092127a92c..c19dac21c468 100644 --- a/drivers/net/team/team.c +++ b/drivers/net/team/team.c @@ -4,6 +4,7 @@ * Copyright (c) 2011 Jiri Pirko */ +#include #include #include #include diff --git a/drivers/net/vrf.c b/drivers/net/vrf.c index f2793ffde191..f8d711a84763 100644 --- a/drivers/net/vrf.c +++ b/drivers/net/vrf.c @@ -9,6 +9,7 @@ * Based on dummy, team and ipvlan drivers */ +#include #include #include #include diff --git a/drivers/net/vsockmon.c b/drivers/net/vsockmon.c index e8563acf98e8..b1bb1b04b664 100644 --- a/drivers/net/vsockmon.c +++ b/drivers/net/vsockmon.c @@ -1,4 +1,5 @@ // SPDX-License-Identifier: GPL-2.0-only +#include #include #include #include diff --git a/drivers/scsi/bnx2fc/bnx2fc_fcoe.c b/drivers/scsi/bnx2fc/bnx2fc_fcoe.c index 6890bbe04a8c..08fb7d5d08b3 100644 --- a/drivers/scsi/bnx2fc/bnx2fc_fcoe.c +++ b/drivers/scsi/bnx2fc/bnx2fc_fcoe.c @@ -16,6 +16,8 @@ #include "bnx2fc.h" +#include + static struct list_head adapter_list; static struct list_head if_list; static u32 adapter_count; diff --git a/drivers/scsi/fcoe/fcoe_transport.c b/drivers/scsi/fcoe/fcoe_transport.c index 6e187d0e71fd..b927b3d84523 100644 --- a/drivers/scsi/fcoe/fcoe_transport.c +++ b/drivers/scsi/fcoe/fcoe_transport.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/staging/fsl-dpaa2/ethsw/ethsw-ethtool.c b/drivers/staging/fsl-dpaa2/ethsw/ethsw-ethtool.c index ace4a6d28562..ad55cd738847 100644 --- a/drivers/staging/fsl-dpaa2/ethsw/ethsw-ethtool.c +++ b/drivers/staging/fsl-dpaa2/ethsw/ethsw-ethtool.c @@ -7,6 +7,8 @@ * */ +#include + #include "ethsw.h" static struct { diff --git a/drivers/staging/wimax/i2400m/usb.c b/drivers/staging/wimax/i2400m/usb.c index 3b84dd7b5567..f250d03ce7c7 100644 --- a/drivers/staging/wimax/i2400m/usb.c +++ b/drivers/staging/wimax/i2400m/usb.c @@ -51,6 +51,7 @@ #include "i2400m-usb.h" #include "linux-wimax-i2400m.h" #include +#include #include #include diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 03433a4c929e..0049e8fe4905 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -34,7 +34,6 @@ #include #include -#include #include #ifdef CONFIG_DCB #include @@ -51,6 +50,7 @@ struct netpoll_info; struct device; +struct ethtool_ops; struct phy_device; struct dsa_port; struct ip_tunnel_parm; diff --git a/include/linux/qed/qed_if.h b/include/linux/qed/qed_if.h index 57fb295ea41a..68d17a4fbf20 100644 --- a/include/linux/qed/qed_if.h +++ b/include/linux/qed/qed_if.h @@ -7,6 +7,7 @@ #ifndef _QED_IF_H #define _QED_IF_H +#include #include #include #include diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index ab249ca5d5d1..78c763dfc99a 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -10,6 +10,7 @@ * Copyright (C) 2018-2020 Intel Corporation */ +#include #include #include #include diff --git a/include/rdma/ib_addr.h b/include/rdma/ib_addr.h index b0e636ac6690..d808dc3d239e 100644 --- a/include/rdma/ib_addr.h +++ b/include/rdma/ib_addr.h @@ -7,6 +7,7 @@ #ifndef IB_ADDR_H #define IB_ADDR_H +#include #include #include #include diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 9bf6c319a670..3883efd588aa 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -12,6 +12,7 @@ #ifndef IB_VERBS_H #define IB_VERBS_H +#include #include #include #include diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 62ebfaa7adcb..48a0ed836b46 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -46,6 +46,7 @@ * Copyright (C) 2011, */ +#include #include #include #include diff --git a/net/sched/sch_cbs.c b/net/sched/sch_cbs.c index 2eaac2ff380f..459cc240eda9 100644 --- a/net/sched/sch_cbs.c +++ b/net/sched/sch_cbs.c @@ -50,6 +50,7 @@ * locredit = max_frame_size * (sendslope / port_transmit_rate) */ +#include #include #include #include diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c index b0ad7687ee2c..26fb8a62996b 100644 --- a/net/sched/sch_taprio.c +++ b/net/sched/sch_taprio.c @@ -6,6 +6,7 @@ * */ +#include #include #include #include diff --git a/net/socket.c b/net/socket.c index 152b1dcf93c6..bfef11ba35b8 100644 --- a/net/socket.c +++ b/net/socket.c @@ -52,6 +52,7 @@ * Based upon Swansea University Computer Society NET3.039 */ +#include #include #include #include -- cgit v1.2.3 From f0a5013e29cbfe55b3e26b7548a741d88a779d91 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Mon, 23 Nov 2020 09:12:28 +0200 Subject: devlink: Add blackhole_nexthop trap Add a packet trap to report packets that were dropped due to a blackhole nexthop. Signed-off-by: Ido Schimmel Reviewed-by: Jiri Pirko Signed-off-by: Jakub Kicinski --- Documentation/networking/devlink/devlink-trap.rst | 4 ++++ include/net/devlink.h | 4 +++- net/core/devlink.c | 1 + 3 files changed, 8 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/Documentation/networking/devlink/devlink-trap.rst b/Documentation/networking/devlink/devlink-trap.rst index ef719ceac299..d875f3e1e9cf 100644 --- a/Documentation/networking/devlink/devlink-trap.rst +++ b/Documentation/networking/devlink/devlink-trap.rst @@ -476,6 +476,10 @@ be added to the following table: * - ``esp_parsing`` - ``drop`` - Traps packets dropped due to an error in the ESP header parsing + * - ``blackhole_nexthop`` + - ``drop`` + - Traps packets that the device decided to drop in case they hit a + blackhole nexthop Driver-specific Packet Traps ============================ diff --git a/include/net/devlink.h b/include/net/devlink.h index 457c537d0ef2..f466819cc477 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -835,6 +835,7 @@ enum devlink_trap_generic_id { DEVLINK_TRAP_GENERIC_ID_DCCP_PARSING, DEVLINK_TRAP_GENERIC_ID_GTP_PARSING, DEVLINK_TRAP_GENERIC_ID_ESP_PARSING, + DEVLINK_TRAP_GENERIC_ID_BLACKHOLE_NEXTHOP, /* Add new generic trap IDs above */ __DEVLINK_TRAP_GENERIC_ID_MAX, @@ -1058,7 +1059,8 @@ enum devlink_trap_group_generic_id { "gtp_parsing" #define DEVLINK_TRAP_GENERIC_NAME_ESP_PARSING \ "esp_parsing" - +#define DEVLINK_TRAP_GENERIC_NAME_BLACKHOLE_NEXTHOP \ + "blackhole_nexthop" #define DEVLINK_TRAP_GROUP_GENERIC_NAME_L2_DROPS \ "l2_drops" diff --git a/net/core/devlink.c b/net/core/devlink.c index e6fb1fdedded..7c05e8603bff 100644 --- a/net/core/devlink.c +++ b/net/core/devlink.c @@ -9490,6 +9490,7 @@ static const struct devlink_trap devlink_trap_generic[] = { DEVLINK_TRAP(DCCP_PARSING, DROP), DEVLINK_TRAP(GTP_PARSING, DROP), DEVLINK_TRAP(ESP_PARSING, DROP), + DEVLINK_TRAP(BLACKHOLE_NEXTHOP, DROP), }; #define DEVLINK_TRAP_GROUP(_id) \ -- cgit v1.2.3 From 025cc2fb6a4e84e9a0552c0017dcd1c24b7ac7da Mon Sep 17 00:00:00 2001 From: Maxim Mikityanskiy Date: Wed, 25 Nov 2020 14:18:10 -0800 Subject: net/tls: Protect from calling tls_dev_del for TLS RX twice tls_device_offload_cleanup_rx doesn't clear tls_ctx->netdev after calling tls_dev_del if TLX TX offload is also enabled. Clearing tls_ctx->netdev gets postponed until tls_device_gc_task. It leaves a time frame when tls_device_down may get called and call tls_dev_del for RX one extra time, confusing the driver, which may lead to a crash. This patch corrects this racy behavior by adding a flag to prevent tls_device_down from calling tls_dev_del the second time. Fixes: e8f69799810c ("net/tls: Add generic NIC offload infrastructure") Signed-off-by: Maxim Mikityanskiy Signed-off-by: Saeed Mahameed Link: https://lore.kernel.org/r/20201125221810.69870-1-saeedm@nvidia.com Signed-off-by: Jakub Kicinski --- include/net/tls.h | 6 ++++++ net/tls/tls_device.c | 5 ++++- 2 files changed, 10 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/tls.h b/include/net/tls.h index cf1473099453..2bdd802212fe 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -199,6 +199,12 @@ enum tls_context_flags { * to be atomic. */ TLS_TX_SYNC_SCHED = 1, + /* tls_dev_del was called for the RX side, device state was released, + * but tls_ctx->netdev might still be kept, because TX-side driver + * resources might not be released yet. Used to prevent the second + * tls_dev_del call in tls_device_down if it happens simultaneously. + */ + TLS_RX_DEV_CLOSED = 2, }; struct cipher_context { diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c index 54d3e161d198..a3ab2d3d4e4e 100644 --- a/net/tls/tls_device.c +++ b/net/tls/tls_device.c @@ -1262,6 +1262,8 @@ void tls_device_offload_cleanup_rx(struct sock *sk) if (tls_ctx->tx_conf != TLS_HW) { dev_put(netdev); tls_ctx->netdev = NULL; + } else { + set_bit(TLS_RX_DEV_CLOSED, &tls_ctx->flags); } out: up_read(&device_offload_lock); @@ -1291,7 +1293,8 @@ static int tls_device_down(struct net_device *netdev) if (ctx->tx_conf == TLS_HW) netdev->tlsdev_ops->tls_dev_del(netdev, ctx, TLS_OFFLOAD_CTX_DIR_TX); - if (ctx->rx_conf == TLS_HW) + if (ctx->rx_conf == TLS_HW && + !test_bit(TLS_RX_DEV_CLOSED, &ctx->flags)) netdev->tlsdev_ops->tls_dev_del(netdev, ctx, TLS_OFFLOAD_CTX_DIR_RX); WRITE_ONCE(ctx->netdev, NULL); -- cgit v1.2.3 From 3c78e9e0d33a27ab8050e4492c03c6a1f8d0ed6b Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 25 Nov 2020 23:50:07 +0100 Subject: netfilter: nftables_offload: set address type in control dissector This patch adds nft_flow_rule_set_addr_type() to set the address type from the nft_payload expression accordingly. If the address type is not set in the control dissector then a rule that matches either on source or destination IP address does not work. After this patch, nft hardware offload generates the flow dissector configuration as tc-flower does to match on an IP address. This patch has been also tested functionally to make sure packets are filtered out by the NIC. This is also getting the code aligned with the existing netfilter flow offload infrastructure which is also setting the control dissector. Fixes: c9626a2cbdb2 ("netfilter: nf_tables: add hardware offload support") Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables_offload.h | 4 ++++ net/netfilter/nf_tables_offload.c | 17 +++++++++++++++++ net/netfilter/nft_payload.c | 4 ++++ 3 files changed, 25 insertions(+) (limited to 'include/net') diff --git a/include/net/netfilter/nf_tables_offload.h b/include/net/netfilter/nf_tables_offload.h index ea7d1d78b92d..bddd34c5bd79 100644 --- a/include/net/netfilter/nf_tables_offload.h +++ b/include/net/netfilter/nf_tables_offload.h @@ -37,6 +37,7 @@ void nft_offload_update_dependency(struct nft_offload_ctx *ctx, struct nft_flow_key { struct flow_dissector_key_basic basic; + struct flow_dissector_key_control control; union { struct flow_dissector_key_ipv4_addrs ipv4; struct flow_dissector_key_ipv6_addrs ipv6; @@ -62,6 +63,9 @@ struct nft_flow_rule { #define NFT_OFFLOAD_F_ACTION (1 << 0) +void nft_flow_rule_set_addr_type(struct nft_flow_rule *flow, + enum flow_dissector_key_id addr_type); + struct nft_rule; struct nft_flow_rule *nft_flow_rule_create(struct net *net, const struct nft_rule *rule); void nft_flow_rule_destroy(struct nft_flow_rule *flow); diff --git a/net/netfilter/nf_tables_offload.c b/net/netfilter/nf_tables_offload.c index 9f625724a20f..9ae14270c543 100644 --- a/net/netfilter/nf_tables_offload.c +++ b/net/netfilter/nf_tables_offload.c @@ -28,6 +28,23 @@ static struct nft_flow_rule *nft_flow_rule_alloc(int num_actions) return flow; } +void nft_flow_rule_set_addr_type(struct nft_flow_rule *flow, + enum flow_dissector_key_id addr_type) +{ + struct nft_flow_match *match = &flow->match; + struct nft_flow_key *mask = &match->mask; + struct nft_flow_key *key = &match->key; + + if (match->dissector.used_keys & BIT(FLOW_DISSECTOR_KEY_CONTROL)) + return; + + key->control.addr_type = addr_type; + mask->control.addr_type = 0xffff; + match->dissector.used_keys |= BIT(FLOW_DISSECTOR_KEY_CONTROL); + match->dissector.offset[FLOW_DISSECTOR_KEY_CONTROL] = + offsetof(struct nft_flow_key, control); +} + struct nft_flow_rule *nft_flow_rule_create(struct net *net, const struct nft_rule *rule) { diff --git a/net/netfilter/nft_payload.c b/net/netfilter/nft_payload.c index dcd3c7b8a367..bbf811d030d5 100644 --- a/net/netfilter/nft_payload.c +++ b/net/netfilter/nft_payload.c @@ -244,6 +244,7 @@ static int nft_payload_offload_ip(struct nft_offload_ctx *ctx, NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV4_ADDRS, ipv4, src, sizeof(struct in_addr), reg); + nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV4_ADDRS); break; case offsetof(struct iphdr, daddr): if (priv->len != sizeof(struct in_addr)) @@ -251,6 +252,7 @@ static int nft_payload_offload_ip(struct nft_offload_ctx *ctx, NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV4_ADDRS, ipv4, dst, sizeof(struct in_addr), reg); + nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV4_ADDRS); break; case offsetof(struct iphdr, protocol): if (priv->len != sizeof(__u8)) @@ -280,6 +282,7 @@ static int nft_payload_offload_ip6(struct nft_offload_ctx *ctx, NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV6_ADDRS, ipv6, src, sizeof(struct in6_addr), reg); + nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV6_ADDRS); break; case offsetof(struct ipv6hdr, daddr): if (priv->len != sizeof(struct in6_addr)) @@ -287,6 +290,7 @@ static int nft_payload_offload_ip6(struct nft_offload_ctx *ctx, NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV6_ADDRS, ipv6, dst, sizeof(struct in6_addr), reg); + nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV6_ADDRS); break; case offsetof(struct ipv6hdr, nexthdr): if (priv->len != sizeof(__u8)) -- cgit v1.2.3 From a5d45bc0dc50f9dd83703510e9804d813a9cac32 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 25 Nov 2020 23:50:17 +0100 Subject: netfilter: nftables_offload: build mask based from the matching bytes Userspace might match on prefix bytes of header fields if they are on the byte boundary, this requires that the mask is adjusted accordingly. Use NFT_OFFLOAD_MATCH_EXACT() for meta since prefix byte matching is not allowed for this type of selector. The bitwise expression might be optimized out by userspace, hence the kernel needs to infer the prefix from the number of payload bytes to match on. This patch adds nft_payload_offload_mask() to calculate the bitmask to match on the prefix. Fixes: c9626a2cbdb2 ("netfilter: nf_tables: add hardware offload support") Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables_offload.h | 3 ++ net/netfilter/nft_cmp.c | 8 ++-- net/netfilter/nft_meta.c | 16 ++++---- net/netfilter/nft_payload.c | 66 +++++++++++++++++++++++-------- 4 files changed, 64 insertions(+), 29 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_tables_offload.h b/include/net/netfilter/nf_tables_offload.h index bddd34c5bd79..1d34fe154fe0 100644 --- a/include/net/netfilter/nf_tables_offload.h +++ b/include/net/netfilter/nf_tables_offload.h @@ -78,6 +78,9 @@ int nft_flow_rule_offload_commit(struct net *net); offsetof(struct nft_flow_key, __base.__field); \ (__reg)->len = __len; \ (__reg)->key = __key; \ + +#define NFT_OFFLOAD_MATCH_EXACT(__key, __base, __field, __len, __reg) \ + NFT_OFFLOAD_MATCH(__key, __base, __field, __len, __reg) \ memset(&(__reg)->mask, 0xff, (__reg)->len); int nft_chain_offload_priority(struct nft_base_chain *basechain); diff --git a/net/netfilter/nft_cmp.c b/net/netfilter/nft_cmp.c index bc079d68a536..00e563a72d3d 100644 --- a/net/netfilter/nft_cmp.c +++ b/net/netfilter/nft_cmp.c @@ -123,11 +123,11 @@ static int __nft_cmp_offload(struct nft_offload_ctx *ctx, u8 *mask = (u8 *)&flow->match.mask; u8 *key = (u8 *)&flow->match.key; - if (priv->op != NFT_CMP_EQ || reg->len != priv->len) + if (priv->op != NFT_CMP_EQ || priv->len > reg->len) return -EOPNOTSUPP; - memcpy(key + reg->offset, &priv->data, priv->len); - memcpy(mask + reg->offset, ®->mask, priv->len); + memcpy(key + reg->offset, &priv->data, reg->len); + memcpy(mask + reg->offset, ®->mask, reg->len); flow->match.dissector.used_keys |= BIT(reg->key); flow->match.dissector.offset[reg->key] = reg->base_offset; @@ -137,7 +137,7 @@ static int __nft_cmp_offload(struct nft_offload_ctx *ctx, nft_reg_load16(priv->data.data) != ARPHRD_ETHER) return -EOPNOTSUPP; - nft_offload_update_dependency(ctx, &priv->data, priv->len); + nft_offload_update_dependency(ctx, &priv->data, reg->len); return 0; } diff --git a/net/netfilter/nft_meta.c b/net/netfilter/nft_meta.c index b37bd02448d8..bf4b3ad5314c 100644 --- a/net/netfilter/nft_meta.c +++ b/net/netfilter/nft_meta.c @@ -724,22 +724,22 @@ static int nft_meta_get_offload(struct nft_offload_ctx *ctx, switch (priv->key) { case NFT_META_PROTOCOL: - NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_BASIC, basic, n_proto, - sizeof(__u16), reg); + NFT_OFFLOAD_MATCH_EXACT(FLOW_DISSECTOR_KEY_BASIC, basic, n_proto, + sizeof(__u16), reg); nft_offload_set_dependency(ctx, NFT_OFFLOAD_DEP_NETWORK); break; case NFT_META_L4PROTO: - NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_BASIC, basic, ip_proto, - sizeof(__u8), reg); + NFT_OFFLOAD_MATCH_EXACT(FLOW_DISSECTOR_KEY_BASIC, basic, ip_proto, + sizeof(__u8), reg); nft_offload_set_dependency(ctx, NFT_OFFLOAD_DEP_TRANSPORT); break; case NFT_META_IIF: - NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_META, meta, - ingress_ifindex, sizeof(__u32), reg); + NFT_OFFLOAD_MATCH_EXACT(FLOW_DISSECTOR_KEY_META, meta, + ingress_ifindex, sizeof(__u32), reg); break; case NFT_META_IIFTYPE: - NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_META, meta, - ingress_iftype, sizeof(__u16), reg); + NFT_OFFLOAD_MATCH_EXACT(FLOW_DISSECTOR_KEY_META, meta, + ingress_iftype, sizeof(__u16), reg); break; default: return -EOPNOTSUPP; diff --git a/net/netfilter/nft_payload.c b/net/netfilter/nft_payload.c index bbf811d030d5..47d4e0e21651 100644 --- a/net/netfilter/nft_payload.c +++ b/net/netfilter/nft_payload.c @@ -165,6 +165,34 @@ nla_put_failure: return -1; } +static bool nft_payload_offload_mask(struct nft_offload_reg *reg, + u32 priv_len, u32 field_len) +{ + unsigned int remainder, delta, k; + struct nft_data mask = {}; + __be32 remainder_mask; + + if (priv_len == field_len) { + memset(®->mask, 0xff, priv_len); + return true; + } else if (priv_len > field_len) { + return false; + } + + memset(&mask, 0xff, field_len); + remainder = priv_len % sizeof(u32); + if (remainder) { + k = priv_len / sizeof(u32); + delta = field_len - priv_len; + remainder_mask = htonl(~((1 << (delta * BITS_PER_BYTE)) - 1)); + mask.data[k] = (__force u32)remainder_mask; + } + + memcpy(®->mask, &mask, field_len); + + return true; +} + static int nft_payload_offload_ll(struct nft_offload_ctx *ctx, struct nft_flow_rule *flow, const struct nft_payload *priv) @@ -173,21 +201,21 @@ static int nft_payload_offload_ll(struct nft_offload_ctx *ctx, switch (priv->offset) { case offsetof(struct ethhdr, h_source): - if (priv->len != ETH_ALEN) + if (!nft_payload_offload_mask(reg, priv->len, ETH_ALEN)) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_ETH_ADDRS, eth_addrs, src, ETH_ALEN, reg); break; case offsetof(struct ethhdr, h_dest): - if (priv->len != ETH_ALEN) + if (!nft_payload_offload_mask(reg, priv->len, ETH_ALEN)) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_ETH_ADDRS, eth_addrs, dst, ETH_ALEN, reg); break; case offsetof(struct ethhdr, h_proto): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_BASIC, basic, @@ -195,14 +223,14 @@ static int nft_payload_offload_ll(struct nft_offload_ctx *ctx, nft_offload_set_dependency(ctx, NFT_OFFLOAD_DEP_NETWORK); break; case offsetof(struct vlan_ethhdr, h_vlan_TCI): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_VLAN, vlan, vlan_tci, sizeof(__be16), reg); break; case offsetof(struct vlan_ethhdr, h_vlan_encapsulated_proto): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_VLAN, vlan, @@ -210,7 +238,7 @@ static int nft_payload_offload_ll(struct nft_offload_ctx *ctx, nft_offload_set_dependency(ctx, NFT_OFFLOAD_DEP_NETWORK); break; case offsetof(struct vlan_ethhdr, h_vlan_TCI) + sizeof(struct vlan_hdr): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_CVLAN, vlan, @@ -218,7 +246,7 @@ static int nft_payload_offload_ll(struct nft_offload_ctx *ctx, break; case offsetof(struct vlan_ethhdr, h_vlan_encapsulated_proto) + sizeof(struct vlan_hdr): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_CVLAN, vlan, @@ -239,7 +267,8 @@ static int nft_payload_offload_ip(struct nft_offload_ctx *ctx, switch (priv->offset) { case offsetof(struct iphdr, saddr): - if (priv->len != sizeof(struct in_addr)) + if (!nft_payload_offload_mask(reg, priv->len, + sizeof(struct in_addr))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV4_ADDRS, ipv4, src, @@ -247,7 +276,8 @@ static int nft_payload_offload_ip(struct nft_offload_ctx *ctx, nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV4_ADDRS); break; case offsetof(struct iphdr, daddr): - if (priv->len != sizeof(struct in_addr)) + if (!nft_payload_offload_mask(reg, priv->len, + sizeof(struct in_addr))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV4_ADDRS, ipv4, dst, @@ -255,7 +285,7 @@ static int nft_payload_offload_ip(struct nft_offload_ctx *ctx, nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV4_ADDRS); break; case offsetof(struct iphdr, protocol): - if (priv->len != sizeof(__u8)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__u8))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_BASIC, basic, ip_proto, @@ -277,7 +307,8 @@ static int nft_payload_offload_ip6(struct nft_offload_ctx *ctx, switch (priv->offset) { case offsetof(struct ipv6hdr, saddr): - if (priv->len != sizeof(struct in6_addr)) + if (!nft_payload_offload_mask(reg, priv->len, + sizeof(struct in6_addr))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV6_ADDRS, ipv6, src, @@ -285,7 +316,8 @@ static int nft_payload_offload_ip6(struct nft_offload_ctx *ctx, nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV6_ADDRS); break; case offsetof(struct ipv6hdr, daddr): - if (priv->len != sizeof(struct in6_addr)) + if (!nft_payload_offload_mask(reg, priv->len, + sizeof(struct in6_addr))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_IPV6_ADDRS, ipv6, dst, @@ -293,7 +325,7 @@ static int nft_payload_offload_ip6(struct nft_offload_ctx *ctx, nft_flow_rule_set_addr_type(flow, FLOW_DISSECTOR_KEY_IPV6_ADDRS); break; case offsetof(struct ipv6hdr, nexthdr): - if (priv->len != sizeof(__u8)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__u8))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_BASIC, basic, ip_proto, @@ -335,14 +367,14 @@ static int nft_payload_offload_tcp(struct nft_offload_ctx *ctx, switch (priv->offset) { case offsetof(struct tcphdr, source): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_PORTS, tp, src, sizeof(__be16), reg); break; case offsetof(struct tcphdr, dest): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_PORTS, tp, dst, @@ -363,14 +395,14 @@ static int nft_payload_offload_udp(struct nft_offload_ctx *ctx, switch (priv->offset) { case offsetof(struct udphdr, source): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_PORTS, tp, src, sizeof(__be16), reg); break; case offsetof(struct udphdr, dest): - if (priv->len != sizeof(__be16)) + if (!nft_payload_offload_mask(reg, priv->len, sizeof(__be16))) return -EOPNOTSUPP; NFT_OFFLOAD_MATCH(FLOW_DISSECTOR_KEY_PORTS, tp, dst, -- cgit v1.2.3 From 6942a284fb3e6bcb8ab03c98ef5cb048e0fbb3e9 Mon Sep 17 00:00:00 2001 From: Vadim Fedorenko Date: Tue, 24 Nov 2020 18:24:46 +0300 Subject: net/tls: make inline helpers protocol-aware Inline functions defined in tls.h have a lot of AES-specific constants. Remove these constants and change argument to struct tls_prot_info to have an access to cipher type in later patches Signed-off-by: Vadim Fedorenko Signed-off-by: Jakub Kicinski --- include/net/tls.h | 26 ++++++++++++-------------- net/tls/tls_device.c | 2 +- net/tls/tls_device_fallback.c | 13 +++++++------ net/tls/tls_sw.c | 12 +++++------- 4 files changed, 25 insertions(+), 28 deletions(-) (limited to 'include/net') diff --git a/include/net/tls.h b/include/net/tls.h index cf1473099453..d04ce73e54c9 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -502,31 +502,30 @@ static inline void tls_advance_record_sn(struct sock *sk, tls_err_abort(sk, EBADMSG); if (prot->version != TLS_1_3_VERSION) - tls_bigint_increment(ctx->iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE, + tls_bigint_increment(ctx->iv + prot->salt_size, prot->iv_size); } static inline void tls_fill_prepend(struct tls_context *ctx, char *buf, size_t plaintext_len, - unsigned char record_type, - int version) + unsigned char record_type) { struct tls_prot_info *prot = &ctx->prot_info; size_t pkt_len, iv_size = prot->iv_size; pkt_len = plaintext_len + prot->tag_size; - if (version != TLS_1_3_VERSION) { + if (prot->version != TLS_1_3_VERSION) { pkt_len += iv_size; memcpy(buf + TLS_NONCE_OFFSET, - ctx->tx.iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE, iv_size); + ctx->tx.iv + prot->salt_size, iv_size); } /* we cover nonce explicit here as well, so buf should be of * size KTLS_DTLS_HEADER_SIZE + KTLS_DTLS_NONCE_EXPLICIT_SIZE */ - buf[0] = version == TLS_1_3_VERSION ? + buf[0] = prot->version == TLS_1_3_VERSION ? TLS_RECORD_TYPE_DATA : record_type; /* Note that VERSION must be TLS_1_2 for both TLS1.2 and TLS1.3 */ buf[1] = TLS_1_2_VERSION_MINOR; @@ -539,18 +538,17 @@ static inline void tls_fill_prepend(struct tls_context *ctx, static inline void tls_make_aad(char *buf, size_t size, char *record_sequence, - int record_sequence_size, unsigned char record_type, - int version) + struct tls_prot_info *prot) { - if (version != TLS_1_3_VERSION) { - memcpy(buf, record_sequence, record_sequence_size); + if (prot->version != TLS_1_3_VERSION) { + memcpy(buf, record_sequence, prot->rec_seq_size); buf += 8; } else { - size += TLS_CIPHER_AES_GCM_128_TAG_SIZE; + size += prot->tag_size; } - buf[0] = version == TLS_1_3_VERSION ? + buf[0] = prot->version == TLS_1_3_VERSION ? TLS_RECORD_TYPE_DATA : record_type; buf[1] = TLS_1_2_VERSION_MAJOR; buf[2] = TLS_1_2_VERSION_MINOR; @@ -558,11 +556,11 @@ static inline void tls_make_aad(char *buf, buf[4] = size & 0xFF; } -static inline void xor_iv_with_seq(int version, char *iv, char *seq) +static inline void xor_iv_with_seq(struct tls_prot_info *prot, char *iv, char *seq) { int i; - if (version == TLS_1_3_VERSION) { + if (prot->version == TLS_1_3_VERSION) { for (i = 0; i < 8; i++) iv[i + 4] ^= seq[i]; } diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c index 54d3e161d198..6f93ad5b7200 100644 --- a/net/tls/tls_device.c +++ b/net/tls/tls_device.c @@ -327,7 +327,7 @@ static int tls_device_record_close(struct sock *sk, /* fill prepend */ tls_fill_prepend(ctx, skb_frag_address(&record->frags[0]), record->len - prot->overhead_size, - record_type, prot->version); + record_type); return ret; } diff --git a/net/tls/tls_device_fallback.c b/net/tls/tls_device_fallback.c index 28895333701e..d946817ed065 100644 --- a/net/tls/tls_device_fallback.c +++ b/net/tls/tls_device_fallback.c @@ -49,7 +49,8 @@ static int tls_enc_record(struct aead_request *aead_req, struct crypto_aead *aead, char *aad, char *iv, __be64 rcd_sn, struct scatter_walk *in, - struct scatter_walk *out, int *in_len) + struct scatter_walk *out, int *in_len, + struct tls_prot_info *prot) { unsigned char buf[TLS_HEADER_SIZE + TLS_CIPHER_AES_GCM_128_IV_SIZE]; struct scatterlist sg_in[3]; @@ -73,8 +74,7 @@ static int tls_enc_record(struct aead_request *aead_req, len -= TLS_CIPHER_AES_GCM_128_IV_SIZE; tls_make_aad(aad, len - TLS_CIPHER_AES_GCM_128_TAG_SIZE, - (char *)&rcd_sn, sizeof(rcd_sn), buf[0], - TLS_1_2_VERSION); + (char *)&rcd_sn, buf[0], prot); memcpy(iv + TLS_CIPHER_AES_GCM_128_SALT_SIZE, buf + TLS_HEADER_SIZE, TLS_CIPHER_AES_GCM_128_IV_SIZE); @@ -140,7 +140,7 @@ static struct aead_request *tls_alloc_aead_request(struct crypto_aead *aead, static int tls_enc_records(struct aead_request *aead_req, struct crypto_aead *aead, struct scatterlist *sg_in, struct scatterlist *sg_out, char *aad, char *iv, - u64 rcd_sn, int len) + u64 rcd_sn, int len, struct tls_prot_info *prot) { struct scatter_walk out, in; int rc; @@ -150,7 +150,7 @@ static int tls_enc_records(struct aead_request *aead_req, do { rc = tls_enc_record(aead_req, aead, aad, iv, - cpu_to_be64(rcd_sn), &in, &out, &len); + cpu_to_be64(rcd_sn), &in, &out, &len, prot); rcd_sn++; } while (rc == 0 && len); @@ -348,7 +348,8 @@ static struct sk_buff *tls_enc_skb(struct tls_context *tls_ctx, payload_len, sync_size, dummy_buf); if (tls_enc_records(aead_req, ctx->aead_send, sg_in, sg_out, aad, iv, - rcd_sn, sync_size + payload_len) < 0) + rcd_sn, sync_size + payload_len, + &tls_ctx->prot_info) < 0) goto free_nskb; complete_skb(nskb, skb, tcp_payload_offset); diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index 2fe9e2cf8659..6bc757a0a0ad 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -505,7 +505,7 @@ static int tls_do_encryption(struct sock *sk, memcpy(&rec->iv_data[iv_offset], tls_ctx->tx.iv, prot->iv_size + prot->salt_size); - xor_iv_with_seq(prot->version, rec->iv_data, tls_ctx->tx.rec_seq); + xor_iv_with_seq(prot, rec->iv_data, tls_ctx->tx.rec_seq); sge->offset += prot->prepend_size; sge->length -= prot->prepend_size; @@ -748,14 +748,13 @@ static int tls_push_record(struct sock *sk, int flags, sg_chain(rec->sg_aead_out, 2, &msg_en->sg.data[i]); tls_make_aad(rec->aad_space, msg_pl->sg.size + prot->tail_size, - tls_ctx->tx.rec_seq, prot->rec_seq_size, - record_type, prot->version); + tls_ctx->tx.rec_seq, record_type, prot); tls_fill_prepend(tls_ctx, page_address(sg_page(&msg_en->sg.data[i])) + msg_en->sg.data[i].offset, msg_pl->sg.size + prot->tail_size, - record_type, prot->version); + record_type); tls_ctx->pending_open_record_frags = false; @@ -1471,13 +1470,12 @@ static int decrypt_internal(struct sock *sk, struct sk_buff *skb, else memcpy(iv + iv_offset, tls_ctx->rx.iv, prot->salt_size); - xor_iv_with_seq(prot->version, iv, tls_ctx->rx.rec_seq); + xor_iv_with_seq(prot, iv, tls_ctx->rx.rec_seq); /* Prepare AAD */ tls_make_aad(aad, rxm->full_len - prot->overhead_size + prot->tail_size, - tls_ctx->rx.rec_seq, prot->rec_seq_size, - ctx->control, prot->version); + tls_ctx->rx.rec_seq, ctx->control, prot); /* Prepare sgin */ sg_init_table(sgin, n_sgin); -- cgit v1.2.3 From 923c40c4651ed8b30cbd9fbac0f0ab612216cccc Mon Sep 17 00:00:00 2001 From: Vadim Fedorenko Date: Tue, 24 Nov 2020 18:24:47 +0300 Subject: net/tls: add CHACHA20-POLY1305 specific defines and structures To provide support for ChaCha-Poly cipher we need to define specific constants and structures. Signed-off-by: Vadim Fedorenko Signed-off-by: Jakub Kicinski --- include/net/tls.h | 1 + include/uapi/linux/tls.h | 15 +++++++++++++++ 2 files changed, 16 insertions(+) (limited to 'include/net') diff --git a/include/net/tls.h b/include/net/tls.h index d04ce73e54c9..e4e9c2ae689e 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -211,6 +211,7 @@ union tls_crypto_context { union { struct tls12_crypto_info_aes_gcm_128 aes_gcm_128; struct tls12_crypto_info_aes_gcm_256 aes_gcm_256; + struct tls12_crypto_info_chacha20_poly1305 chacha20_poly1305; }; }; diff --git a/include/uapi/linux/tls.h b/include/uapi/linux/tls.h index bcd2869ed472..0d54baea1d8d 100644 --- a/include/uapi/linux/tls.h +++ b/include/uapi/linux/tls.h @@ -77,6 +77,13 @@ #define TLS_CIPHER_AES_CCM_128_TAG_SIZE 16 #define TLS_CIPHER_AES_CCM_128_REC_SEQ_SIZE 8 +#define TLS_CIPHER_CHACHA20_POLY1305 54 +#define TLS_CIPHER_CHACHA20_POLY1305_IV_SIZE 12 +#define TLS_CIPHER_CHACHA20_POLY1305_KEY_SIZE 32 +#define TLS_CIPHER_CHACHA20_POLY1305_SALT_SIZE 0 +#define TLS_CIPHER_CHACHA20_POLY1305_TAG_SIZE 16 +#define TLS_CIPHER_CHACHA20_POLY1305_REC_SEQ_SIZE 8 + #define TLS_SET_RECORD_TYPE 1 #define TLS_GET_RECORD_TYPE 2 @@ -109,6 +116,14 @@ struct tls12_crypto_info_aes_ccm_128 { unsigned char rec_seq[TLS_CIPHER_AES_CCM_128_REC_SEQ_SIZE]; }; +struct tls12_crypto_info_chacha20_poly1305 { + struct tls_crypto_info info; + unsigned char iv[TLS_CIPHER_CHACHA20_POLY1305_IV_SIZE]; + unsigned char key[TLS_CIPHER_CHACHA20_POLY1305_KEY_SIZE]; + unsigned char salt[TLS_CIPHER_CHACHA20_POLY1305_SALT_SIZE]; + unsigned char rec_seq[TLS_CIPHER_CHACHA20_POLY1305_REC_SEQ_SIZE]; +}; + enum { TLS_INFO_UNSPEC, TLS_INFO_VERSION, -- cgit v1.2.3 From a6acbe62353834dc1913d96e1ace140bc9aafca3 Mon Sep 17 00:00:00 2001 From: Vadim Fedorenko Date: Tue, 24 Nov 2020 18:24:48 +0300 Subject: net/tls: add CHACHA20-POLY1305 specific behavior RFC 7905 defines special behavior for ChaCha-Poly TLS sessions. The differences are in the calculation of nonce and the absence of explicit IV. This behavior is like TLSv1.3 partly. Signed-off-by: Vadim Fedorenko Signed-off-by: Jakub Kicinski --- include/net/tls.h | 9 ++++++--- net/tls/tls_sw.c | 6 ++++-- 2 files changed, 10 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/include/net/tls.h b/include/net/tls.h index e4e9c2ae689e..b2637edc2726 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -502,7 +502,8 @@ static inline void tls_advance_record_sn(struct sock *sk, if (tls_bigint_increment(ctx->rec_seq, prot->rec_seq_size)) tls_err_abort(sk, EBADMSG); - if (prot->version != TLS_1_3_VERSION) + if (prot->version != TLS_1_3_VERSION && + prot->cipher_type != TLS_CIPHER_CHACHA20_POLY1305) tls_bigint_increment(ctx->iv + prot->salt_size, prot->iv_size); } @@ -516,7 +517,8 @@ static inline void tls_fill_prepend(struct tls_context *ctx, size_t pkt_len, iv_size = prot->iv_size; pkt_len = plaintext_len + prot->tag_size; - if (prot->version != TLS_1_3_VERSION) { + if (prot->version != TLS_1_3_VERSION && + prot->cipher_type != TLS_CIPHER_CHACHA20_POLY1305) { pkt_len += iv_size; memcpy(buf + TLS_NONCE_OFFSET, @@ -561,7 +563,8 @@ static inline void xor_iv_with_seq(struct tls_prot_info *prot, char *iv, char *s { int i; - if (prot->version == TLS_1_3_VERSION) { + if (prot->version == TLS_1_3_VERSION || + prot->cipher_type == TLS_CIPHER_CHACHA20_POLY1305) { for (i = 0; i < 8; i++) iv[i + 4] ^= seq[i]; } diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index 6bc757a0a0ad..b4eefdb9c30a 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -1464,7 +1464,8 @@ static int decrypt_internal(struct sock *sk, struct sk_buff *skb, kfree(mem); return err; } - if (prot->version == TLS_1_3_VERSION) + if (prot->version == TLS_1_3_VERSION || + prot->cipher_type == TLS_CIPHER_CHACHA20_POLY1305) memcpy(iv + iv_offset, tls_ctx->rx.iv, crypto_aead_ivsize(ctx->aead_recv)); else @@ -2068,7 +2069,8 @@ static int tls_read_size(struct strparser *strp, struct sk_buff *skb) data_len = ((header[4] & 0xFF) | (header[3] << 8)); cipher_overhead = prot->tag_size; - if (prot->version != TLS_1_3_VERSION) + if (prot->version != TLS_1_3_VERSION && + prot->cipher_type != TLS_CIPHER_CHACHA20_POLY1305) cipher_overhead += prot->iv_size; if (data_len > TLS_MAX_PAYLOAD_SIZE + cipher_overhead + -- cgit v1.2.3 From fa6d639930ee5cd3f932cc314f3407f07a06582d Mon Sep 17 00:00:00 2001 From: wenxu Date: Wed, 25 Nov 2020 12:01:22 +0800 Subject: net/sched: act_mirred: refactor the handle of xmit This one is prepare for the next patch. Signed-off-by: wenxu Signed-off-by: Jakub Kicinski --- include/net/sch_generic.h | 5 ----- net/sched/act_mirred.c | 21 +++++++++++++++------ 2 files changed, 15 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index d8fd8676fc72..dd74f06bbb0b 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -1281,9 +1281,4 @@ void mini_qdisc_pair_init(struct mini_Qdisc_pair *miniqp, struct Qdisc *qdisc, void mini_qdisc_pair_block_init(struct mini_Qdisc_pair *miniqp, struct tcf_block *block); -static inline int skb_tc_reinsert(struct sk_buff *skb, struct tcf_result *res) -{ - return res->ingress ? netif_receive_skb(skb) : dev_queue_xmit(skb); -} - #endif diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c index e24b7e2331cd..17d00958047e 100644 --- a/net/sched/act_mirred.c +++ b/net/sched/act_mirred.c @@ -205,6 +205,18 @@ release_idr: return err; } +static int tcf_mirred_forward(bool want_ingress, struct sk_buff *skb) +{ + int err; + + if (!want_ingress) + err = dev_queue_xmit(skb); + else + err = netif_receive_skb(skb); + + return err; +} + static int tcf_mirred_act(struct sk_buff *skb, const struct tc_action *a, struct tcf_result *res) { @@ -287,18 +299,15 @@ static int tcf_mirred_act(struct sk_buff *skb, const struct tc_action *a, /* let's the caller reinsert the packet, if possible */ if (use_reinsert) { res->ingress = want_ingress; - if (skb_tc_reinsert(skb, res)) + err = tcf_mirred_forward(res->ingress, skb); + if (err) tcf_action_inc_overlimit_qstats(&m->common); __this_cpu_dec(mirred_rec_level); return TC_ACT_CONSUMED; } } - if (!want_ingress) - err = dev_queue_xmit(skb2); - else - err = netif_receive_skb(skb2); - + err = tcf_mirred_forward(want_ingress, skb2); if (err) { out: tcf_action_inc_overlimit_qstats(&m->common); -- cgit v1.2.3 From c129412f74e99b609f0a8e95fc3915af1fd40f34 Mon Sep 17 00:00:00 2001 From: wenxu Date: Wed, 25 Nov 2020 12:01:23 +0800 Subject: net/sched: sch_frag: add generic packet fragment support. Currently kernel tc subsystem can do conntrack in cat_ct. But when several fragment packets go through the act_ct, function tcf_ct_handle_fragments will defrag the packets to a big one. But the last action will redirect mirred to a device which maybe lead the reassembly big packet over the mtu of target device. This patch add support for a xmit hook to mirred, that gets executed before xmiting the packet. Then, when act_ct gets loaded, it configs that hook. The frag xmit hook maybe reused by other modules. Signed-off-by: wenxu Acked-by: Cong Wang Acked-by: Jamal Hadi Salim Signed-off-by: Jakub Kicinski --- include/net/act_api.h | 6 ++ include/net/sch_generic.h | 2 + net/sched/Makefile | 1 + net/sched/act_api.c | 16 +++++ net/sched/act_ct.c | 3 + net/sched/act_mirred.c | 2 +- net/sched/sch_frag.c | 150 ++++++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 179 insertions(+), 1 deletion(-) create mode 100644 net/sched/sch_frag.c (limited to 'include/net') diff --git a/include/net/act_api.h b/include/net/act_api.h index 87214927314a..55dab604861f 100644 --- a/include/net/act_api.h +++ b/include/net/act_api.h @@ -239,6 +239,12 @@ int tcf_action_check_ctrlact(int action, struct tcf_proto *tp, struct netlink_ext_ack *newchain); struct tcf_chain *tcf_action_set_ctrlact(struct tc_action *a, int action, struct tcf_chain *newchain); + +#ifdef CONFIG_INET +DECLARE_STATIC_KEY_FALSE(tcf_frag_xmit_count); +#endif + +int tcf_dev_queue_xmit(struct sk_buff *skb, int (*xmit)(struct sk_buff *skb)); #endif /* CONFIG_NET_CLS_ACT */ static inline void tcf_action_stats_update(struct tc_action *a, u64 bytes, diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index dd74f06bbb0b..162ed6249e8b 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -1281,4 +1281,6 @@ void mini_qdisc_pair_init(struct mini_Qdisc_pair *miniqp, struct Qdisc *qdisc, void mini_qdisc_pair_block_init(struct mini_Qdisc_pair *miniqp, struct tcf_block *block); +int sch_frag_xmit_hook(struct sk_buff *skb, int (*xmit)(struct sk_buff *skb)); + #endif diff --git a/net/sched/Makefile b/net/sched/Makefile index 66bbf9a98f9e..dd14ef413fda 100644 --- a/net/sched/Makefile +++ b/net/sched/Makefile @@ -5,6 +5,7 @@ obj-y := sch_generic.o sch_mq.o +obj-$(CONFIG_INET) += sch_frag.o obj-$(CONFIG_NET_SCHED) += sch_api.o sch_blackhole.o obj-$(CONFIG_NET_CLS) += cls_api.o obj-$(CONFIG_NET_CLS_ACT) += act_api.o diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 99db1c77426b..2e85b636b27b 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -22,6 +22,22 @@ #include #include +#ifdef CONFIG_INET +DEFINE_STATIC_KEY_FALSE(tcf_frag_xmit_count); +EXPORT_SYMBOL_GPL(tcf_frag_xmit_count); +#endif + +int tcf_dev_queue_xmit(struct sk_buff *skb, int (*xmit)(struct sk_buff *skb)) +{ +#ifdef CONFIG_INET + if (static_branch_unlikely(&tcf_frag_xmit_count)) + return sch_frag_xmit_hook(skb, xmit); +#endif + + return xmit(skb); +} +EXPORT_SYMBOL_GPL(tcf_dev_queue_xmit); + static void tcf_action_goto_chain_exec(const struct tc_action *a, struct tcf_result *res) { diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index aba3cd85f284..61092c5d646c 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -1541,6 +1541,8 @@ static int __init ct_init_module(void) if (err) goto err_register; + static_branch_inc(&tcf_frag_xmit_count); + return 0; err_register: @@ -1552,6 +1554,7 @@ err_tbl_init: static void __exit ct_cleanup_module(void) { + static_branch_dec(&tcf_frag_xmit_count); tcf_unregister_action(&act_ct_ops, &ct_net_ops); tcf_ct_flow_tables_uninit(); destroy_workqueue(act_ct_wq); diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c index 17d00958047e..7153c67f641e 100644 --- a/net/sched/act_mirred.c +++ b/net/sched/act_mirred.c @@ -210,7 +210,7 @@ static int tcf_mirred_forward(bool want_ingress, struct sk_buff *skb) int err; if (!want_ingress) - err = dev_queue_xmit(skb); + err = tcf_dev_queue_xmit(skb, dev_queue_xmit); else err = netif_receive_skb(skb); diff --git a/net/sched/sch_frag.c b/net/sched/sch_frag.c new file mode 100644 index 000000000000..e1e77d3fb6c0 --- /dev/null +++ b/net/sched/sch_frag.c @@ -0,0 +1,150 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +#include +#include +#include +#include +#include + +struct sch_frag_data { + unsigned long dst; + struct qdisc_skb_cb cb; + __be16 inner_protocol; + u16 vlan_tci; + __be16 vlan_proto; + unsigned int l2_len; + u8 l2_data[VLAN_ETH_HLEN]; + int (*xmit)(struct sk_buff *skb); +}; + +static DEFINE_PER_CPU(struct sch_frag_data, sch_frag_data_storage); + +static int sch_frag_xmit(struct net *net, struct sock *sk, struct sk_buff *skb) +{ + struct sch_frag_data *data = this_cpu_ptr(&sch_frag_data_storage); + + if (skb_cow_head(skb, data->l2_len) < 0) { + kfree_skb(skb); + return -ENOMEM; + } + + __skb_dst_copy(skb, data->dst); + *qdisc_skb_cb(skb) = data->cb; + skb->inner_protocol = data->inner_protocol; + if (data->vlan_tci & VLAN_CFI_MASK) + __vlan_hwaccel_put_tag(skb, data->vlan_proto, + data->vlan_tci & ~VLAN_CFI_MASK); + else + __vlan_hwaccel_clear_tag(skb); + + /* Reconstruct the MAC header. */ + skb_push(skb, data->l2_len); + memcpy(skb->data, &data->l2_data, data->l2_len); + skb_postpush_rcsum(skb, skb->data, data->l2_len); + skb_reset_mac_header(skb); + + return data->xmit(skb); +} + +static void sch_frag_prepare_frag(struct sk_buff *skb, + int (*xmit)(struct sk_buff *skb)) +{ + unsigned int hlen = skb_network_offset(skb); + struct sch_frag_data *data; + + data = this_cpu_ptr(&sch_frag_data_storage); + data->dst = skb->_skb_refdst; + data->cb = *qdisc_skb_cb(skb); + data->xmit = xmit; + data->inner_protocol = skb->inner_protocol; + if (skb_vlan_tag_present(skb)) + data->vlan_tci = skb_vlan_tag_get(skb) | VLAN_CFI_MASK; + else + data->vlan_tci = 0; + data->vlan_proto = skb->vlan_proto; + data->l2_len = hlen; + memcpy(&data->l2_data, skb->data, hlen); + + memset(IPCB(skb), 0, sizeof(struct inet_skb_parm)); + skb_pull(skb, hlen); +} + +static unsigned int +sch_frag_dst_get_mtu(const struct dst_entry *dst) +{ + return dst->dev->mtu; +} + +static struct dst_ops sch_frag_dst_ops = { + .family = AF_UNSPEC, + .mtu = sch_frag_dst_get_mtu, +}; + +static int sch_fragment(struct net *net, struct sk_buff *skb, + u16 mru, int (*xmit)(struct sk_buff *skb)) +{ + int ret = -1; + + if (skb_network_offset(skb) > VLAN_ETH_HLEN) { + net_warn_ratelimited("L2 header too long to fragment\n"); + goto err; + } + + if (skb_protocol(skb, true) == htons(ETH_P_IP)) { + struct dst_entry sch_frag_dst; + unsigned long orig_dst; + + sch_frag_prepare_frag(skb, xmit); + dst_init(&sch_frag_dst, &sch_frag_dst_ops, NULL, 1, + DST_OBSOLETE_NONE, DST_NOCOUNT); + sch_frag_dst.dev = skb->dev; + + orig_dst = skb->_skb_refdst; + skb_dst_set_noref(skb, &sch_frag_dst); + IPCB(skb)->frag_max_size = mru; + + ret = ip_do_fragment(net, skb->sk, skb, sch_frag_xmit); + refdst_drop(orig_dst); + } else if (skb_protocol(skb, true) == htons(ETH_P_IPV6)) { + unsigned long orig_dst; + struct rt6_info sch_frag_rt; + + sch_frag_prepare_frag(skb, xmit); + memset(&sch_frag_rt, 0, sizeof(sch_frag_rt)); + dst_init(&sch_frag_rt.dst, &sch_frag_dst_ops, NULL, 1, + DST_OBSOLETE_NONE, DST_NOCOUNT); + sch_frag_rt.dst.dev = skb->dev; + + orig_dst = skb->_skb_refdst; + skb_dst_set_noref(skb, &sch_frag_rt.dst); + IP6CB(skb)->frag_max_size = mru; + + ret = ipv6_stub->ipv6_fragment(net, skb->sk, skb, + sch_frag_xmit); + refdst_drop(orig_dst); + } else { + net_warn_ratelimited("Fail frag %s: eth=%x, MRU=%d, MTU=%d\n", + netdev_name(skb->dev), + ntohs(skb_protocol(skb, true)), mru, + skb->dev->mtu); + goto err; + } + + return ret; +err: + kfree_skb(skb); + return ret; +} + +int sch_frag_xmit_hook(struct sk_buff *skb, int (*xmit)(struct sk_buff *skb)) +{ + u16 mru = qdisc_skb_cb(skb)->mru; + int err; + + if (mru && skb->len > mru + skb->dev->hard_header_len) + err = sch_fragment(dev_net(skb->dev), skb, mru, xmit); + else + err = xmit(skb); + + return err; +} +EXPORT_SYMBOL_GPL(sch_frag_xmit_hook); -- cgit v1.2.3 From 7fd3253a7de6a317a0683f83739479fb880bffc8 Mon Sep 17 00:00:00 2001 From: Björn Töpel Date: Mon, 30 Nov 2020 19:51:56 +0100 Subject: net: Introduce preferred busy-polling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The existing busy-polling mode, enabled by the SO_BUSY_POLL socket option or system-wide using the /proc/sys/net/core/busy_read knob, is an opportunistic. That means that if the NAPI context is not scheduled, it will poll it. If, after busy-polling, the budget is exceeded the busy-polling logic will schedule the NAPI onto the regular softirq handling. One implication of the behavior above is that a busy/heavy loaded NAPI context will never enter/allow for busy-polling. Some applications prefer that most NAPI processing would be done by busy-polling. This series adds a new socket option, SO_PREFER_BUSY_POLL, that works in concert with the napi_defer_hard_irqs and gro_flush_timeout knobs. The napi_defer_hard_irqs and gro_flush_timeout knobs were introduced in commit 6f8b12d661d0 ("net: napi: add hard irqs deferral feature"), and allows for a user to defer interrupts to be enabled and instead schedule the NAPI context from a watchdog timer. When a user enables the SO_PREFER_BUSY_POLL, again with the other knobs enabled, and the NAPI context is being processed by a softirq, the softirq NAPI processing will exit early to allow the busy-polling to be performed. If the application stops performing busy-polling via a system call, the watchdog timer defined by gro_flush_timeout will timeout, and regular softirq handling will resume. In summary; Heavy traffic applications that prefer busy-polling over softirq processing should use this option. Example usage: $ echo 2 | sudo tee /sys/class/net/ens785f1/napi_defer_hard_irqs $ echo 200000 | sudo tee /sys/class/net/ens785f1/gro_flush_timeout Note that the timeout should be larger than the userspace processing window, otherwise the watchdog will timeout and fall back to regular softirq processing. Enable the SO_BUSY_POLL/SO_PREFER_BUSY_POLL options on your socket. Signed-off-by: Björn Töpel Signed-off-by: Daniel Borkmann Reviewed-by: Jakub Kicinski Link: https://lore.kernel.org/bpf/20201130185205.196029-2-bjorn.topel@gmail.com --- arch/alpha/include/uapi/asm/socket.h | 2 + arch/mips/include/uapi/asm/socket.h | 2 + arch/parisc/include/uapi/asm/socket.h | 2 + arch/sparc/include/uapi/asm/socket.h | 2 + fs/eventpoll.c | 2 +- include/linux/netdevice.h | 35 +++++++++------- include/net/busy_poll.h | 5 ++- include/net/sock.h | 4 ++ include/uapi/asm-generic/socket.h | 2 + net/core/dev.c | 78 ++++++++++++++++++++++++++++------- net/core/sock.c | 9 ++++ 11 files changed, 111 insertions(+), 32 deletions(-) (limited to 'include/net') diff --git a/arch/alpha/include/uapi/asm/socket.h b/arch/alpha/include/uapi/asm/socket.h index de6c4df61082..538359642554 100644 --- a/arch/alpha/include/uapi/asm/socket.h +++ b/arch/alpha/include/uapi/asm/socket.h @@ -124,6 +124,8 @@ #define SO_DETACH_REUSEPORT_BPF 68 +#define SO_PREFER_BUSY_POLL 69 + #if !defined(__KERNEL__) #if __BITS_PER_LONG == 64 diff --git a/arch/mips/include/uapi/asm/socket.h b/arch/mips/include/uapi/asm/socket.h index d0a9ed2ca2d6..e406e73b5e6e 100644 --- a/arch/mips/include/uapi/asm/socket.h +++ b/arch/mips/include/uapi/asm/socket.h @@ -135,6 +135,8 @@ #define SO_DETACH_REUSEPORT_BPF 68 +#define SO_PREFER_BUSY_POLL 69 + #if !defined(__KERNEL__) #if __BITS_PER_LONG == 64 diff --git a/arch/parisc/include/uapi/asm/socket.h b/arch/parisc/include/uapi/asm/socket.h index 10173c32195e..1bc46200889d 100644 --- a/arch/parisc/include/uapi/asm/socket.h +++ b/arch/parisc/include/uapi/asm/socket.h @@ -116,6 +116,8 @@ #define SO_DETACH_REUSEPORT_BPF 0x4042 +#define SO_PREFER_BUSY_POLL 0x4043 + #if !defined(__KERNEL__) #if __BITS_PER_LONG == 64 diff --git a/arch/sparc/include/uapi/asm/socket.h b/arch/sparc/include/uapi/asm/socket.h index 8029b681fc7c..99688cf673a4 100644 --- a/arch/sparc/include/uapi/asm/socket.h +++ b/arch/sparc/include/uapi/asm/socket.h @@ -117,6 +117,8 @@ #define SO_DETACH_REUSEPORT_BPF 0x0047 +#define SO_PREFER_BUSY_POLL 0x0048 + #if !defined(__KERNEL__) diff --git a/fs/eventpoll.c b/fs/eventpoll.c index 4df61129566d..e11fab3a0b9e 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -397,7 +397,7 @@ static void ep_busy_loop(struct eventpoll *ep, int nonblock) unsigned int napi_id = READ_ONCE(ep->napi_id); if ((napi_id >= MIN_NAPI_ID) && net_busy_loop_on()) - napi_busy_loop(napi_id, nonblock ? NULL : ep_busy_loop_end, ep); + napi_busy_loop(napi_id, nonblock ? NULL : ep_busy_loop_end, ep, false); } static inline void ep_reset_busy_poll_napi_id(struct eventpoll *ep) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 7ce648a564f7..52d1cc2bd8a7 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -350,23 +350,25 @@ struct napi_struct { }; enum { - NAPI_STATE_SCHED, /* Poll is scheduled */ - NAPI_STATE_MISSED, /* reschedule a napi */ - NAPI_STATE_DISABLE, /* Disable pending */ - NAPI_STATE_NPSVC, /* Netpoll - don't dequeue from poll_list */ - NAPI_STATE_LISTED, /* NAPI added to system lists */ - NAPI_STATE_NO_BUSY_POLL,/* Do not add in napi_hash, no busy polling */ - NAPI_STATE_IN_BUSY_POLL,/* sk_busy_loop() owns this NAPI */ + NAPI_STATE_SCHED, /* Poll is scheduled */ + NAPI_STATE_MISSED, /* reschedule a napi */ + NAPI_STATE_DISABLE, /* Disable pending */ + NAPI_STATE_NPSVC, /* Netpoll - don't dequeue from poll_list */ + NAPI_STATE_LISTED, /* NAPI added to system lists */ + NAPI_STATE_NO_BUSY_POLL, /* Do not add in napi_hash, no busy polling */ + NAPI_STATE_IN_BUSY_POLL, /* sk_busy_loop() owns this NAPI */ + NAPI_STATE_PREFER_BUSY_POLL, /* prefer busy-polling over softirq processing*/ }; enum { - NAPIF_STATE_SCHED = BIT(NAPI_STATE_SCHED), - NAPIF_STATE_MISSED = BIT(NAPI_STATE_MISSED), - NAPIF_STATE_DISABLE = BIT(NAPI_STATE_DISABLE), - NAPIF_STATE_NPSVC = BIT(NAPI_STATE_NPSVC), - NAPIF_STATE_LISTED = BIT(NAPI_STATE_LISTED), - NAPIF_STATE_NO_BUSY_POLL = BIT(NAPI_STATE_NO_BUSY_POLL), - NAPIF_STATE_IN_BUSY_POLL = BIT(NAPI_STATE_IN_BUSY_POLL), + NAPIF_STATE_SCHED = BIT(NAPI_STATE_SCHED), + NAPIF_STATE_MISSED = BIT(NAPI_STATE_MISSED), + NAPIF_STATE_DISABLE = BIT(NAPI_STATE_DISABLE), + NAPIF_STATE_NPSVC = BIT(NAPI_STATE_NPSVC), + NAPIF_STATE_LISTED = BIT(NAPI_STATE_LISTED), + NAPIF_STATE_NO_BUSY_POLL = BIT(NAPI_STATE_NO_BUSY_POLL), + NAPIF_STATE_IN_BUSY_POLL = BIT(NAPI_STATE_IN_BUSY_POLL), + NAPIF_STATE_PREFER_BUSY_POLL = BIT(NAPI_STATE_PREFER_BUSY_POLL), }; enum gro_result { @@ -437,6 +439,11 @@ static inline bool napi_disable_pending(struct napi_struct *n) return test_bit(NAPI_STATE_DISABLE, &n->state); } +static inline bool napi_prefer_busy_poll(struct napi_struct *n) +{ + return test_bit(NAPI_STATE_PREFER_BUSY_POLL, &n->state); +} + bool napi_schedule_prep(struct napi_struct *n); /** diff --git a/include/net/busy_poll.h b/include/net/busy_poll.h index b001fa91c14e..0292b8353d7e 100644 --- a/include/net/busy_poll.h +++ b/include/net/busy_poll.h @@ -43,7 +43,7 @@ bool sk_busy_loop_end(void *p, unsigned long start_time); void napi_busy_loop(unsigned int napi_id, bool (*loop_end)(void *, unsigned long), - void *loop_end_arg); + void *loop_end_arg, bool prefer_busy_poll); #else /* CONFIG_NET_RX_BUSY_POLL */ static inline unsigned long net_busy_loop_on(void) @@ -105,7 +105,8 @@ static inline void sk_busy_loop(struct sock *sk, int nonblock) unsigned int napi_id = READ_ONCE(sk->sk_napi_id); if (napi_id >= MIN_NAPI_ID) - napi_busy_loop(napi_id, nonblock ? NULL : sk_busy_loop_end, sk); + napi_busy_loop(napi_id, nonblock ? NULL : sk_busy_loop_end, sk, + READ_ONCE(sk->sk_prefer_busy_poll)); #endif } diff --git a/include/net/sock.h b/include/net/sock.h index a5c6ae78df77..d49b89b071b6 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -301,6 +301,7 @@ struct bpf_local_storage; * @sk_ack_backlog: current listen backlog * @sk_max_ack_backlog: listen backlog set in listen() * @sk_uid: user id of owner + * @sk_prefer_busy_poll: prefer busypolling over softirq processing * @sk_priority: %SO_PRIORITY setting * @sk_type: socket type (%SOCK_STREAM, etc) * @sk_protocol: which protocol this socket belongs in this network family @@ -479,6 +480,9 @@ struct sock { u32 sk_ack_backlog; u32 sk_max_ack_backlog; kuid_t sk_uid; +#ifdef CONFIG_NET_RX_BUSY_POLL + u8 sk_prefer_busy_poll; +#endif struct pid *sk_peer_pid; const struct cred *sk_peer_cred; long sk_rcvtimeo; diff --git a/include/uapi/asm-generic/socket.h b/include/uapi/asm-generic/socket.h index 77f7c1638eb1..7dd02408b7ce 100644 --- a/include/uapi/asm-generic/socket.h +++ b/include/uapi/asm-generic/socket.h @@ -119,6 +119,8 @@ #define SO_DETACH_REUSEPORT_BPF 68 +#define SO_PREFER_BUSY_POLL 69 + #if !defined(__KERNEL__) #if __BITS_PER_LONG == 64 || (defined(__x86_64__) && defined(__ILP32__)) diff --git a/net/core/dev.c b/net/core/dev.c index 60d325bda0d7..6f8d2cffb7c5 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6458,7 +6458,8 @@ bool napi_complete_done(struct napi_struct *n, int work_done) WARN_ON_ONCE(!(val & NAPIF_STATE_SCHED)); - new = val & ~(NAPIF_STATE_MISSED | NAPIF_STATE_SCHED); + new = val & ~(NAPIF_STATE_MISSED | NAPIF_STATE_SCHED | + NAPIF_STATE_PREFER_BUSY_POLL); /* If STATE_MISSED was set, leave STATE_SCHED set, * because we will call napi->poll() one more time. @@ -6497,8 +6498,29 @@ static struct napi_struct *napi_by_id(unsigned int napi_id) #define BUSY_POLL_BUDGET 8 -static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock) +static void __busy_poll_stop(struct napi_struct *napi, bool skip_schedule) { + if (!skip_schedule) { + gro_normal_list(napi); + __napi_schedule(napi); + return; + } + + if (napi->gro_bitmask) { + /* flush too old packets + * If HZ < 1000, flush all packets. + */ + napi_gro_flush(napi, HZ >= 1000); + } + + gro_normal_list(napi); + clear_bit(NAPI_STATE_SCHED, &napi->state); +} + +static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock, bool prefer_busy_poll) +{ + bool skip_schedule = false; + unsigned long timeout; int rc; /* Busy polling means there is a high chance device driver hard irq @@ -6515,6 +6537,15 @@ static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock) local_bh_disable(); + if (prefer_busy_poll) { + napi->defer_hard_irqs_count = READ_ONCE(napi->dev->napi_defer_hard_irqs); + timeout = READ_ONCE(napi->dev->gro_flush_timeout); + if (napi->defer_hard_irqs_count && timeout) { + hrtimer_start(&napi->timer, ns_to_ktime(timeout), HRTIMER_MODE_REL_PINNED); + skip_schedule = true; + } + } + /* All we really want here is to re-enable device interrupts. * Ideally, a new ndo_busy_poll_stop() could avoid another round. */ @@ -6525,19 +6556,14 @@ static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock) */ trace_napi_poll(napi, rc, BUSY_POLL_BUDGET); netpoll_poll_unlock(have_poll_lock); - if (rc == BUSY_POLL_BUDGET) { - /* As the whole budget was spent, we still own the napi so can - * safely handle the rx_list. - */ - gro_normal_list(napi); - __napi_schedule(napi); - } + if (rc == BUSY_POLL_BUDGET) + __busy_poll_stop(napi, skip_schedule); local_bh_enable(); } void napi_busy_loop(unsigned int napi_id, bool (*loop_end)(void *, unsigned long), - void *loop_end_arg) + void *loop_end_arg, bool prefer_busy_poll) { unsigned long start_time = loop_end ? busy_loop_current_time() : 0; int (*napi_poll)(struct napi_struct *napi, int budget); @@ -6565,12 +6591,18 @@ restart: * we avoid dirtying napi->state as much as we can. */ if (val & (NAPIF_STATE_DISABLE | NAPIF_STATE_SCHED | - NAPIF_STATE_IN_BUSY_POLL)) + NAPIF_STATE_IN_BUSY_POLL)) { + if (prefer_busy_poll) + set_bit(NAPI_STATE_PREFER_BUSY_POLL, &napi->state); goto count; + } if (cmpxchg(&napi->state, val, val | NAPIF_STATE_IN_BUSY_POLL | - NAPIF_STATE_SCHED) != val) + NAPIF_STATE_SCHED) != val) { + if (prefer_busy_poll) + set_bit(NAPI_STATE_PREFER_BUSY_POLL, &napi->state); goto count; + } have_poll_lock = netpoll_poll_lock(napi); napi_poll = napi->poll; } @@ -6588,7 +6620,7 @@ count: if (unlikely(need_resched())) { if (napi_poll) - busy_poll_stop(napi, have_poll_lock); + busy_poll_stop(napi, have_poll_lock, prefer_busy_poll); preempt_enable(); rcu_read_unlock(); cond_resched(); @@ -6599,7 +6631,7 @@ count: cpu_relax(); } if (napi_poll) - busy_poll_stop(napi, have_poll_lock); + busy_poll_stop(napi, have_poll_lock, prefer_busy_poll); preempt_enable(); out: rcu_read_unlock(); @@ -6650,8 +6682,10 @@ static enum hrtimer_restart napi_watchdog(struct hrtimer *timer) * NAPI_STATE_MISSED, since we do not react to a device IRQ. */ if (!napi_disable_pending(napi) && - !test_and_set_bit(NAPI_STATE_SCHED, &napi->state)) + !test_and_set_bit(NAPI_STATE_SCHED, &napi->state)) { + clear_bit(NAPI_STATE_PREFER_BUSY_POLL, &napi->state); __napi_schedule_irqoff(napi); + } return HRTIMER_NORESTART; } @@ -6709,6 +6743,7 @@ void napi_disable(struct napi_struct *n) hrtimer_cancel(&n->timer); + clear_bit(NAPI_STATE_PREFER_BUSY_POLL, &n->state); clear_bit(NAPI_STATE_DISABLE, &n->state); } EXPORT_SYMBOL(napi_disable); @@ -6781,6 +6816,19 @@ static int napi_poll(struct napi_struct *n, struct list_head *repoll) goto out_unlock; } + /* The NAPI context has more processing work, but busy-polling + * is preferred. Exit early. + */ + if (napi_prefer_busy_poll(n)) { + if (napi_complete_done(n, work)) { + /* If timeout is not set, we need to make sure + * that the NAPI is re-scheduled. + */ + napi_schedule(n); + } + goto out_unlock; + } + if (n->gro_bitmask) { /* flush too old packets * If HZ < 1000, flush all packets. diff --git a/net/core/sock.c b/net/core/sock.c index 727ea1cc633c..e05f2e52b5a8 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1159,6 +1159,12 @@ set_sndbuf: sk->sk_ll_usec = val; } break; + case SO_PREFER_BUSY_POLL: + if (valbool && !capable(CAP_NET_ADMIN)) + ret = -EPERM; + else + WRITE_ONCE(sk->sk_prefer_busy_poll, valbool); + break; #endif case SO_MAX_PACING_RATE: @@ -1523,6 +1529,9 @@ int sock_getsockopt(struct socket *sock, int level, int optname, case SO_BUSY_POLL: v.val = sk->sk_ll_usec; break; + case SO_PREFER_BUSY_POLL: + v.val = READ_ONCE(sk->sk_prefer_busy_poll); + break; #endif case SO_MAX_PACING_RATE: -- cgit v1.2.3 From 7c951cafc0cb2e575f1d58677b95ac387ac0a5bd Mon Sep 17 00:00:00 2001 From: Björn Töpel Date: Mon, 30 Nov 2020 19:51:57 +0100 Subject: net: Add SO_BUSY_POLL_BUDGET socket option MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This option lets a user set a per socket NAPI budget for busy-polling. If the options is not set, it will use the default of 8. Signed-off-by: Björn Töpel Signed-off-by: Daniel Borkmann Reviewed-by: Jakub Kicinski Link: https://lore.kernel.org/bpf/20201130185205.196029-3-bjorn.topel@gmail.com --- arch/alpha/include/uapi/asm/socket.h | 1 + arch/mips/include/uapi/asm/socket.h | 1 + arch/parisc/include/uapi/asm/socket.h | 1 + arch/sparc/include/uapi/asm/socket.h | 1 + fs/eventpoll.c | 3 ++- include/net/busy_poll.h | 7 +++++-- include/net/sock.h | 2 ++ include/uapi/asm-generic/socket.h | 1 + net/core/dev.c | 21 ++++++++++----------- net/core/sock.c | 10 ++++++++++ 10 files changed, 34 insertions(+), 14 deletions(-) (limited to 'include/net') diff --git a/arch/alpha/include/uapi/asm/socket.h b/arch/alpha/include/uapi/asm/socket.h index 538359642554..57420356ce4c 100644 --- a/arch/alpha/include/uapi/asm/socket.h +++ b/arch/alpha/include/uapi/asm/socket.h @@ -125,6 +125,7 @@ #define SO_DETACH_REUSEPORT_BPF 68 #define SO_PREFER_BUSY_POLL 69 +#define SO_BUSY_POLL_BUDGET 70 #if !defined(__KERNEL__) diff --git a/arch/mips/include/uapi/asm/socket.h b/arch/mips/include/uapi/asm/socket.h index e406e73b5e6e..2d949969313b 100644 --- a/arch/mips/include/uapi/asm/socket.h +++ b/arch/mips/include/uapi/asm/socket.h @@ -136,6 +136,7 @@ #define SO_DETACH_REUSEPORT_BPF 68 #define SO_PREFER_BUSY_POLL 69 +#define SO_BUSY_POLL_BUDGET 70 #if !defined(__KERNEL__) diff --git a/arch/parisc/include/uapi/asm/socket.h b/arch/parisc/include/uapi/asm/socket.h index 1bc46200889d..f60904329bbc 100644 --- a/arch/parisc/include/uapi/asm/socket.h +++ b/arch/parisc/include/uapi/asm/socket.h @@ -117,6 +117,7 @@ #define SO_DETACH_REUSEPORT_BPF 0x4042 #define SO_PREFER_BUSY_POLL 0x4043 +#define SO_BUSY_POLL_BUDGET 0x4044 #if !defined(__KERNEL__) diff --git a/arch/sparc/include/uapi/asm/socket.h b/arch/sparc/include/uapi/asm/socket.h index 99688cf673a4..848a22fbac20 100644 --- a/arch/sparc/include/uapi/asm/socket.h +++ b/arch/sparc/include/uapi/asm/socket.h @@ -118,6 +118,7 @@ #define SO_DETACH_REUSEPORT_BPF 0x0047 #define SO_PREFER_BUSY_POLL 0x0048 +#define SO_BUSY_POLL_BUDGET 0x0049 #if !defined(__KERNEL__) diff --git a/fs/eventpoll.c b/fs/eventpoll.c index e11fab3a0b9e..73c346e503d7 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -397,7 +397,8 @@ static void ep_busy_loop(struct eventpoll *ep, int nonblock) unsigned int napi_id = READ_ONCE(ep->napi_id); if ((napi_id >= MIN_NAPI_ID) && net_busy_loop_on()) - napi_busy_loop(napi_id, nonblock ? NULL : ep_busy_loop_end, ep, false); + napi_busy_loop(napi_id, nonblock ? NULL : ep_busy_loop_end, ep, false, + BUSY_POLL_BUDGET); } static inline void ep_reset_busy_poll_napi_id(struct eventpoll *ep) diff --git a/include/net/busy_poll.h b/include/net/busy_poll.h index 0292b8353d7e..2f8f51807b83 100644 --- a/include/net/busy_poll.h +++ b/include/net/busy_poll.h @@ -23,6 +23,8 @@ */ #define MIN_NAPI_ID ((unsigned int)(NR_CPUS + 1)) +#define BUSY_POLL_BUDGET 8 + #ifdef CONFIG_NET_RX_BUSY_POLL struct napi_struct; @@ -43,7 +45,7 @@ bool sk_busy_loop_end(void *p, unsigned long start_time); void napi_busy_loop(unsigned int napi_id, bool (*loop_end)(void *, unsigned long), - void *loop_end_arg, bool prefer_busy_poll); + void *loop_end_arg, bool prefer_busy_poll, u16 budget); #else /* CONFIG_NET_RX_BUSY_POLL */ static inline unsigned long net_busy_loop_on(void) @@ -106,7 +108,8 @@ static inline void sk_busy_loop(struct sock *sk, int nonblock) if (napi_id >= MIN_NAPI_ID) napi_busy_loop(napi_id, nonblock ? NULL : sk_busy_loop_end, sk, - READ_ONCE(sk->sk_prefer_busy_poll)); + READ_ONCE(sk->sk_prefer_busy_poll), + READ_ONCE(sk->sk_busy_poll_budget) ?: BUSY_POLL_BUDGET); #endif } diff --git a/include/net/sock.h b/include/net/sock.h index d49b89b071b6..77ba2c2737db 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -302,6 +302,7 @@ struct bpf_local_storage; * @sk_max_ack_backlog: listen backlog set in listen() * @sk_uid: user id of owner * @sk_prefer_busy_poll: prefer busypolling over softirq processing + * @sk_busy_poll_budget: napi processing budget when busypolling * @sk_priority: %SO_PRIORITY setting * @sk_type: socket type (%SOCK_STREAM, etc) * @sk_protocol: which protocol this socket belongs in this network family @@ -482,6 +483,7 @@ struct sock { kuid_t sk_uid; #ifdef CONFIG_NET_RX_BUSY_POLL u8 sk_prefer_busy_poll; + u16 sk_busy_poll_budget; #endif struct pid *sk_peer_pid; const struct cred *sk_peer_cred; diff --git a/include/uapi/asm-generic/socket.h b/include/uapi/asm-generic/socket.h index 7dd02408b7ce..4dcd13d097a9 100644 --- a/include/uapi/asm-generic/socket.h +++ b/include/uapi/asm-generic/socket.h @@ -120,6 +120,7 @@ #define SO_DETACH_REUSEPORT_BPF 68 #define SO_PREFER_BUSY_POLL 69 +#define SO_BUSY_POLL_BUDGET 70 #if !defined(__KERNEL__) diff --git a/net/core/dev.c b/net/core/dev.c index 6f8d2cffb7c5..7a1e5936c67f 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6496,8 +6496,6 @@ static struct napi_struct *napi_by_id(unsigned int napi_id) #if defined(CONFIG_NET_RX_BUSY_POLL) -#define BUSY_POLL_BUDGET 8 - static void __busy_poll_stop(struct napi_struct *napi, bool skip_schedule) { if (!skip_schedule) { @@ -6517,7 +6515,8 @@ static void __busy_poll_stop(struct napi_struct *napi, bool skip_schedule) clear_bit(NAPI_STATE_SCHED, &napi->state); } -static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock, bool prefer_busy_poll) +static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock, bool prefer_busy_poll, + u16 budget) { bool skip_schedule = false; unsigned long timeout; @@ -6549,21 +6548,21 @@ static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock, bool /* All we really want here is to re-enable device interrupts. * Ideally, a new ndo_busy_poll_stop() could avoid another round. */ - rc = napi->poll(napi, BUSY_POLL_BUDGET); + rc = napi->poll(napi, budget); /* We can't gro_normal_list() here, because napi->poll() might have * rearmed the napi (napi_complete_done()) in which case it could * already be running on another CPU. */ - trace_napi_poll(napi, rc, BUSY_POLL_BUDGET); + trace_napi_poll(napi, rc, budget); netpoll_poll_unlock(have_poll_lock); - if (rc == BUSY_POLL_BUDGET) + if (rc == budget) __busy_poll_stop(napi, skip_schedule); local_bh_enable(); } void napi_busy_loop(unsigned int napi_id, bool (*loop_end)(void *, unsigned long), - void *loop_end_arg, bool prefer_busy_poll) + void *loop_end_arg, bool prefer_busy_poll, u16 budget) { unsigned long start_time = loop_end ? busy_loop_current_time() : 0; int (*napi_poll)(struct napi_struct *napi, int budget); @@ -6606,8 +6605,8 @@ restart: have_poll_lock = netpoll_poll_lock(napi); napi_poll = napi->poll; } - work = napi_poll(napi, BUSY_POLL_BUDGET); - trace_napi_poll(napi, work, BUSY_POLL_BUDGET); + work = napi_poll(napi, budget); + trace_napi_poll(napi, work, budget); gro_normal_list(napi); count: if (work > 0) @@ -6620,7 +6619,7 @@ count: if (unlikely(need_resched())) { if (napi_poll) - busy_poll_stop(napi, have_poll_lock, prefer_busy_poll); + busy_poll_stop(napi, have_poll_lock, prefer_busy_poll, budget); preempt_enable(); rcu_read_unlock(); cond_resched(); @@ -6631,7 +6630,7 @@ count: cpu_relax(); } if (napi_poll) - busy_poll_stop(napi, have_poll_lock, prefer_busy_poll); + busy_poll_stop(napi, have_poll_lock, prefer_busy_poll, budget); preempt_enable(); out: rcu_read_unlock(); diff --git a/net/core/sock.c b/net/core/sock.c index e05f2e52b5a8..d422a6808405 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1165,6 +1165,16 @@ set_sndbuf: else WRITE_ONCE(sk->sk_prefer_busy_poll, valbool); break; + case SO_BUSY_POLL_BUDGET: + if (val > READ_ONCE(sk->sk_busy_poll_budget) && !capable(CAP_NET_ADMIN)) { + ret = -EPERM; + } else { + if (val < 0 || val > U16_MAX) + ret = -EINVAL; + else + WRITE_ONCE(sk->sk_busy_poll_budget, val); + } + break; #endif case SO_MAX_PACING_RATE: -- cgit v1.2.3 From b02e5a0ebb172c8276cea3151942aac681f7a4a6 Mon Sep 17 00:00:00 2001 From: Björn Töpel Date: Mon, 30 Nov 2020 19:52:01 +0100 Subject: xsk: Propagate napi_id to XDP socket Rx path MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add napi_id to the xdp_rxq_info structure, and make sure the XDP socket pick up the napi_id in the Rx path. The napi_id is used to find the corresponding NAPI structure for socket busy polling. Signed-off-by: Björn Töpel Signed-off-by: Daniel Borkmann Acked-by: Ilias Apalodimas Acked-by: Michael S. Tsirkin Acked-by: Tariq Toukan Link: https://lore.kernel.org/bpf/20201130185205.196029-7-bjorn.topel@gmail.com --- drivers/net/ethernet/amazon/ena/ena_netdev.c | 2 +- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 +- drivers/net/ethernet/cavium/thunder/nicvf_queues.c | 2 +- drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 2 +- drivers/net/ethernet/intel/i40e/i40e_txrx.c | 2 +- drivers/net/ethernet/intel/ice/ice_base.c | 4 ++-- drivers/net/ethernet/intel/ice/ice_txrx.c | 2 +- drivers/net/ethernet/intel/igb/igb_main.c | 2 +- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 2 +- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 2 +- drivers/net/ethernet/marvell/mvneta.c | 2 +- drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 4 ++-- drivers/net/ethernet/mellanox/mlx4/en_rx.c | 2 +- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 2 +- drivers/net/ethernet/netronome/nfp/nfp_net_common.c | 2 +- drivers/net/ethernet/qlogic/qede/qede_main.c | 2 +- drivers/net/ethernet/sfc/rx_common.c | 2 +- drivers/net/ethernet/socionext/netsec.c | 2 +- drivers/net/ethernet/ti/cpsw_priv.c | 2 +- drivers/net/hyperv/netvsc.c | 2 +- drivers/net/tun.c | 2 +- drivers/net/veth.c | 12 ++++++++---- drivers/net/virtio_net.c | 2 +- drivers/net/xen-netfront.c | 2 +- include/net/busy_poll.h | 19 +++++++++++++++---- include/net/xdp.h | 3 ++- net/core/dev.c | 2 +- net/core/xdp.c | 3 ++- net/xdp/xsk.c | 1 + 29 files changed, 54 insertions(+), 36 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index e8131dadc22c..6ad59f0068f6 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -416,7 +416,7 @@ static int ena_xdp_register_rxq_info(struct ena_ring *rx_ring) { int rc; - rc = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev, rx_ring->qid); + rc = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev, rx_ring->qid, 0); if (rc) { netif_err(rx_ring->adapter, ifup, rx_ring->netdev, diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 7975f59735d6..725d929eddb1 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -2884,7 +2884,7 @@ static int bnxt_alloc_rx_rings(struct bnxt *bp) if (rc) return rc; - rc = xdp_rxq_info_reg(&rxr->xdp_rxq, bp->dev, i); + rc = xdp_rxq_info_reg(&rxr->xdp_rxq, bp->dev, i, 0); if (rc < 0) return rc; diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_queues.c b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c index 7a141ce32e86..f782e6af45e9 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_queues.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c @@ -770,7 +770,7 @@ static void nicvf_rcv_queue_config(struct nicvf *nic, struct queue_set *qs, rq->caching = 1; /* Driver have no proper error path for failed XDP RX-queue info reg */ - WARN_ON(xdp_rxq_info_reg(&rq->xdp_rxq, nic->netdev, qidx) < 0); + WARN_ON(xdp_rxq_info_reg(&rq->xdp_rxq, nic->netdev, qidx, 0) < 0); /* Send a mailbox msg to PF to config RQ */ mbx.rq.msg = NIC_MBOX_MSG_RQ_CFG; diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c index cf9400a9886d..40953980e846 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c +++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c @@ -3334,7 +3334,7 @@ static int dpaa2_eth_setup_rx_flow(struct dpaa2_eth_priv *priv, return 0; err = xdp_rxq_info_reg(&fq->channel->xdp_rxq, priv->net_dev, - fq->flowid); + fq->flowid, 0); if (err) { dev_err(dev, "xdp_rxq_info_reg failed\n"); return err; diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c index c21548c71bb1..9f73cd7aee09 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c +++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c @@ -1447,7 +1447,7 @@ int i40e_setup_rx_descriptors(struct i40e_ring *rx_ring) /* XDP RX-queue info only needed for RX rings exposed to XDP */ if (rx_ring->vsi->type == I40E_VSI_MAIN) { err = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev, - rx_ring->queue_index); + rx_ring->queue_index, rx_ring->q_vector->napi.napi_id); if (err < 0) return err; } diff --git a/drivers/net/ethernet/intel/ice/ice_base.c b/drivers/net/ethernet/intel/ice/ice_base.c index fe4320e2d1f2..3124a3bf519a 100644 --- a/drivers/net/ethernet/intel/ice/ice_base.c +++ b/drivers/net/ethernet/intel/ice/ice_base.c @@ -306,7 +306,7 @@ int ice_setup_rx_ctx(struct ice_ring *ring) if (!xdp_rxq_info_is_reg(&ring->xdp_rxq)) /* coverity[check_return] */ xdp_rxq_info_reg(&ring->xdp_rxq, ring->netdev, - ring->q_index); + ring->q_index, ring->q_vector->napi.napi_id); ring->xsk_pool = ice_xsk_pool(ring); if (ring->xsk_pool) { @@ -333,7 +333,7 @@ int ice_setup_rx_ctx(struct ice_ring *ring) /* coverity[check_return] */ xdp_rxq_info_reg(&ring->xdp_rxq, ring->netdev, - ring->q_index); + ring->q_index, ring->q_vector->napi.napi_id); err = xdp_rxq_info_reg_mem_model(&ring->xdp_rxq, MEM_TYPE_PAGE_SHARED, diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c index eae75260fe20..77d5eae6b4c2 100644 --- a/drivers/net/ethernet/intel/ice/ice_txrx.c +++ b/drivers/net/ethernet/intel/ice/ice_txrx.c @@ -483,7 +483,7 @@ int ice_setup_rx_ring(struct ice_ring *rx_ring) if (rx_ring->vsi->type == ICE_VSI_PF && !xdp_rxq_info_is_reg(&rx_ring->xdp_rxq)) if (xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev, - rx_ring->q_index)) + rx_ring->q_index, rx_ring->q_vector->napi.napi_id)) goto err; return 0; diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index 5fc2c381da55..6a4ef4934fcf 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -4352,7 +4352,7 @@ int igb_setup_rx_resources(struct igb_ring *rx_ring) /* XDP RX-queue info */ if (xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev, - rx_ring->queue_index) < 0) + rx_ring->queue_index, 0) < 0) goto err; return 0; diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index 45ae33e15303..50e6b8b6ba7b 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -6577,7 +6577,7 @@ int ixgbe_setup_rx_resources(struct ixgbe_adapter *adapter, /* XDP RX-queue info */ if (xdp_rxq_info_reg(&rx_ring->xdp_rxq, adapter->netdev, - rx_ring->queue_index) < 0) + rx_ring->queue_index, rx_ring->q_vector->napi.napi_id) < 0) goto err; rx_ring->xdp_prog = adapter->xdp_prog; diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index 82fce27f682b..4061cd7db5dd 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -3493,7 +3493,7 @@ int ixgbevf_setup_rx_resources(struct ixgbevf_adapter *adapter, /* XDP RX-queue info */ if (xdp_rxq_info_reg(&rx_ring->xdp_rxq, adapter->netdev, - rx_ring->queue_index) < 0) + rx_ring->queue_index, 0) < 0) goto err; rx_ring->xdp_prog = adapter->xdp_prog; diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index 183530ed4d1d..ba6dcb19bb1d 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -3227,7 +3227,7 @@ static int mvneta_create_page_pool(struct mvneta_port *pp, return err; } - err = xdp_rxq_info_reg(&rxq->xdp_rxq, pp->dev, rxq->id); + err = xdp_rxq_info_reg(&rxq->xdp_rxq, pp->dev, rxq->id, 0); if (err < 0) goto err_free_pp; diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c index 3069e192d773..5504cbc24970 100644 --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c @@ -2614,11 +2614,11 @@ static int mvpp2_rxq_init(struct mvpp2_port *port, mvpp2_rxq_status_update(port, rxq->id, 0, rxq->size); if (priv->percpu_pools) { - err = xdp_rxq_info_reg(&rxq->xdp_rxq_short, port->dev, rxq->id); + err = xdp_rxq_info_reg(&rxq->xdp_rxq_short, port->dev, rxq->id, 0); if (err < 0) goto err_free_dma; - err = xdp_rxq_info_reg(&rxq->xdp_rxq_long, port->dev, rxq->id); + err = xdp_rxq_info_reg(&rxq->xdp_rxq_long, port->dev, rxq->id, 0); if (err < 0) goto err_unregister_rxq_short; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c index b0f79a5151cf..40775cb8fb2a 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c @@ -283,7 +283,7 @@ int mlx4_en_create_rx_ring(struct mlx4_en_priv *priv, ring->log_stride = ffs(ring->stride) - 1; ring->buf_size = ring->size * ring->stride + TXBB_SIZE; - if (xdp_rxq_info_reg(&ring->xdp_rxq, priv->dev, queue_index) < 0) + if (xdp_rxq_info_reg(&ring->xdp_rxq, priv->dev, queue_index, 0) < 0) goto err_ring; tmp = size * roundup_pow_of_two(MLX4_EN_MAX_RX_FRAGS * diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 527c5f12c5af..427fc376fe1a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -434,7 +434,7 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, rq_xdp_ix = rq->ix; if (xsk) rq_xdp_ix += params->num_channels * MLX5E_RQ_GROUP_XSK; - err = xdp_rxq_info_reg(&rq->xdp_rxq, rq->netdev, rq_xdp_ix); + err = xdp_rxq_info_reg(&rq->xdp_rxq, rq->netdev, rq_xdp_ix, 0); if (err < 0) goto err_rq_xdp_prog; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c index b150da43adb2..b4acf2f41e84 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c @@ -2533,7 +2533,7 @@ nfp_net_rx_ring_alloc(struct nfp_net_dp *dp, struct nfp_net_rx_ring *rx_ring) if (dp->netdev) { err = xdp_rxq_info_reg(&rx_ring->xdp_rxq, dp->netdev, - rx_ring->idx); + rx_ring->idx, rx_ring->r_vec->napi.napi_id); if (err < 0) return err; } diff --git a/drivers/net/ethernet/qlogic/qede/qede_main.c b/drivers/net/ethernet/qlogic/qede/qede_main.c index 05e3a3b60269..9cf960a6d007 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_main.c +++ b/drivers/net/ethernet/qlogic/qede/qede_main.c @@ -1762,7 +1762,7 @@ static void qede_init_fp(struct qede_dev *edev) /* Driver have no error path from here */ WARN_ON(xdp_rxq_info_reg(&fp->rxq->xdp_rxq, edev->ndev, - fp->rxq->rxq_id) < 0); + fp->rxq->rxq_id, 0) < 0); if (xdp_rxq_info_reg_mem_model(&fp->rxq->xdp_rxq, MEM_TYPE_PAGE_ORDER0, diff --git a/drivers/net/ethernet/sfc/rx_common.c b/drivers/net/ethernet/sfc/rx_common.c index 19cf7cac1e6e..68fc7d317693 100644 --- a/drivers/net/ethernet/sfc/rx_common.c +++ b/drivers/net/ethernet/sfc/rx_common.c @@ -262,7 +262,7 @@ void efx_init_rx_queue(struct efx_rx_queue *rx_queue) /* Initialise XDP queue information */ rc = xdp_rxq_info_reg(&rx_queue->xdp_rxq_info, efx->net_dev, - rx_queue->core_index); + rx_queue->core_index, 0); if (rc) { netif_err(efx, rx_err, efx->net_dev, diff --git a/drivers/net/ethernet/socionext/netsec.c b/drivers/net/ethernet/socionext/netsec.c index 1503cc9ec6e2..27d3c9d9210e 100644 --- a/drivers/net/ethernet/socionext/netsec.c +++ b/drivers/net/ethernet/socionext/netsec.c @@ -1304,7 +1304,7 @@ static int netsec_setup_rx_dring(struct netsec_priv *priv) goto err_out; } - err = xdp_rxq_info_reg(&dring->xdp_rxq, priv->ndev, 0); + err = xdp_rxq_info_reg(&dring->xdp_rxq, priv->ndev, 0, priv->napi.napi_id); if (err) goto err_out; diff --git a/drivers/net/ethernet/ti/cpsw_priv.c b/drivers/net/ethernet/ti/cpsw_priv.c index 31c5e36ff706..6dd73bd0f458 100644 --- a/drivers/net/ethernet/ti/cpsw_priv.c +++ b/drivers/net/ethernet/ti/cpsw_priv.c @@ -1186,7 +1186,7 @@ static int cpsw_ndev_create_xdp_rxq(struct cpsw_priv *priv, int ch) pool = cpsw->page_pool[ch]; rxq = &priv->xdp_rxq[ch]; - ret = xdp_rxq_info_reg(rxq, priv->ndev, ch); + ret = xdp_rxq_info_reg(rxq, priv->ndev, ch, 0); if (ret) return ret; diff --git a/drivers/net/hyperv/netvsc.c b/drivers/net/hyperv/netvsc.c index 0c3de94b5178..fa8341f8359a 100644 --- a/drivers/net/hyperv/netvsc.c +++ b/drivers/net/hyperv/netvsc.c @@ -1499,7 +1499,7 @@ struct netvsc_device *netvsc_device_add(struct hv_device *device, u64_stats_init(&nvchan->tx_stats.syncp); u64_stats_init(&nvchan->rx_stats.syncp); - ret = xdp_rxq_info_reg(&nvchan->xdp_rxq, ndev, i); + ret = xdp_rxq_info_reg(&nvchan->xdp_rxq, ndev, i, 0); if (ret) { netdev_err(ndev, "xdp_rxq_info_reg fail: %d\n", ret); diff --git a/drivers/net/tun.c b/drivers/net/tun.c index 3d45d56172cb..8867d39db6ac 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -780,7 +780,7 @@ static int tun_attach(struct tun_struct *tun, struct file *file, } else { /* Setup XDP RX-queue info, for new tfile getting attached */ err = xdp_rxq_info_reg(&tfile->xdp_rxq, - tun->dev, tfile->queue_index); + tun->dev, tfile->queue_index, 0); if (err < 0) goto out; err = xdp_rxq_info_reg_mem_model(&tfile->xdp_rxq, diff --git a/drivers/net/veth.c b/drivers/net/veth.c index 8c737668008a..9bd37c7151f8 100644 --- a/drivers/net/veth.c +++ b/drivers/net/veth.c @@ -884,7 +884,6 @@ static int veth_napi_add(struct net_device *dev) for (i = 0; i < dev->real_num_rx_queues; i++) { struct veth_rq *rq = &priv->rq[i]; - netif_napi_add(dev, &rq->xdp_napi, veth_poll, NAPI_POLL_WEIGHT); napi_enable(&rq->xdp_napi); } @@ -926,7 +925,8 @@ static int veth_enable_xdp(struct net_device *dev) for (i = 0; i < dev->real_num_rx_queues; i++) { struct veth_rq *rq = &priv->rq[i]; - err = xdp_rxq_info_reg(&rq->xdp_rxq, dev, i); + netif_napi_add(dev, &rq->xdp_napi, veth_poll, NAPI_POLL_WEIGHT); + err = xdp_rxq_info_reg(&rq->xdp_rxq, dev, i, rq->xdp_napi.napi_id); if (err < 0) goto err_rxq_reg; @@ -952,8 +952,12 @@ static int veth_enable_xdp(struct net_device *dev) err_reg_mem: xdp_rxq_info_unreg(&priv->rq[i].xdp_rxq); err_rxq_reg: - for (i--; i >= 0; i--) - xdp_rxq_info_unreg(&priv->rq[i].xdp_rxq); + for (i--; i >= 0; i--) { + struct veth_rq *rq = &priv->rq[i]; + + xdp_rxq_info_unreg(&rq->xdp_rxq); + netif_napi_del(&rq->xdp_napi); + } return err; } diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 21b71148c532..052975ea0af4 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -1485,7 +1485,7 @@ static int virtnet_open(struct net_device *dev) if (!try_fill_recv(vi, &vi->rq[i], GFP_KERNEL)) schedule_delayed_work(&vi->refill, 0); - err = xdp_rxq_info_reg(&vi->rq[i].xdp_rxq, dev, i); + err = xdp_rxq_info_reg(&vi->rq[i].xdp_rxq, dev, i, vi->rq[i].napi.napi_id); if (err < 0) return err; diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c index 920cac4385bf..b01848ef4649 100644 --- a/drivers/net/xen-netfront.c +++ b/drivers/net/xen-netfront.c @@ -2014,7 +2014,7 @@ static int xennet_create_page_pool(struct netfront_queue *queue) } err = xdp_rxq_info_reg(&queue->xdp_rxq, queue->info->netdev, - queue->id); + queue->id, 0); if (err) { netdev_err(queue->info->netdev, "xdp_rxq_info_reg failed\n"); goto err_free_pp; diff --git a/include/net/busy_poll.h b/include/net/busy_poll.h index 2f8f51807b83..45b3e04b99d3 100644 --- a/include/net/busy_poll.h +++ b/include/net/busy_poll.h @@ -135,14 +135,25 @@ static inline void sk_mark_napi_id(struct sock *sk, const struct sk_buff *skb) sk_rx_queue_set(sk, skb); } -/* variant used for unconnected sockets */ -static inline void sk_mark_napi_id_once(struct sock *sk, - const struct sk_buff *skb) +static inline void __sk_mark_napi_id_once_xdp(struct sock *sk, unsigned int napi_id) { #ifdef CONFIG_NET_RX_BUSY_POLL if (!READ_ONCE(sk->sk_napi_id)) - WRITE_ONCE(sk->sk_napi_id, skb->napi_id); + WRITE_ONCE(sk->sk_napi_id, napi_id); #endif } +/* variant used for unconnected sockets */ +static inline void sk_mark_napi_id_once(struct sock *sk, + const struct sk_buff *skb) +{ + __sk_mark_napi_id_once_xdp(sk, skb->napi_id); +} + +static inline void sk_mark_napi_id_once_xdp(struct sock *sk, + const struct xdp_buff *xdp) +{ + __sk_mark_napi_id_once_xdp(sk, xdp->rxq->napi_id); +} + #endif /* _LINUX_NET_BUSY_POLL_H */ diff --git a/include/net/xdp.h b/include/net/xdp.h index 7d48b2ae217a..700ad5db7f5d 100644 --- a/include/net/xdp.h +++ b/include/net/xdp.h @@ -59,6 +59,7 @@ struct xdp_rxq_info { u32 queue_index; u32 reg_state; struct xdp_mem_info mem; + unsigned int napi_id; } ____cacheline_aligned; /* perf critical, avoid false-sharing */ struct xdp_txq_info { @@ -226,7 +227,7 @@ static inline void xdp_release_frame(struct xdp_frame *xdpf) } int xdp_rxq_info_reg(struct xdp_rxq_info *xdp_rxq, - struct net_device *dev, u32 queue_index); + struct net_device *dev, u32 queue_index, unsigned int napi_id); void xdp_rxq_info_unreg(struct xdp_rxq_info *xdp_rxq); void xdp_rxq_info_unused(struct xdp_rxq_info *xdp_rxq); bool xdp_rxq_info_is_reg(struct xdp_rxq_info *xdp_rxq); diff --git a/net/core/dev.c b/net/core/dev.c index 7a1e5936c67f..3b6b0e175fe7 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -9810,7 +9810,7 @@ static int netif_alloc_rx_queues(struct net_device *dev) rx[i].dev = dev; /* XDP RX-queue setup */ - err = xdp_rxq_info_reg(&rx[i].xdp_rxq, dev, i); + err = xdp_rxq_info_reg(&rx[i].xdp_rxq, dev, i, 0); if (err < 0) goto err_rxq_info; } diff --git a/net/core/xdp.c b/net/core/xdp.c index 3d330ebda893..17ffd33c6b18 100644 --- a/net/core/xdp.c +++ b/net/core/xdp.c @@ -158,7 +158,7 @@ static void xdp_rxq_info_init(struct xdp_rxq_info *xdp_rxq) /* Returns 0 on success, negative on failure */ int xdp_rxq_info_reg(struct xdp_rxq_info *xdp_rxq, - struct net_device *dev, u32 queue_index) + struct net_device *dev, u32 queue_index, unsigned int napi_id) { if (xdp_rxq->reg_state == REG_STATE_UNUSED) { WARN(1, "Driver promised not to register this"); @@ -179,6 +179,7 @@ int xdp_rxq_info_reg(struct xdp_rxq_info *xdp_rxq, xdp_rxq_info_init(xdp_rxq); xdp_rxq->dev = dev; xdp_rxq->queue_index = queue_index; + xdp_rxq->napi_id = napi_id; xdp_rxq->reg_state = REG_STATE_REGISTERED; return 0; diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index a8501a5477cf..7588e599a048 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -233,6 +233,7 @@ static int xsk_rcv(struct xdp_sock *xs, struct xdp_buff *xdp, if (xs->dev != xdp->rxq->dev || xs->queue_id != xdp->rxq->queue_index) return -EINVAL; + sk_mark_napi_id_once_xdp(&xs->sk, xdp); len = xdp->data_end - xdp->data; return xdp->rxq->mem.type == MEM_TYPE_XSK_BUFF_POOL ? -- cgit v1.2.3 From ad80b0fc6e7f56bb1b09af86749ff3014477cfe6 Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Fri, 27 Nov 2020 11:10:22 +0100 Subject: mptcp: open code mptcp variant for lock_sock This allows invoking an additional callback under the socket spin lock. Will be used by the next patches to avoid additional spin lock contention. Acked-by: Florian Westphal Signed-off-by: Paolo Abeni Reviewed-by: Mat Martineau Signed-off-by: Jakub Kicinski --- include/net/sock.h | 1 + net/core/sock.c | 2 +- net/mptcp/protocol.h | 13 +++++++++++++ 3 files changed, 15 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 80469c2c448d..f59764614e30 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1590,6 +1590,7 @@ static inline void lock_sock(struct sock *sk) lock_sock_nested(sk, 0); } +void __lock_sock(struct sock *sk); void __release_sock(struct sock *sk); void release_sock(struct sock *sk); diff --git a/net/core/sock.c b/net/core/sock.c index 9badbe7bb4e4..f0f096852876 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2486,7 +2486,7 @@ bool sk_page_frag_refill(struct sock *sk, struct page_frag *pfrag) } EXPORT_SYMBOL(sk_page_frag_refill); -static void __lock_sock(struct sock *sk) +void __lock_sock(struct sock *sk) __releases(&sk->sk_lock.slock) __acquires(&sk->sk_lock.slock) { diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 82d5626323b1..6abac8238de3 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -253,6 +253,19 @@ struct mptcp_sock { } rcvq_space; }; +#define mptcp_lock_sock(___sk, cb) do { \ + struct sock *__sk = (___sk); /* silence macro reuse warning */ \ + might_sleep(); \ + spin_lock_bh(&__sk->sk_lock.slock); \ + if (__sk->sk_lock.owned) \ + __lock_sock(__sk); \ + cb; \ + __sk->sk_lock.owned = 1; \ + spin_unlock(&__sk->sk_lock.slock); \ + mutex_acquire(&__sk->sk_lock.dep_map, 0, 0, _RET_IP_); \ + local_bh_enable(); \ +} while (0) + #define mptcp_for_each_subflow(__msk, __subflow) \ list_for_each_entry(__subflow, &((__msk)->conn_list), node) -- cgit v1.2.3 From f7583f02a538bdb6d5191e82bacc7b85a27d4b0a Mon Sep 17 00:00:00 2001 From: Wang Shanker Date: Mon, 16 Nov 2020 12:17:24 +0800 Subject: netfilter: nfnl_acct: remove data from struct net This patch removes nfnl_acct_list from struct net to reduce the default memory footprint for the netns structure. Signed-off-by: Miao Wang Signed-off-by: Pablo Neira Ayuso --- include/net/net_namespace.h | 3 --- net/netfilter/nfnetlink_acct.c | 38 +++++++++++++++++++++++++++++--------- 2 files changed, 29 insertions(+), 12 deletions(-) (limited to 'include/net') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 22bc07f4b043..dc20a47e3828 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -151,9 +151,6 @@ struct net { #endif struct sock *nfnl; struct sock *nfnl_stash; -#if IS_ENABLED(CONFIG_NETFILTER_NETLINK_ACCT) - struct list_head nfnl_acct_list; -#endif #if IS_ENABLED(CONFIG_NF_CT_NETLINK_TIMEOUT) struct list_head nfct_timeout_list; #endif diff --git a/net/netfilter/nfnetlink_acct.c b/net/netfilter/nfnetlink_acct.c index 5e511df8d709..0fa1653b5f19 100644 --- a/net/netfilter/nfnetlink_acct.c +++ b/net/netfilter/nfnetlink_acct.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include @@ -41,6 +42,17 @@ struct nfacct_filter { u32 mask; }; +struct nfnl_acct_net { + struct list_head nfnl_acct_list; +}; + +static unsigned int nfnl_acct_net_id __read_mostly; + +static inline struct nfnl_acct_net *nfnl_acct_pernet(struct net *net) +{ + return net_generic(net, nfnl_acct_net_id); +} + #define NFACCT_F_QUOTA (NFACCT_F_QUOTA_PKTS | NFACCT_F_QUOTA_BYTES) #define NFACCT_OVERQUOTA_BIT 2 /* NFACCT_F_OVERQUOTA */ @@ -49,6 +61,7 @@ static int nfnl_acct_new(struct net *net, struct sock *nfnl, const struct nlattr * const tb[], struct netlink_ext_ack *extack) { + struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net); struct nf_acct *nfacct, *matching = NULL; char *acct_name; unsigned int size = 0; @@ -61,7 +74,7 @@ static int nfnl_acct_new(struct net *net, struct sock *nfnl, if (strlen(acct_name) == 0) return -EINVAL; - list_for_each_entry(nfacct, &net->nfnl_acct_list, head) { + list_for_each_entry(nfacct, &nfnl_acct_net->nfnl_acct_list, head) { if (strncmp(nfacct->name, acct_name, NFACCT_NAME_MAX) != 0) continue; @@ -123,7 +136,7 @@ static int nfnl_acct_new(struct net *net, struct sock *nfnl, be64_to_cpu(nla_get_be64(tb[NFACCT_PKTS]))); } refcount_set(&nfacct->refcnt, 1); - list_add_tail_rcu(&nfacct->head, &net->nfnl_acct_list); + list_add_tail_rcu(&nfacct->head, &nfnl_acct_net->nfnl_acct_list); return 0; } @@ -188,6 +201,7 @@ static int nfnl_acct_dump(struct sk_buff *skb, struct netlink_callback *cb) { struct net *net = sock_net(skb->sk); + struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net); struct nf_acct *cur, *last; const struct nfacct_filter *filter = cb->data; @@ -199,7 +213,7 @@ nfnl_acct_dump(struct sk_buff *skb, struct netlink_callback *cb) cb->args[1] = 0; rcu_read_lock(); - list_for_each_entry_rcu(cur, &net->nfnl_acct_list, head) { + list_for_each_entry_rcu(cur, &nfnl_acct_net->nfnl_acct_list, head) { if (last) { if (cur != last) continue; @@ -269,6 +283,7 @@ static int nfnl_acct_get(struct net *net, struct sock *nfnl, const struct nlattr * const tb[], struct netlink_ext_ack *extack) { + struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net); int ret = -ENOENT; struct nf_acct *cur; char *acct_name; @@ -288,7 +303,7 @@ static int nfnl_acct_get(struct net *net, struct sock *nfnl, return -EINVAL; acct_name = nla_data(tb[NFACCT_NAME]); - list_for_each_entry(cur, &net->nfnl_acct_list, head) { + list_for_each_entry(cur, &nfnl_acct_net->nfnl_acct_list, head) { struct sk_buff *skb2; if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX)!= 0) @@ -342,19 +357,20 @@ static int nfnl_acct_del(struct net *net, struct sock *nfnl, const struct nlattr * const tb[], struct netlink_ext_ack *extack) { + struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net); struct nf_acct *cur, *tmp; int ret = -ENOENT; char *acct_name; if (!tb[NFACCT_NAME]) { - list_for_each_entry_safe(cur, tmp, &net->nfnl_acct_list, head) + list_for_each_entry_safe(cur, tmp, &nfnl_acct_net->nfnl_acct_list, head) nfnl_acct_try_del(cur); return 0; } acct_name = nla_data(tb[NFACCT_NAME]); - list_for_each_entry(cur, &net->nfnl_acct_list, head) { + list_for_each_entry(cur, &nfnl_acct_net->nfnl_acct_list, head) { if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX) != 0) continue; @@ -402,10 +418,11 @@ MODULE_ALIAS_NFNL_SUBSYS(NFNL_SUBSYS_ACCT); struct nf_acct *nfnl_acct_find_get(struct net *net, const char *acct_name) { + struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net); struct nf_acct *cur, *acct = NULL; rcu_read_lock(); - list_for_each_entry_rcu(cur, &net->nfnl_acct_list, head) { + list_for_each_entry_rcu(cur, &nfnl_acct_net->nfnl_acct_list, head) { if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX)!= 0) continue; @@ -488,16 +505,17 @@ EXPORT_SYMBOL_GPL(nfnl_acct_overquota); static int __net_init nfnl_acct_net_init(struct net *net) { - INIT_LIST_HEAD(&net->nfnl_acct_list); + INIT_LIST_HEAD(&nfnl_acct_pernet(net)->nfnl_acct_list); return 0; } static void __net_exit nfnl_acct_net_exit(struct net *net) { + struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net); struct nf_acct *cur, *tmp; - list_for_each_entry_safe(cur, tmp, &net->nfnl_acct_list, head) { + list_for_each_entry_safe(cur, tmp, &nfnl_acct_net->nfnl_acct_list, head) { list_del_rcu(&cur->head); if (refcount_dec_and_test(&cur->refcnt)) @@ -508,6 +526,8 @@ static void __net_exit nfnl_acct_net_exit(struct net *net) static struct pernet_operations nfnl_acct_ops = { .init = nfnl_acct_net_init, .exit = nfnl_acct_net_exit, + .id = &nfnl_acct_net_id, + .size = sizeof(struct nfnl_acct_net), }; static int __init nfnl_acct_init(void) -- cgit v1.2.3 From 04295878beac396dae47ba93141cae0d9386e7ef Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Sat, 21 Nov 2020 12:11:51 +0100 Subject: netfilter: use actual socket sk for REJECT action True to the message of commit v5.10-rc1-105-g46d6c5ae953c, _do_ actually make use of state->sk when possible, such as in the REJECT modules. Reported-by: Minqiang Chen Cc: Jason A. Donenfeld Signed-off-by: Jan Engelhardt Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/ipv4/nf_reject.h | 4 ++-- include/net/netfilter/ipv6/nf_reject.h | 5 ++--- net/ipv4/netfilter/ipt_REJECT.c | 3 ++- net/ipv4/netfilter/nf_reject_ipv4.c | 6 +++--- net/ipv4/netfilter/nft_reject_ipv4.c | 3 ++- net/ipv6/netfilter/ip6t_REJECT.c | 2 +- net/ipv6/netfilter/nf_reject_ipv6.c | 5 +++-- net/ipv6/netfilter/nft_reject_ipv6.c | 3 ++- net/netfilter/nft_reject_inet.c | 6 ++++-- 9 files changed, 21 insertions(+), 16 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/ipv4/nf_reject.h b/include/net/netfilter/ipv4/nf_reject.h index 0d8ff84a2588..c653fcb88354 100644 --- a/include/net/netfilter/ipv4/nf_reject.h +++ b/include/net/netfilter/ipv4/nf_reject.h @@ -8,8 +8,8 @@ #include void nf_send_unreach(struct sk_buff *skb_in, int code, int hook); -void nf_send_reset(struct net *net, struct sk_buff *oldskb, int hook); - +void nf_send_reset(struct net *net, struct sock *, struct sk_buff *oldskb, + int hook); const struct tcphdr *nf_reject_ip_tcphdr_get(struct sk_buff *oldskb, struct tcphdr *_oth, int hook); struct iphdr *nf_reject_iphdr_put(struct sk_buff *nskb, diff --git a/include/net/netfilter/ipv6/nf_reject.h b/include/net/netfilter/ipv6/nf_reject.h index edcf6d1cd316..d729344ba644 100644 --- a/include/net/netfilter/ipv6/nf_reject.h +++ b/include/net/netfilter/ipv6/nf_reject.h @@ -7,9 +7,8 @@ void nf_send_unreach6(struct net *net, struct sk_buff *skb_in, unsigned char code, unsigned int hooknum); - -void nf_send_reset6(struct net *net, struct sk_buff *oldskb, int hook); - +void nf_send_reset6(struct net *net, struct sock *sk, struct sk_buff *oldskb, + int hook); const struct tcphdr *nf_reject_ip6_tcphdr_get(struct sk_buff *oldskb, struct tcphdr *otcph, unsigned int *otcplen, int hook); diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c index e16b98ee6266..4b8840734762 100644 --- a/net/ipv4/netfilter/ipt_REJECT.c +++ b/net/ipv4/netfilter/ipt_REJECT.c @@ -56,7 +56,8 @@ reject_tg(struct sk_buff *skb, const struct xt_action_param *par) nf_send_unreach(skb, ICMP_PKT_FILTERED, hook); break; case IPT_TCP_RESET: - nf_send_reset(xt_net(par), skb, hook); + nf_send_reset(xt_net(par), par->state->sk, skb, hook); + break; case IPT_ICMP_ECHOREPLY: /* Doesn't happen. */ break; diff --git a/net/ipv4/netfilter/nf_reject_ipv4.c b/net/ipv4/netfilter/nf_reject_ipv4.c index 4109055cdea6..4eed5afca392 100644 --- a/net/ipv4/netfilter/nf_reject_ipv4.c +++ b/net/ipv4/netfilter/nf_reject_ipv4.c @@ -234,7 +234,8 @@ static int nf_reject_fill_skb_dst(struct sk_buff *skb_in) } /* Send RST reply */ -void nf_send_reset(struct net *net, struct sk_buff *oldskb, int hook) +void nf_send_reset(struct net *net, struct sock *sk, struct sk_buff *oldskb, + int hook) { struct net_device *br_indev __maybe_unused; struct sk_buff *nskb; @@ -267,8 +268,7 @@ void nf_send_reset(struct net *net, struct sk_buff *oldskb, int hook) niph = nf_reject_iphdr_put(nskb, oldskb, IPPROTO_TCP, ip4_dst_hoplimit(skb_dst(nskb))); nf_reject_ip_tcphdr_put(nskb, oldskb, oth); - - if (ip_route_me_harder(net, nskb->sk, nskb, RTN_UNSPEC)) + if (ip_route_me_harder(net, sk, nskb, RTN_UNSPEC)) goto free_nskb; niph = ip_hdr(nskb); diff --git a/net/ipv4/netfilter/nft_reject_ipv4.c b/net/ipv4/netfilter/nft_reject_ipv4.c index e408f813f5d8..ff437e4ed6db 100644 --- a/net/ipv4/netfilter/nft_reject_ipv4.c +++ b/net/ipv4/netfilter/nft_reject_ipv4.c @@ -27,7 +27,8 @@ static void nft_reject_ipv4_eval(const struct nft_expr *expr, nf_send_unreach(pkt->skb, priv->icmp_code, nft_hook(pkt)); break; case NFT_REJECT_TCP_RST: - nf_send_reset(nft_net(pkt), pkt->skb, nft_hook(pkt)); + nf_send_reset(nft_net(pkt), pkt->xt.state->sk, pkt->skb, + nft_hook(pkt)); break; default: break; diff --git a/net/ipv6/netfilter/ip6t_REJECT.c b/net/ipv6/netfilter/ip6t_REJECT.c index 3ac5485049f0..a35019d2e480 100644 --- a/net/ipv6/netfilter/ip6t_REJECT.c +++ b/net/ipv6/netfilter/ip6t_REJECT.c @@ -61,7 +61,7 @@ reject_tg6(struct sk_buff *skb, const struct xt_action_param *par) /* Do nothing */ break; case IP6T_TCP_RESET: - nf_send_reset6(net, skb, xt_hooknum(par)); + nf_send_reset6(net, par->state->sk, skb, xt_hooknum(par)); break; case IP6T_ICMP6_POLICY_FAIL: nf_send_unreach6(net, skb, ICMPV6_POLICY_FAIL, xt_hooknum(par)); diff --git a/net/ipv6/netfilter/nf_reject_ipv6.c b/net/ipv6/netfilter/nf_reject_ipv6.c index aa35e6e37c1f..570d1d76c44d 100644 --- a/net/ipv6/netfilter/nf_reject_ipv6.c +++ b/net/ipv6/netfilter/nf_reject_ipv6.c @@ -275,7 +275,8 @@ static int nf_reject6_fill_skb_dst(struct sk_buff *skb_in) return 0; } -void nf_send_reset6(struct net *net, struct sk_buff *oldskb, int hook) +void nf_send_reset6(struct net *net, struct sock *sk, struct sk_buff *oldskb, + int hook) { struct net_device *br_indev __maybe_unused; struct sk_buff *nskb; @@ -367,7 +368,7 @@ void nf_send_reset6(struct net *net, struct sk_buff *oldskb, int hook) dev_queue_xmit(nskb); } else #endif - ip6_local_out(net, nskb->sk, nskb); + ip6_local_out(net, sk, nskb); } EXPORT_SYMBOL_GPL(nf_send_reset6); diff --git a/net/ipv6/netfilter/nft_reject_ipv6.c b/net/ipv6/netfilter/nft_reject_ipv6.c index c1098a1968e1..7969d1f3018d 100644 --- a/net/ipv6/netfilter/nft_reject_ipv6.c +++ b/net/ipv6/netfilter/nft_reject_ipv6.c @@ -28,7 +28,8 @@ static void nft_reject_ipv6_eval(const struct nft_expr *expr, nft_hook(pkt)); break; case NFT_REJECT_TCP_RST: - nf_send_reset6(nft_net(pkt), pkt->skb, nft_hook(pkt)); + nf_send_reset6(nft_net(pkt), pkt->xt.state->sk, pkt->skb, + nft_hook(pkt)); break; default: break; diff --git a/net/netfilter/nft_reject_inet.c b/net/netfilter/nft_reject_inet.c index 32f3ea398ddf..95090186ee90 100644 --- a/net/netfilter/nft_reject_inet.c +++ b/net/netfilter/nft_reject_inet.c @@ -28,7 +28,8 @@ static void nft_reject_inet_eval(const struct nft_expr *expr, nft_hook(pkt)); break; case NFT_REJECT_TCP_RST: - nf_send_reset(nft_net(pkt), pkt->skb, nft_hook(pkt)); + nf_send_reset(nft_net(pkt), pkt->xt.state->sk, + pkt->skb, nft_hook(pkt)); break; case NFT_REJECT_ICMPX_UNREACH: nf_send_unreach(pkt->skb, @@ -44,7 +45,8 @@ static void nft_reject_inet_eval(const struct nft_expr *expr, priv->icmp_code, nft_hook(pkt)); break; case NFT_REJECT_TCP_RST: - nf_send_reset6(nft_net(pkt), pkt->skb, nft_hook(pkt)); + nf_send_reset6(nft_net(pkt), pkt->xt.state->sk, + pkt->skb, nft_hook(pkt)); break; case NFT_REJECT_ICMPX_UNREACH: nf_send_unreach6(nft_net(pkt), pkt->skb, -- cgit v1.2.3 From ba0581749fec389e55c9d761f2716f8fcbefced5 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Tue, 1 Dec 2020 15:22:59 +0100 Subject: net, xdp, xsk: fix __sk_mark_napi_id_once napi_id error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Stephen reported the following build error for !CONFIG_NET_RX_BUSY_POLL built kernels: In file included from fs/select.c:32: include/net/busy_poll.h: In function 'sk_mark_napi_id_once': include/net/busy_poll.h:150:36: error: 'const struct sk_buff' has no member named 'napi_id' 150 | __sk_mark_napi_id_once_xdp(sk, skb->napi_id); | ^~ Fix it by wrapping a CONFIG_NET_RX_BUSY_POLL around the helpers. Fixes: b02e5a0ebb17 ("xsk: Propagate napi_id to XDP socket Rx path") Reported-by: Stephen Rothwell Signed-off-by: Stephen Rothwell Signed-off-by: Daniel Borkmann Cc: Björn Töpel Link: https://lore.kernel.org/linux-next/20201201190746.7d3357fb@canb.auug.org.au --- include/net/busy_poll.h | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/busy_poll.h b/include/net/busy_poll.h index 45b3e04b99d3..73af4a64a599 100644 --- a/include/net/busy_poll.h +++ b/include/net/busy_poll.h @@ -135,7 +135,7 @@ static inline void sk_mark_napi_id(struct sock *sk, const struct sk_buff *skb) sk_rx_queue_set(sk, skb); } -static inline void __sk_mark_napi_id_once_xdp(struct sock *sk, unsigned int napi_id) +static inline void __sk_mark_napi_id_once(struct sock *sk, unsigned int napi_id) { #ifdef CONFIG_NET_RX_BUSY_POLL if (!READ_ONCE(sk->sk_napi_id)) @@ -147,13 +147,17 @@ static inline void __sk_mark_napi_id_once_xdp(struct sock *sk, unsigned int napi static inline void sk_mark_napi_id_once(struct sock *sk, const struct sk_buff *skb) { - __sk_mark_napi_id_once_xdp(sk, skb->napi_id); +#ifdef CONFIG_NET_RX_BUSY_POLL + __sk_mark_napi_id_once(sk, skb->napi_id); +#endif } static inline void sk_mark_napi_id_once_xdp(struct sock *sk, const struct xdp_buff *xdp) { - __sk_mark_napi_id_once_xdp(sk, xdp->rxq->napi_id); +#ifdef CONFIG_NET_RX_BUSY_POLL + __sk_mark_napi_id_once(sk, xdp->rxq->napi_id); +#endif } #endif /* _LINUX_NET_BUSY_POLL_H */ -- cgit v1.2.3 From 0fca55ed988a694f5896f36de2a8f18715a78279 Mon Sep 17 00:00:00 2001 From: Vlad Buslov Date: Fri, 27 Nov 2020 17:12:05 +0200 Subject: net: sched: remove redundant 'rtnl_held' argument Functions tfilter_notify_chain() and tcf_get_next_proto() are always called with rtnl lock held in current implementation. Moreover, attempting to call them without rtnl lock would cause a warning down the call chain in function __tcf_get_next_proto() that requires the lock to be held by callers. Remove the 'rtnl_held' argument in order to simplify the code and make rtnl lock requirement explicit. Signed-off-by: Vlad Buslov Link: https://lore.kernel.org/r/20201127151205.23492-1-vladbu@nvidia.com Signed-off-by: Jakub Kicinski --- include/net/pkt_cls.h | 2 +- net/sched/cls_api.c | 18 ++++++++---------- net/sched/sch_api.c | 4 ++-- 3 files changed, 11 insertions(+), 13 deletions(-) (limited to 'include/net') diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index 133f9ad4d4f9..0f2a9c44171c 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -48,7 +48,7 @@ void tcf_chain_put_by_act(struct tcf_chain *chain); struct tcf_chain *tcf_get_next_chain(struct tcf_block *block, struct tcf_chain *chain); struct tcf_proto *tcf_get_next_proto(struct tcf_chain *chain, - struct tcf_proto *tp, bool rtnl_held); + struct tcf_proto *tp); void tcf_block_netif_keep_dst(struct tcf_block *block); int tcf_block_get(struct tcf_block **p_block, struct tcf_proto __rcu **p_filter_chain, struct Qdisc *q, diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index ff3e943febaa..37b77bd30974 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -991,13 +991,12 @@ __tcf_get_next_proto(struct tcf_chain *chain, struct tcf_proto *tp) */ struct tcf_proto * -tcf_get_next_proto(struct tcf_chain *chain, struct tcf_proto *tp, - bool rtnl_held) +tcf_get_next_proto(struct tcf_chain *chain, struct tcf_proto *tp) { struct tcf_proto *tp_next = __tcf_get_next_proto(chain, tp); if (tp) - tcf_proto_put(tp, rtnl_held, NULL); + tcf_proto_put(tp, true, NULL); return tp_next; } @@ -1924,15 +1923,14 @@ static int tfilter_del_notify(struct net *net, struct sk_buff *oskb, static void tfilter_notify_chain(struct net *net, struct sk_buff *oskb, struct tcf_block *block, struct Qdisc *q, u32 parent, struct nlmsghdr *n, - struct tcf_chain *chain, int event, - bool rtnl_held) + struct tcf_chain *chain, int event) { struct tcf_proto *tp; - for (tp = tcf_get_next_proto(chain, NULL, rtnl_held); - tp; tp = tcf_get_next_proto(chain, tp, rtnl_held)) + for (tp = tcf_get_next_proto(chain, NULL); + tp; tp = tcf_get_next_proto(chain, tp)) tfilter_notify(net, oskb, n, tp, block, - q, parent, NULL, event, false, rtnl_held); + q, parent, NULL, event, false, true); } static void tfilter_put(struct tcf_proto *tp, void *fh) @@ -2262,7 +2260,7 @@ static int tc_del_tfilter(struct sk_buff *skb, struct nlmsghdr *n, if (prio == 0) { tfilter_notify_chain(net, skb, block, q, parent, n, - chain, RTM_DELTFILTER, rtnl_held); + chain, RTM_DELTFILTER); tcf_chain_flush(chain, rtnl_held); err = 0; goto errout; @@ -2895,7 +2893,7 @@ replay: break; case RTM_DELCHAIN: tfilter_notify_chain(net, skb, block, q, parent, n, - chain, RTM_DELTFILTER, true); + chain, RTM_DELTFILTER); /* Flush the chain first as the user requested chain removal. */ tcf_chain_flush(chain, true); /* In case the chain was successfully deleted, put a reference diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 1a2d2471b078..51cb553e4317 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -1943,8 +1943,8 @@ static int tc_bind_class_walker(struct Qdisc *q, unsigned long cl, chain = tcf_get_next_chain(block, chain)) { struct tcf_proto *tp; - for (tp = tcf_get_next_proto(chain, NULL, true); - tp; tp = tcf_get_next_proto(chain, tp, true)) { + for (tp = tcf_get_next_proto(chain, NULL); + tp; tp = tcf_get_next_proto(chain, tp)) { struct tcf_bind_args arg = {}; arg.w.fn = tcf_node_bind; -- cgit v1.2.3 From 22ec19f3aee327806c37c9fa1188741574bc6445 Mon Sep 17 00:00:00 2001 From: Danielle Ratson Date: Sun, 29 Nov 2020 14:54:05 +0200 Subject: bridge: switchdev: Notify about VLAN protocol changes Drivers that support bridge offload need to be notified about changes to the bridge's VLAN protocol so that they could react accordingly and potentially veto the change. Add a new switchdev attribute to communicate the change to drivers. Signed-off-by: Danielle Ratson Reviewed-by: Petr Machata Acked-by: Nikolay Aleksandrov Signed-off-by: Ido Schimmel Reviewed-by: Ivan Vecera Signed-off-by: Jakub Kicinski --- include/net/switchdev.h | 2 ++ net/bridge/br_vlan.c | 16 ++++++++++++++-- 2 files changed, 16 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/switchdev.h b/include/net/switchdev.h index 53e8b4994296..99cd538d6519 100644 --- a/include/net/switchdev.h +++ b/include/net/switchdev.h @@ -38,6 +38,7 @@ enum switchdev_attr_id { SWITCHDEV_ATTR_ID_PORT_MROUTER, SWITCHDEV_ATTR_ID_BRIDGE_AGEING_TIME, SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING, + SWITCHDEV_ATTR_ID_BRIDGE_VLAN_PROTOCOL, SWITCHDEV_ATTR_ID_BRIDGE_MC_DISABLED, SWITCHDEV_ATTR_ID_BRIDGE_MROUTER, #if IS_ENABLED(CONFIG_BRIDGE_MRP) @@ -58,6 +59,7 @@ struct switchdev_attr { bool mrouter; /* PORT_MROUTER */ clock_t ageing_time; /* BRIDGE_AGEING_TIME */ bool vlan_filtering; /* BRIDGE_VLAN_FILTERING */ + u16 vlan_protocol; /* BRIDGE_VLAN_PROTOCOL */ bool mc_disabled; /* MC_DISABLED */ #if IS_ENABLED(CONFIG_BRIDGE_MRP) u8 mrp_port_state; /* MRP_PORT_STATE */ diff --git a/net/bridge/br_vlan.c b/net/bridge/br_vlan.c index 11f54a7c0d1d..d07008678d32 100644 --- a/net/bridge/br_vlan.c +++ b/net/bridge/br_vlan.c @@ -854,15 +854,25 @@ EXPORT_SYMBOL_GPL(br_vlan_get_proto); int __br_vlan_set_proto(struct net_bridge *br, __be16 proto) { + struct switchdev_attr attr = { + .orig_dev = br->dev, + .id = SWITCHDEV_ATTR_ID_BRIDGE_VLAN_PROTOCOL, + .flags = SWITCHDEV_F_SKIP_EOPNOTSUPP, + .u.vlan_protocol = ntohs(proto), + }; int err = 0; struct net_bridge_port *p; struct net_bridge_vlan *vlan; struct net_bridge_vlan_group *vg; - __be16 oldproto; + __be16 oldproto = br->vlan_proto; if (br->vlan_proto == proto) return 0; + err = switchdev_port_attr_set(br->dev, &attr); + if (err && err != -EOPNOTSUPP) + return err; + /* Add VLANs for the new proto to the device filter. */ list_for_each_entry(p, &br->port_list, list) { vg = nbp_vlan_group(p); @@ -873,7 +883,6 @@ int __br_vlan_set_proto(struct net_bridge *br, __be16 proto) } } - oldproto = br->vlan_proto; br->vlan_proto = proto; recalculate_group_addr(br); @@ -889,6 +898,9 @@ int __br_vlan_set_proto(struct net_bridge *br, __be16 proto) return 0; err_filt: + attr.u.vlan_protocol = ntohs(oldproto); + switchdev_port_attr_set(br->dev, &attr); + list_for_each_entry_continue_reverse(vlan, &vg->vlan_list, vlist) vlan_vid_del(p->dev, proto, vlan->vid); -- cgit v1.2.3 From 2867e1eac61016f59b3d730e3f7aa488e186e917 Mon Sep 17 00:00:00 2001 From: Toke Høiland-Jørgensen Date: Mon, 30 Nov 2020 19:37:05 +0100 Subject: inet_ecn: Fix endianness of checksum update when setting ECT(1) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When adding support for propagating ECT(1) marking in IP headers it seems I suffered from endianness-confusion in the checksum update calculation: In fact the ECN field is in the *lower* bits of the first 16-bit word of the IP header when calculating in network byte order. This means that the addition performed to update the checksum field was wrong; let's fix that. Fixes: b723748750ec ("tunnel: Propagate ECT(1) when decapsulating as recommended by RFC6040") Reported-by: Jonathan Morton Tested-by: Pete Heist Signed-off-by: Toke Høiland-Jørgensen Link: https://lore.kernel.org/r/20201130183705.17540-1-toke@redhat.com Signed-off-by: Jakub Kicinski --- include/net/inet_ecn.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/inet_ecn.h b/include/net/inet_ecn.h index e1eaf1780288..563457fec557 100644 --- a/include/net/inet_ecn.h +++ b/include/net/inet_ecn.h @@ -107,7 +107,7 @@ static inline int IP_ECN_set_ect1(struct iphdr *iph) if ((iph->tos & INET_ECN_MASK) != INET_ECN_ECT_0) return 0; - check += (__force u16)htons(0x100); + check += (__force u16)htons(0x1); iph->check = (__force __sum16)(check + (check>=0xFFFF)); iph->tos ^= INET_ECN_MASK; -- cgit v1.2.3 From 7ea851d19b23593d7601ecb8091d529f4f475bf5 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Mon, 30 Nov 2020 16:36:30 +0100 Subject: tcp: merge 'init_req' and 'route_req' functions The Multipath-TCP standard (RFC 8684) says that an MPTCP host should send a TCP reset if the token in a MP_JOIN request is unknown. At this time we don't do this, the 3whs completes and the 'new subflow' is reset afterwards. There are two ways to allow MPTCP to send the reset. 1. override 'send_synack' callback and emit the rst from there. The drawback is that the request socket gets inserted into the listeners queue just to get removed again right away. 2. Send the reset from the 'route_req' function instead. This avoids the 'add&remove request socket', but route_req lacks the skb that is required to send the TCP reset. Instead of just adding the skb to that function for MPTCP sake alone, Paolo suggested to merge init_req and route_req functions. This saves one indirection from syn processing path and provides the skb to the merged function at the same time. 'send reset on unknown mptcp join token' is added in next patch. Suggested-by: Paolo Abeni Cc: Eric Dumazet Signed-off-by: Florian Westphal Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 9 ++++----- net/ipv4/tcp_input.c | 9 ++------- net/ipv4/tcp_ipv4.c | 9 +++++++-- net/ipv6/tcp_ipv6.c | 9 +++++++-- net/mptcp/subflow.c | 36 ++++++++++++++++++++++++------------ 5 files changed, 44 insertions(+), 28 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index d5d22c411918..4525d6256321 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -2007,15 +2007,14 @@ struct tcp_request_sock_ops { const struct sock *sk, const struct sk_buff *skb); #endif - void (*init_req)(struct request_sock *req, - const struct sock *sk_listener, - struct sk_buff *skb); #ifdef CONFIG_SYN_COOKIES __u32 (*cookie_init_seq)(const struct sk_buff *skb, __u16 *mss); #endif - struct dst_entry *(*route_req)(const struct sock *sk, struct flowi *fl, - const struct request_sock *req); + struct dst_entry *(*route_req)(const struct sock *sk, + struct sk_buff *skb, + struct flowi *fl, + struct request_sock *req); u32 (*init_seq)(const struct sk_buff *skb); u32 (*init_ts_off)(const struct net *net, const struct sk_buff *skb); int (*send_synack)(const struct sock *sk, struct dst_entry *dst, diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index fb3a7750f623..9e8a6c1aa019 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -6799,18 +6799,13 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops, /* Note: tcp_v6_init_req() might override ir_iif for link locals */ inet_rsk(req)->ir_iif = inet_request_bound_dev_if(sk, skb); - af_ops->init_req(req, sk, skb); - - if (security_inet_conn_request(sk, skb, req)) + dst = af_ops->route_req(sk, skb, &fl, req); + if (!dst) goto drop_and_free; if (tmp_opt.tstamp_ok) tcp_rsk(req)->ts_off = af_ops->init_ts_off(net, skb); - dst = af_ops->route_req(sk, &fl, req); - if (!dst) - goto drop_and_free; - if (!want_cookie && !isn) { /* Kill the following clause, if you dislike this way. */ if (!net->ipv4.sysctl_tcp_syncookies && diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index e4b31e70bd30..af2338294598 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1444,9 +1444,15 @@ static void tcp_v4_init_req(struct request_sock *req, } static struct dst_entry *tcp_v4_route_req(const struct sock *sk, + struct sk_buff *skb, struct flowi *fl, - const struct request_sock *req) + struct request_sock *req) { + tcp_v4_init_req(req, sk, skb); + + if (security_inet_conn_request(sk, skb, req)) + return NULL; + return inet_csk_route_req(sk, &fl->u.ip4, req); } @@ -1466,7 +1472,6 @@ const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = { .req_md5_lookup = tcp_v4_md5_lookup, .calc_md5_hash = tcp_v4_md5_hash_skb, #endif - .init_req = tcp_v4_init_req, #ifdef CONFIG_SYN_COOKIES .cookie_init_seq = cookie_v4_init_sequence, #endif diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 992cbf3eb9e3..1a1510513739 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -828,9 +828,15 @@ static void tcp_v6_init_req(struct request_sock *req, } static struct dst_entry *tcp_v6_route_req(const struct sock *sk, + struct sk_buff *skb, struct flowi *fl, - const struct request_sock *req) + struct request_sock *req) { + tcp_v6_init_req(req, sk, skb); + + if (security_inet_conn_request(sk, skb, req)) + return NULL; + return inet6_csk_route_req(sk, &fl->u.ip6, req, IPPROTO_TCP); } @@ -851,7 +857,6 @@ const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = { .req_md5_lookup = tcp_v6_md5_lookup, .calc_md5_hash = tcp_v6_md5_hash_skb, #endif - .init_req = tcp_v6_init_req, #ifdef CONFIG_SYN_COOKIES .cookie_init_seq = cookie_v6_init_sequence, #endif diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index 023c856424b7..727e607f40d2 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -228,27 +228,39 @@ int mptcp_subflow_init_cookie_req(struct request_sock *req, } EXPORT_SYMBOL_GPL(mptcp_subflow_init_cookie_req); -static void subflow_v4_init_req(struct request_sock *req, - const struct sock *sk_listener, - struct sk_buff *skb) +static struct dst_entry *subflow_v4_route_req(const struct sock *sk, + struct sk_buff *skb, + struct flowi *fl, + struct request_sock *req) { + struct dst_entry *dst; + tcp_rsk(req)->is_mptcp = 1; - tcp_request_sock_ipv4_ops.init_req(req, sk_listener, skb); + dst = tcp_request_sock_ipv4_ops.route_req(sk, skb, fl, req); + if (!dst) + return NULL; - subflow_init_req(req, sk_listener, skb); + subflow_init_req(req, sk, skb); + return dst; } #if IS_ENABLED(CONFIG_MPTCP_IPV6) -static void subflow_v6_init_req(struct request_sock *req, - const struct sock *sk_listener, - struct sk_buff *skb) +static struct dst_entry *subflow_v6_route_req(const struct sock *sk, + struct sk_buff *skb, + struct flowi *fl, + struct request_sock *req) { + struct dst_entry *dst; + tcp_rsk(req)->is_mptcp = 1; - tcp_request_sock_ipv6_ops.init_req(req, sk_listener, skb); + dst = tcp_request_sock_ipv6_ops.route_req(sk, skb, fl, req); + if (!dst) + return NULL; - subflow_init_req(req, sk_listener, skb); + subflow_init_req(req, sk, skb); + return dst; } #endif @@ -1388,7 +1400,7 @@ void __init mptcp_subflow_init(void) panic("MPTCP: failed to init subflow request sock ops\n"); subflow_request_sock_ipv4_ops = tcp_request_sock_ipv4_ops; - subflow_request_sock_ipv4_ops.init_req = subflow_v4_init_req; + subflow_request_sock_ipv4_ops.route_req = subflow_v4_route_req; subflow_specific = ipv4_specific; subflow_specific.conn_request = subflow_v4_conn_request; @@ -1397,7 +1409,7 @@ void __init mptcp_subflow_init(void) #if IS_ENABLED(CONFIG_MPTCP_IPV6) subflow_request_sock_ipv6_ops = tcp_request_sock_ipv6_ops; - subflow_request_sock_ipv6_ops.init_req = subflow_v6_init_req; + subflow_request_sock_ipv6_ops.route_req = subflow_v6_route_req; subflow_v6_specific = ipv6_specific; subflow_v6_specific.conn_request = subflow_v6_conn_request; -- cgit v1.2.3 From cb81110997d1f5097f29dd8e49d32a1fc55cbf86 Mon Sep 17 00:00:00 2001 From: Prankur gupta Date: Wed, 2 Dec 2020 13:31:51 -0800 Subject: bpf: Adds support for setting window clamp Adds a new bpf_setsockopt for TCP sockets, TCP_BPF_WINDOW_CLAMP, which sets the maximum receiver window size. It will be useful for limiting receiver window based on RTT. Signed-off-by: Prankur gupta Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20201202213152.435886-2-prankgup@fb.com --- include/net/tcp.h | 1 + net/core/filter.c | 3 +++ net/ipv4/tcp.c | 25 ++++++++++++++++--------- 3 files changed, 20 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 4aba0f069b05..347a76f176b4 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -406,6 +406,7 @@ void tcp_syn_ack_timeout(const struct request_sock *req); int tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int nonblock, int flags, int *addr_len); int tcp_set_rcvlowat(struct sock *sk, int val); +int tcp_set_window_clamp(struct sock *sk, int val); void tcp_data_ready(struct sock *sk); #ifdef CONFIG_MMU int tcp_mmap(struct file *file, struct socket *sock, diff --git a/net/core/filter.c b/net/core/filter.c index 21d91dcf0260..77001a35768f 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -4910,6 +4910,9 @@ static int _bpf_setsockopt(struct sock *sk, int level, int optname, tp->notsent_lowat = val; sk->sk_write_space(sk); break; + case TCP_WINDOW_CLAMP: + ret = tcp_set_window_clamp(sk, val); + break; default: ret = -EINVAL; } diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index b2bc3d7fe9e8..17379f6dd955 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -3022,6 +3022,21 @@ int tcp_sock_set_keepcnt(struct sock *sk, int val) } EXPORT_SYMBOL(tcp_sock_set_keepcnt); +int tcp_set_window_clamp(struct sock *sk, int val) +{ + struct tcp_sock *tp = tcp_sk(sk); + + if (!val) { + if (sk->sk_state != TCP_CLOSE) + return -EINVAL; + tp->window_clamp = 0; + } else { + tp->window_clamp = val < SOCK_MIN_RCVBUF / 2 ? + SOCK_MIN_RCVBUF / 2 : val; + } + return 0; +} + /* * Socket option code for TCP. */ @@ -3235,15 +3250,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level, int optname, break; case TCP_WINDOW_CLAMP: - if (!val) { - if (sk->sk_state != TCP_CLOSE) { - err = -EINVAL; - break; - } - tp->window_clamp = 0; - } else - tp->window_clamp = val < SOCK_MIN_RCVBUF / 2 ? - SOCK_MIN_RCVBUF / 2 : val; + err = tcp_set_window_clamp(sk, val); break; case TCP_QUICKACK: -- cgit v1.2.3 From bcd684aace34fedbd473fbd9b21ed06b0c2d2212 Mon Sep 17 00:00:00 2001 From: Bongsu Jeon Date: Thu, 3 Dec 2020 07:31:47 +0900 Subject: net/nfc/nci: Support NCI 2.x initial sequence implement the NCI 2.x initial sequence to support NCI 2.x NFCC. Since NCI 2.0, CORE_RESET and CORE_INIT sequence have been changed. If NFCEE supports NCI 2.x, then NCI 2.x initial sequence will work. In NCI 1.0, Initial sequence and payloads are as below: (DH) (NFCC) | -- CORE_RESET_CMD --> | | <-- CORE_RESET_RSP -- | | -- CORE_INIT_CMD --> | | <-- CORE_INIT_RSP -- | CORE_RESET_RSP payloads are Status, NCI version, Configuration Status. CORE_INIT_CMD payloads are empty. CORE_INIT_RSP payloads are Status, NFCC Features, Number of Supported RF Interfaces, Supported RF Interface, Max Logical Connections, Max Routing table Size, Max Control Packet Payload Size, Max Size for Large Parameters, Manufacturer ID, Manufacturer Specific Information. In NCI 2.0, Initial Sequence and Parameters are as below: (DH) (NFCC) | -- CORE_RESET_CMD --> | | <-- CORE_RESET_RSP -- | | <-- CORE_RESET_NTF -- | | -- CORE_INIT_CMD --> | | <-- CORE_INIT_RSP -- | CORE_RESET_RSP payloads are Status. CORE_RESET_NTF payloads are Reset Trigger, Configuration Status, NCI Version, Manufacturer ID, Manufacturer Specific Information Length, Manufacturer Specific Information. CORE_INIT_CMD payloads are Feature1, Feature2. CORE_INIT_RSP payloads are Status, NFCC Features, Max Logical Connections, Max Routing Table Size, Max Control Packet Payload Size, Max Data Packet Payload Size of the Static HCI Connection, Number of Credits of the Static HCI Connection, Max NFC-V RF Frame Size, Number of Supported RF Interfaces, Supported RF Interfaces. Signed-off-by: Bongsu Jeon Link: https://lore.kernel.org/r/20201202223147.3472-1-bongsu.jeon@samsung.com Signed-off-by: Jakub Kicinski --- include/net/nfc/nci.h | 34 +++++++++++++++++++++ net/nfc/nci/core.c | 18 ++++++++++-- net/nfc/nci/ntf.c | 21 +++++++++++++ net/nfc/nci/rsp.c | 81 ++++++++++++++++++++++++++++++++++++++++++--------- 4 files changed, 138 insertions(+), 16 deletions(-) (limited to 'include/net') diff --git a/include/net/nfc/nci.h b/include/net/nfc/nci.h index 0550e0380b8d..e82f55f543bb 100644 --- a/include/net/nfc/nci.h +++ b/include/net/nfc/nci.h @@ -25,6 +25,8 @@ #define NCI_MAX_PARAM_LEN 251 #define NCI_MAX_PAYLOAD_SIZE 255 #define NCI_MAX_PACKET_SIZE 258 +#define NCI_MAX_LARGE_PARAMS_NCI_v2 15 +#define NCI_VER_2_MASK 0x20 /* NCI Status Codes */ #define NCI_STATUS_OK 0x00 @@ -131,6 +133,9 @@ #define NCI_LF_CON_BITR_F_212 0x02 #define NCI_LF_CON_BITR_F_424 0x04 +/* NCI 2.x Feature Enable Bit */ +#define NCI_FEATURE_DISABLE 0x00 + /* NCI Reset types */ #define NCI_RESET_TYPE_KEEP_CONFIG 0x00 #define NCI_RESET_TYPE_RESET_CONFIG 0x01 @@ -220,6 +225,11 @@ struct nci_core_reset_cmd { } __packed; #define NCI_OP_CORE_INIT_CMD nci_opcode_pack(NCI_GID_CORE, 0x01) +/* To support NCI 2.x */ +struct nci_core_init_v2_cmd { + u8 feature1; + u8 feature2; +}; #define NCI_OP_CORE_SET_CONFIG_CMD nci_opcode_pack(NCI_GID_CORE, 0x02) struct set_config_param { @@ -334,6 +344,20 @@ struct nci_core_init_rsp_2 { __le32 manufact_specific_info; } __packed; +/* To support NCI ver 2.x */ +struct nci_core_init_rsp_nci_ver2 { + u8 status; + __le32 nfcc_features; + u8 max_logical_connections; + __le16 max_routing_table_size; + u8 max_ctrl_pkt_payload_len; + u8 max_data_pkt_hci_payload_len; + u8 number_of_hci_credit; + __le16 max_nfc_v_frame_size; + u8 num_supported_rf_interfaces; + u8 supported_rf_interfaces[]; +} __packed; + #define NCI_OP_CORE_SET_CONFIG_RSP nci_opcode_pack(NCI_GID_CORE, 0x02) struct nci_core_set_config_rsp { __u8 status; @@ -372,6 +396,16 @@ struct nci_nfcee_discover_rsp { /* --------------------------- */ /* ---- NCI Notifications ---- */ /* --------------------------- */ +#define NCI_OP_CORE_RESET_NTF nci_opcode_pack(NCI_GID_CORE, 0x00) +struct nci_core_reset_ntf { + u8 reset_trigger; + u8 config_status; + u8 nci_ver; + u8 manufact_id; + u8 manufacturer_specific_len; + __le32 manufact_specific_info; +} __packed; + #define NCI_OP_CORE_CONN_CREDITS_NTF nci_opcode_pack(NCI_GID_CORE, 0x06) struct conn_credit_entry { __u8 conn_id; diff --git a/net/nfc/nci/core.c b/net/nfc/nci/core.c index 4953ee5146e1..e64727e1a72f 100644 --- a/net/nfc/nci/core.c +++ b/net/nfc/nci/core.c @@ -165,7 +165,12 @@ static void nci_reset_req(struct nci_dev *ndev, unsigned long opt) static void nci_init_req(struct nci_dev *ndev, unsigned long opt) { - nci_send_cmd(ndev, NCI_OP_CORE_INIT_CMD, 0, NULL); + u8 plen = 0; + + if (opt) + plen = sizeof(struct nci_core_init_v2_cmd); + + nci_send_cmd(ndev, NCI_OP_CORE_INIT_CMD, plen, (void *)opt); } static void nci_init_complete_req(struct nci_dev *ndev, unsigned long opt) @@ -497,7 +502,16 @@ static int nci_open_device(struct nci_dev *ndev) } if (!rc) { - rc = __nci_request(ndev, nci_init_req, 0, + struct nci_core_init_v2_cmd nci_init_v2_cmd = { + .feature1 = NCI_FEATURE_DISABLE, + .feature2 = NCI_FEATURE_DISABLE + }; + unsigned long opt = 0; + + if (!(ndev->nci_ver & NCI_VER_2_MASK)) + opt = (unsigned long)&nci_init_v2_cmd; + + rc = __nci_request(ndev, nci_init_req, opt, msecs_to_jiffies(NCI_INIT_TIMEOUT)); } diff --git a/net/nfc/nci/ntf.c b/net/nfc/nci/ntf.c index 33e1170817f0..98af04c86b2c 100644 --- a/net/nfc/nci/ntf.c +++ b/net/nfc/nci/ntf.c @@ -27,6 +27,23 @@ /* Handle NCI Notification packets */ +static void nci_core_reset_ntf_packet(struct nci_dev *ndev, + struct sk_buff *skb) +{ + /* Handle NCI 2.x core reset notification */ + struct nci_core_reset_ntf *ntf = (void *)skb->data; + + ndev->nci_ver = ntf->nci_ver; + pr_debug("nci_ver 0x%x, config_status 0x%x\n", + ntf->nci_ver, ntf->config_status); + + ndev->manufact_id = ntf->manufact_id; + ndev->manufact_specific_info = + __le32_to_cpu(ntf->manufact_specific_info); + + nci_req_complete(ndev, NCI_STATUS_OK); +} + static void nci_core_conn_credits_ntf_packet(struct nci_dev *ndev, struct sk_buff *skb) { @@ -756,6 +773,10 @@ void nci_ntf_packet(struct nci_dev *ndev, struct sk_buff *skb) } switch (ntf_opcode) { + case NCI_OP_CORE_RESET_NTF: + nci_core_reset_ntf_packet(ndev, skb); + break; + case NCI_OP_CORE_CONN_CREDITS_NTF: nci_core_conn_credits_ntf_packet(ndev, skb); break; diff --git a/net/nfc/nci/rsp.c b/net/nfc/nci/rsp.c index a48297b79f34..e9605922a322 100644 --- a/net/nfc/nci/rsp.c +++ b/net/nfc/nci/rsp.c @@ -31,16 +31,19 @@ static void nci_core_reset_rsp_packet(struct nci_dev *ndev, struct sk_buff *skb) pr_debug("status 0x%x\n", rsp->status); - if (rsp->status == NCI_STATUS_OK) { - ndev->nci_ver = rsp->nci_ver; - pr_debug("nci_ver 0x%x, config_status 0x%x\n", - rsp->nci_ver, rsp->config_status); - } + /* Handle NCI 1.x ver */ + if (skb->len != 1) { + if (rsp->status == NCI_STATUS_OK) { + ndev->nci_ver = rsp->nci_ver; + pr_debug("nci_ver 0x%x, config_status 0x%x\n", + rsp->nci_ver, rsp->config_status); + } - nci_req_complete(ndev, rsp->status); + nci_req_complete(ndev, rsp->status); + } } -static void nci_core_init_rsp_packet(struct nci_dev *ndev, struct sk_buff *skb) +static u8 nci_core_init_rsp_packet_v1(struct nci_dev *ndev, struct sk_buff *skb) { struct nci_core_init_rsp_1 *rsp_1 = (void *) skb->data; struct nci_core_init_rsp_2 *rsp_2; @@ -48,16 +51,14 @@ static void nci_core_init_rsp_packet(struct nci_dev *ndev, struct sk_buff *skb) pr_debug("status 0x%x\n", rsp_1->status); if (rsp_1->status != NCI_STATUS_OK) - goto exit; + return rsp_1->status; ndev->nfcc_features = __le32_to_cpu(rsp_1->nfcc_features); ndev->num_supported_rf_interfaces = rsp_1->num_supported_rf_interfaces; - if (ndev->num_supported_rf_interfaces > - NCI_MAX_SUPPORTED_RF_INTERFACES) { - ndev->num_supported_rf_interfaces = - NCI_MAX_SUPPORTED_RF_INTERFACES; - } + ndev->num_supported_rf_interfaces = + min((int)ndev->num_supported_rf_interfaces, + NCI_MAX_SUPPORTED_RF_INTERFACES); memcpy(ndev->supported_rf_interfaces, rsp_1->supported_rf_interfaces, @@ -77,6 +78,58 @@ static void nci_core_init_rsp_packet(struct nci_dev *ndev, struct sk_buff *skb) ndev->manufact_specific_info = __le32_to_cpu(rsp_2->manufact_specific_info); + return NCI_STATUS_OK; +} + +static u8 nci_core_init_rsp_packet_v2(struct nci_dev *ndev, struct sk_buff *skb) +{ + struct nci_core_init_rsp_nci_ver2 *rsp = (void *)skb->data; + u8 *supported_rf_interface = rsp->supported_rf_interfaces; + u8 rf_interface_idx = 0; + u8 rf_extension_cnt = 0; + + pr_debug("status %x\n", rsp->status); + + if (rsp->status != NCI_STATUS_OK) + return rsp->status; + + ndev->nfcc_features = __le32_to_cpu(rsp->nfcc_features); + ndev->num_supported_rf_interfaces = rsp->num_supported_rf_interfaces; + + ndev->num_supported_rf_interfaces = + min((int)ndev->num_supported_rf_interfaces, + NCI_MAX_SUPPORTED_RF_INTERFACES); + + while (rf_interface_idx < ndev->num_supported_rf_interfaces) { + ndev->supported_rf_interfaces[rf_interface_idx++] = *supported_rf_interface++; + + /* skip rf extension parameters */ + rf_extension_cnt = *supported_rf_interface++; + supported_rf_interface += rf_extension_cnt; + } + + ndev->max_logical_connections = rsp->max_logical_connections; + ndev->max_routing_table_size = + __le16_to_cpu(rsp->max_routing_table_size); + ndev->max_ctrl_pkt_payload_len = + rsp->max_ctrl_pkt_payload_len; + ndev->max_size_for_large_params = NCI_MAX_LARGE_PARAMS_NCI_v2; + + return NCI_STATUS_OK; +} + +static void nci_core_init_rsp_packet(struct nci_dev *ndev, struct sk_buff *skb) +{ + u8 status = 0; + + if (!(ndev->nci_ver & NCI_VER_2_MASK)) + status = nci_core_init_rsp_packet_v1(ndev, skb); + else + status = nci_core_init_rsp_packet_v2(ndev, skb); + + if (status != NCI_STATUS_OK) + goto exit; + pr_debug("nfcc_features 0x%x\n", ndev->nfcc_features); pr_debug("num_supported_rf_interfaces %d\n", @@ -103,7 +156,7 @@ static void nci_core_init_rsp_packet(struct nci_dev *ndev, struct sk_buff *skb) ndev->manufact_specific_info); exit: - nci_req_complete(ndev, rsp_1->status); + nci_req_complete(ndev, status); } static void nci_core_set_config_rsp_packet(struct nci_dev *ndev, -- cgit v1.2.3 From 353021588cb57db15d52f9b157ad6f2251250b50 Mon Sep 17 00:00:00 2001 From: Reo Shiseki Date: Thu, 19 Nov 2020 16:37:11 +0900 Subject: Bluetooth: fix typo in struct name Signed-off-by: Reo Shiseki Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/mgmt.h | 4 ++-- net/bluetooth/mgmt.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/mgmt.h b/include/net/bluetooth/mgmt.h index 6b55155e05e9..d8367850e8cd 100644 --- a/include/net/bluetooth/mgmt.h +++ b/include/net/bluetooth/mgmt.h @@ -621,7 +621,7 @@ struct mgmt_cp_set_appearance { #define MGMT_SET_APPEARANCE_SIZE 2 #define MGMT_OP_GET_PHY_CONFIGURATION 0x0044 -struct mgmt_rp_get_phy_confguration { +struct mgmt_rp_get_phy_configuration { __le32 supported_phys; __le32 configurable_phys; __le32 selected_phys; @@ -658,7 +658,7 @@ struct mgmt_rp_get_phy_confguration { MGMT_PHY_LE_CODED_RX) #define MGMT_OP_SET_PHY_CONFIGURATION 0x0045 -struct mgmt_cp_set_phy_confguration { +struct mgmt_cp_set_phy_configuration { __le32 selected_phys; } __packed; #define MGMT_SET_PHY_CONFIGURATION_SIZE 4 diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 12d7b368b428..3dfed4efa078 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -3387,7 +3387,7 @@ static int set_appearance(struct sock *sk, struct hci_dev *hdev, void *data, static int get_phy_configuration(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { - struct mgmt_rp_get_phy_confguration rp; + struct mgmt_rp_get_phy_configuration rp; bt_dev_dbg(hdev, "sock %p", sk); @@ -3451,7 +3451,7 @@ unlock: static int set_phy_configuration(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { - struct mgmt_cp_set_phy_confguration *cp = data; + struct mgmt_cp_set_phy_configuration *cp = data; struct hci_cp_le_set_default_phy cp_phy; struct mgmt_pending_cmd *cmd; struct hci_request req; -- cgit v1.2.3 From c4f1f408168cd6a83d973e98e1cd1888e4d3d907 Mon Sep 17 00:00:00 2001 From: Howard Chung Date: Thu, 26 Nov 2020 12:22:21 +0800 Subject: Bluetooth: Interleave with allowlist scan This patch implements the interleaving between allowlist scan and no-filter scan. It'll be used to save power when at least one monitor is registered and at least one pending connection or one device to be scanned for. The durations of the allowlist scan and the no-filter scan are controlled by MGMT command: Set Default System Configuration. The default values are set randomly for now. Signed-off-by: Howard Chung Reviewed-by: Alain Michaud Reviewed-by: Manish Mandlik Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/hci_core.h | 10 +++ net/bluetooth/hci_core.c | 3 + net/bluetooth/hci_request.c | 128 ++++++++++++++++++++++++++++++++++++--- net/bluetooth/mgmt_config.c | 10 +++ 4 files changed, 144 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 9873e1c8cd16..cfede18709d8 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -361,6 +361,8 @@ struct hci_dev { __u8 ssp_debug_mode; __u8 hw_error_code; __u32 clock; + __u16 advmon_allowlist_duration; + __u16 advmon_no_filter_duration; __u16 devid_source; __u16 devid_vendor; @@ -542,6 +544,14 @@ struct hci_dev { struct delayed_work rpa_expired; bdaddr_t rpa; + enum { + INTERLEAVE_SCAN_NONE, + INTERLEAVE_SCAN_NO_FILTER, + INTERLEAVE_SCAN_ALLOWLIST + } interleave_scan_state; + + struct delayed_work interleave_scan; + #if IS_ENABLED(CONFIG_BT_LEDS) struct led_trigger *power_led; #endif diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index c4aa2cbb9269..20506b31492d 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3592,6 +3592,9 @@ struct hci_dev *hci_alloc_dev(void) hdev->cur_adv_instance = 0x00; hdev->adv_instance_timeout = 0; + hdev->advmon_allowlist_duration = 300; + hdev->advmon_no_filter_duration = 500; + hdev->sniff_max_interval = 800; hdev->sniff_min_interval = 80; diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index fb87882fb71a..d6bf1517ddae 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -378,6 +378,53 @@ void __hci_req_write_fast_connectable(struct hci_request *req, bool enable) hci_req_add(req, HCI_OP_WRITE_PAGE_SCAN_TYPE, 1, &type); } +static void start_interleave_scan(struct hci_dev *hdev) +{ + hdev->interleave_scan_state = INTERLEAVE_SCAN_NO_FILTER; + queue_delayed_work(hdev->req_workqueue, + &hdev->interleave_scan, 0); +} + +static bool is_interleave_scanning(struct hci_dev *hdev) +{ + return hdev->interleave_scan_state != INTERLEAVE_SCAN_NONE; +} + +static void cancel_interleave_scan(struct hci_dev *hdev) +{ + bt_dev_dbg(hdev, "cancelling interleave scan"); + + cancel_delayed_work_sync(&hdev->interleave_scan); + + hdev->interleave_scan_state = INTERLEAVE_SCAN_NONE; +} + +/* Return true if interleave_scan wasn't started until exiting this function, + * otherwise, return false + */ +static bool __hci_update_interleaved_scan(struct hci_dev *hdev) +{ + /* If there is at least one ADV monitors and one pending LE connection + * or one device to be scanned for, we should alternate between + * allowlist scan and one without any filters to save power. + */ + bool use_interleaving = hci_is_adv_monitoring(hdev) && + !(list_empty(&hdev->pend_le_conns) && + list_empty(&hdev->pend_le_reports)); + bool is_interleaving = is_interleave_scanning(hdev); + + if (use_interleaving && !is_interleaving) { + start_interleave_scan(hdev); + bt_dev_dbg(hdev, "starting interleave scan"); + return true; + } + + if (!use_interleaving && is_interleaving) + cancel_interleave_scan(hdev); + + return false; +} + /* This function controls the background scanning based on hdev->pend_le_conns * list. If there are pending LE connection we start the background scanning, * otherwise we stop it. @@ -450,8 +497,7 @@ static void __hci_update_background_scan(struct hci_request *req) hci_req_add_le_scan_disable(req, false); hci_req_add_le_passive_scan(req); - - BT_DBG("%s starting background scanning", hdev->name); + bt_dev_dbg(hdev, "starting background scanning"); } } @@ -848,12 +894,17 @@ static u8 update_white_list(struct hci_request *req) return 0x00; } - /* Once the controller offloading of advertisement monitor is in place, - * the if condition should include the support of MSFT extension - * support. If suspend is ongoing, whitelist should be the default to - * prevent waking by random advertisements. + /* Use the allowlist unless the following conditions are all true: + * - We are not currently suspending + * - There are 1 or more ADV monitors registered + * - Interleaved scanning is not currently using the allowlist + * + * Once the controller offloading of advertisement monitor is in place, + * the above condition should include the support of MSFT extension + * support. */ - if (!idr_is_empty(&hdev->adv_monitors_idr) && !hdev->suspended) + if (!idr_is_empty(&hdev->adv_monitors_idr) && !hdev->suspended && + hdev->interleave_scan_state != INTERLEAVE_SCAN_ALLOWLIST) return 0x00; /* Select filter policy to use white list */ @@ -1006,6 +1057,10 @@ void hci_req_add_le_passive_scan(struct hci_request *req) &own_addr_type)) return; + if (__hci_update_interleaved_scan(hdev)) + return; + + bt_dev_dbg(hdev, "interleave state %d", hdev->interleave_scan_state); /* Adding or removing entries from the white list must * happen before enabling scanning. The controller does * not allow white list modification while scanning. @@ -1870,6 +1925,62 @@ unlock: hci_dev_unlock(hdev); } +static int hci_req_add_le_interleaved_scan(struct hci_request *req, + unsigned long opt) +{ + struct hci_dev *hdev = req->hdev; + int ret = 0; + + hci_dev_lock(hdev); + + if (hci_dev_test_flag(hdev, HCI_LE_SCAN)) + hci_req_add_le_scan_disable(req, false); + hci_req_add_le_passive_scan(req); + + switch (hdev->interleave_scan_state) { + case INTERLEAVE_SCAN_ALLOWLIST: + bt_dev_dbg(hdev, "next state: allowlist"); + hdev->interleave_scan_state = INTERLEAVE_SCAN_NO_FILTER; + break; + case INTERLEAVE_SCAN_NO_FILTER: + bt_dev_dbg(hdev, "next state: no filter"); + hdev->interleave_scan_state = INTERLEAVE_SCAN_ALLOWLIST; + break; + case INTERLEAVE_SCAN_NONE: + BT_ERR("unexpected error"); + ret = -1; + } + + hci_dev_unlock(hdev); + + return ret; +} + +static void interleave_scan_work(struct work_struct *work) +{ + struct hci_dev *hdev = container_of(work, struct hci_dev, + interleave_scan.work); + u8 status; + unsigned long timeout; + + if (hdev->interleave_scan_state == INTERLEAVE_SCAN_ALLOWLIST) { + timeout = msecs_to_jiffies(hdev->advmon_allowlist_duration); + } else if (hdev->interleave_scan_state == INTERLEAVE_SCAN_NO_FILTER) { + timeout = msecs_to_jiffies(hdev->advmon_no_filter_duration); + } else { + bt_dev_err(hdev, "unexpected error"); + return; + } + + hci_req_sync(hdev, hci_req_add_le_interleaved_scan, 0, + HCI_CMD_TIMEOUT, &status); + + /* Don't continue interleaving if it was canceled */ + if (is_interleave_scanning(hdev)) + queue_delayed_work(hdev->req_workqueue, + &hdev->interleave_scan, timeout); +} + int hci_get_random_address(struct hci_dev *hdev, bool require_privacy, bool use_rpa, struct adv_info *adv_instance, u8 *own_addr_type, bdaddr_t *rand_addr) @@ -3297,6 +3408,7 @@ void hci_request_setup(struct hci_dev *hdev) INIT_DELAYED_WORK(&hdev->le_scan_disable, le_scan_disable_work); INIT_DELAYED_WORK(&hdev->le_scan_restart, le_scan_restart_work); INIT_DELAYED_WORK(&hdev->adv_instance_expire, adv_timeout_expire); + INIT_DELAYED_WORK(&hdev->interleave_scan, interleave_scan_work); } void hci_request_cancel_all(struct hci_dev *hdev) @@ -3316,4 +3428,6 @@ void hci_request_cancel_all(struct hci_dev *hdev) cancel_delayed_work_sync(&hdev->adv_instance_expire); hdev->adv_instance_timeout = 0; } + + cancel_interleave_scan(hdev); } diff --git a/net/bluetooth/mgmt_config.c b/net/bluetooth/mgmt_config.c index b30b571f8caf..2d3ad288c78a 100644 --- a/net/bluetooth/mgmt_config.c +++ b/net/bluetooth/mgmt_config.c @@ -67,6 +67,8 @@ int read_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, HDEV_PARAM_U16(0x001a, le_supv_timeout), HDEV_PARAM_U16_JIFFIES_TO_MSECS(0x001b, def_le_autoconnect_timeout), + HDEV_PARAM_U16(0x001d, advmon_allowlist_duration), + HDEV_PARAM_U16(0x001e, advmon_no_filter_duration), }; struct mgmt_rp_read_def_system_config *rp = (void *)params; @@ -138,6 +140,8 @@ int set_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, case 0x0019: case 0x001a: case 0x001b: + case 0x001d: + case 0x001e: if (len != sizeof(u16)) { bt_dev_warn(hdev, "invalid length %d, exp %zu for type %d", len, sizeof(u16), type); @@ -251,6 +255,12 @@ int set_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, hdev->def_le_autoconnect_timeout = msecs_to_jiffies(TLV_GET_LE16(buffer)); break; + case 0x0001d: + hdev->advmon_allowlist_duration = TLV_GET_LE16(buffer); + break; + case 0x0001e: + hdev->advmon_no_filter_duration = TLV_GET_LE16(buffer); + break; default: bt_dev_warn(hdev, "unsupported parameter %u", type); break; -- cgit v1.2.3 From 80af16a3e473f0789d205810733a513279e5b6f9 Mon Sep 17 00:00:00 2001 From: Howard Chung Date: Thu, 26 Nov 2020 12:22:25 +0800 Subject: Bluetooth: Add toggle to switch off interleave scan This patch add a configurable parameter to switch off the interleave scan feature. Signed-off-by: Howard Chung Reviewed-by: Alain Michaud Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/hci_core.h | 1 + net/bluetooth/hci_core.c | 1 + net/bluetooth/hci_request.c | 3 ++- net/bluetooth/mgmt_config.c | 41 ++++++++++++++++++++++++++++++++-------- 4 files changed, 37 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index cfede18709d8..63c6d656564a 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -363,6 +363,7 @@ struct hci_dev { __u32 clock; __u16 advmon_allowlist_duration; __u16 advmon_no_filter_duration; + __u8 enable_advmon_interleave_scan; __u16 devid_source; __u16 devid_vendor; diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 20506b31492d..8cfcf43eb08f 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3594,6 +3594,7 @@ struct hci_dev *hci_alloc_dev(void) hdev->advmon_allowlist_duration = 300; hdev->advmon_no_filter_duration = 500; + hdev->enable_advmon_interleave_scan = 0x00; /* Default to disable */ hdev->sniff_max_interval = 800; hdev->sniff_min_interval = 80; diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index 0c326e32e240..d0d0fbbb3fa5 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -1057,7 +1057,8 @@ void hci_req_add_le_passive_scan(struct hci_request *req) &own_addr_type)) return; - if (__hci_update_interleaved_scan(hdev)) + if (hdev->enable_advmon_interleave_scan && + __hci_update_interleaved_scan(hdev)) return; bt_dev_dbg(hdev, "interleave state %d", hdev->interleave_scan_state); diff --git a/net/bluetooth/mgmt_config.c b/net/bluetooth/mgmt_config.c index 282fbf82f319..1deb0ca7a929 100644 --- a/net/bluetooth/mgmt_config.c +++ b/net/bluetooth/mgmt_config.c @@ -17,12 +17,24 @@ __le16 value; \ } __packed _param_name_ +#define HDEV_PARAM_U8(_param_name_) \ + struct {\ + struct mgmt_tlv entry; \ + __u8 value; \ + } __packed _param_name_ + #define TLV_SET_U16(_param_code_, _param_name_) \ { \ { cpu_to_le16(_param_code_), sizeof(__u16) }, \ cpu_to_le16(hdev->_param_name_) \ } +#define TLV_SET_U8(_param_code_, _param_name_) \ + { \ + { cpu_to_le16(_param_code_), sizeof(__u8) }, \ + hdev->_param_name_ \ + } + #define TLV_SET_U16_JIFFIES_TO_MSECS(_param_code_, _param_name_) \ { \ { cpu_to_le16(_param_code_), sizeof(__u16) }, \ @@ -65,6 +77,7 @@ int read_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, HDEV_PARAM_U16(def_le_autoconnect_timeout); HDEV_PARAM_U16(advmon_allowlist_duration); HDEV_PARAM_U16(advmon_no_filter_duration); + HDEV_PARAM_U8(enable_advmon_interleave_scan); } __packed rp = { TLV_SET_U16(0x0000, def_page_scan_type), TLV_SET_U16(0x0001, def_page_scan_int), @@ -97,6 +110,7 @@ int read_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, def_le_autoconnect_timeout), TLV_SET_U16(0x001d, advmon_allowlist_duration), TLV_SET_U16(0x001e, advmon_no_filter_duration), + TLV_SET_U8(0x001f, enable_advmon_interleave_scan), }; bt_dev_dbg(hdev, "sock %p", sk); @@ -109,6 +123,7 @@ int read_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, #define TO_TLV(x) ((struct mgmt_tlv *)(x)) #define TLV_GET_LE16(tlv) le16_to_cpu(*((__le16 *)(TO_TLV(tlv)->value))) +#define TLV_GET_U8(tlv) (*((__u8 *)(TO_TLV(tlv)->value))) int set_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, u16 data_len) @@ -125,6 +140,7 @@ int set_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, /* First pass to validate the tlv */ while (buffer_left >= sizeof(struct mgmt_tlv)) { const u8 len = TO_TLV(buffer)->length; + size_t exp_type_len; const u16 exp_len = sizeof(struct mgmt_tlv) + len; const u16 type = le16_to_cpu(TO_TLV(buffer)->type); @@ -170,20 +186,26 @@ int set_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, case 0x001b: case 0x001d: case 0x001e: - if (len != sizeof(u16)) { - bt_dev_warn(hdev, "invalid length %d, exp %zu for type %d", - len, sizeof(u16), type); - - return mgmt_cmd_status(sk, hdev->id, - MGMT_OP_SET_DEF_SYSTEM_CONFIG, - MGMT_STATUS_INVALID_PARAMS); - } + exp_type_len = sizeof(u16); + break; + case 0x001f: + exp_type_len = sizeof(u8); break; default: + exp_type_len = 0; bt_dev_warn(hdev, "unsupported parameter %u", type); break; } + if (exp_type_len && len != exp_type_len) { + bt_dev_warn(hdev, "invalid length %d, exp %zu for type %d", + len, exp_type_len, type); + + return mgmt_cmd_status(sk, hdev->id, + MGMT_OP_SET_DEF_SYSTEM_CONFIG, + MGMT_STATUS_INVALID_PARAMS); + } + buffer_left -= exp_len; buffer += exp_len; } @@ -289,6 +311,9 @@ int set_def_system_config(struct sock *sk, struct hci_dev *hdev, void *data, case 0x0001e: hdev->advmon_no_filter_duration = TLV_GET_LE16(buffer); break; + case 0x0001f: + hdev->enable_advmon_interleave_scan = TLV_GET_U8(buffer); + break; default: bt_dev_warn(hdev, "unsupported parameter %u", type); break; -- cgit v1.2.3 From 31aab5c22e14c1c10110281d7f74b5e554f731b7 Mon Sep 17 00:00:00 2001 From: Daniel Winkler Date: Thu, 3 Dec 2020 12:12:48 -0800 Subject: Bluetooth: Add helper to set adv data We wish to handle advertising data separately from advertising parameters in our new MGMT requests. This change adds a helper that allows the advertising data and scan response to be updated for an existing advertising instance. Reviewed-by: Sonny Sasaka Signed-off-by: Daniel Winkler Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/hci_core.h | 3 +++ net/bluetooth/hci_core.c | 31 +++++++++++++++++++++++++++++++ 2 files changed, 34 insertions(+) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 63c6d656564a..4bf0b6a04aea 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -1302,6 +1302,9 @@ int hci_add_adv_instance(struct hci_dev *hdev, u8 instance, u32 flags, u16 adv_data_len, u8 *adv_data, u16 scan_rsp_len, u8 *scan_rsp_data, u16 timeout, u16 duration); +int hci_set_adv_instance_data(struct hci_dev *hdev, u8 instance, + u16 adv_data_len, u8 *adv_data, + u16 scan_rsp_len, u8 *scan_rsp_data); int hci_remove_adv_instance(struct hci_dev *hdev, u8 instance); void hci_adv_instances_set_rpa_expired(struct hci_dev *hdev, bool rpa_expired); diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 8cfcf43eb08f..46ec523d96a7 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3005,6 +3005,37 @@ int hci_add_adv_instance(struct hci_dev *hdev, u8 instance, u32 flags, return 0; } +/* This function requires the caller holds hdev->lock */ +int hci_set_adv_instance_data(struct hci_dev *hdev, u8 instance, + u16 adv_data_len, u8 *adv_data, + u16 scan_rsp_len, u8 *scan_rsp_data) +{ + struct adv_info *adv_instance; + + adv_instance = hci_find_adv_instance(hdev, instance); + + /* If advertisement doesn't exist, we can't modify its data */ + if (!adv_instance) + return -ENOENT; + + if (adv_data_len) { + memset(adv_instance->adv_data, 0, + sizeof(adv_instance->adv_data)); + memcpy(adv_instance->adv_data, adv_data, adv_data_len); + adv_instance->adv_data_len = adv_data_len; + } + + if (scan_rsp_len) { + memset(adv_instance->scan_rsp_data, 0, + sizeof(adv_instance->scan_rsp_data)); + memcpy(adv_instance->scan_rsp_data, + scan_rsp_data, scan_rsp_len); + adv_instance->scan_rsp_len = scan_rsp_len; + } + + return 0; +} + /* This function requires the caller holds hdev->lock */ void hci_adv_monitors_clear(struct hci_dev *hdev) { -- cgit v1.2.3 From 12410572833a283ce92fcf9679ca8a2f372097ee Mon Sep 17 00:00:00 2001 From: Daniel Winkler Date: Thu, 3 Dec 2020 12:12:49 -0800 Subject: Bluetooth: Break add adv into two mgmt commands This patch adds support for the new advertising add interface, with the first command setting advertising parameters and the second to set advertising data. The set parameters command allows the caller to leave some fields "unset", with a params bitfield defining which params were purposefully set. Unset parameters will be given defaults when calling hci_add_adv_instance. The data passed to the param mgmt command is allowed to be flexible, so in the future if bluetoothd passes a larger structure with new params, the mgmt command will ignore the unknown members at the end. This change has been validated on both hatch (extended advertising) and kukui (no extended advertising) chromebooks running bluetoothd that support this new interface. I ran the following manual tests: - Set several (3) advertisements using modified test_advertisement.py - For each, validate correct data and parameters in btmon trace - Verified both for software rotation and extended adv Automatic test suite also run, testing many (25) scenarios of single and multi-advertising for data/parameter correctness. Reviewed-by: Sonny Sasaka Signed-off-by: Daniel Winkler Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/hci_core.h | 2 + include/net/bluetooth/mgmt.h | 34 ++++ net/bluetooth/hci_event.c | 1 + net/bluetooth/mgmt.c | 381 +++++++++++++++++++++++++++++++++++++-- 4 files changed, 407 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 4bf0b6a04aea..3c2e205f226a 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -238,6 +238,8 @@ struct adv_info { #define HCI_MAX_ADV_INSTANCES 5 #define HCI_DEFAULT_ADV_DURATION 2 +#define HCI_ADV_TX_POWER_NO_PREFERENCE 0x7F + struct adv_pattern { struct list_head list; __u8 ad_type; diff --git a/include/net/bluetooth/mgmt.h b/include/net/bluetooth/mgmt.h index d8367850e8cd..2e18e4173e2f 100644 --- a/include/net/bluetooth/mgmt.h +++ b/include/net/bluetooth/mgmt.h @@ -574,6 +574,10 @@ struct mgmt_rp_add_advertising { #define MGMT_ADV_FLAG_SEC_CODED BIT(9) #define MGMT_ADV_FLAG_CAN_SET_TX_POWER BIT(10) #define MGMT_ADV_FLAG_HW_OFFLOAD BIT(11) +#define MGMT_ADV_PARAM_DURATION BIT(12) +#define MGMT_ADV_PARAM_TIMEOUT BIT(13) +#define MGMT_ADV_PARAM_INTERVALS BIT(14) +#define MGMT_ADV_PARAM_TX_POWER BIT(15) #define MGMT_ADV_FLAG_SEC_MASK (MGMT_ADV_FLAG_SEC_1M | MGMT_ADV_FLAG_SEC_2M | \ MGMT_ADV_FLAG_SEC_CODED) @@ -782,6 +786,36 @@ struct mgmt_rp_remove_adv_monitor { __le16 monitor_handle; } __packed; +#define MGMT_OP_ADD_EXT_ADV_PARAMS 0x0054 +struct mgmt_cp_add_ext_adv_params { + __u8 instance; + __le32 flags; + __le16 duration; + __le16 timeout; + __le32 min_interval; + __le32 max_interval; + __s8 tx_power; +} __packed; +#define MGMT_ADD_EXT_ADV_PARAMS_MIN_SIZE 18 +struct mgmt_rp_add_ext_adv_params { + __u8 instance; + __s8 tx_power; + __u8 max_adv_data_len; + __u8 max_scan_rsp_len; +} __packed; + +#define MGMT_OP_ADD_EXT_ADV_DATA 0x0055 +struct mgmt_cp_add_ext_adv_data { + __u8 instance; + __u8 adv_data_len; + __u8 scan_rsp_len; + __u8 data[]; +} __packed; +#define MGMT_ADD_EXT_ADV_DATA_SIZE 3 +struct mgmt_rp_add_ext_adv_data { + __u8 instance; +} __packed; + #define MGMT_EV_CMD_COMPLETE 0x0001 struct mgmt_ev_cmd_complete { __le16 opcode; diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 8281a5ce0f73..f193e73ef47c 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -1752,6 +1752,7 @@ static void hci_cc_set_ext_adv_param(struct hci_dev *hdev, struct sk_buff *skb) } /* Update adv data as tx power is known now */ hci_req_update_adv_data(hdev, hdev->cur_adv_instance); + hci_dev_unlock(hdev); } diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 3dfed4efa078..ec6b520be368 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -122,6 +122,8 @@ static const u16 mgmt_commands[] = { MGMT_OP_READ_ADV_MONITOR_FEATURES, MGMT_OP_ADD_ADV_PATTERNS_MONITOR, MGMT_OP_REMOVE_ADV_MONITOR, + MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_OP_ADD_EXT_ADV_DATA, }; static const u16 mgmt_events[] = { @@ -7203,6 +7205,10 @@ static u32 get_supported_adv_flags(struct hci_dev *hdev) flags |= MGMT_ADV_FLAG_MANAGED_FLAGS; flags |= MGMT_ADV_FLAG_APPEARANCE; flags |= MGMT_ADV_FLAG_LOCAL_NAME; + flags |= MGMT_ADV_PARAM_DURATION; + flags |= MGMT_ADV_PARAM_TIMEOUT; + flags |= MGMT_ADV_PARAM_INTERVALS; + flags |= MGMT_ADV_PARAM_TX_POWER; /* In extended adv TX_POWER returned from Set Adv Param * will be always valid. @@ -7377,6 +7383,31 @@ static bool tlv_data_is_valid(struct hci_dev *hdev, u32 adv_flags, u8 *data, return true; } +static bool requested_adv_flags_are_valid(struct hci_dev *hdev, u32 adv_flags) +{ + u32 supported_flags, phy_flags; + + /* The current implementation only supports a subset of the specified + * flags. Also need to check mutual exclusiveness of sec flags. + */ + supported_flags = get_supported_adv_flags(hdev); + phy_flags = adv_flags & MGMT_ADV_FLAG_SEC_MASK; + if (adv_flags & ~supported_flags || + ((phy_flags && (phy_flags ^ (phy_flags & -phy_flags))))) + return false; + + return true; +} + +static bool adv_busy(struct hci_dev *hdev) +{ + return (pending_find(MGMT_OP_ADD_ADVERTISING, hdev) || + pending_find(MGMT_OP_REMOVE_ADVERTISING, hdev) || + pending_find(MGMT_OP_SET_LE, hdev) || + pending_find(MGMT_OP_ADD_EXT_ADV_PARAMS, hdev) || + pending_find(MGMT_OP_ADD_EXT_ADV_DATA, hdev)); +} + static void add_advertising_complete(struct hci_dev *hdev, u8 status, u16 opcode) { @@ -7391,6 +7422,8 @@ static void add_advertising_complete(struct hci_dev *hdev, u8 status, hci_dev_lock(hdev); cmd = pending_find(MGMT_OP_ADD_ADVERTISING, hdev); + if (!cmd) + cmd = pending_find(MGMT_OP_ADD_EXT_ADV_DATA, hdev); list_for_each_entry_safe(adv_instance, n, &hdev->adv_instances, list) { if (!adv_instance->pending) @@ -7435,7 +7468,6 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, struct mgmt_cp_add_advertising *cp = data; struct mgmt_rp_add_advertising rp; u32 flags; - u32 supported_flags, phy_flags; u8 status; u16 timeout, duration; unsigned int prev_instance_cnt = hdev->adv_instance_cnt; @@ -7471,13 +7503,7 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, timeout = __le16_to_cpu(cp->timeout); duration = __le16_to_cpu(cp->duration); - /* The current implementation only supports a subset of the specified - * flags. Also need to check mutual exclusiveness of sec flags. - */ - supported_flags = get_supported_adv_flags(hdev); - phy_flags = flags & MGMT_ADV_FLAG_SEC_MASK; - if (flags & ~supported_flags || - ((phy_flags && (phy_flags ^ (phy_flags & -phy_flags))))) + if (!requested_adv_flags_are_valid(hdev, flags)) return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, MGMT_STATUS_INVALID_PARAMS); @@ -7489,9 +7515,7 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, goto unlock; } - if (pending_find(MGMT_OP_ADD_ADVERTISING, hdev) || - pending_find(MGMT_OP_REMOVE_ADVERTISING, hdev) || - pending_find(MGMT_OP_SET_LE, hdev)) { + if (adv_busy(hdev)) { err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, MGMT_STATUS_BUSY); goto unlock; @@ -7582,6 +7606,337 @@ unlock: return err; } +static void add_ext_adv_params_complete(struct hci_dev *hdev, u8 status, + u16 opcode) +{ + struct mgmt_pending_cmd *cmd; + struct mgmt_cp_add_ext_adv_params *cp; + struct mgmt_rp_add_ext_adv_params rp; + struct adv_info *adv_instance; + u32 flags; + + BT_DBG("%s", hdev->name); + + hci_dev_lock(hdev); + + cmd = pending_find(MGMT_OP_ADD_EXT_ADV_PARAMS, hdev); + if (!cmd) + goto unlock; + + cp = cmd->param; + adv_instance = hci_find_adv_instance(hdev, cp->instance); + if (!adv_instance) + goto unlock; + + rp.instance = cp->instance; + rp.tx_power = adv_instance->tx_power; + + /* While we're at it, inform userspace of the available space for this + * advertisement, given the flags that will be used. + */ + flags = __le32_to_cpu(cp->flags); + rp.max_adv_data_len = tlv_data_max_len(hdev, flags, true); + rp.max_scan_rsp_len = tlv_data_max_len(hdev, flags, false); + + if (status) { + /* If this advertisement was previously advertising and we + * failed to update it, we signal that it has been removed and + * delete its structure + */ + if (!adv_instance->pending) + mgmt_advertising_removed(cmd->sk, hdev, cp->instance); + + hci_remove_adv_instance(hdev, cp->instance); + + mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, + mgmt_status(status)); + + } else { + mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, + mgmt_status(status), &rp, sizeof(rp)); + } + +unlock: + if (cmd) + mgmt_pending_remove(cmd); + + hci_dev_unlock(hdev); +} + +static int add_ext_adv_params(struct sock *sk, struct hci_dev *hdev, + void *data, u16 data_len) +{ + struct mgmt_cp_add_ext_adv_params *cp = data; + struct mgmt_rp_add_ext_adv_params rp; + struct mgmt_pending_cmd *cmd = NULL; + struct adv_info *adv_instance; + struct hci_request req; + u32 flags, min_interval, max_interval; + u16 timeout, duration; + u8 status; + s8 tx_power; + int err; + + BT_DBG("%s", hdev->name); + + status = mgmt_le_support(hdev); + if (status) + return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, + status); + + if (cp->instance < 1 || cp->instance > hdev->le_num_of_adv_sets) + return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_STATUS_INVALID_PARAMS); + + /* The purpose of breaking add_advertising into two separate MGMT calls + * for params and data is to allow more parameters to be added to this + * structure in the future. For this reason, we verify that we have the + * bare minimum structure we know of when the interface was defined. Any + * extra parameters we don't know about will be ignored in this request. + */ + if (data_len < MGMT_ADD_EXT_ADV_PARAMS_MIN_SIZE) + return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, + MGMT_STATUS_INVALID_PARAMS); + + flags = __le32_to_cpu(cp->flags); + + if (!requested_adv_flags_are_valid(hdev, flags)) + return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_STATUS_INVALID_PARAMS); + + hci_dev_lock(hdev); + + /* In new interface, we require that we are powered to register */ + if (!hdev_is_powered(hdev)) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_STATUS_REJECTED); + goto unlock; + } + + if (adv_busy(hdev)) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_STATUS_BUSY); + goto unlock; + } + + /* Parse defined parameters from request, use defaults otherwise */ + timeout = (flags & MGMT_ADV_PARAM_TIMEOUT) ? + __le16_to_cpu(cp->timeout) : 0; + + duration = (flags & MGMT_ADV_PARAM_DURATION) ? + __le16_to_cpu(cp->duration) : + hdev->def_multi_adv_rotation_duration; + + min_interval = (flags & MGMT_ADV_PARAM_INTERVALS) ? + __le32_to_cpu(cp->min_interval) : + hdev->le_adv_min_interval; + + max_interval = (flags & MGMT_ADV_PARAM_INTERVALS) ? + __le32_to_cpu(cp->max_interval) : + hdev->le_adv_max_interval; + + tx_power = (flags & MGMT_ADV_PARAM_TX_POWER) ? + cp->tx_power : + HCI_ADV_TX_POWER_NO_PREFERENCE; + + /* Create advertising instance with no advertising or response data */ + err = hci_add_adv_instance(hdev, cp->instance, flags, + 0, NULL, 0, NULL, timeout, duration); + + if (err < 0) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_STATUS_FAILED); + goto unlock; + } + + hdev->cur_adv_instance = cp->instance; + /* Submit request for advertising params if ext adv available */ + if (ext_adv_capable(hdev)) { + hci_req_init(&req, hdev); + adv_instance = hci_find_adv_instance(hdev, cp->instance); + + /* Updating parameters of an active instance will return a + * Command Disallowed error, so we must first disable the + * instance if it is active. + */ + if (!adv_instance->pending) + __hci_req_disable_ext_adv_instance(&req, cp->instance); + + __hci_req_setup_ext_adv_instance(&req, cp->instance); + + err = hci_req_run(&req, add_ext_adv_params_complete); + + if (!err) + cmd = mgmt_pending_add(sk, MGMT_OP_ADD_EXT_ADV_PARAMS, + hdev, data, data_len); + if (!cmd) { + err = -ENOMEM; + hci_remove_adv_instance(hdev, cp->instance); + goto unlock; + } + + } else { + rp.instance = cp->instance; + rp.tx_power = HCI_ADV_TX_POWER_NO_PREFERENCE; + rp.max_adv_data_len = tlv_data_max_len(hdev, flags, true); + rp.max_scan_rsp_len = tlv_data_max_len(hdev, flags, false); + err = mgmt_cmd_complete(sk, hdev->id, + MGMT_OP_ADD_EXT_ADV_PARAMS, + MGMT_STATUS_SUCCESS, &rp, sizeof(rp)); + } + +unlock: + hci_dev_unlock(hdev); + + return err; +} + +static int add_ext_adv_data(struct sock *sk, struct hci_dev *hdev, void *data, + u16 data_len) +{ + struct mgmt_cp_add_ext_adv_data *cp = data; + struct mgmt_rp_add_ext_adv_data rp; + u8 schedule_instance = 0; + struct adv_info *next_instance; + struct adv_info *adv_instance; + int err = 0; + struct mgmt_pending_cmd *cmd; + struct hci_request req; + + BT_DBG("%s", hdev->name); + + hci_dev_lock(hdev); + + adv_instance = hci_find_adv_instance(hdev, cp->instance); + + if (!adv_instance) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_INVALID_PARAMS); + goto unlock; + } + + /* In new interface, we require that we are powered to register */ + if (!hdev_is_powered(hdev)) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_REJECTED); + goto clear_new_instance; + } + + if (adv_busy(hdev)) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_BUSY); + goto clear_new_instance; + } + + /* Validate new data */ + if (!tlv_data_is_valid(hdev, adv_instance->flags, cp->data, + cp->adv_data_len, true) || + !tlv_data_is_valid(hdev, adv_instance->flags, cp->data + + cp->adv_data_len, cp->scan_rsp_len, false)) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_INVALID_PARAMS); + goto clear_new_instance; + } + + /* Set the data in the advertising instance */ + hci_set_adv_instance_data(hdev, cp->instance, cp->adv_data_len, + cp->data, cp->scan_rsp_len, + cp->data + cp->adv_data_len); + + /* We're good to go, update advertising data, parameters, and start + * advertising. + */ + + hci_req_init(&req, hdev); + + hci_req_add(&req, HCI_OP_READ_LOCAL_NAME, 0, NULL); + + if (ext_adv_capable(hdev)) { + __hci_req_update_adv_data(&req, cp->instance); + __hci_req_update_scan_rsp_data(&req, cp->instance); + __hci_req_enable_ext_advertising(&req, cp->instance); + + } else { + /* If using software rotation, determine next instance to use */ + + if (hdev->cur_adv_instance == cp->instance) { + /* If the currently advertised instance is being changed + * then cancel the current advertising and schedule the + * next instance. If there is only one instance then the + * overridden advertising data will be visible right + * away + */ + cancel_adv_timeout(hdev); + + next_instance = hci_get_next_instance(hdev, + cp->instance); + if (next_instance) + schedule_instance = next_instance->instance; + } else if (!hdev->adv_instance_timeout) { + /* Immediately advertise the new instance if no other + * instance is currently being advertised. + */ + schedule_instance = cp->instance; + } + + /* If the HCI_ADVERTISING flag is set or there is no instance to + * be advertised then we have no HCI communication to make. + * Simply return. + */ + if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || + !schedule_instance) { + if (adv_instance->pending) { + mgmt_advertising_added(sk, hdev, cp->instance); + adv_instance->pending = false; + } + rp.instance = cp->instance; + err = mgmt_cmd_complete(sk, hdev->id, + MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_SUCCESS, &rp, + sizeof(rp)); + goto unlock; + } + + err = __hci_req_schedule_adv_instance(&req, schedule_instance, + true); + } + + cmd = mgmt_pending_add(sk, MGMT_OP_ADD_EXT_ADV_DATA, hdev, data, + data_len); + if (!cmd) { + err = -ENOMEM; + goto clear_new_instance; + } + + if (!err) + err = hci_req_run(&req, add_advertising_complete); + + if (err < 0) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_FAILED); + mgmt_pending_remove(cmd); + goto clear_new_instance; + } + + /* We were successful in updating data, so trigger advertising_added + * event if this is an instance that wasn't previously advertising. If + * a failure occurs in the requests we initiated, we will remove the + * instance again in add_advertising_complete + */ + if (adv_instance->pending) + mgmt_advertising_added(sk, hdev, cp->instance); + + goto unlock; + +clear_new_instance: + hci_remove_adv_instance(hdev, cp->instance); + +unlock: + hci_dev_unlock(hdev); + + return err; +} + static void remove_advertising_complete(struct hci_dev *hdev, u8 status, u16 opcode) { @@ -7856,6 +8211,10 @@ static const struct hci_mgmt_handler mgmt_handlers[] = { { add_adv_patterns_monitor,MGMT_ADD_ADV_PATTERNS_MONITOR_SIZE, HCI_MGMT_VAR_LEN }, { remove_adv_monitor, MGMT_REMOVE_ADV_MONITOR_SIZE }, + { add_ext_adv_params, MGMT_ADD_EXT_ADV_PARAMS_MIN_SIZE, + HCI_MGMT_VAR_LEN }, + { add_ext_adv_data, MGMT_ADD_EXT_ADV_DATA_SIZE, + HCI_MGMT_VAR_LEN }, }; void mgmt_index_added(struct hci_dev *hdev) -- cgit v1.2.3 From 9bf9f4b6301ffbd51674e1168f8eeed214d2cf99 Mon Sep 17 00:00:00 2001 From: Daniel Winkler Date: Thu, 3 Dec 2020 12:12:50 -0800 Subject: Bluetooth: Use intervals and tx power from mgmt cmds This patch takes the min/max intervals and tx power optionally provided in mgmt interface, stores them in the advertisement struct, and uses them when configuring the hci requests. While tx power is not used if extended advertising is unavailable, software rotation will use the min and max advertising intervals specified by the client. This change is validated manually by ensuring the min/max intervals are propagated to the controller on both hatch (extended advertising) and kukui (no extended advertising) chromebooks, and that tx power is propagated correctly on hatch. These tests are performed with multiple advertisements simultaneously. Reviewed-by: Sonny Sasaka Signed-off-by: Daniel Winkler Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/hci_core.h | 5 ++++- net/bluetooth/hci_core.c | 8 +++++--- net/bluetooth/hci_request.c | 29 +++++++++++++++++++---------- net/bluetooth/mgmt.c | 8 ++++++-- 4 files changed, 34 insertions(+), 16 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 3c2e205f226a..88988d4fd347 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -230,6 +230,8 @@ struct adv_info { __u16 scan_rsp_len; __u8 scan_rsp_data[HCI_MAX_AD_LENGTH]; __s8 tx_power; + __u32 min_interval; + __u32 max_interval; bdaddr_t random_addr; bool rpa_expired; struct delayed_work rpa_expired_cb; @@ -1303,7 +1305,8 @@ struct adv_info *hci_get_next_instance(struct hci_dev *hdev, u8 instance); int hci_add_adv_instance(struct hci_dev *hdev, u8 instance, u32 flags, u16 adv_data_len, u8 *adv_data, u16 scan_rsp_len, u8 *scan_rsp_data, - u16 timeout, u16 duration); + u16 timeout, u16 duration, s8 tx_power, + u32 min_interval, u32 max_interval); int hci_set_adv_instance_data(struct hci_dev *hdev, u8 instance, u16 adv_data_len, u8 *adv_data, u16 scan_rsp_len, u8 *scan_rsp_data); diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 46ec523d96a7..8855d07534ed 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -2951,7 +2951,8 @@ static void adv_instance_rpa_expired(struct work_struct *work) int hci_add_adv_instance(struct hci_dev *hdev, u8 instance, u32 flags, u16 adv_data_len, u8 *adv_data, u16 scan_rsp_len, u8 *scan_rsp_data, - u16 timeout, u16 duration) + u16 timeout, u16 duration, s8 tx_power, + u32 min_interval, u32 max_interval) { struct adv_info *adv_instance; @@ -2979,6 +2980,9 @@ int hci_add_adv_instance(struct hci_dev *hdev, u8 instance, u32 flags, adv_instance->flags = flags; adv_instance->adv_data_len = adv_data_len; adv_instance->scan_rsp_len = scan_rsp_len; + adv_instance->min_interval = min_interval; + adv_instance->max_interval = max_interval; + adv_instance->tx_power = tx_power; if (adv_data_len) memcpy(adv_instance->adv_data, adv_data, adv_data_len); @@ -2995,8 +2999,6 @@ int hci_add_adv_instance(struct hci_dev *hdev, u8 instance, u32 flags, else adv_instance->duration = duration; - adv_instance->tx_power = HCI_TX_POWER_INVALID; - INIT_DELAYED_WORK(&adv_instance->rpa_expired_cb, adv_instance_rpa_expired); diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index d0d0fbbb3fa5..80dc451d6e12 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -1488,6 +1488,7 @@ static bool is_advertising_allowed(struct hci_dev *hdev, bool connectable) void __hci_req_enable_advertising(struct hci_request *req) { struct hci_dev *hdev = req->hdev; + struct adv_info *adv_instance; struct hci_cp_le_set_adv_param cp; u8 own_addr_type, enable = 0x01; bool connectable; @@ -1495,6 +1496,7 @@ void __hci_req_enable_advertising(struct hci_request *req) u32 flags; flags = get_adv_instance_flags(hdev, hdev->cur_adv_instance); + adv_instance = hci_find_adv_instance(hdev, hdev->cur_adv_instance); /* If the "connectable" instance flag was not set, then choose between * ADV_IND and ADV_NONCONN_IND based on the global connectable setting. @@ -1526,11 +1528,16 @@ void __hci_req_enable_advertising(struct hci_request *req) memset(&cp, 0, sizeof(cp)); - if (connectable) { - cp.type = LE_ADV_IND; - + if (adv_instance) { + adv_min_interval = adv_instance->min_interval; + adv_max_interval = adv_instance->max_interval; + } else { adv_min_interval = hdev->le_adv_min_interval; adv_max_interval = hdev->le_adv_max_interval; + } + + if (connectable) { + cp.type = LE_ADV_IND; } else { if (adv_cur_instance_is_scannable(hdev)) cp.type = LE_ADV_SCAN_IND; @@ -1541,9 +1548,6 @@ void __hci_req_enable_advertising(struct hci_request *req) hci_dev_test_flag(hdev, HCI_LIMITED_DISCOVERABLE)) { adv_min_interval = DISCOV_LE_FAST_ADV_INT_MIN; adv_max_interval = DISCOV_LE_FAST_ADV_INT_MAX; - } else { - adv_min_interval = hdev->le_adv_min_interval; - adv_max_interval = hdev->le_adv_max_interval; } } @@ -2119,9 +2123,15 @@ int __hci_req_setup_ext_adv_instance(struct hci_request *req, u8 instance) memset(&cp, 0, sizeof(cp)); - /* In ext adv set param interval is 3 octets */ - hci_cpu_to_le24(hdev->le_adv_min_interval, cp.min_interval); - hci_cpu_to_le24(hdev->le_adv_max_interval, cp.max_interval); + if (adv_instance) { + hci_cpu_to_le24(adv_instance->min_interval, cp.min_interval); + hci_cpu_to_le24(adv_instance->max_interval, cp.max_interval); + cp.tx_power = adv_instance->tx_power; + } else { + hci_cpu_to_le24(hdev->le_adv_min_interval, cp.min_interval); + hci_cpu_to_le24(hdev->le_adv_max_interval, cp.max_interval); + cp.tx_power = HCI_ADV_TX_POWER_NO_PREFERENCE; + } secondary_adv = (flags & MGMT_ADV_FLAG_SEC_MASK); @@ -2144,7 +2154,6 @@ int __hci_req_setup_ext_adv_instance(struct hci_request *req, u8 instance) cp.own_addr_type = own_addr_type; cp.channel_map = hdev->le_adv_channel_map; - cp.tx_power = 127; cp.handle = instance; if (flags & MGMT_ADV_FLAG_SEC_2M) { diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ec6b520be368..668a62c8181e 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -7533,7 +7533,10 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, cp->adv_data_len, cp->data, cp->scan_rsp_len, cp->data + cp->adv_data_len, - timeout, duration); + timeout, duration, + HCI_ADV_TX_POWER_NO_PREFERENCE, + hdev->le_adv_min_interval, + hdev->le_adv_max_interval); if (err < 0) { err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, MGMT_STATUS_FAILED); @@ -7741,7 +7744,8 @@ static int add_ext_adv_params(struct sock *sk, struct hci_dev *hdev, /* Create advertising instance with no advertising or response data */ err = hci_add_adv_instance(hdev, cp->instance, flags, - 0, NULL, 0, NULL, timeout, duration); + 0, NULL, 0, NULL, timeout, duration, + tx_power, min_interval, max_interval); if (err < 0) { err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_PARAMS, -- cgit v1.2.3 From 7c395ea521e6c8d77f643be61bf2f0f3a1f5b3e8 Mon Sep 17 00:00:00 2001 From: Daniel Winkler Date: Thu, 3 Dec 2020 12:12:51 -0800 Subject: Bluetooth: Query LE tx power on startup Queries tx power via HCI_LE_Read_Transmit_Power command when the hci device is initialized, and stores resulting min/max LE power in hdev struct. If command isn't available (< BT5 support), min/max values both default to HCI_TX_POWER_INVALID. This patch is manually verified by ensuring BT5 devices correctly query and receive controller tx power range. Reviewed-by: Sonny Sasaka Signed-off-by: Daniel Winkler Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/hci.h | 7 +++++++ include/net/bluetooth/hci_core.h | 2 ++ net/bluetooth/hci_core.c | 8 ++++++++ net/bluetooth/hci_event.c | 18 ++++++++++++++++++ 4 files changed, 35 insertions(+) (limited to 'include/net') diff --git a/include/net/bluetooth/hci.h b/include/net/bluetooth/hci.h index c8e67042a3b1..c1504aa3d9cf 100644 --- a/include/net/bluetooth/hci.h +++ b/include/net/bluetooth/hci.h @@ -1797,6 +1797,13 @@ struct hci_cp_le_set_adv_set_rand_addr { bdaddr_t bdaddr; } __packed; +#define HCI_OP_LE_READ_TRANSMIT_POWER 0x204b +struct hci_rp_le_read_transmit_power { + __u8 status; + __s8 min_le_tx_power; + __s8 max_le_tx_power; +} __packed; + #define HCI_OP_LE_READ_BUFFER_SIZE_V2 0x2060 struct hci_rp_le_read_buffer_size_v2 { __u8 status; diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 88988d4fd347..677a8c50b2ad 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -384,6 +384,8 @@ struct hci_dev { __u16 def_page_timeout; __u16 def_multi_adv_rotation_duration; __u16 def_le_autoconnect_timeout; + __s8 min_le_tx_power; + __s8 max_le_tx_power; __u16 pkt_type; __u16 esco_type; diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 8855d07534ed..9d2c9a1c552f 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -741,6 +741,12 @@ static int hci_init3_req(struct hci_request *req, unsigned long opt) hci_req_add(req, HCI_OP_LE_READ_ADV_TX_POWER, 0, NULL); } + if (hdev->commands[38] & 0x80) { + /* Read LE Min/Max Tx Power*/ + hci_req_add(req, HCI_OP_LE_READ_TRANSMIT_POWER, + 0, NULL); + } + if (hdev->commands[26] & 0x40) { /* Read LE White List Size */ hci_req_add(req, HCI_OP_LE_READ_WHITE_LIST_SIZE, @@ -3660,6 +3666,8 @@ struct hci_dev *hci_alloc_dev(void) hdev->le_num_of_adv_sets = HCI_MAX_ADV_INSTANCES; hdev->def_multi_adv_rotation_duration = HCI_DEFAULT_ADV_DURATION; hdev->def_le_autoconnect_timeout = HCI_LE_AUTOCONN_TIMEOUT; + hdev->min_le_tx_power = HCI_TX_POWER_INVALID; + hdev->max_le_tx_power = HCI_TX_POWER_INVALID; hdev->rpa_timeout = HCI_DEFAULT_RPA_TIMEOUT; hdev->discov_interleaved_timeout = DISCOV_INTERLEAVED_TIMEOUT; diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index f193e73ef47c..67668be3461e 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -1202,6 +1202,20 @@ static void hci_cc_le_set_adv_set_random_addr(struct hci_dev *hdev, hci_dev_unlock(hdev); } +static void hci_cc_le_read_transmit_power(struct hci_dev *hdev, + struct sk_buff *skb) +{ + struct hci_rp_le_read_transmit_power *rp = (void *)skb->data; + + BT_DBG("%s status 0x%2.2x", hdev->name, rp->status); + + if (rp->status) + return; + + hdev->min_le_tx_power = rp->min_le_tx_power; + hdev->max_le_tx_power = rp->max_le_tx_power; +} + static void hci_cc_le_set_adv_enable(struct hci_dev *hdev, struct sk_buff *skb) { __u8 *sent, status = *((__u8 *) skb->data); @@ -3582,6 +3596,10 @@ static void hci_cmd_complete_evt(struct hci_dev *hdev, struct sk_buff *skb, hci_cc_le_set_adv_set_random_addr(hdev, skb); break; + case HCI_OP_LE_READ_TRANSMIT_POWER: + hci_cc_le_read_transmit_power(hdev, skb); + break; + default: BT_DBG("%s opcode 0x%4.4x", hdev->name, *opcode); break; -- cgit v1.2.3 From 4d9b952857533b61c662d59dc413094b0c4c8231 Mon Sep 17 00:00:00 2001 From: Daniel Winkler Date: Thu, 3 Dec 2020 12:12:52 -0800 Subject: Bluetooth: Change MGMT security info CMD to be more generic For advertising, we wish to know the LE tx power capabilities of the controller in userspace, so this patch edits the Security Info MGMT command to be more generic, such that other various controller capabilities can be included in the EIR data. This change also includes the LE min and max tx power into this newly-named command. The change was tested by manually verifying that the MGMT command returns the tx power range as expected in userspace. Reviewed-by: Sonny Sasaka Signed-off-by: Daniel Winkler Signed-off-by: Marcel Holtmann Signed-off-by: Johan Hedberg --- include/net/bluetooth/mgmt.h | 15 ++++++++++----- net/bluetooth/mgmt.c | 43 +++++++++++++++++++++++++++++-------------- 2 files changed, 39 insertions(+), 19 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/mgmt.h b/include/net/bluetooth/mgmt.h index 2e18e4173e2f..f9a6638e20b3 100644 --- a/include/net/bluetooth/mgmt.h +++ b/include/net/bluetooth/mgmt.h @@ -686,11 +686,16 @@ struct mgmt_cp_set_blocked_keys { #define MGMT_OP_SET_WIDEBAND_SPEECH 0x0047 -#define MGMT_OP_READ_SECURITY_INFO 0x0048 -#define MGMT_READ_SECURITY_INFO_SIZE 0 -struct mgmt_rp_read_security_info { - __le16 sec_len; - __u8 sec[]; +#define MGMT_CAP_SEC_FLAGS 0x01 +#define MGMT_CAP_MAX_ENC_KEY_SIZE 0x02 +#define MGMT_CAP_SMP_MAX_ENC_KEY_SIZE 0x03 +#define MGMT_CAP_LE_TX_PWR 0x04 + +#define MGMT_OP_READ_CONTROLLER_CAP 0x0048 +#define MGMT_READ_CONTROLLER_CAP_SIZE 0 +struct mgmt_rp_read_controller_cap { + __le16 cap_len; + __u8 cap[0]; } __packed; #define MGMT_OP_READ_EXP_FEATURES_INFO 0x0049 diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 668a62c8181e..754489e4e065 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -110,7 +110,7 @@ static const u16 mgmt_commands[] = { MGMT_OP_SET_APPEARANCE, MGMT_OP_SET_BLOCKED_KEYS, MGMT_OP_SET_WIDEBAND_SPEECH, - MGMT_OP_READ_SECURITY_INFO, + MGMT_OP_READ_CONTROLLER_CAP, MGMT_OP_READ_EXP_FEATURES_INFO, MGMT_OP_SET_EXP_FEATURE, MGMT_OP_READ_DEF_SYSTEM_CONFIG, @@ -176,7 +176,7 @@ static const u16 mgmt_untrusted_commands[] = { MGMT_OP_READ_CONFIG_INFO, MGMT_OP_READ_EXT_INDEX_LIST, MGMT_OP_READ_EXT_INFO, - MGMT_OP_READ_SECURITY_INFO, + MGMT_OP_READ_CONTROLLER_CAP, MGMT_OP_READ_EXP_FEATURES_INFO, MGMT_OP_READ_DEF_SYSTEM_CONFIG, MGMT_OP_READ_DEF_RUNTIME_CONFIG, @@ -3710,13 +3710,14 @@ unlock: return err; } -static int read_security_info(struct sock *sk, struct hci_dev *hdev, - void *data, u16 data_len) +static int read_controller_cap(struct sock *sk, struct hci_dev *hdev, + void *data, u16 data_len) { - char buf[16]; - struct mgmt_rp_read_security_info *rp = (void *)buf; - u16 sec_len = 0; + char buf[20]; + struct mgmt_rp_read_controller_cap *rp = (void *)buf; + u16 cap_len = 0; u8 flags = 0; + u8 tx_power_range[2]; bt_dev_dbg(hdev, "sock %p", sk); @@ -3740,23 +3741,37 @@ static int read_security_info(struct sock *sk, struct hci_dev *hdev, flags |= 0x08; /* Encryption key size enforcement (LE) */ - sec_len = eir_append_data(rp->sec, sec_len, 0x01, &flags, 1); + cap_len = eir_append_data(rp->cap, cap_len, MGMT_CAP_SEC_FLAGS, + &flags, 1); /* When the Read Simple Pairing Options command is supported, then * also max encryption key size information is provided. */ if (hdev->commands[41] & 0x08) - sec_len = eir_append_le16(rp->sec, sec_len, 0x02, + cap_len = eir_append_le16(rp->cap, cap_len, + MGMT_CAP_MAX_ENC_KEY_SIZE, hdev->max_enc_key_size); - sec_len = eir_append_le16(rp->sec, sec_len, 0x03, SMP_MAX_ENC_KEY_SIZE); + cap_len = eir_append_le16(rp->cap, cap_len, + MGMT_CAP_SMP_MAX_ENC_KEY_SIZE, + SMP_MAX_ENC_KEY_SIZE); + + /* Append the min/max LE tx power parameters if we were able to fetch + * it from the controller + */ + if (hdev->commands[38] & 0x80) { + memcpy(&tx_power_range[0], &hdev->min_le_tx_power, 1); + memcpy(&tx_power_range[1], &hdev->max_le_tx_power, 1); + cap_len = eir_append_data(rp->cap, cap_len, MGMT_CAP_LE_TX_PWR, + tx_power_range, 2); + } - rp->sec_len = cpu_to_le16(sec_len); + rp->cap_len = cpu_to_le16(cap_len); hci_dev_unlock(hdev); - return mgmt_cmd_complete(sk, hdev->id, MGMT_OP_READ_SECURITY_INFO, 0, - rp, sizeof(*rp) + sec_len); + return mgmt_cmd_complete(sk, hdev->id, MGMT_OP_READ_CONTROLLER_CAP, 0, + rp, sizeof(*rp) + cap_len); } #ifdef CONFIG_BT_FEATURE_DEBUG @@ -8193,7 +8208,7 @@ static const struct hci_mgmt_handler mgmt_handlers[] = { { set_blocked_keys, MGMT_OP_SET_BLOCKED_KEYS_SIZE, HCI_MGMT_VAR_LEN }, { set_wideband_speech, MGMT_SETTING_SIZE }, - { read_security_info, MGMT_READ_SECURITY_INFO_SIZE, + { read_controller_cap, MGMT_READ_CONTROLLER_CAP_SIZE, HCI_MGMT_UNTRUSTED }, { read_exp_features_info, MGMT_READ_EXP_FEATURES_INFO_SIZE, HCI_MGMT_UNTRUSTED | -- cgit v1.2.3 From 007ab5345545aba2f9cbe4c096cc35d2fd3275ac Mon Sep 17 00:00:00 2001 From: Jarod Wilson Date: Sat, 5 Dec 2020 12:22:29 -0500 Subject: bonding: fix feature flag setting at init time Don't try to adjust XFRM support flags if the bond device isn't yet registered. Bad things can currently happen when netdev_change_features() is called without having wanted_features fully filled in yet. This code runs both on post-module-load mode changes, as well as at module init time, and when run at module init time, it is before register_netdevice() has been called and filled in wanted_features. The empty wanted_features led to features also getting emptied out, which was definitely not the intended behavior, so prevent that from happening. Originally, I'd hoped to stop adjusting wanted_features at all in the bonding driver, as it's documented as being something only the network core should touch, but we actually do need to do this to properly update both the features and wanted_features fields when changing the bond type, or we get to a situation where ethtool sees: esp-hw-offload: off [requested on] I do think we should be using netdev_update_features instead of netdev_change_features here though, so we only send notifiers when the features actually changed. Fixes: a3b658cfb664 ("bonding: allow xfrm offload setup post-module-load") Reported-by: Ivan Vecera Suggested-by: Ivan Vecera Cc: Jay Vosburgh Cc: Veaceslav Falico Cc: Andy Gospodarek Signed-off-by: Jarod Wilson Link: https://lore.kernel.org/r/20201205172229.576587-1-jarod@redhat.com Signed-off-by: Jakub Kicinski --- drivers/net/bonding/bond_options.c | 22 +++++++++++++++------- include/net/bonding.h | 2 -- 2 files changed, 15 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/drivers/net/bonding/bond_options.c b/drivers/net/bonding/bond_options.c index 9abfaae1c6f7..a4e4e15f574d 100644 --- a/drivers/net/bonding/bond_options.c +++ b/drivers/net/bonding/bond_options.c @@ -745,6 +745,19 @@ const struct bond_option *bond_opt_get(unsigned int option) return &bond_opts[option]; } +static void bond_set_xfrm_features(struct net_device *bond_dev, u64 mode) +{ + if (!IS_ENABLED(CONFIG_XFRM_OFFLOAD)) + return; + + if (mode == BOND_MODE_ACTIVEBACKUP) + bond_dev->wanted_features |= BOND_XFRM_FEATURES; + else + bond_dev->wanted_features &= ~BOND_XFRM_FEATURES; + + netdev_update_features(bond_dev); +} + static int bond_option_mode_set(struct bonding *bond, const struct bond_opt_value *newval) { @@ -767,13 +780,8 @@ static int bond_option_mode_set(struct bonding *bond, if (newval->value == BOND_MODE_ALB) bond->params.tlb_dynamic_lb = 1; -#ifdef CONFIG_XFRM_OFFLOAD - if (newval->value == BOND_MODE_ACTIVEBACKUP) - bond->dev->wanted_features |= BOND_XFRM_FEATURES; - else - bond->dev->wanted_features &= ~BOND_XFRM_FEATURES; - netdev_change_features(bond->dev); -#endif /* CONFIG_XFRM_OFFLOAD */ + if (bond->dev->reg_state == NETREG_REGISTERED) + bond_set_xfrm_features(bond->dev, newval->value); /* don't cache arp_validate between modes */ bond->params.arp_validate = BOND_ARP_VALIDATE_NONE; diff --git a/include/net/bonding.h b/include/net/bonding.h index d9d0ff3b0ad3..adc3da776970 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -86,10 +86,8 @@ #define bond_for_each_slave_rcu(bond, pos, iter) \ netdev_for_each_lower_private_rcu((bond)->dev, pos, iter) -#ifdef CONFIG_XFRM_OFFLOAD #define BOND_XFRM_FEATURES (NETIF_F_HW_ESP | NETIF_F_HW_ESP_TX_CSUM | \ NETIF_F_GSO_ESP) -#endif /* CONFIG_XFRM_OFFLOAD */ #ifdef CONFIG_NET_POLL_CONTROLLER extern atomic_t netpoll_block_tx; -- cgit v1.2.3 From 917d80d376ffbaa9725fde9e3c0282f63643f278 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 8 Dec 2020 18:25:53 +0100 Subject: netfilter: nft_dynset: fix timeouts later than 23 days Use nf_msecs_to_jiffies64 and nf_jiffies64_to_msecs as provided by 8e1102d5a159 ("netfilter: nf_tables: support timeouts larger than 23 days"), otherwise ruleset listing breaks. Fixes: a8b1e36d0d1d ("netfilter: nft_dynset: fix element timeout for HZ != 1000") Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 4 ++++ net/netfilter/nf_tables_api.c | 4 ++-- net/netfilter/nft_dynset.c | 8 +++++--- 3 files changed, 11 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 55b4cadf290a..c1c0a4ff92ae 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -1524,4 +1524,8 @@ void __init nft_chain_route_init(void); void nft_chain_route_fini(void); void nf_tables_trans_destroy_flush_work(void); + +int nf_msecs_to_jiffies64(const struct nlattr *nla, u64 *result); +__be64 nf_jiffies64_to_msecs(u64 input); + #endif /* _NET_NF_TABLES_H */ diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 23abf1578594..c2f59879a48d 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -3719,7 +3719,7 @@ cont: return 0; } -static int nf_msecs_to_jiffies64(const struct nlattr *nla, u64 *result) +int nf_msecs_to_jiffies64(const struct nlattr *nla, u64 *result) { u64 ms = be64_to_cpu(nla_get_be64(nla)); u64 max = (u64)(~((u64)0)); @@ -3733,7 +3733,7 @@ static int nf_msecs_to_jiffies64(const struct nlattr *nla, u64 *result) return 0; } -static __be64 nf_jiffies64_to_msecs(u64 input) +__be64 nf_jiffies64_to_msecs(u64 input) { return cpu_to_be64(jiffies64_to_msecs(input)); } diff --git a/net/netfilter/nft_dynset.c b/net/netfilter/nft_dynset.c index 64ca13a1885b..9af4f93c7f0e 100644 --- a/net/netfilter/nft_dynset.c +++ b/net/netfilter/nft_dynset.c @@ -157,8 +157,10 @@ static int nft_dynset_init(const struct nft_ctx *ctx, if (tb[NFTA_DYNSET_TIMEOUT] != NULL) { if (!(set->flags & NFT_SET_TIMEOUT)) return -EINVAL; - timeout = msecs_to_jiffies(be64_to_cpu(nla_get_be64( - tb[NFTA_DYNSET_TIMEOUT]))); + + err = nf_msecs_to_jiffies64(tb[NFTA_DYNSET_TIMEOUT], &timeout); + if (err) + return err; } priv->sreg_key = nft_parse_register(tb[NFTA_DYNSET_SREG_KEY]); @@ -267,7 +269,7 @@ static int nft_dynset_dump(struct sk_buff *skb, const struct nft_expr *expr) if (nla_put_string(skb, NFTA_DYNSET_SET_NAME, priv->set->name)) goto nla_put_failure; if (nla_put_be64(skb, NFTA_DYNSET_TIMEOUT, - cpu_to_be64(jiffies_to_msecs(priv->timeout)), + nf_jiffies64_to_msecs(priv->timeout), NFTA_DYNSET_PAD)) goto nla_put_failure; if (priv->expr && nft_expr_dump(skb, NFTA_DYNSET_EXPR, priv->expr)) -- cgit v1.2.3 From 998f17296234aa8d3676b4a13962eb39f4ad24e0 Mon Sep 17 00:00:00 2001 From: Toke Høiland-Jørgensen Date: Wed, 9 Dec 2020 14:57:37 +0100 Subject: xdp: Remove the xdp_attachment_flags_ok() callback MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since commit 7f0a838254bd ("bpf, xdp: Maintain info on attached XDP BPF programs in net_device"), the XDP program attachment info is now maintained in the core code. This interacts badly with the xdp_attachment_flags_ok() check that prevents unloading an XDP program with different load flags than it was loaded with. In practice, two kinds of failures are seen: - An XDP program loaded without specifying a mode (and which then ends up in driver mode) cannot be unloaded if the program mode is specified on unload. - The dev_xdp_uninstall() hook always calls the driver callback with the mode set to the type of the program but an empty flags argument, which means the flags_ok() check prevents the program from being removed, leading to bpf prog reference leaks. The original reason this check was added was to avoid ambiguity when multiple programs were loaded. With the way the checks are done in the core now, this is quite simple to enforce in the core code, so let's add a check there and get rid of the xdp_attachment_flags_ok() callback entirely. Fixes: 7f0a838254bd ("bpf, xdp: Maintain info on attached XDP BPF programs in net_device") Signed-off-by: Toke Høiland-Jørgensen Signed-off-by: Daniel Borkmann Acked-by: Jakub Kicinski Link: https://lore.kernel.org/bpf/160752225751.110217.10267659521308669050.stgit@toke.dk --- .../net/ethernet/netronome/nfp/nfp_net_common.c | 6 ------ drivers/net/ethernet/ti/cpsw_priv.c | 3 --- drivers/net/netdevsim/bpf.c | 3 --- include/net/xdp.h | 2 -- net/core/dev.c | 22 ++++++++++++++++++++-- net/core/xdp.c | 12 ------------ 6 files changed, 20 insertions(+), 28 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c index b150da43adb2..437226866ce8 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c @@ -3562,9 +3562,6 @@ static int nfp_net_xdp_setup_drv(struct nfp_net *nn, struct netdev_bpf *bpf) struct nfp_net_dp *dp; int err; - if (!xdp_attachment_flags_ok(&nn->xdp, bpf)) - return -EBUSY; - if (!prog == !nn->dp.xdp_prog) { WRITE_ONCE(nn->dp.xdp_prog, prog); xdp_attachment_setup(&nn->xdp, bpf); @@ -3593,9 +3590,6 @@ static int nfp_net_xdp_setup_hw(struct nfp_net *nn, struct netdev_bpf *bpf) { int err; - if (!xdp_attachment_flags_ok(&nn->xdp_hw, bpf)) - return -EBUSY; - err = nfp_app_xdp_offload(nn->app, nn, bpf->prog, bpf->extack); if (err) return err; diff --git a/drivers/net/ethernet/ti/cpsw_priv.c b/drivers/net/ethernet/ti/cpsw_priv.c index 31c5e36ff706..424e644724e4 100644 --- a/drivers/net/ethernet/ti/cpsw_priv.c +++ b/drivers/net/ethernet/ti/cpsw_priv.c @@ -1265,9 +1265,6 @@ static int cpsw_xdp_prog_setup(struct cpsw_priv *priv, struct netdev_bpf *bpf) if (!priv->xdpi.prog && !prog) return 0; - if (!xdp_attachment_flags_ok(&priv->xdpi, bpf)) - return -EBUSY; - WRITE_ONCE(priv->xdp_prog, prog); xdp_attachment_setup(&priv->xdpi, bpf); diff --git a/drivers/net/netdevsim/bpf.c b/drivers/net/netdevsim/bpf.c index 2e90512f3bbe..85546664bdd5 100644 --- a/drivers/net/netdevsim/bpf.c +++ b/drivers/net/netdevsim/bpf.c @@ -190,9 +190,6 @@ nsim_xdp_set_prog(struct netdevsim *ns, struct netdev_bpf *bpf, { int err; - if (!xdp_attachment_flags_ok(xdp, bpf)) - return -EBUSY; - if (bpf->command == XDP_SETUP_PROG && !ns->bpf_xdpdrv_accept) { NSIM_EA(bpf->extack, "driver XDP disabled in DebugFS"); return -EOPNOTSUPP; diff --git a/include/net/xdp.h b/include/net/xdp.h index 3814fb631d52..9dab2bc6f187 100644 --- a/include/net/xdp.h +++ b/include/net/xdp.h @@ -240,8 +240,6 @@ struct xdp_attachment_info { }; struct netdev_bpf; -bool xdp_attachment_flags_ok(struct xdp_attachment_info *info, - struct netdev_bpf *bpf); void xdp_attachment_setup(struct xdp_attachment_info *info, struct netdev_bpf *bpf); diff --git a/net/core/dev.c b/net/core/dev.c index 8588ade790cb..38412e70f761 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -8917,6 +8917,17 @@ static struct bpf_prog *dev_xdp_prog(struct net_device *dev, return dev->xdp_state[mode].prog; } +static u8 dev_xdp_prog_count(struct net_device *dev) +{ + u8 count = 0; + int i; + + for (i = 0; i < __MAX_XDP_MODE; i++) + if (dev->xdp_state[i].prog || dev->xdp_state[i].link) + count++; + return count; +} + u32 dev_xdp_prog_id(struct net_device *dev, enum bpf_xdp_mode mode) { struct bpf_prog *prog = dev_xdp_prog(dev, mode); @@ -9007,6 +9018,7 @@ static int dev_xdp_attach(struct net_device *dev, struct netlink_ext_ack *extack struct bpf_xdp_link *link, struct bpf_prog *new_prog, struct bpf_prog *old_prog, u32 flags) { + unsigned int num_modes = hweight32(flags & XDP_FLAGS_MODES); struct bpf_prog *cur_prog; enum bpf_xdp_mode mode; bpf_op_t bpf_op; @@ -9022,11 +9034,17 @@ static int dev_xdp_attach(struct net_device *dev, struct netlink_ext_ack *extack NL_SET_ERR_MSG(extack, "Invalid XDP flags for BPF link attachment"); return -EINVAL; } - /* just one XDP mode bit should be set, zero defaults to SKB mode */ - if (hweight32(flags & XDP_FLAGS_MODES) > 1) { + /* just one XDP mode bit should be set, zero defaults to drv/skb mode */ + if (num_modes > 1) { NL_SET_ERR_MSG(extack, "Only one XDP mode flag can be set"); return -EINVAL; } + /* avoid ambiguity if offload + drv/skb mode progs are both loaded */ + if (!num_modes && dev_xdp_prog_count(dev) > 1) { + NL_SET_ERR_MSG(extack, + "More than one program loaded, unset mode is ambiguous"); + return -EINVAL; + } /* old_prog != NULL implies XDP_FLAGS_REPLACE is set */ if (old_prog && !(flags & XDP_FLAGS_REPLACE)) { NL_SET_ERR_MSG(extack, "XDP_FLAGS_REPLACE is not specified"); diff --git a/net/core/xdp.c b/net/core/xdp.c index 491ad569a79c..d900cebc0acd 100644 --- a/net/core/xdp.c +++ b/net/core/xdp.c @@ -403,18 +403,6 @@ void __xdp_release_frame(void *data, struct xdp_mem_info *mem) } EXPORT_SYMBOL_GPL(__xdp_release_frame); -bool xdp_attachment_flags_ok(struct xdp_attachment_info *info, - struct netdev_bpf *bpf) -{ - if (info->prog && (bpf->flags ^ info->flags) & XDP_FLAGS_MODES) { - NL_SET_ERR_MSG(bpf->extack, - "program loaded with different flags"); - return false; - } - return true; -} -EXPORT_SYMBOL_GPL(xdp_attachment_flags_ok); - void xdp_attachment_setup(struct xdp_attachment_info *info, struct netdev_bpf *bpf) { -- cgit v1.2.3 From 22fb85ffaefb80a22c815008a500273b3f61bba3 Mon Sep 17 00:00:00 2001 From: Geliang Tang Date: Wed, 9 Dec 2020 15:51:20 -0800 Subject: mptcp: add port support for ADD_ADDR suboption writing In rfc8684, the length of ADD_ADDR suboption with IPv4 address and port is 18 octets, but mptcp_write_options is 32-bit aligned, so we need to pad it to 20 octets. All the other port related option lengths need to be added up 2 octets similarly. This patch added a new field 'port' in mptcp_out_options. When this field is set with a port number, we need to add up 4 octets for the ADD_ADDR suboption, and put the port number into the suboption. Signed-off-by: Geliang Tang Signed-off-by: Mat Martineau Signed-off-by: David S. Miller --- include/net/mptcp.h | 1 + net/mptcp/options.c | 30 +++++++++++++++++++++++++++--- net/mptcp/protocol.h | 10 +++++----- 3 files changed, 33 insertions(+), 8 deletions(-) (limited to 'include/net') diff --git a/include/net/mptcp.h b/include/net/mptcp.h index b6cf07143a8a..5694370be3d4 100644 --- a/include/net/mptcp.h +++ b/include/net/mptcp.h @@ -46,6 +46,7 @@ struct mptcp_out_options { #endif }; u8 addr_id; + u16 port; u64 ahmac; u8 rm_id; u8 join_id; diff --git a/net/mptcp/options.c b/net/mptcp/options.c index ab86f897c08b..f841128a86c6 100644 --- a/net/mptcp/options.c +++ b/net/mptcp/options.c @@ -1083,6 +1083,9 @@ mp_capable_done: len = TCPOLEN_MPTCP_ADD_ADDR6_BASE; #endif + if (opts->port) + len += TCPOLEN_MPTCP_PORT_LEN; + if (opts->ahmac) { len += sizeof(opts->ahmac); echo = 0; @@ -1100,9 +1103,30 @@ mp_capable_done: ptr += 4; } #endif - if (opts->ahmac) { - put_unaligned_be64(opts->ahmac, ptr); - ptr += 2; + + if (!opts->port) { + if (opts->ahmac) { + put_unaligned_be64(opts->ahmac, ptr); + ptr += 2; + } + } else { + if (opts->ahmac) { + u8 *bptr = (u8 *)ptr; + + put_unaligned_be16(opts->port, bptr); + bptr += 2; + put_unaligned_be64(opts->ahmac, bptr); + bptr += 8; + put_unaligned_be16(TCPOPT_NOP << 8 | + TCPOPT_NOP, bptr); + + ptr += 3; + } else { + put_unaligned_be32(opts->port << 16 | + TCPOPT_NOP << 8 | + TCPOPT_NOP, ptr); + ptr += 1; + } } } diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index fc56e730fb35..9032174b446a 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -49,14 +49,14 @@ #define TCPOLEN_MPTCP_DSS_MAP64 14 #define TCPOLEN_MPTCP_DSS_CHECKSUM 2 #define TCPOLEN_MPTCP_ADD_ADDR 16 -#define TCPOLEN_MPTCP_ADD_ADDR_PORT 18 +#define TCPOLEN_MPTCP_ADD_ADDR_PORT 20 #define TCPOLEN_MPTCP_ADD_ADDR_BASE 8 -#define TCPOLEN_MPTCP_ADD_ADDR_BASE_PORT 10 +#define TCPOLEN_MPTCP_ADD_ADDR_BASE_PORT 12 #define TCPOLEN_MPTCP_ADD_ADDR6 28 -#define TCPOLEN_MPTCP_ADD_ADDR6_PORT 30 +#define TCPOLEN_MPTCP_ADD_ADDR6_PORT 32 #define TCPOLEN_MPTCP_ADD_ADDR6_BASE 20 -#define TCPOLEN_MPTCP_ADD_ADDR6_BASE_PORT 22 -#define TCPOLEN_MPTCP_PORT_LEN 2 +#define TCPOLEN_MPTCP_ADD_ADDR6_BASE_PORT 24 +#define TCPOLEN_MPTCP_PORT_LEN 4 #define TCPOLEN_MPTCP_RM_ADDR_BASE 4 /* MPTCP MP_JOIN flags */ -- cgit v1.2.3 From 081e1e7ece05c5eb8bbaf28dc20970cf49edf5d5 Mon Sep 17 00:00:00 2001 From: Shaul Triebitz Date: Sun, 29 Nov 2020 17:30:43 +0200 Subject: mac80211: he: remove non-bss-conf fields from bss_conf ack_enabled and multi_sta_back_32bit are station capabilities and should not be in the bss_conf structure. Signed-off-by: Shaul Triebitz Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201129172929.69a7f7753444.I405c4b5245145e24577512c477f19131d4036489@changeid Signed-off-by: Johannes Berg --- include/net/mac80211.h | 2 -- net/mac80211/mlme.c | 8 -------- 2 files changed, 10 deletions(-) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 05c7524bab26..1328b7166460 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -635,9 +635,7 @@ struct ieee80211_fils_discovery { struct ieee80211_bss_conf { const u8 *bssid; u8 htc_trig_based_pkt_ext; - bool multi_sta_back_32bit; bool uora_exists; - bool ack_enabled; u8 uora_ocw_range; u16 frame_time_rts_th; bool he_support; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 6adfcb9c06dc..b0afb61d9e84 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -3494,14 +3494,6 @@ static bool ieee80211_assoc_success(struct ieee80211_sub_if_data *sdata, le32_get_bits(elems->he_operation->he_oper_params, IEEE80211_HE_OPERATION_RTS_THRESHOLD_MASK); - bss_conf->multi_sta_back_32bit = - sta->sta.he_cap.he_cap_elem.mac_cap_info[2] & - IEEE80211_HE_MAC_CAP2_32BIT_BA_BITMAP; - - bss_conf->ack_enabled = - sta->sta.he_cap.he_cap_elem.mac_cap_info[2] & - IEEE80211_HE_MAC_CAP2_ACK_EN; - bss_conf->uora_exists = !!elems->uora_element; if (elems->uora_element) bss_conf->uora_ocw_range = elems->uora_element[0]; -- cgit v1.2.3 From 4271d4bde0a23edc53097339fc185d0c09c75819 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 29 Nov 2020 17:30:45 +0200 Subject: mac80211: support MIC error/replay detected counters driver update Support the driver incrementing MIC error and replay detected counters when having detected a bad frame, if it drops it directly instead of relying on mac80211 to do the checks. These are then exposed to userspace, though currently only in some cases and in debugfs. Signed-off-by: Johannes Berg Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201129172929.fb59be9c6de8.Ife2260887366f585afadd78c983ebea93d2bb54b@changeid Signed-off-by: Johannes Berg --- include/net/mac80211.h | 20 ++++++++++++++++++++ net/mac80211/key.c | 49 +++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 69 insertions(+) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 1328b7166460..2d01280c0564 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -5321,6 +5321,26 @@ ieee80211_gtk_rekey_add(struct ieee80211_vif *vif, void ieee80211_gtk_rekey_notify(struct ieee80211_vif *vif, const u8 *bssid, const u8 *replay_ctr, gfp_t gfp); +/** + * ieee80211_key_mic_failure - increment MIC failure counter for the key + * + * Note: this is really only safe if no other RX function is called + * at the same time. + * + * @keyconf: the key in question + */ +void ieee80211_key_mic_failure(struct ieee80211_key_conf *keyconf); + +/** + * ieee80211_key_replay - increment replay counter for the key + * + * Note: this is really only safe if no other RX function is called + * at the same time. + * + * @keyconf: the key in question + */ +void ieee80211_key_replay(struct ieee80211_key_conf *keyconf); + /** * ieee80211_wake_queue - wake specific queue * @hw: pointer as obtained from ieee80211_alloc_hw(). diff --git a/net/mac80211/key.c b/net/mac80211/key.c index 8c5f829ff6d7..a4817aa4b171 100644 --- a/net/mac80211/key.c +++ b/net/mac80211/key.c @@ -1300,3 +1300,52 @@ ieee80211_gtk_rekey_add(struct ieee80211_vif *vif, return &key->conf; } EXPORT_SYMBOL_GPL(ieee80211_gtk_rekey_add); + +void ieee80211_key_mic_failure(struct ieee80211_key_conf *keyconf) +{ + struct ieee80211_key *key; + + key = container_of(keyconf, struct ieee80211_key, conf); + + switch (key->conf.cipher) { + case WLAN_CIPHER_SUITE_AES_CMAC: + case WLAN_CIPHER_SUITE_BIP_CMAC_256: + key->u.aes_cmac.icverrors++; + break; + case WLAN_CIPHER_SUITE_BIP_GMAC_128: + case WLAN_CIPHER_SUITE_BIP_GMAC_256: + key->u.aes_gmac.icverrors++; + break; + default: + /* ignore the others for now, we don't keep counters now */ + break; + } +} +EXPORT_SYMBOL_GPL(ieee80211_key_mic_failure); + +void ieee80211_key_replay(struct ieee80211_key_conf *keyconf) +{ + struct ieee80211_key *key; + + key = container_of(keyconf, struct ieee80211_key, conf); + + switch (key->conf.cipher) { + case WLAN_CIPHER_SUITE_CCMP: + case WLAN_CIPHER_SUITE_CCMP_256: + key->u.ccmp.replays++; + break; + case WLAN_CIPHER_SUITE_AES_CMAC: + case WLAN_CIPHER_SUITE_BIP_CMAC_256: + key->u.aes_cmac.replays++; + break; + case WLAN_CIPHER_SUITE_BIP_GMAC_128: + case WLAN_CIPHER_SUITE_BIP_GMAC_256: + key->u.aes_gmac.replays++; + break; + case WLAN_CIPHER_SUITE_GCMP: + case WLAN_CIPHER_SUITE_GCMP_256: + key->u.gcmp.replays++; + break; + } +} +EXPORT_SYMBOL_GPL(ieee80211_key_replay); -- cgit v1.2.3 From d6587602c59974a2eda35e8ed70a4f5970380be8 Mon Sep 17 00:00:00 2001 From: Ilan Peer Date: Sun, 29 Nov 2020 17:30:46 +0200 Subject: cfg80211: Parse SAE H2E only membership selector This extends the support for drivers that rebuild IEs in the FW (same as with HT/VHT/HE). Signed-off-by: Ilan Peer Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201129172929.4012647275f3.I1a93ae71c57ef0b6f58f99d47fce919d19d65ff0@changeid Signed-off-by: Johannes Berg --- include/linux/ieee80211.h | 1 + include/net/cfg80211.h | 3 ++- net/wireless/nl80211.c | 2 ++ 3 files changed, 5 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h index 5e8cc9c3d45a..cbd294239430 100644 --- a/include/linux/ieee80211.h +++ b/include/linux/ieee80211.h @@ -1261,6 +1261,7 @@ struct ieee80211_mgmt { #define BSS_MEMBERSHIP_SELECTOR_HT_PHY 127 #define BSS_MEMBERSHIP_SELECTOR_VHT_PHY 126 #define BSS_MEMBERSHIP_SELECTOR_HE_PHY 122 +#define BSS_MEMBERSHIP_SELECTOR_SAE_H2E 123 /* mgmt header + 1 byte category code */ #define IEEE80211_MIN_ACTION_SIZE offsetof(struct ieee80211_mgmt, u.action.u) diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 78c763dfc99a..f3dfb26b50b9 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1187,6 +1187,7 @@ enum cfg80211_ap_settings_flags { * @vht_required: stations must support VHT * @twt_responder: Enable Target Wait Time * @he_required: stations must support HE + * @sae_h2e_required: stations must support direct H2E technique in SAE * @flags: flags, as defined in enum cfg80211_ap_settings_flags * @he_obss_pd: OBSS Packet Detection settings * @he_bss_color: BSS Color settings @@ -1218,7 +1219,7 @@ struct cfg80211_ap_settings { const struct ieee80211_vht_cap *vht_cap; const struct ieee80211_he_cap_elem *he_cap; const struct ieee80211_he_operation *he_oper; - bool ht_required, vht_required, he_required; + bool ht_required, vht_required, he_required, sae_h2e_required; bool twt_responder; u32 flags; struct ieee80211_he_obss_pd he_obss_pd; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 57f615a6d2f3..8e7320a868d3 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -5017,6 +5017,8 @@ static void nl80211_check_ap_rate_selectors(struct cfg80211_ap_settings *params, params->vht_required = true; if (rates[2 + i] == BSS_MEMBERSHIP_SELECTOR_HE_PHY) params->he_required = true; + if (rates[2 + i] == BSS_MEMBERSHIP_SELECTOR_SAE_H2E) + params->sae_h2e_required = true; } } -- cgit v1.2.3 From 669b84134a2be14d333d4f82b65943d467404f87 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 29 Nov 2020 17:30:55 +0200 Subject: cfg80211: include block-tx flag in channel switch started event In the NL80211_CMD_CH_SWITCH_STARTED_NOTIFY event, include the NL80211_ATTR_CH_SWITCH_BLOCK_TX flag attribute if block-tx was requested by the AP. Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201129172929.8953ef22cc64.Ifee9cab337a4369938545920ba5590559e91327a@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 3 ++- include/uapi/linux/nl80211.h | 3 ++- net/mac80211/cfg.c | 2 +- net/mac80211/mlme.c | 2 +- net/wireless/nl80211.c | 17 +++++++++++------ 5 files changed, 17 insertions(+), 10 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index f3dfb26b50b9..d9b67eed4f75 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -7532,6 +7532,7 @@ void cfg80211_ch_switch_notify(struct net_device *dev, * @dev: the device on which the channel switch started * @chandef: the future channel definition * @count: the number of TBTTs until the channel switch happens + * @quiet: whether or not immediate quiet was requested by the AP * * Inform the userspace about the channel switch that has just * started, so that it can take appropriate actions (eg. starting @@ -7539,7 +7540,7 @@ void cfg80211_ch_switch_notify(struct net_device *dev, */ void cfg80211_ch_switch_started_notify(struct net_device *dev, struct cfg80211_chan_def *chandef, - u8 count); + u8 count, bool quiet); /** * ieee80211_operating_class_to_band - convert operating class to band diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index 3e0d4a038ab6..83c860395dd6 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -2079,7 +2079,8 @@ enum nl80211_commands { * until the channel switch event. * @NL80211_ATTR_CH_SWITCH_BLOCK_TX: flag attribute specifying that transmission * must be blocked on the current channel (before the channel switch - * operation). + * operation). Also included in the channel switch started event if quiet + * was requested by the AP. * @NL80211_ATTR_CSA_IES: Nested set of attributes containing the IE information * for the time while performing a channel switch. * @NL80211_ATTR_CNTDWN_OFFS_BEACON: An array of offsets (u16) to the channel diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index c0d0b15c10fd..7da343efd090 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -3450,7 +3450,7 @@ __ieee80211_channel_switch(struct wiphy *wiphy, struct net_device *dev, IEEE80211_QUEUE_STOP_REASON_CSA); cfg80211_ch_switch_started_notify(sdata->dev, &sdata->csa_chandef, - params->count); + params->count, params->block_tx); if (changed) { ieee80211_bss_info_change_notify(sdata, changed); diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 67829667d394..d4da9822a111 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1509,7 +1509,7 @@ ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, mutex_unlock(&local->mtx); cfg80211_ch_switch_started_notify(sdata->dev, &csa_ie.chandef, - csa_ie.count); + csa_ie.count, csa_ie.mode); if (local->ops->channel_switch) { /* use driver's channel switch callback */ diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 4a7ef3b584be..c8d31181a660 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -17062,7 +17062,7 @@ static void nl80211_ch_switch_notify(struct cfg80211_registered_device *rdev, struct cfg80211_chan_def *chandef, gfp_t gfp, enum nl80211_commands notif, - u8 count) + u8 count, bool quiet) { struct sk_buff *msg; void *hdr; @@ -17083,9 +17083,13 @@ static void nl80211_ch_switch_notify(struct cfg80211_registered_device *rdev, if (nl80211_send_chandef(msg, chandef)) goto nla_put_failure; - if ((notif == NL80211_CMD_CH_SWITCH_STARTED_NOTIFY) && - (nla_put_u32(msg, NL80211_ATTR_CH_SWITCH_COUNT, count))) + if (notif == NL80211_CMD_CH_SWITCH_STARTED_NOTIFY) { + if (nla_put_u32(msg, NL80211_ATTR_CH_SWITCH_COUNT, count)) goto nla_put_failure; + if (quiet && + nla_put_flag(msg, NL80211_ATTR_CH_SWITCH_BLOCK_TX)) + goto nla_put_failure; + } genlmsg_end(msg, hdr); @@ -17118,13 +17122,13 @@ void cfg80211_ch_switch_notify(struct net_device *dev, cfg80211_sched_dfs_chan_update(rdev); nl80211_ch_switch_notify(rdev, dev, chandef, GFP_KERNEL, - NL80211_CMD_CH_SWITCH_NOTIFY, 0); + NL80211_CMD_CH_SWITCH_NOTIFY, 0, false); } EXPORT_SYMBOL(cfg80211_ch_switch_notify); void cfg80211_ch_switch_started_notify(struct net_device *dev, struct cfg80211_chan_def *chandef, - u8 count) + u8 count, bool quiet) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct wiphy *wiphy = wdev->wiphy; @@ -17133,7 +17137,8 @@ void cfg80211_ch_switch_started_notify(struct net_device *dev, trace_cfg80211_ch_switch_started_notify(dev, chandef); nl80211_ch_switch_notify(rdev, dev, chandef, GFP_KERNEL, - NL80211_CMD_CH_SWITCH_STARTED_NOTIFY, count); + NL80211_CMD_CH_SWITCH_STARTED_NOTIFY, + count, quiet); } EXPORT_SYMBOL(cfg80211_ch_switch_started_notify); -- cgit v1.2.3 From 539a36ba2f07110e6d05eb795c2b6fd6a7b4b881 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 6 Dec 2020 14:54:40 +0200 Subject: cfg80211: remove struct ieee80211_he_bss_color We don't really use this struct, we're now using struct cfg80211_he_bss_color instead. Change the one place in mac80211 that's using the old name to use struct assignment instead of memcpy() and thus remove the wrong sizeof while at it. Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201206145305.f6698d97ae4e.Iba2dffcb79c4ab80bde7407609806010b55edfdf@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 13 ------------- net/mac80211/cfg.c | 3 +-- 2 files changed, 1 insertion(+), 15 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index d9b67eed4f75..1a79d6baa254 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -302,19 +302,6 @@ struct cfg80211_he_bss_color { bool partial; }; -/** - * struct ieee80211_he_bss_color - AP settings for BSS coloring - * - * @color: the current color. - * @disabled: is the feature disabled. - * @partial: define the AID equation. - */ -struct ieee80211_he_bss_color { - u8 color; - bool disabled; - bool partial; -}; - /** * struct ieee80211_sta_ht_cap - STA's HT capabilities * diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 7da343efd090..1344ce27353d 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -1124,8 +1124,7 @@ static int ieee80211_start_ap(struct wiphy *wiphy, struct net_device *dev, sdata->vif.bss_conf.twt_responder = params->twt_responder; memcpy(&sdata->vif.bss_conf.he_obss_pd, ¶ms->he_obss_pd, sizeof(struct ieee80211_he_obss_pd)); - memcpy(&sdata->vif.bss_conf.he_bss_color, ¶ms->he_bss_color, - sizeof(struct ieee80211_he_bss_color)); + sdata->vif.bss_conf.he_bss_color = params->he_bss_color; sdata->vif.bss_conf.s1g = params->chandef.chan->band == NL80211_BAND_S1GHZ; -- cgit v1.2.3 From 3bb02143ff55fec55558da4ad48425bf368eb8ed Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 6 Dec 2020 14:54:42 +0200 Subject: cfg80211: support immediate reconnect request hint There are cases where it's necessary to disconnect, but an immediate reconnection is desired. Support a hint to userspace that this is the case, by including a new attribute in the deauth or disassoc event. Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201206145305.58d33941fb9d.I0e7168c205c7949529c8e3b86f3c9b12c01a7017@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 4 +++- include/uapi/linux/nl80211.h | 6 ++++++ net/mac80211/mlme.c | 5 +++-- net/wireless/mlme.c | 26 +++++++++++++++----------- net/wireless/nl80211.c | 23 +++++++++++++++-------- net/wireless/nl80211.h | 8 +++++--- net/wireless/trace.h | 12 ++++++++---- 7 files changed, 55 insertions(+), 29 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 1a79d6baa254..f7470eac2fc8 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -6406,13 +6406,15 @@ void cfg80211_abandon_assoc(struct net_device *dev, struct cfg80211_bss *bss); * @dev: network device * @buf: 802.11 frame (header + body) * @len: length of the frame data + * @reconnect: immediate reconnect is desired (include the nl80211 attribute) * * This function is called whenever deauthentication has been processed in * station mode. This includes both received deauthentication frames and * locally generated ones. This function may sleep. The caller must hold the * corresponding wdev's mutex. */ -void cfg80211_tx_mlme_mgmt(struct net_device *dev, const u8 *buf, size_t len); +void cfg80211_tx_mlme_mgmt(struct net_device *dev, const u8 *buf, size_t len, + bool reconnect); /** * cfg80211_rx_unprot_mlme_mgmt - notification of unprotected mlme mgmt frame diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index 83c860395dd6..c5b729e91068 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -2535,6 +2535,10 @@ enum nl80211_commands { * This is a u8 attribute that encapsulates one of the values from * &enum nl80211_sae_pwe_mechanism. * + * @NL80211_ATTR_RECONNECT_REQUESTED: flag attribute, used with deauth and + * disassoc events to indicate that an immediate reconnect to the AP + * is desired. + * * @NUM_NL80211_ATTR: total number of nl80211_attrs available * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use @@ -3026,6 +3030,8 @@ enum nl80211_attrs { NL80211_ATTR_SAE_PWE, + NL80211_ATTR_RECONNECT_REQUESTED, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index d4da9822a111..5a2828dccfa5 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2741,7 +2741,7 @@ static void ieee80211_report_disconnect(struct ieee80211_sub_if_data *sdata, }; if (tx) - cfg80211_tx_mlme_mgmt(sdata->dev, buf, len); + cfg80211_tx_mlme_mgmt(sdata->dev, buf, len, false); else cfg80211_rx_mlme_mgmt(sdata->dev, buf, len); @@ -4721,7 +4721,8 @@ void ieee80211_mgd_quiesce(struct ieee80211_sub_if_data *sdata) if (ifmgd->auth_data) ieee80211_destroy_auth_data(sdata, false); cfg80211_tx_mlme_mgmt(sdata->dev, frame_buf, - IEEE80211_DEAUTH_FRAME_LEN); + IEEE80211_DEAUTH_FRAME_LEN, + false); } /* This is a bit of a hack - we should find a better and more generic diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 0ac820780437..e1e90761dc00 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -4,7 +4,7 @@ * * Copyright (c) 2009, Jouni Malinen * Copyright (c) 2015 Intel Deutschland GmbH - * Copyright (C) 2019 Intel Corporation + * Copyright (C) 2019-2020 Intel Corporation */ #include @@ -81,7 +81,8 @@ static void cfg80211_process_auth(struct wireless_dev *wdev, } static void cfg80211_process_deauth(struct wireless_dev *wdev, - const u8 *buf, size_t len) + const u8 *buf, size_t len, + bool reconnect) { struct cfg80211_registered_device *rdev = wiphy_to_rdev(wdev->wiphy); struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; @@ -89,7 +90,7 @@ static void cfg80211_process_deauth(struct wireless_dev *wdev, u16 reason_code = le16_to_cpu(mgmt->u.deauth.reason_code); bool from_ap = !ether_addr_equal(mgmt->sa, wdev->netdev->dev_addr); - nl80211_send_deauth(rdev, wdev->netdev, buf, len, GFP_KERNEL); + nl80211_send_deauth(rdev, wdev->netdev, buf, len, reconnect, GFP_KERNEL); if (!wdev->current_bss || !ether_addr_equal(wdev->current_bss->pub.bssid, bssid)) @@ -100,7 +101,8 @@ static void cfg80211_process_deauth(struct wireless_dev *wdev, } static void cfg80211_process_disassoc(struct wireless_dev *wdev, - const u8 *buf, size_t len) + const u8 *buf, size_t len, + bool reconnect) { struct cfg80211_registered_device *rdev = wiphy_to_rdev(wdev->wiphy); struct ieee80211_mgmt *mgmt = (struct ieee80211_mgmt *)buf; @@ -108,7 +110,8 @@ static void cfg80211_process_disassoc(struct wireless_dev *wdev, u16 reason_code = le16_to_cpu(mgmt->u.disassoc.reason_code); bool from_ap = !ether_addr_equal(mgmt->sa, wdev->netdev->dev_addr); - nl80211_send_disassoc(rdev, wdev->netdev, buf, len, GFP_KERNEL); + nl80211_send_disassoc(rdev, wdev->netdev, buf, len, reconnect, + GFP_KERNEL); if (WARN_ON(!wdev->current_bss || !ether_addr_equal(wdev->current_bss->pub.bssid, bssid))) @@ -133,9 +136,9 @@ void cfg80211_rx_mlme_mgmt(struct net_device *dev, const u8 *buf, size_t len) if (ieee80211_is_auth(mgmt->frame_control)) cfg80211_process_auth(wdev, buf, len); else if (ieee80211_is_deauth(mgmt->frame_control)) - cfg80211_process_deauth(wdev, buf, len); + cfg80211_process_deauth(wdev, buf, len, false); else if (ieee80211_is_disassoc(mgmt->frame_control)) - cfg80211_process_disassoc(wdev, buf, len); + cfg80211_process_disassoc(wdev, buf, len, false); } EXPORT_SYMBOL(cfg80211_rx_mlme_mgmt); @@ -180,22 +183,23 @@ void cfg80211_abandon_assoc(struct net_device *dev, struct cfg80211_bss *bss) } EXPORT_SYMBOL(cfg80211_abandon_assoc); -void cfg80211_tx_mlme_mgmt(struct net_device *dev, const u8 *buf, size_t len) +void cfg80211_tx_mlme_mgmt(struct net_device *dev, const u8 *buf, size_t len, + bool reconnect) { struct wireless_dev *wdev = dev->ieee80211_ptr; struct ieee80211_mgmt *mgmt = (void *)buf; ASSERT_WDEV_LOCK(wdev); - trace_cfg80211_tx_mlme_mgmt(dev, buf, len); + trace_cfg80211_tx_mlme_mgmt(dev, buf, len, reconnect); if (WARN_ON(len < 2)) return; if (ieee80211_is_deauth(mgmt->frame_control)) - cfg80211_process_deauth(wdev, buf, len); + cfg80211_process_deauth(wdev, buf, len, reconnect); else - cfg80211_process_disassoc(wdev, buf, len); + cfg80211_process_disassoc(wdev, buf, len, reconnect); } EXPORT_SYMBOL(cfg80211_tx_mlme_mgmt); diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 910872974f2d..390e6e0f23ac 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -718,6 +718,7 @@ static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { [NL80211_ATTR_SAE_PWE] = NLA_POLICY_RANGE(NLA_U8, NL80211_SAE_PWE_HUNT_AND_PECK, NL80211_SAE_PWE_BOTH), + [NL80211_ATTR_RECONNECT_REQUESTED] = { .type = NLA_REJECT }, }; /* policy for the key attributes */ @@ -15855,7 +15856,7 @@ static void nl80211_send_mlme_event(struct cfg80211_registered_device *rdev, const u8 *buf, size_t len, enum nl80211_commands cmd, gfp_t gfp, int uapsd_queues, const u8 *req_ies, - size_t req_ies_len) + size_t req_ies_len, bool reconnect) { struct sk_buff *msg; void *hdr; @@ -15877,6 +15878,9 @@ static void nl80211_send_mlme_event(struct cfg80211_registered_device *rdev, nla_put(msg, NL80211_ATTR_REQ_IE, req_ies_len, req_ies))) goto nla_put_failure; + if (reconnect && nla_put_flag(msg, NL80211_ATTR_RECONNECT_REQUESTED)) + goto nla_put_failure; + if (uapsd_queues >= 0) { struct nlattr *nla_wmm = nla_nest_start_noflag(msg, NL80211_ATTR_STA_WME); @@ -15905,7 +15909,8 @@ void nl80211_send_rx_auth(struct cfg80211_registered_device *rdev, size_t len, gfp_t gfp) { nl80211_send_mlme_event(rdev, netdev, buf, len, - NL80211_CMD_AUTHENTICATE, gfp, -1, NULL, 0); + NL80211_CMD_AUTHENTICATE, gfp, -1, NULL, 0, + false); } void nl80211_send_rx_assoc(struct cfg80211_registered_device *rdev, @@ -15915,23 +15920,25 @@ void nl80211_send_rx_assoc(struct cfg80211_registered_device *rdev, { nl80211_send_mlme_event(rdev, netdev, buf, len, NL80211_CMD_ASSOCIATE, gfp, uapsd_queues, - req_ies, req_ies_len); + req_ies, req_ies_len, false); } void nl80211_send_deauth(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *buf, - size_t len, gfp_t gfp) + size_t len, bool reconnect, gfp_t gfp) { nl80211_send_mlme_event(rdev, netdev, buf, len, - NL80211_CMD_DEAUTHENTICATE, gfp, -1, NULL, 0); + NL80211_CMD_DEAUTHENTICATE, gfp, -1, NULL, 0, + reconnect); } void nl80211_send_disassoc(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *buf, - size_t len, gfp_t gfp) + size_t len, bool reconnect, gfp_t gfp) { nl80211_send_mlme_event(rdev, netdev, buf, len, - NL80211_CMD_DISASSOCIATE, gfp, -1, NULL, 0); + NL80211_CMD_DISASSOCIATE, gfp, -1, NULL, 0, + reconnect); } void cfg80211_rx_unprot_mlme_mgmt(struct net_device *dev, const u8 *buf, @@ -15962,7 +15969,7 @@ void cfg80211_rx_unprot_mlme_mgmt(struct net_device *dev, const u8 *buf, trace_cfg80211_rx_unprot_mlme_mgmt(dev, buf, len); nl80211_send_mlme_event(rdev, dev, buf, len, cmd, GFP_ATOMIC, -1, - NULL, 0); + NULL, 0, false); } EXPORT_SYMBOL(cfg80211_rx_unprot_mlme_mgmt); diff --git a/net/wireless/nl80211.h b/net/wireless/nl80211.h index d3e8e426c486..a3f387770f1b 100644 --- a/net/wireless/nl80211.h +++ b/net/wireless/nl80211.h @@ -1,7 +1,7 @@ /* SPDX-License-Identifier: GPL-2.0 */ /* * Portions of this file - * Copyright (C) 2018 Intel Corporation + * Copyright (C) 2018, 2020 Intel Corporation */ #ifndef __NET_WIRELESS_NL80211_H #define __NET_WIRELESS_NL80211_H @@ -69,10 +69,12 @@ void nl80211_send_rx_assoc(struct cfg80211_registered_device *rdev, const u8 *req_ies, size_t req_ies_len); void nl80211_send_deauth(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *buf, size_t len, gfp_t gfp); + const u8 *buf, size_t len, + bool reconnect, gfp_t gfp); void nl80211_send_disassoc(struct cfg80211_registered_device *rdev, struct net_device *netdev, - const u8 *buf, size_t len, gfp_t gfp); + const u8 *buf, size_t len, + bool reconnect, gfp_t gfp); void nl80211_send_auth_timeout(struct cfg80211_registered_device *rdev, struct net_device *netdev, const u8 *addr, gfp_t gfp); diff --git a/net/wireless/trace.h b/net/wireless/trace.h index 817c6fef13be..d75cd23aea02 100644 --- a/net/wireless/trace.h +++ b/net/wireless/trace.h @@ -2679,19 +2679,23 @@ DEFINE_EVENT(netdev_frame_event, cfg80211_rx_mlme_mgmt, ); TRACE_EVENT(cfg80211_tx_mlme_mgmt, - TP_PROTO(struct net_device *netdev, const u8 *buf, int len), - TP_ARGS(netdev, buf, len), + TP_PROTO(struct net_device *netdev, const u8 *buf, int len, + bool reconnect), + TP_ARGS(netdev, buf, len, reconnect), TP_STRUCT__entry( NETDEV_ENTRY __dynamic_array(u8, frame, len) + __field(int, reconnect) ), TP_fast_assign( NETDEV_ASSIGN; memcpy(__get_dynamic_array(frame), buf, len); + __entry->reconnect = reconnect; ), - TP_printk(NETDEV_PR_FMT ", ftype:0x%.2x", + TP_printk(NETDEV_PR_FMT ", ftype:0x%.2x reconnect:%d", NETDEV_PR_ARG, - le16_to_cpup((__le16 *)__get_dynamic_array(frame))) + le16_to_cpup((__le16 *)__get_dynamic_array(frame)), + __entry->reconnect) ); DECLARE_EVENT_CLASS(netdev_mac_evt, -- cgit v1.2.3 From 3f8a39ff28078e4b56d94e8676f49d9975f82e51 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Sun, 6 Dec 2020 14:54:43 +0200 Subject: mac80211: support driver-based disconnect with reconnect hint Support the driver indicating that a disconnection needs to be performed, and pass through the reconnect hint in this case. Signed-off-by: Johannes Berg Signed-off-by: Luca Coelho Link: https://lore.kernel.org/r/iwlwifi.20201206145305.5c8dab7a22a0.I58459fdf6968b16c90cab9c574f0f04ca22b0c79@changeid Signed-off-by: Johannes Berg --- include/net/mac80211.h | 11 ++++++++ net/mac80211/ieee80211_i.h | 4 ++- net/mac80211/mlme.c | 69 ++++++++++++++++++++++++++++++++++------------ net/mac80211/trace.h | 23 +++++++++++++++- 4 files changed, 87 insertions(+), 20 deletions(-) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 2d01280c0564..95e8484eb4f1 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -5899,6 +5899,17 @@ void ieee80211_beacon_loss(struct ieee80211_vif *vif); */ void ieee80211_connection_loss(struct ieee80211_vif *vif); +/** + * ieee80211_disconnect - request disconnection + * + * @vif: &struct ieee80211_vif pointer from the add_interface callback. + * @reconnect: immediate reconnect is desired + * + * Request disconnection from the current network and, if enabled, send a + * hint to the higher layers that immediate reconnect is desired. + */ +void ieee80211_disconnect(struct ieee80211_vif *vif, bool reconnect); + /** * ieee80211_resume_disconnect - disconnect from AP after resume * diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index cde2e3f4fbcd..cd8275e4b2cd 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -452,7 +452,9 @@ struct ieee80211_if_managed { unsigned long probe_timeout; int probe_send_count; bool nullfunc_failed; - bool connection_loss; + u8 connection_loss:1, + driver_disconnect:1, + reconnect:1; struct cfg80211_bss *associated; struct ieee80211_mgd_auth_data *auth_data; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 5a2828dccfa5..3e124eee91d4 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2732,7 +2732,7 @@ EXPORT_SYMBOL(ieee80211_ap_probereq_get); static void ieee80211_report_disconnect(struct ieee80211_sub_if_data *sdata, const u8 *buf, size_t len, bool tx, - u16 reason) + u16 reason, bool reconnect) { struct ieee80211_event event = { .type = MLME_EVENT, @@ -2741,7 +2741,7 @@ static void ieee80211_report_disconnect(struct ieee80211_sub_if_data *sdata, }; if (tx) - cfg80211_tx_mlme_mgmt(sdata->dev, buf, len, false); + cfg80211_tx_mlme_mgmt(sdata->dev, buf, len, reconnect); else cfg80211_rx_mlme_mgmt(sdata->dev, buf, len); @@ -2763,13 +2763,18 @@ static void __ieee80211_disconnect(struct ieee80211_sub_if_data *sdata) tx = !sdata->csa_block_tx; - /* AP is probably out of range (or not reachable for another reason) so - * remove the bss struct for that AP. - */ - cfg80211_unlink_bss(local->hw.wiphy, ifmgd->associated); + if (!ifmgd->driver_disconnect) { + /* + * AP is probably out of range (or not reachable for another + * reason) so remove the bss struct for that AP. + */ + cfg80211_unlink_bss(local->hw.wiphy, ifmgd->associated); + } ieee80211_set_disassoc(sdata, IEEE80211_STYPE_DEAUTH, - WLAN_REASON_DISASSOC_DUE_TO_INACTIVITY, + ifmgd->driver_disconnect ? + WLAN_REASON_DEAUTH_LEAVING : + WLAN_REASON_DISASSOC_DUE_TO_INACTIVITY, tx, frame_buf); mutex_lock(&local->mtx); sdata->vif.csa_active = false; @@ -2782,7 +2787,9 @@ static void __ieee80211_disconnect(struct ieee80211_sub_if_data *sdata) mutex_unlock(&local->mtx); ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), tx, - WLAN_REASON_DISASSOC_DUE_TO_INACTIVITY); + WLAN_REASON_DISASSOC_DUE_TO_INACTIVITY, + ifmgd->reconnect); + ifmgd->reconnect = false; sdata_unlock(sdata); } @@ -2801,6 +2808,13 @@ static void ieee80211_beacon_connection_loss_work(struct work_struct *work) sdata_info(sdata, "Connection to AP %pM lost\n", ifmgd->bssid); __ieee80211_disconnect(sdata); + ifmgd->connection_loss = false; + } else if (ifmgd->driver_disconnect) { + sdata_info(sdata, + "Driver requested disconnection from AP %pM\n", + ifmgd->bssid); + __ieee80211_disconnect(sdata); + ifmgd->driver_disconnect = false; } else { ieee80211_mgd_probe_ap(sdata, true); } @@ -2839,6 +2853,21 @@ void ieee80211_connection_loss(struct ieee80211_vif *vif) } EXPORT_SYMBOL(ieee80211_connection_loss); +void ieee80211_disconnect(struct ieee80211_vif *vif, bool reconnect) +{ + struct ieee80211_sub_if_data *sdata = vif_to_sdata(vif); + struct ieee80211_hw *hw = &sdata->local->hw; + + trace_api_disconnect(sdata, reconnect); + + if (WARN_ON(sdata->vif.type != NL80211_IFTYPE_STATION)) + return; + + sdata->u.mgd.driver_disconnect = true; + sdata->u.mgd.reconnect = reconnect; + ieee80211_queue_work(hw, &sdata->u.mgd.beacon_connection_loss_work); +} +EXPORT_SYMBOL(ieee80211_disconnect); static void ieee80211_destroy_auth_data(struct ieee80211_sub_if_data *sdata, bool assoc) @@ -3142,7 +3171,7 @@ static void ieee80211_rx_mgmt_deauth(struct ieee80211_sub_if_data *sdata, ieee80211_set_disassoc(sdata, 0, 0, false, NULL); ieee80211_report_disconnect(sdata, (u8 *)mgmt, len, false, - reason_code); + reason_code, false); return; } @@ -3191,7 +3220,8 @@ static void ieee80211_rx_mgmt_disassoc(struct ieee80211_sub_if_data *sdata, ieee80211_set_disassoc(sdata, 0, 0, false, NULL); - ieee80211_report_disconnect(sdata, (u8 *)mgmt, len, false, reason_code); + ieee80211_report_disconnect(sdata, (u8 *)mgmt, len, false, reason_code, + false); } static void ieee80211_get_rates(struct ieee80211_supported_band *sband, @@ -4204,7 +4234,8 @@ static void ieee80211_rx_mgmt_beacon(struct ieee80211_sub_if_data *sdata, true, deauth_buf); ieee80211_report_disconnect(sdata, deauth_buf, sizeof(deauth_buf), true, - WLAN_REASON_DEAUTH_LEAVING); + WLAN_REASON_DEAUTH_LEAVING, + false); return; } @@ -4349,7 +4380,7 @@ static void ieee80211_sta_connection_lost(struct ieee80211_sub_if_data *sdata, tx, frame_buf); ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - reason); + reason, false); } static int ieee80211_auth(struct ieee80211_sub_if_data *sdata) @@ -5436,7 +5467,8 @@ int ieee80211_mgd_auth(struct ieee80211_sub_if_data *sdata, ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - WLAN_REASON_UNSPECIFIED); + WLAN_REASON_UNSPECIFIED, + false); } sdata_info(sdata, "authenticate with %pM\n", req->bss->bssid); @@ -5508,7 +5540,8 @@ int ieee80211_mgd_assoc(struct ieee80211_sub_if_data *sdata, ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - WLAN_REASON_UNSPECIFIED); + WLAN_REASON_UNSPECIFIED, + false); } if (ifmgd->auth_data && !ifmgd->auth_data->done) { @@ -5807,7 +5840,7 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, ieee80211_destroy_auth_data(sdata, false); ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - req->reason_code); + req->reason_code, false); return 0; } @@ -5827,7 +5860,7 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, ieee80211_destroy_assoc_data(sdata, false, true); ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - req->reason_code); + req->reason_code, false); return 0; } @@ -5842,7 +5875,7 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, req->reason_code, tx, frame_buf); ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - req->reason_code); + req->reason_code, false); return 0; } @@ -5875,7 +5908,7 @@ int ieee80211_mgd_disassoc(struct ieee80211_sub_if_data *sdata, frame_buf); ieee80211_report_disconnect(sdata, frame_buf, sizeof(frame_buf), true, - req->reason_code); + req->reason_code, false); return 0; } diff --git a/net/mac80211/trace.h b/net/mac80211/trace.h index 89723907a094..601322e16957 100644 --- a/net/mac80211/trace.h +++ b/net/mac80211/trace.h @@ -2,7 +2,7 @@ /* * Portions of this file * Copyright(c) 2016-2017 Intel Deutschland GmbH -* Copyright (C) 2018 - 2019 Intel Corporation +* Copyright (C) 2018 - 2020 Intel Corporation */ #if !defined(__MAC80211_DRIVER_TRACE) || defined(TRACE_HEADER_MULTI_READ) @@ -2086,6 +2086,27 @@ TRACE_EVENT(api_connection_loss, ) ); +TRACE_EVENT(api_disconnect, + TP_PROTO(struct ieee80211_sub_if_data *sdata, bool reconnect), + + TP_ARGS(sdata, reconnect), + + TP_STRUCT__entry( + VIF_ENTRY + __field(int, reconnect) + ), + + TP_fast_assign( + VIF_ASSIGN; + __entry->reconnect = reconnect; + ), + + TP_printk( + VIF_PR_FMT " reconnect:%d", + VIF_PR_ARG, __entry->reconnect + ) +); + TRACE_EVENT(api_cqm_rssi_notify, TP_PROTO(struct ieee80211_sub_if_data *sdata, enum nl80211_cqm_rssi_threshold_event rssi_event, -- cgit v1.2.3 From 6bdb68cef7bf57cdb3f8d1498623556d6823ff3a Mon Sep 17 00:00:00 2001 From: Carl Huang Date: Thu, 3 Dec 2020 05:37:26 -0500 Subject: nl80211: add common API to configure SAR power limitations NL80211_CMD_SET_SAR_SPECS is added to configure SAR from user space. NL80211_ATTR_SAR_SPEC is used to pass the SAR power specification when used with NL80211_CMD_SET_SAR_SPECS. Wireless driver needs to register SAR type, supported frequency ranges to wiphy, so user space can query it. The index in frequency range is used to specify which sub band the power limitation applies to. The SAR type is for compatibility, so later other SAR mechanism can be implemented without breaking the user space SAR applications. Normal process is user space queries the SAR capability, and gets the index of supported frequency ranges and associates the power limitation with this index and sends to kernel. Here is an example of message send to kernel: 8c 00 00 00 08 00 01 00 00 00 00 00 38 00 2b 81 08 00 01 00 00 00 00 00 2c 00 02 80 14 00 00 80 08 00 02 00 00 00 00 00 08 00 01 00 38 00 00 00 14 00 01 80 08 00 02 00 01 00 00 00 08 00 01 00 48 00 00 00 NL80211_CMD_SET_SAR_SPECS: 0x8c NL80211_ATTR_WIPHY: 0x01(phy idx is 0) NL80211_ATTR_SAR_SPEC: 0x812b (NLA_NESTED) NL80211_SAR_ATTR_TYPE: 0x00 (NL80211_SAR_TYPE_POWER) NL80211_SAR_ATTR_SPECS: 0x8002 (NLA_NESTED) freq range 0 power: 0x38 in 0.25dbm unit (14dbm) freq range 1 power: 0x48 in 0.25dbm unit (18dbm) Signed-off-by: Carl Huang Reviewed-by: Brian Norris Reviewed-by: Abhishek Kumar Link: https://lore.kernel.org/r/20201203103728.3034-2-cjhuang@codeaurora.org [minor edits, NLA parse cleanups] Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 52 ++++++++++++ include/uapi/linux/nl80211.h | 105 +++++++++++++++++++++++++ net/wireless/nl80211.c | 183 +++++++++++++++++++++++++++++++++++++++++++ net/wireless/rdev-ops.h | 12 +++ net/wireless/trace.h | 19 +++++ 5 files changed, 371 insertions(+) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index f7470eac2fc8..9a4bbccddc7f 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1732,6 +1732,54 @@ struct station_info { u8 connected_to_as; }; +/** + * struct cfg80211_sar_sub_specs - sub specs limit + * @power: power limitation in 0.25dbm + * @freq_range_index: index the power limitation applies to + */ +struct cfg80211_sar_sub_specs { + s32 power; + u32 freq_range_index; +}; + +/** + * struct cfg80211_sar_specs - sar limit specs + * @type: it's set with power in 0.25dbm or other types + * @num_sub_specs: number of sar sub specs + * @sub_specs: memory to hold the sar sub specs + */ +struct cfg80211_sar_specs { + enum nl80211_sar_type type; + u32 num_sub_specs; + struct cfg80211_sar_sub_specs sub_specs[]; +}; + + +/** + * @struct cfg80211_sar_chan_ranges - sar frequency ranges + * @start_freq: start range edge frequency + * @end_freq: end range edge frequency + */ +struct cfg80211_sar_freq_ranges { + u32 start_freq; + u32 end_freq; +}; + +/** + * struct cfg80211_sar_capa - sar limit capability + * @type: it's set via power in 0.25dbm or other types + * @num_freq_ranges: number of frequency ranges + * @freq_ranges: memory to hold the freq ranges. + * + * Note: WLAN driver may append new ranges or split an existing + * range to small ones and then append them. + */ +struct cfg80211_sar_capa { + enum nl80211_sar_type type; + u32 num_freq_ranges; + const struct cfg80211_sar_freq_ranges *freq_ranges; +}; + #if IS_ENABLED(CONFIG_CFG80211) /** * cfg80211_get_station - retrieve information about a given station @@ -4249,6 +4297,8 @@ struct cfg80211_ops { struct cfg80211_tid_config *tid_conf); int (*reset_tid_config)(struct wiphy *wiphy, struct net_device *dev, const u8 *peer, u8 tids); + int (*set_sar_specs)(struct wiphy *wiphy, + struct cfg80211_sar_specs *sar); }; /* @@ -5017,6 +5067,8 @@ struct wiphy { u8 max_data_retry_count; + const struct cfg80211_sar_capa *sar_capa; + char priv[] __aligned(NETDEV_ALIGN); }; diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index c5b729e91068..40832d13c2f1 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -1178,6 +1178,10 @@ * includes the contents of the frame. %NL80211_ATTR_ACK flag is included * if the recipient acknowledged the frame. * + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is + * passed using %NL80211_ATTR_SAR_SPEC. %NL80211_ATTR_WIPHY is used to + * specify the wiphy index to be applied to. + * * @NL80211_CMD_MAX: highest used command number * @__NL80211_CMD_AFTER_LAST: internal use */ @@ -1408,6 +1412,8 @@ enum nl80211_commands { NL80211_CMD_CONTROL_PORT_FRAME_TX_STATUS, + NL80211_CMD_SET_SAR_SPECS, + /* add new commands above here */ /* used to define NL80211_CMD_MAX below */ @@ -2535,6 +2541,11 @@ enum nl80211_commands { * This is a u8 attribute that encapsulates one of the values from * &enum nl80211_sae_pwe_mechanism. * + * @NL80211_ATTR_SAR_SPEC: SAR power limitation specification when + * used with %NL80211_CMD_SET_SAR_SPECS. The message contains fields + * of %nl80211_sar_attrs which specifies the sar type and related + * sar specs. Sar specs contains array of %nl80211_sar_specs_attrs. + * * @NL80211_ATTR_RECONNECT_REQUESTED: flag attribute, used with deauth and * disassoc events to indicate that an immediate reconnect to the AP * is desired. @@ -3032,6 +3043,8 @@ enum nl80211_attrs { NL80211_ATTR_RECONNECT_REQUESTED, + NL80211_ATTR_SAR_SPEC, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, @@ -7163,4 +7176,96 @@ enum nl80211_sae_pwe_mechanism { NL80211_SAE_PWE_HASH_TO_ELEMENT, NL80211_SAE_PWE_BOTH, }; + +/** + * enum nl80211_sar_type - type of SAR specs + * + * @NL80211_SAR_TYPE_POWER: power limitation specified in 0.25dBm unit + * + */ +enum nl80211_sar_type { + NL80211_SAR_TYPE_POWER, + + /* add new type here */ + + /* Keep last */ + NUM_NL80211_SAR_TYPE, +}; + +/** + * enum nl80211_sar_attrs - Attributes for SAR spec + * + * @NL80211_SAR_ATTR_TYPE: the SAR type as defined in &enum nl80211_sar_type. + * + * @NL80211_SAR_ATTR_SPECS: Nested array of SAR power + * limit specifications. Each specification contains a set + * of %nl80211_sar_specs_attrs. + * + * For SET operation, it contains array of %NL80211_SAR_ATTR_SPECS_POWER + * and %NL80211_SAR_ATTR_SPECS_RANGE_INDEX. + * + * For sar_capa dump, it contains array of + * %NL80211_SAR_ATTR_SPECS_START_FREQ + * and %NL80211_SAR_ATTR_SPECS_END_FREQ. + * + * @__NL80211_SAR_ATTR_LAST: Internal + * @NL80211_SAR_ATTR_MAX: highest sar attribute + * + * These attributes are used with %NL80211_CMD_SET_SAR_SPEC + */ +enum nl80211_sar_attrs { + __NL80211_SAR_ATTR_INVALID, + + NL80211_SAR_ATTR_TYPE, + NL80211_SAR_ATTR_SPECS, + + __NL80211_SAR_ATTR_LAST, + NL80211_SAR_ATTR_MAX = __NL80211_SAR_ATTR_LAST - 1, +}; + +/** + * enum nl80211_sar_specs_attrs - Attributes for SAR power limit specs + * + * @NL80211_SAR_ATTR_SPECS_POWER: Required (s32)value to specify the actual + * power limit value in units of 0.25 dBm if type is + * NL80211_SAR_TYPE_POWER. (i.e., a value of 44 represents 11 dBm). + * 0 means userspace doesn't have SAR limitation on this associated range. + * + * @NL80211_SAR_ATTR_SPECS_RANGE_INDEX: Required (u32) value to specify the + * index of exported freq range table and the associated power limitation + * is applied to this range. + * + * Userspace isn't required to set all the ranges advertised by WLAN driver, + * and userspace can skip some certain ranges. These skipped ranges don't + * have SAR limitations, and they are same as setting the + * %NL80211_SAR_ATTR_SPECS_POWER to any unreasonable high value because any + * value higher than regulatory allowed value just means SAR power + * limitation is removed, but it's required to set at least one range. + * It's not allowed to set duplicated range in one SET operation. + * + * Every SET operation overwrites previous SET operation. + * + * @NL80211_SAR_ATTR_SPECS_START_FREQ: Required (u32) value to specify the start + * frequency of this range edge when registering SAR capability to wiphy. + * It's not a channel center frequency. The unit is kHz. + * + * @NL80211_SAR_ATTR_SPECS_END_FREQ: Required (u32) value to specify the end + * frequency of this range edge when registering SAR capability to wiphy. + * It's not a channel center frequency. The unit is kHz. + * + * @__NL80211_SAR_ATTR_SPECS_LAST: Internal + * @NL80211_SAR_ATTR_SPECS_MAX: highest sar specs attribute + */ +enum nl80211_sar_specs_attrs { + __NL80211_SAR_ATTR_SPECS_INVALID, + + NL80211_SAR_ATTR_SPECS_POWER, + NL80211_SAR_ATTR_SPECS_RANGE_INDEX, + NL80211_SAR_ATTR_SPECS_START_FREQ, + NL80211_SAR_ATTR_SPECS_END_FREQ, + + __NL80211_SAR_ATTR_SPECS_LAST, + NL80211_SAR_ATTR_SPECS_MAX = __NL80211_SAR_ATTR_SPECS_LAST - 1, +}; + #endif /* __LINUX_NL80211_H */ diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 390e6e0f23ac..7db6079fab04 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -399,6 +399,18 @@ nl80211_unsol_bcast_probe_resp_policy[NL80211_UNSOL_BCAST_PROBE_RESP_ATTR_MAX + .len = IEEE80211_MAX_DATA_LEN } }; +static const struct nla_policy +sar_specs_policy[NL80211_SAR_ATTR_SPECS_MAX + 1] = { + [NL80211_SAR_ATTR_SPECS_POWER] = { .type = NLA_S32 }, + [NL80211_SAR_ATTR_SPECS_RANGE_INDEX] = {.type = NLA_U32 }, +}; + +static const struct nla_policy +sar_policy[NL80211_SAR_ATTR_MAX + 1] = { + [NL80211_SAR_ATTR_TYPE] = NLA_POLICY_MAX(NLA_U32, NUM_NL80211_SAR_TYPE), + [NL80211_SAR_ATTR_SPECS] = NLA_POLICY_NESTED_ARRAY(sar_specs_policy), +}; + static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { [0] = { .strict_start_type = NL80211_ATTR_HE_OBSS_PD }, [NL80211_ATTR_WIPHY] = { .type = NLA_U32 }, @@ -719,6 +731,7 @@ static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { NLA_POLICY_RANGE(NLA_U8, NL80211_SAE_PWE_HUNT_AND_PECK, NL80211_SAE_PWE_BOTH), [NL80211_ATTR_RECONNECT_REQUESTED] = { .type = NLA_REJECT }, + [NL80211_ATTR_SAR_SPEC] = NLA_POLICY_NESTED(sar_policy), }; /* policy for the key attributes */ @@ -2095,6 +2108,56 @@ fail: return -ENOBUFS; } +static int +nl80211_put_sar_specs(struct cfg80211_registered_device *rdev, + struct sk_buff *msg) +{ + struct nlattr *sar_capa, *specs, *sub_freq_range; + u8 num_freq_ranges; + int i; + + if (!rdev->wiphy.sar_capa) + return 0; + + num_freq_ranges = rdev->wiphy.sar_capa->num_freq_ranges; + + sar_capa = nla_nest_start(msg, NL80211_ATTR_SAR_SPEC); + if (!sar_capa) + return -ENOSPC; + + if (nla_put_u32(msg, NL80211_SAR_ATTR_TYPE, rdev->wiphy.sar_capa->type)) + goto fail; + + specs = nla_nest_start(msg, NL80211_SAR_ATTR_SPECS); + if (!specs) + goto fail; + + /* report supported freq_ranges */ + for (i = 0; i < num_freq_ranges; i++) { + sub_freq_range = nla_nest_start(msg, i + 1); + if (!sub_freq_range) + goto fail; + + if (nla_put_u32(msg, NL80211_SAR_ATTR_SPECS_START_FREQ, + rdev->wiphy.sar_capa->freq_ranges[i].start_freq)) + goto fail; + + if (nla_put_u32(msg, NL80211_SAR_ATTR_SPECS_END_FREQ, + rdev->wiphy.sar_capa->freq_ranges[i].end_freq)) + goto fail; + + nla_nest_end(msg, sub_freq_range); + } + + nla_nest_end(msg, specs); + nla_nest_end(msg, sar_capa); + + return 0; +fail: + nla_nest_cancel(msg, sar_capa); + return -ENOBUFS; +} + struct nl80211_dump_wiphy_state { s64 filter_wiphy; long start; @@ -2344,6 +2407,8 @@ static int nl80211_send_wiphy(struct cfg80211_registered_device *rdev, CMD(set_multicast_to_unicast, SET_MULTICAST_TO_UNICAST); CMD(update_connect_params, UPDATE_CONNECT_PARAMS); CMD(update_ft_ies, UPDATE_FT_IES); + if (rdev->wiphy.sar_capa) + CMD(set_sar_specs, SET_SAR_SPECS); } #undef CMD @@ -2669,6 +2734,11 @@ static int nl80211_send_wiphy(struct cfg80211_registered_device *rdev, if (nl80211_put_tid_config_support(rdev, msg)) goto nla_put_failure; + state->split_start++; + break; + case 16: + if (nl80211_put_sar_specs(rdev, msg)) + goto nla_put_failure; /* done */ state->split_start = 0; @@ -14668,6 +14738,111 @@ static void nl80211_post_doit(const struct genl_ops *ops, struct sk_buff *skb, } } +static int nl80211_set_sar_sub_specs(struct cfg80211_registered_device *rdev, + struct cfg80211_sar_specs *sar_specs, + struct nlattr *spec[], int index) +{ + u32 range_index, i; + + if (!sar_specs || !spec) + return -EINVAL; + + if (!spec[NL80211_SAR_ATTR_SPECS_POWER] || + !spec[NL80211_SAR_ATTR_SPECS_RANGE_INDEX]) + return -EINVAL; + + range_index = nla_get_u32(spec[NL80211_SAR_ATTR_SPECS_RANGE_INDEX]); + + /* check if range_index exceeds num_freq_ranges */ + if (range_index >= rdev->wiphy.sar_capa->num_freq_ranges) + return -EINVAL; + + /* check if range_index duplicates */ + for (i = 0; i < index; i++) { + if (sar_specs->sub_specs[i].freq_range_index == range_index) + return -EINVAL; + } + + sar_specs->sub_specs[index].power = + nla_get_s32(spec[NL80211_SAR_ATTR_SPECS_POWER]); + + sar_specs->sub_specs[index].freq_range_index = range_index; + + return 0; +} + +static int nl80211_set_sar_specs(struct sk_buff *skb, struct genl_info *info) +{ + struct cfg80211_registered_device *rdev = info->user_ptr[0]; + struct nlattr *spec[NL80211_SAR_ATTR_SPECS_MAX + 1]; + struct nlattr *tb[NL80211_SAR_ATTR_MAX + 1]; + struct cfg80211_sar_specs *sar_spec; + enum nl80211_sar_type type; + struct nlattr *spec_list; + u32 specs; + int rem, err; + + if (!rdev->wiphy.sar_capa || !rdev->ops->set_sar_specs) + return -EOPNOTSUPP; + + if (!info->attrs[NL80211_ATTR_SAR_SPEC]) + return -EINVAL; + + nla_parse_nested(tb, NL80211_SAR_ATTR_MAX, + info->attrs[NL80211_ATTR_SAR_SPEC], + NULL, NULL); + + if (!tb[NL80211_SAR_ATTR_TYPE] || !tb[NL80211_SAR_ATTR_SPECS]) + return -EINVAL; + + type = nla_get_u32(tb[NL80211_SAR_ATTR_TYPE]); + if (type != rdev->wiphy.sar_capa->type) + return -EINVAL; + + specs = 0; + nla_for_each_nested(spec_list, tb[NL80211_SAR_ATTR_SPECS], rem) + specs++; + + if (specs > rdev->wiphy.sar_capa->num_freq_ranges) + return -EINVAL; + + sar_spec = kzalloc(sizeof(*sar_spec) + + specs * sizeof(struct cfg80211_sar_sub_specs), + GFP_KERNEL); + if (!sar_spec) + return -ENOMEM; + + sar_spec->type = type; + specs = 0; + nla_for_each_nested(spec_list, tb[NL80211_SAR_ATTR_SPECS], rem) { + nla_parse_nested(spec, NL80211_SAR_ATTR_SPECS_MAX, + spec_list, NULL, NULL); + + switch (type) { + case NL80211_SAR_TYPE_POWER: + if (nl80211_set_sar_sub_specs(rdev, sar_spec, + spec, specs)) { + err = -EINVAL; + goto error; + } + break; + default: + err = -EINVAL; + goto error; + } + specs++; + } + + sar_spec->num_sub_specs = specs; + + rdev->cur_cmd_info = info; + err = rdev_set_sar_specs(rdev, sar_spec); + rdev->cur_cmd_info = NULL; +error: + kfree(sar_spec); + return err; +} + static const struct genl_ops nl80211_ops[] = { { .cmd = NL80211_CMD_GET_WIPHY, @@ -15521,6 +15696,14 @@ static const struct genl_small_ops nl80211_small_ops[] = { .internal_flags = NL80211_FLAG_NEED_NETDEV | NL80211_FLAG_NEED_RTNL, }, + { + .cmd = NL80211_CMD_SET_SAR_SPECS, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .doit = nl80211_set_sar_specs, + .flags = GENL_UNS_ADMIN_PERM, + .internal_flags = NL80211_FLAG_NEED_WIPHY | + NL80211_FLAG_NEED_RTNL, + }, }; static struct genl_family nl80211_fam __ro_after_init = { diff --git a/net/wireless/rdev-ops.h b/net/wireless/rdev-ops.h index 5e2f349c92a8..8b1358d04ca2 100644 --- a/net/wireless/rdev-ops.h +++ b/net/wireless/rdev-ops.h @@ -1346,4 +1346,16 @@ static inline int rdev_reset_tid_config(struct cfg80211_registered_device *rdev, return ret; } +static inline int rdev_set_sar_specs(struct cfg80211_registered_device *rdev, + struct cfg80211_sar_specs *sar) +{ + int ret; + + trace_rdev_set_sar_specs(&rdev->wiphy, sar); + ret = rdev->ops->set_sar_specs(&rdev->wiphy, sar); + trace_rdev_return_int(&rdev->wiphy, ret); + + return ret; +} + #endif /* __CFG80211_RDEV_OPS */ diff --git a/net/wireless/trace.h b/net/wireless/trace.h index d75cd23aea02..76b777d5903f 100644 --- a/net/wireless/trace.h +++ b/net/wireless/trace.h @@ -3546,6 +3546,25 @@ TRACE_EVENT(rdev_reset_tid_config, TP_printk(WIPHY_PR_FMT ", " NETDEV_PR_FMT ", peer: " MAC_PR_FMT ", tids: 0x%x", WIPHY_PR_ARG, NETDEV_PR_ARG, MAC_PR_ARG(peer), __entry->tids) ); + +TRACE_EVENT(rdev_set_sar_specs, + TP_PROTO(struct wiphy *wiphy, struct cfg80211_sar_specs *sar), + TP_ARGS(wiphy, sar), + TP_STRUCT__entry( + WIPHY_ENTRY + __field(u16, type) + __field(u16, num) + ), + TP_fast_assign( + WIPHY_ASSIGN; + __entry->type = sar->type; + __entry->num = sar->num_sub_specs; + + ), + TP_printk(WIPHY_PR_FMT ", Set type:%d, num_specs:%d", + WIPHY_PR_ARG, __entry->type, __entry->num) +); + #endif /* !__RDEV_OPS_TRACE || TRACE_HEADER_MULTI_READ */ #undef TRACE_INCLUDE_PATH -- cgit v1.2.3 From c534e093d865d926d042e0a3f228d1152627ccab Mon Sep 17 00:00:00 2001 From: Carl Huang Date: Thu, 3 Dec 2020 05:37:27 -0500 Subject: mac80211: add ieee80211_set_sar_specs This change registers ieee80211_set_sar_specs to mac80211_config_ops, so cfg80211 can call it. Signed-off-by: Carl Huang Reviewed-by: Brian Norris Reviewed-by: Abhishek Kumar Link: https://lore.kernel.org/r/20201203103728.3034-3-cjhuang@codeaurora.org Signed-off-by: Johannes Berg --- include/net/mac80211.h | 2 ++ net/mac80211/cfg.c | 12 ++++++++++++ 2 files changed, 14 insertions(+) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index 95e8484eb4f1..d315740581f1 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -4195,6 +4195,8 @@ struct ieee80211_ops { struct ieee80211_vif *vif); void (*sta_set_4addr)(struct ieee80211_hw *hw, struct ieee80211_vif *vif, struct ieee80211_sta *sta, bool enabled); + int (*set_sar_specs)(struct ieee80211_hw *hw, + const struct cfg80211_sar_specs *sar); }; /** diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index e0f4662b3abf..c4c70e30ad7f 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -4073,6 +4073,17 @@ static int ieee80211_reset_tid_config(struct wiphy *wiphy, return ret; } +static int ieee80211_set_sar_specs(struct wiphy *wiphy, + struct cfg80211_sar_specs *sar) +{ + struct ieee80211_local *local = wiphy_priv(wiphy); + + if (!local->ops->set_sar_specs) + return -EOPNOTSUPP; + + return local->ops->set_sar_specs(&local->hw, sar); +} + const struct cfg80211_ops mac80211_config_ops = { .add_virtual_intf = ieee80211_add_iface, .del_virtual_intf = ieee80211_del_iface, @@ -4175,4 +4186,5 @@ const struct cfg80211_ops mac80211_config_ops = { .probe_mesh_link = ieee80211_probe_mesh_link, .set_tid_config = ieee80211_set_tid_config, .reset_tid_config = ieee80211_reset_tid_config, + .set_sar_specs = ieee80211_set_sar_specs, }; -- cgit v1.2.3 From 86d21fc7474563cb5d054ff001d8ad7b69206717 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 10 Dec 2020 14:43:23 +0100 Subject: netfilter: ctnetlink: add timeout and protoinfo to destroy events DESTROY events do not include the remaining timeout. Add the timeout if the entry was removed explicitly. This can happen when a conntrack gets deleted prematurely, e.g. due to a tcp reset, module removal, netdev notifier (nat/masquerade device went down), ctnetlink and so on. Add the protocol state too for the destroy message to check for abnormal state on connection termination. Joint work with Pablo. Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_conntrack_l4proto.h | 2 +- net/netfilter/nf_conntrack_netlink.c | 31 ++++++++++++++++++---------- net/netfilter/nf_conntrack_proto_dccp.c | 13 +++++++++--- net/netfilter/nf_conntrack_proto_sctp.c | 13 ++++++++---- net/netfilter/nf_conntrack_proto_tcp.c | 13 ++++++++---- 5 files changed, 49 insertions(+), 23 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h index 9be7320b994f..96f9cf81f46b 100644 --- a/include/net/netfilter/nf_conntrack_l4proto.h +++ b/include/net/netfilter/nf_conntrack_l4proto.h @@ -32,7 +32,7 @@ struct nf_conntrack_l4proto { /* convert protoinfo to nfnetink attributes */ int (*to_nlattr)(struct sk_buff *skb, struct nlattr *nla, - struct nf_conn *ct); + struct nf_conn *ct, bool destroy); /* convert nfnetlink attributes to protoinfo */ int (*from_nlattr)(struct nlattr *tb[], struct nf_conn *ct); diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 3d0fd33be018..84caf3316946 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -167,10 +167,14 @@ nla_put_failure: return -1; } -static int ctnetlink_dump_timeout(struct sk_buff *skb, const struct nf_conn *ct) +static int ctnetlink_dump_timeout(struct sk_buff *skb, const struct nf_conn *ct, + bool skip_zero) { long timeout = nf_ct_expires(ct) / HZ; + if (skip_zero && timeout == 0) + return 0; + if (nla_put_be32(skb, CTA_TIMEOUT, htonl(timeout))) goto nla_put_failure; return 0; @@ -179,7 +183,8 @@ nla_put_failure: return -1; } -static int ctnetlink_dump_protoinfo(struct sk_buff *skb, struct nf_conn *ct) +static int ctnetlink_dump_protoinfo(struct sk_buff *skb, struct nf_conn *ct, + bool destroy) { const struct nf_conntrack_l4proto *l4proto; struct nlattr *nest_proto; @@ -193,7 +198,7 @@ static int ctnetlink_dump_protoinfo(struct sk_buff *skb, struct nf_conn *ct) if (!nest_proto) goto nla_put_failure; - ret = l4proto->to_nlattr(skb, nest_proto, ct); + ret = l4proto->to_nlattr(skb, nest_proto, ct, destroy); nla_nest_end(skb, nest_proto); @@ -537,8 +542,8 @@ static int ctnetlink_dump_info(struct sk_buff *skb, struct nf_conn *ct) return -1; if (!test_bit(IPS_OFFLOAD_BIT, &ct->status) && - (ctnetlink_dump_timeout(skb, ct) < 0 || - ctnetlink_dump_protoinfo(skb, ct) < 0)) + (ctnetlink_dump_timeout(skb, ct, false) < 0 || + ctnetlink_dump_protoinfo(skb, ct, false) < 0)) return -1; return 0; @@ -780,15 +785,19 @@ ctnetlink_conntrack_event(unsigned int events, struct nf_ct_event *item) goto nla_put_failure; if (events & (1 << IPCT_DESTROY)) { + if (ctnetlink_dump_timeout(skb, ct, true) < 0) + goto nla_put_failure; + if (ctnetlink_dump_acct(skb, ct, type) < 0 || - ctnetlink_dump_timestamp(skb, ct) < 0) + ctnetlink_dump_timestamp(skb, ct) < 0 || + ctnetlink_dump_protoinfo(skb, ct, true) < 0) goto nla_put_failure; } else { - if (ctnetlink_dump_timeout(skb, ct) < 0) + if (ctnetlink_dump_timeout(skb, ct, false) < 0) goto nla_put_failure; - if (events & (1 << IPCT_PROTOINFO) - && ctnetlink_dump_protoinfo(skb, ct) < 0) + if (events & (1 << IPCT_PROTOINFO) && + ctnetlink_dump_protoinfo(skb, ct, false) < 0) goto nla_put_failure; if ((events & (1 << IPCT_HELPER) || nfct_help(ct)) @@ -2720,10 +2729,10 @@ static int __ctnetlink_glue_build(struct sk_buff *skb, struct nf_conn *ct) if (ctnetlink_dump_status(skb, ct) < 0) goto nla_put_failure; - if (ctnetlink_dump_timeout(skb, ct) < 0) + if (ctnetlink_dump_timeout(skb, ct, false) < 0) goto nla_put_failure; - if (ctnetlink_dump_protoinfo(skb, ct) < 0) + if (ctnetlink_dump_protoinfo(skb, ct, false) < 0) goto nla_put_failure; if (ctnetlink_dump_helpinfo(skb, ct) < 0) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index b3f4a334f9d7..db7479db8512 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -589,7 +589,7 @@ static void dccp_print_conntrack(struct seq_file *s, struct nf_conn *ct) #if IS_ENABLED(CONFIG_NF_CT_NETLINK) static int dccp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, - struct nf_conn *ct) + struct nf_conn *ct, bool destroy) { struct nlattr *nest_parms; @@ -597,15 +597,22 @@ static int dccp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, nest_parms = nla_nest_start(skb, CTA_PROTOINFO_DCCP); if (!nest_parms) goto nla_put_failure; - if (nla_put_u8(skb, CTA_PROTOINFO_DCCP_STATE, ct->proto.dccp.state) || - nla_put_u8(skb, CTA_PROTOINFO_DCCP_ROLE, + if (nla_put_u8(skb, CTA_PROTOINFO_DCCP_STATE, ct->proto.dccp.state)) + goto nla_put_failure; + + if (destroy) + goto skip_state; + + if (nla_put_u8(skb, CTA_PROTOINFO_DCCP_ROLE, ct->proto.dccp.role[IP_CT_DIR_ORIGINAL]) || nla_put_be64(skb, CTA_PROTOINFO_DCCP_HANDSHAKE_SEQ, cpu_to_be64(ct->proto.dccp.handshake_seq), CTA_PROTOINFO_DCCP_PAD)) goto nla_put_failure; +skip_state: nla_nest_end(skb, nest_parms); spin_unlock_bh(&ct->lock); + return 0; nla_put_failure: diff --git a/net/netfilter/nf_conntrack_proto_sctp.c b/net/netfilter/nf_conntrack_proto_sctp.c index 810cca24b399..fb8dc02e502f 100644 --- a/net/netfilter/nf_conntrack_proto_sctp.c +++ b/net/netfilter/nf_conntrack_proto_sctp.c @@ -543,7 +543,7 @@ static bool sctp_can_early_drop(const struct nf_conn *ct) #include static int sctp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, - struct nf_conn *ct) + struct nf_conn *ct, bool destroy) { struct nlattr *nest_parms; @@ -552,15 +552,20 @@ static int sctp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, if (!nest_parms) goto nla_put_failure; - if (nla_put_u8(skb, CTA_PROTOINFO_SCTP_STATE, ct->proto.sctp.state) || - nla_put_be32(skb, CTA_PROTOINFO_SCTP_VTAG_ORIGINAL, + if (nla_put_u8(skb, CTA_PROTOINFO_SCTP_STATE, ct->proto.sctp.state)) + goto nla_put_failure; + + if (destroy) + goto skip_state; + + if (nla_put_be32(skb, CTA_PROTOINFO_SCTP_VTAG_ORIGINAL, ct->proto.sctp.vtag[IP_CT_DIR_ORIGINAL]) || nla_put_be32(skb, CTA_PROTOINFO_SCTP_VTAG_REPLY, ct->proto.sctp.vtag[IP_CT_DIR_REPLY])) goto nla_put_failure; +skip_state: spin_unlock_bh(&ct->lock); - nla_nest_end(skb, nest_parms); return 0; diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index 811c6c9b59e1..1d7e1c595546 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -1186,7 +1186,7 @@ static bool tcp_can_early_drop(const struct nf_conn *ct) #include static int tcp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, - struct nf_conn *ct) + struct nf_conn *ct, bool destroy) { struct nlattr *nest_parms; struct nf_ct_tcp_flags tmp = {}; @@ -1196,8 +1196,13 @@ static int tcp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, if (!nest_parms) goto nla_put_failure; - if (nla_put_u8(skb, CTA_PROTOINFO_TCP_STATE, ct->proto.tcp.state) || - nla_put_u8(skb, CTA_PROTOINFO_TCP_WSCALE_ORIGINAL, + if (nla_put_u8(skb, CTA_PROTOINFO_TCP_STATE, ct->proto.tcp.state)) + goto nla_put_failure; + + if (destroy) + goto skip_state; + + if (nla_put_u8(skb, CTA_PROTOINFO_TCP_WSCALE_ORIGINAL, ct->proto.tcp.seen[0].td_scale) || nla_put_u8(skb, CTA_PROTOINFO_TCP_WSCALE_REPLY, ct->proto.tcp.seen[1].td_scale)) @@ -1212,8 +1217,8 @@ static int tcp_to_nlattr(struct sk_buff *skb, struct nlattr *nla, if (nla_put(skb, CTA_PROTOINFO_TCP_FLAGS_REPLY, sizeof(struct nf_ct_tcp_flags), &tmp)) goto nla_put_failure; +skip_state: spin_unlock_bh(&ct->lock); - nla_nest_end(skb, nest_parms); return 0; -- cgit v1.2.3 From 8cfd9b0f8515e7c361bba27e2a2684cbd427fe01 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 7 Dec 2020 17:37:01 +0100 Subject: netfilter: nftables: generalize set expressions support Currently, the set infrastucture allows for one single expressions per element. This patch extends the existing infrastructure to allow for up to two expressions. This is not updating the netlink API yet, this is coming as an initial preparation patch. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 5 ++- net/netfilter/nf_tables_api.c | 90 ++++++++++++++++++++++++++++----------- net/netfilter/nft_dynset.c | 3 +- 3 files changed, 70 insertions(+), 28 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 55b4cadf290a..aad7e1381200 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -396,6 +396,8 @@ struct nft_set_type { }; #define to_set_type(o) container_of(o, struct nft_set_type, ops) +#define NFT_SET_EXPR_MAX 2 + /** * struct nft_set - nf_tables set instance * @@ -448,13 +450,14 @@ struct nft_set { u16 policy; u16 udlen; unsigned char *udata; - struct nft_expr *expr; /* runtime data below here */ const struct nft_set_ops *ops ____cacheline_aligned; u16 flags:14, genmask:2; u8 klen; u8 dlen; + u8 num_exprs; + struct nft_expr *exprs[NFT_SET_EXPR_MAX]; unsigned char data[] __attribute__((aligned(__alignof__(u64)))); }; diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 65aa98fc5eb6..ade10cd23acc 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -3841,9 +3841,9 @@ static int nf_tables_fill_set(struct sk_buff *skb, const struct nft_ctx *ctx, nla_nest_end(skb, nest); - if (set->expr) { + if (set->num_exprs == 1) { nest = nla_nest_start_noflag(skb, NFTA_SET_EXPR); - if (nf_tables_fill_expr_info(skb, set->expr) < 0) + if (nf_tables_fill_expr_info(skb, set->exprs[0]) < 0) goto nla_put_failure; nla_nest_end(skb, nest); @@ -4279,6 +4279,8 @@ static int nf_tables_newset(struct net *net, struct sock *nlsk, err = PTR_ERR(expr); goto err_set_alloc_name; } + set->exprs[0] = expr; + set->num_exprs++; } udata = NULL; @@ -4296,7 +4298,6 @@ static int nf_tables_newset(struct net *net, struct sock *nlsk, set->dtype = dtype; set->objtype = objtype; set->dlen = desc.dlen; - set->expr = expr; set->flags = flags; set->size = desc.size; set->policy = policy; @@ -4325,8 +4326,8 @@ static int nf_tables_newset(struct net *net, struct sock *nlsk, err_set_trans: ops->destroy(set); err_set_init: - if (expr) - nft_expr_destroy(&ctx, expr); + for (i = 0; i < set->num_exprs; i++) + nft_expr_destroy(&ctx, set->exprs[i]); err_set_alloc_name: kfree(set->name); err_set_name: @@ -4336,11 +4337,13 @@ err_set_name: static void nft_set_destroy(const struct nft_ctx *ctx, struct nft_set *set) { + int i; + if (WARN_ON(set->use > 0)) return; - if (set->expr) - nft_expr_destroy(ctx, set->expr); + for (i = 0; i < set->num_exprs; i++) + nft_expr_destroy(ctx, set->exprs[i]); set->ops->destroy(set); kfree(set->name); @@ -5139,9 +5142,39 @@ static void nf_tables_set_elem_destroy(const struct nft_ctx *ctx, kfree(elem); } +static int nft_set_elem_expr_clone(const struct nft_ctx *ctx, + struct nft_set *set, + struct nft_expr *expr_array[]) +{ + struct nft_expr *expr; + int err, i, k; + + for (i = 0; i < set->num_exprs; i++) { + expr = kzalloc(set->exprs[i]->ops->size, GFP_KERNEL); + if (!expr) + goto err_expr; + + err = nft_expr_clone(expr, set->exprs[i]); + if (err < 0) { + nft_expr_destroy(ctx, expr); + goto err_expr; + } + expr_array[i] = expr; + } + + return 0; + +err_expr: + for (k = i - 1; k >= 0; k++) + nft_expr_destroy(ctx, expr_array[i]); + + return -ENOMEM; +} + static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, const struct nlattr *attr, u32 nlmsg_flags) { + struct nft_expr *expr_array[NFT_SET_EXPR_MAX] = {}; struct nlattr *nla[NFTA_SET_ELEM_MAX + 1]; u8 genmask = nft_genmask_next(ctx->net); struct nft_set_ext_tmpl tmpl; @@ -5149,7 +5182,6 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, struct nft_set_elem elem; struct nft_set_binding *binding; struct nft_object *obj = NULL; - struct nft_expr *expr = NULL; struct nft_userdata *udata; struct nft_data_desc desc; enum nft_registers dreg; @@ -5158,7 +5190,7 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, u64 timeout; u64 expiration; u8 ulen; - int err; + int err, i; err = nla_parse_nested_deprecated(nla, NFTA_SET_ELEM_MAX, attr, nft_set_elem_policy, NULL); @@ -5216,23 +5248,27 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, return err; } - if (nla[NFTA_SET_ELEM_EXPR] != NULL) { + if (nla[NFTA_SET_ELEM_EXPR]) { + struct nft_expr *expr; + + if (set->num_exprs != 1) + return -EOPNOTSUPP; + expr = nft_set_elem_expr_alloc(ctx, set, nla[NFTA_SET_ELEM_EXPR]); if (IS_ERR(expr)) return PTR_ERR(expr); - err = -EOPNOTSUPP; - if (set->expr && set->expr->ops != expr->ops) - goto err_set_elem_expr; - } else if (set->expr) { - expr = kzalloc(set->expr->ops->size, GFP_KERNEL); - if (!expr) - return -ENOMEM; + expr_array[0] = expr; - err = nft_expr_clone(expr, set->expr); - if (err < 0) + if (set->exprs[0] && set->exprs[0]->ops != expr->ops) { + err = -EOPNOTSUPP; goto err_set_elem_expr; + } + } else if (set->num_exprs > 0) { + err = nft_set_elem_expr_clone(ctx, set, expr_array); + if (err < 0) + goto err_set_elem_expr_clone; } err = nft_setelem_parse_key(ctx, set, &elem.key.val, @@ -5257,9 +5293,9 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, nft_set_ext_add(&tmpl, NFT_SET_EXT_TIMEOUT); } - if (expr) + if (set->num_exprs == 1) nft_set_ext_add_length(&tmpl, NFT_SET_EXT_EXPR, - expr->ops->size); + expr_array[0]->ops->size); if (nla[NFTA_SET_ELEM_OBJREF] != NULL) { if (!(set->flags & NFT_SET_OBJECT)) { @@ -5341,10 +5377,12 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, *nft_set_ext_obj(ext) = obj; obj->use++; } - if (expr) { + if (set->num_exprs == 1) { + struct nft_expr *expr = expr_array[0]; + memcpy(nft_set_ext_expr(ext), expr, expr->ops->size); kfree(expr); - expr = NULL; + expr_array[0] = NULL; } trans = nft_trans_elem_alloc(ctx, NFT_MSG_NEWSETELEM, set); @@ -5406,9 +5444,9 @@ err_parse_key_end: err_parse_key: nft_data_release(&elem.key.val, NFT_DATA_VALUE); err_set_elem_expr: - if (expr != NULL) - nft_expr_destroy(ctx, expr); - + for (i = 0; i < set->num_exprs && expr_array[i]; i++) + nft_expr_destroy(ctx, expr_array[i]); +err_set_elem_expr_clone: return err; } diff --git a/net/netfilter/nft_dynset.c b/net/netfilter/nft_dynset.c index 64ca13a1885b..4353e47c30fc 100644 --- a/net/netfilter/nft_dynset.c +++ b/net/netfilter/nft_dynset.c @@ -188,7 +188,8 @@ static int nft_dynset_init(const struct nft_ctx *ctx, if (IS_ERR(priv->expr)) return PTR_ERR(priv->expr); - if (set->expr && set->expr->ops != priv->expr->ops) { + if (set->num_exprs == 1 && + set->exprs[0]->ops != priv->expr->ops) { err = -EOPNOTSUPP; goto err_expr_free; } -- cgit v1.2.3 From 92b211a28992b82a693547e3fe5ff97646961785 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 7 Dec 2020 17:37:05 +0100 Subject: netfilter: nftables: move nft_expr before nft_set Move the nft_expr structure definition before nft_set. Expressions are used by rules and sets, remove unnecessary forward declarations. This comes as preparation to support for multiple expressions per set element. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 54 +++++++++++++++++++-------------------- 1 file changed, 26 insertions(+), 28 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index aad7e1381200..0f4ae16a0c42 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -305,8 +305,33 @@ struct nft_set_estimate { enum nft_set_class space; }; +#define NFT_EXPR_MAXATTR 16 +#define NFT_EXPR_SIZE(size) (sizeof(struct nft_expr) + \ + ALIGN(size, __alignof__(struct nft_expr))) + +/** + * struct nft_expr - nf_tables expression + * + * @ops: expression ops + * @data: expression private data + */ +struct nft_expr { + const struct nft_expr_ops *ops; + unsigned char data[] + __attribute__((aligned(__alignof__(u64)))); +}; + +static inline void *nft_expr_priv(const struct nft_expr *expr) +{ + return (void *)expr->data; +} + +int nft_expr_clone(struct nft_expr *dst, struct nft_expr *src); +void nft_expr_destroy(const struct nft_ctx *ctx, struct nft_expr *expr); +int nft_expr_dump(struct sk_buff *skb, unsigned int attr, + const struct nft_expr *expr); + struct nft_set_ext; -struct nft_expr; /** * struct nft_set_ops - nf_tables set operations @@ -797,7 +822,6 @@ struct nft_offload_ctx; * @validate: validate expression, called during loop detection * @data: extra data to attach to this expression operation */ -struct nft_expr; struct nft_expr_ops { void (*eval)(const struct nft_expr *expr, struct nft_regs *regs, @@ -833,32 +857,6 @@ struct nft_expr_ops { void *data; }; -#define NFT_EXPR_MAXATTR 16 -#define NFT_EXPR_SIZE(size) (sizeof(struct nft_expr) + \ - ALIGN(size, __alignof__(struct nft_expr))) - -/** - * struct nft_expr - nf_tables expression - * - * @ops: expression ops - * @data: expression private data - */ -struct nft_expr { - const struct nft_expr_ops *ops; - unsigned char data[] - __attribute__((aligned(__alignof__(u64)))); -}; - -static inline void *nft_expr_priv(const struct nft_expr *expr) -{ - return (void *)expr->data; -} - -int nft_expr_clone(struct nft_expr *dst, struct nft_expr *src); -void nft_expr_destroy(const struct nft_ctx *ctx, struct nft_expr *expr); -int nft_expr_dump(struct sk_buff *skb, unsigned int attr, - const struct nft_expr *expr); - /** * struct nft_rule - nf_tables rule * -- cgit v1.2.3 From 563125a73ac30d7036ae69ca35c40500562c1de4 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 9 Dec 2020 20:10:27 +0100 Subject: netfilter: nftables: generalize set extension to support for several expressions This patch replaces NFT_SET_EXPR by NFT_SET_EXT_EXPRESSIONS. This new extension allows to attach several expressions to one set element (not only one single expression as NFT_SET_EXPR provides). This patch prepares for support for several expressions per set element in the netlink userspace API. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 36 ++++++++++--- net/netfilter/nf_tables_api.c | 85 +++++++++++++++++++++++-------- net/netfilter/nft_dynset.c | 103 ++++++++++++++++++++++++++++++-------- net/netfilter/nft_set_hash.c | 27 +++++++--- 4 files changed, 196 insertions(+), 55 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 0f4ae16a0c42..bba56f2ce60c 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -421,6 +421,20 @@ struct nft_set_type { }; #define to_set_type(o) container_of(o, struct nft_set_type, ops) +struct nft_set_elem_expr { + u8 size; + unsigned char data[] + __attribute__((aligned(__alignof__(struct nft_expr)))); +}; + +#define nft_setelem_expr_at(__elem_expr, __offset) \ + ((struct nft_expr *)&__elem_expr->data[__offset]) + +#define nft_setelem_expr_foreach(__expr, __elem_expr, __size) \ + for (__expr = nft_setelem_expr_at(__elem_expr, 0), __size = 0; \ + __size < (__elem_expr)->size; \ + __size += (__expr)->ops->size, __expr = ((void *)(__expr)) + (__expr)->ops->size) + #define NFT_SET_EXPR_MAX 2 /** @@ -547,7 +561,7 @@ void nf_tables_destroy_set(const struct nft_ctx *ctx, struct nft_set *set); * @NFT_SET_EXT_TIMEOUT: element timeout * @NFT_SET_EXT_EXPIRATION: element expiration time * @NFT_SET_EXT_USERDATA: user data associated with the element - * @NFT_SET_EXT_EXPR: expression assiociated with the element + * @NFT_SET_EXT_EXPRESSIONS: expressions assiciated with the element * @NFT_SET_EXT_OBJREF: stateful object reference associated with element * @NFT_SET_EXT_NUM: number of extension types */ @@ -559,7 +573,7 @@ enum nft_set_extensions { NFT_SET_EXT_TIMEOUT, NFT_SET_EXT_EXPIRATION, NFT_SET_EXT_USERDATA, - NFT_SET_EXT_EXPR, + NFT_SET_EXT_EXPRESSIONS, NFT_SET_EXT_OBJREF, NFT_SET_EXT_NUM }; @@ -677,9 +691,9 @@ static inline struct nft_userdata *nft_set_ext_userdata(const struct nft_set_ext return nft_set_ext(ext, NFT_SET_EXT_USERDATA); } -static inline struct nft_expr *nft_set_ext_expr(const struct nft_set_ext *ext) +static inline struct nft_set_elem_expr *nft_set_ext_expr(const struct nft_set_ext *ext) { - return nft_set_ext(ext, NFT_SET_EXT_EXPR); + return nft_set_ext(ext, NFT_SET_EXT_EXPRESSIONS); } static inline bool nft_set_elem_expired(const struct nft_set_ext *ext) @@ -909,11 +923,17 @@ static inline void nft_set_elem_update_expr(const struct nft_set_ext *ext, struct nft_regs *regs, const struct nft_pktinfo *pkt) { + struct nft_set_elem_expr *elem_expr; struct nft_expr *expr; - - if (__nft_set_ext_exists(ext, NFT_SET_EXT_EXPR)) { - expr = nft_set_ext_expr(ext); - expr->ops->eval(expr, regs, pkt); + u32 size; + + if (__nft_set_ext_exists(ext, NFT_SET_EXT_EXPRESSIONS)) { + elem_expr = nft_set_ext_expr(ext); + nft_setelem_expr_foreach(expr, elem_expr, size) { + expr->ops->eval(expr, regs, pkt); + if (regs->verdict.code == NFT_BREAK) + return; + } } } diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index ade10cd23acc..a3d5014dd246 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -4496,8 +4496,8 @@ const struct nft_set_ext_type nft_set_ext_types[] = { [NFT_SET_EXT_DATA] = { .align = __alignof__(u32), }, - [NFT_SET_EXT_EXPR] = { - .align = __alignof__(struct nft_expr), + [NFT_SET_EXT_EXPRESSIONS] = { + .align = __alignof__(struct nft_set_elem_expr), }, [NFT_SET_EXT_OBJREF] = { .len = sizeof(struct nft_object *), @@ -4573,6 +4573,29 @@ static int nft_ctx_init_from_elemattr(struct nft_ctx *ctx, struct net *net, return 0; } +static int nft_set_elem_expr_dump(struct sk_buff *skb, + const struct nft_set *set, + const struct nft_set_ext *ext) +{ + struct nft_set_elem_expr *elem_expr; + u32 size, num_exprs = 0; + struct nft_expr *expr; + + elem_expr = nft_set_ext_expr(ext); + nft_setelem_expr_foreach(expr, elem_expr, size) + num_exprs++; + + if (num_exprs == 1) { + expr = nft_setelem_expr_at(elem_expr, 0); + if (nft_expr_dump(skb, NFTA_SET_ELEM_EXPR, expr) < 0) + return -1; + + return 0; + } + + return 0; +} + static int nf_tables_fill_setelem(struct sk_buff *skb, const struct nft_set *set, const struct nft_set_elem *elem) @@ -4600,8 +4623,8 @@ static int nf_tables_fill_setelem(struct sk_buff *skb, set->dlen) < 0) goto nla_put_failure; - if (nft_set_ext_exists(ext, NFT_SET_EXT_EXPR) && - nft_expr_dump(skb, NFTA_SET_ELEM_EXPR, nft_set_ext_expr(ext)) < 0) + if (nft_set_ext_exists(ext, NFT_SET_EXT_EXPRESSIONS) && + nft_set_elem_expr_dump(skb, set, ext)) goto nla_put_failure; if (nft_set_ext_exists(ext, NFT_SET_EXT_OBJREF) && @@ -5096,8 +5119,8 @@ void *nft_set_elem_init(const struct nft_set *set, return elem; } -static void nft_set_elem_expr_destroy(const struct nft_ctx *ctx, - struct nft_expr *expr) +static void __nft_set_elem_expr_destroy(const struct nft_ctx *ctx, + struct nft_expr *expr) { if (expr->ops->destroy_clone) { expr->ops->destroy_clone(ctx, expr); @@ -5107,6 +5130,16 @@ static void nft_set_elem_expr_destroy(const struct nft_ctx *ctx, } } +static void nft_set_elem_expr_destroy(const struct nft_ctx *ctx, + struct nft_set_elem_expr *elem_expr) +{ + struct nft_expr *expr; + u32 size; + + nft_setelem_expr_foreach(expr, elem_expr, size) + __nft_set_elem_expr_destroy(ctx, expr); +} + void nft_set_elem_destroy(const struct nft_set *set, void *elem, bool destroy_expr) { @@ -5119,7 +5152,7 @@ void nft_set_elem_destroy(const struct nft_set *set, void *elem, nft_data_release(nft_set_ext_key(ext), NFT_DATA_VALUE); if (nft_set_ext_exists(ext, NFT_SET_EXT_DATA)) nft_data_release(nft_set_ext_data(ext), set->dtype); - if (destroy_expr && nft_set_ext_exists(ext, NFT_SET_EXT_EXPR)) + if (destroy_expr && nft_set_ext_exists(ext, NFT_SET_EXT_EXPRESSIONS)) nft_set_elem_expr_destroy(&ctx, nft_set_ext_expr(ext)); if (nft_set_ext_exists(ext, NFT_SET_EXT_OBJREF)) @@ -5136,7 +5169,7 @@ static void nf_tables_set_elem_destroy(const struct nft_ctx *ctx, { struct nft_set_ext *ext = nft_set_elem_ext(set, elem); - if (nft_set_ext_exists(ext, NFT_SET_EXT_EXPR)) + if (nft_set_ext_exists(ext, NFT_SET_EXT_EXPRESSIONS)) nft_set_elem_expr_destroy(ctx, nft_set_ext_expr(ext)); kfree(elem); @@ -5171,6 +5204,18 @@ err_expr: return -ENOMEM; } +static void nft_set_elem_expr_setup(const struct nft_set_ext *ext, int i, + struct nft_expr *expr_array[]) +{ + struct nft_set_elem_expr *elem_expr = nft_set_ext_expr(ext); + struct nft_expr *expr = nft_setelem_expr_at(elem_expr, elem_expr->size); + + memcpy(expr, expr_array[i], expr_array[i]->ops->size); + elem_expr->size += expr_array[i]->ops->size; + kfree(expr_array[i]); + expr_array[i] = NULL; +} + static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, const struct nlattr *attr, u32 nlmsg_flags) { @@ -5186,11 +5231,11 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, struct nft_data_desc desc; enum nft_registers dreg; struct nft_trans *trans; - u32 flags = 0; + u32 flags = 0, size = 0; u64 timeout; u64 expiration; - u8 ulen; int err, i; + u8 ulen; err = nla_parse_nested_deprecated(nla, NFTA_SET_ELEM_MAX, attr, nft_set_elem_policy, NULL); @@ -5293,9 +5338,14 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, nft_set_ext_add(&tmpl, NFT_SET_EXT_TIMEOUT); } - if (set->num_exprs == 1) - nft_set_ext_add_length(&tmpl, NFT_SET_EXT_EXPR, - expr_array[0]->ops->size); + if (set->num_exprs) { + for (i = 0; i < set->num_exprs; i++) + size += expr_array[i]->ops->size; + + nft_set_ext_add_length(&tmpl, NFT_SET_EXT_EXPRESSIONS, + sizeof(struct nft_set_elem_expr) + + size); + } if (nla[NFTA_SET_ELEM_OBJREF] != NULL) { if (!(set->flags & NFT_SET_OBJECT)) { @@ -5377,13 +5427,8 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, *nft_set_ext_obj(ext) = obj; obj->use++; } - if (set->num_exprs == 1) { - struct nft_expr *expr = expr_array[0]; - - memcpy(nft_set_ext_expr(ext), expr, expr->ops->size); - kfree(expr); - expr_array[0] = NULL; - } + for (i = 0; i < set->num_exprs; i++) + nft_set_elem_expr_setup(ext, i, expr_array); trans = nft_trans_elem_alloc(ctx, NFT_MSG_NEWSETELEM, set); if (trans == NULL) diff --git a/net/netfilter/nft_dynset.c b/net/netfilter/nft_dynset.c index 4353e47c30fc..d9e609b2e5d4 100644 --- a/net/netfilter/nft_dynset.c +++ b/net/netfilter/nft_dynset.c @@ -19,11 +19,30 @@ struct nft_dynset { enum nft_registers sreg_key:8; enum nft_registers sreg_data:8; bool invert; + u8 num_exprs; u64 timeout; - struct nft_expr *expr; + struct nft_expr *expr_array[NFT_SET_EXPR_MAX]; struct nft_set_binding binding; }; +static int nft_dynset_expr_setup(const struct nft_dynset *priv, + const struct nft_set_ext *ext) +{ + struct nft_set_elem_expr *elem_expr = nft_set_ext_expr(ext); + struct nft_expr *expr; + int i; + + for (i = 0; i < priv->num_exprs; i++) { + expr = nft_setelem_expr_at(elem_expr, elem_expr->size); + if (nft_expr_clone(expr, priv->expr_array[i]) < 0) + return -1; + + elem_expr->size += priv->expr_array[i]->ops->size; + } + + return 0; +} + static void *nft_dynset_new(struct nft_set *set, const struct nft_expr *expr, struct nft_regs *regs) { @@ -44,8 +63,7 @@ static void *nft_dynset_new(struct nft_set *set, const struct nft_expr *expr, goto err1; ext = nft_set_elem_ext(set, elem); - if (priv->expr != NULL && - nft_expr_clone(nft_set_ext_expr(ext), priv->expr) < 0) + if (priv->num_exprs && nft_dynset_expr_setup(priv, ext) < 0) goto err2; return elem; @@ -90,6 +108,41 @@ void nft_dynset_eval(const struct nft_expr *expr, regs->verdict.code = NFT_BREAK; } +static void nft_dynset_ext_add_expr(struct nft_dynset *priv) +{ + u8 size = 0; + int i; + + for (i = 0; i < priv->num_exprs; i++) + size += priv->expr_array[i]->ops->size; + + nft_set_ext_add_length(&priv->tmpl, NFT_SET_EXT_EXPRESSIONS, + sizeof(struct nft_set_elem_expr) + size); +} + +static struct nft_expr * +nft_dynset_expr_alloc(const struct nft_ctx *ctx, const struct nft_set *set, + const struct nlattr *attr, int pos) +{ + struct nft_expr *expr; + int err; + + expr = nft_set_elem_expr_alloc(ctx, set, attr); + if (IS_ERR(expr)) + return expr; + + if (set->exprs[pos] && set->exprs[pos]->ops != expr->ops) { + err = -EOPNOTSUPP; + goto err_dynset_expr; + } + + return expr; + +err_dynset_expr: + nft_expr_destroy(ctx, expr); + return ERR_PTR(err); +} + static const struct nla_policy nft_dynset_policy[NFTA_DYNSET_MAX + 1] = { [NFTA_DYNSET_SET_NAME] = { .type = NLA_STRING, .len = NFT_SET_MAXNAMELEN - 1 }, @@ -110,7 +163,7 @@ static int nft_dynset_init(const struct nft_ctx *ctx, u8 genmask = nft_genmask_next(ctx->net); struct nft_set *set; u64 timeout; - int err; + int err, i; lockdep_assert_held(&ctx->net->nft.commit_mutex); @@ -179,17 +232,23 @@ static int nft_dynset_init(const struct nft_ctx *ctx, } else if (set->flags & NFT_SET_MAP) return -EINVAL; - if (tb[NFTA_DYNSET_EXPR] != NULL) { + if (tb[NFTA_DYNSET_EXPR]) { + struct nft_expr *dynset_expr; + if (!(set->flags & NFT_SET_EVAL)) return -EINVAL; - priv->expr = nft_set_elem_expr_alloc(ctx, set, - tb[NFTA_DYNSET_EXPR]); - if (IS_ERR(priv->expr)) - return PTR_ERR(priv->expr); + dynset_expr = nft_dynset_expr_alloc(ctx, set, + tb[NFTA_DYNSET_EXPR], 0); + if (IS_ERR(dynset_expr)) + return PTR_ERR(dynset_expr); - if (set->num_exprs == 1 && - set->exprs[0]->ops != priv->expr->ops) { + priv->num_exprs++; + priv->expr_array[0] = dynset_expr; + + if (set->num_exprs > 1 || + (set->num_exprs == 1 && + dynset_expr->ops != set->exprs[0]->ops)) { err = -EOPNOTSUPP; goto err_expr_free; } @@ -199,9 +258,10 @@ static int nft_dynset_init(const struct nft_ctx *ctx, nft_set_ext_add_length(&priv->tmpl, NFT_SET_EXT_KEY, set->klen); if (set->flags & NFT_SET_MAP) nft_set_ext_add_length(&priv->tmpl, NFT_SET_EXT_DATA, set->dlen); - if (priv->expr != NULL) - nft_set_ext_add_length(&priv->tmpl, NFT_SET_EXT_EXPR, - priv->expr->ops->size); + + if (priv->num_exprs) + nft_dynset_ext_add_expr(priv); + if (set->flags & NFT_SET_TIMEOUT) { if (timeout || set->timeout) nft_set_ext_add(&priv->tmpl, NFT_SET_EXT_EXPIRATION); @@ -220,8 +280,8 @@ static int nft_dynset_init(const struct nft_ctx *ctx, return 0; err_expr_free: - if (priv->expr != NULL) - nft_expr_destroy(ctx, priv->expr); + for (i = 0; i < priv->num_exprs; i++) + nft_expr_destroy(ctx, priv->expr_array[i]); return err; } @@ -246,9 +306,10 @@ static void nft_dynset_destroy(const struct nft_ctx *ctx, const struct nft_expr *expr) { struct nft_dynset *priv = nft_expr_priv(expr); + int i; - if (priv->expr != NULL) - nft_expr_destroy(ctx, priv->expr); + for (i = 0; i < priv->num_exprs; i++) + nft_expr_destroy(ctx, priv->expr_array[i]); nf_tables_destroy_set(ctx, priv->set); } @@ -271,8 +332,10 @@ static int nft_dynset_dump(struct sk_buff *skb, const struct nft_expr *expr) cpu_to_be64(jiffies_to_msecs(priv->timeout)), NFTA_DYNSET_PAD)) goto nla_put_failure; - if (priv->expr && nft_expr_dump(skb, NFTA_DYNSET_EXPR, priv->expr)) - goto nla_put_failure; + if (priv->num_exprs == 1) { + if (nft_expr_dump(skb, NFTA_DYNSET_EXPR, priv->expr_array[0])) + goto nla_put_failure; + } if (nla_put_be32(skb, NFTA_DYNSET_FLAGS, htonl(flags))) goto nla_put_failure; return 0; diff --git a/net/netfilter/nft_set_hash.c b/net/netfilter/nft_set_hash.c index 4d3f147e8d8d..bf618b7ec1ae 100644 --- a/net/netfilter/nft_set_hash.c +++ b/net/netfilter/nft_set_hash.c @@ -293,6 +293,22 @@ cont: rhashtable_walk_exit(&hti); } +static bool nft_rhash_expr_needs_gc_run(const struct nft_set *set, + struct nft_set_ext *ext) +{ + struct nft_set_elem_expr *elem_expr = nft_set_ext_expr(ext); + struct nft_expr *expr; + u32 size; + + nft_setelem_expr_foreach(expr, elem_expr, size) { + if (expr->ops->gc && + expr->ops->gc(read_pnet(&set->net), expr)) + return true; + } + + return false; +} + static void nft_rhash_gc(struct work_struct *work) { struct nft_set *set; @@ -314,16 +330,13 @@ static void nft_rhash_gc(struct work_struct *work) continue; } - if (nft_set_ext_exists(&he->ext, NFT_SET_EXT_EXPR)) { - struct nft_expr *expr = nft_set_ext_expr(&he->ext); + if (nft_set_ext_exists(&he->ext, NFT_SET_EXT_EXPRESSIONS) && + nft_rhash_expr_needs_gc_run(set, &he->ext)) + goto needs_gc_run; - if (expr->ops->gc && - expr->ops->gc(read_pnet(&set->net), expr)) - goto gc; - } if (!nft_set_elem_expired(&he->ext)) continue; -gc: +needs_gc_run: if (nft_set_elem_mark_busy(&he->ext)) continue; -- cgit v1.2.3 From 0b9b241406818a871c6d25390aa487dba966d548 Mon Sep 17 00:00:00 2001 From: SeongJae Park Date: Fri, 11 Dec 2020 12:24:05 +0100 Subject: inet: frags: batch fqdir destroy works On a few of our systems, I found frequent 'unshare(CLONE_NEWNET)' calls make the number of active slab objects including 'sock_inode_cache' type rapidly and continuously increase. As a result, memory pressure occurs. In more detail, I made an artificial reproducer that resembles the workload that we found the problem and reproduce the problem faster. It merely repeats 'unshare(CLONE_NEWNET)' 50,000 times in a loop. It takes about 2 minutes. On 40 CPU cores / 70GB DRAM machine, the available memory continuously reduced in a fast speed (about 120MB per second, 15GB in total within the 2 minutes). Note that the issue don't reproduce on every machine. On my 6 CPU cores machine, the problem didn't reproduce. 'cleanup_net()' and 'fqdir_work_fn()' are functions that deallocate the relevant memory objects. They are asynchronously invoked by the work queues and internally use 'rcu_barrier()' to ensure safe destructions. 'cleanup_net()' works in a batched maneer in a single thread worker, while 'fqdir_work_fn()' works for each 'fqdir_exit()' call in the 'system_wq'. Therefore, 'fqdir_work_fn()' called frequently under the workload and made the contention for 'rcu_barrier()' high. In more detail, the global mutex, 'rcu_state.barrier_mutex' became the bottleneck. This commit avoids such contention by doing the 'rcu_barrier()' and subsequent lightweight works in a batched manner, as similar to that of 'cleanup_net()'. The fqdir hashtable destruction, which is done before the 'rcu_barrier()', is still allowed to run in parallel for fast processing, but this commit makes it to use a dedicated work queue instead of the 'system_wq', to make sure that the number of threads is bounded. Signed-off-by: SeongJae Park Reviewed-by: Eric Dumazet Link: https://lore.kernel.org/r/20201211112405.31158-1-sjpark@amazon.com Signed-off-by: Jakub Kicinski --- include/net/inet_frag.h | 1 + net/ipv4/inet_fragment.c | 47 +++++++++++++++++++++++++++++++++++++++-------- 2 files changed, 40 insertions(+), 8 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_frag.h b/include/net/inet_frag.h index bac79e817776..48cc5795ceda 100644 --- a/include/net/inet_frag.h +++ b/include/net/inet_frag.h @@ -21,6 +21,7 @@ struct fqdir { /* Keep atomic mem on separate cachelines in structs that include it */ atomic_long_t mem ____cacheline_aligned_in_smp; struct work_struct destroy_work; + struct llist_node free_list; }; /** diff --git a/net/ipv4/inet_fragment.c b/net/ipv4/inet_fragment.c index 10d31733297d..05cd198d7a6b 100644 --- a/net/ipv4/inet_fragment.c +++ b/net/ipv4/inet_fragment.c @@ -145,12 +145,16 @@ static void inet_frags_free_cb(void *ptr, void *arg) inet_frag_destroy(fq); } -static void fqdir_work_fn(struct work_struct *work) +static LLIST_HEAD(fqdir_free_list); + +static void fqdir_free_fn(struct work_struct *work) { - struct fqdir *fqdir = container_of(work, struct fqdir, destroy_work); - struct inet_frags *f = fqdir->f; + struct llist_node *kill_list; + struct fqdir *fqdir, *tmp; + struct inet_frags *f; - rhashtable_free_and_destroy(&fqdir->rhashtable, inet_frags_free_cb, NULL); + /* Atomically snapshot the list of fqdirs to free */ + kill_list = llist_del_all(&fqdir_free_list); /* We need to make sure all ongoing call_rcu(..., inet_frag_destroy_rcu) * have completed, since they need to dereference fqdir. @@ -158,10 +162,25 @@ static void fqdir_work_fn(struct work_struct *work) */ rcu_barrier(); - if (refcount_dec_and_test(&f->refcnt)) - complete(&f->completion); + llist_for_each_entry_safe(fqdir, tmp, kill_list, free_list) { + f = fqdir->f; + if (refcount_dec_and_test(&f->refcnt)) + complete(&f->completion); - kfree(fqdir); + kfree(fqdir); + } +} + +static DECLARE_WORK(fqdir_free_work, fqdir_free_fn); + +static void fqdir_work_fn(struct work_struct *work) +{ + struct fqdir *fqdir = container_of(work, struct fqdir, destroy_work); + + rhashtable_free_and_destroy(&fqdir->rhashtable, inet_frags_free_cb, NULL); + + if (llist_add(&fqdir->free_list, &fqdir_free_list)) + queue_work(system_wq, &fqdir_free_work); } int fqdir_init(struct fqdir **fqdirp, struct inet_frags *f, struct net *net) @@ -184,10 +203,22 @@ int fqdir_init(struct fqdir **fqdirp, struct inet_frags *f, struct net *net) } EXPORT_SYMBOL(fqdir_init); +static struct workqueue_struct *inet_frag_wq; + +static int __init inet_frag_wq_init(void) +{ + inet_frag_wq = create_workqueue("inet_frag_wq"); + if (!inet_frag_wq) + panic("Could not create inet frag workq"); + return 0; +} + +pure_initcall(inet_frag_wq_init); + void fqdir_exit(struct fqdir *fqdir) { INIT_WORK(&fqdir->destroy_work, fqdir_work_fn); - queue_work(system_wq, &fqdir->destroy_work); + queue_work(inet_frag_wq, &fqdir->destroy_work); } EXPORT_SYMBOL(fqdir_exit); -- cgit v1.2.3 From 049fe386d35353398eee40ba8d76ab62cb5a24e5 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 10 Dec 2020 14:25:03 -0800 Subject: tcp: parse mptcp options contained in reset packets Because TCP-level resets only affect the subflow, there is a MPTCP option to indicate that the MPTCP-level connection should be closed immediately without a mptcp-level fin exchange. This is the 'MPTCP fast close option'. It can be carried on ack segments or TCP resets. In the latter case, its needed to parse mptcp options also for reset packets so that MPTCP can act accordingly. Next patch will add receive side fastclose support in MPTCP. Acked-by: Matthieu Baerts Signed-off-by: Florian Westphal Signed-off-by: Mat Martineau Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 2 +- net/ipv4/tcp_input.c | 13 ++++++++----- net/ipv4/tcp_minisocks.c | 2 +- 3 files changed, 10 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index a62fb7f8a1e3..b1a05f8b35f0 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -611,7 +611,7 @@ void tcp_skb_collapse_tstamp(struct sk_buff *skb, /* tcp_input.c */ void tcp_rearm_rto(struct sock *sk); void tcp_synack_rtt_meas(struct sock *sk, struct request_sock *req); -void tcp_reset(struct sock *sk); +void tcp_reset(struct sock *sk, struct sk_buff *skb); void tcp_skb_mark_lost_uncond_verify(struct tcp_sock *tp, struct sk_buff *skb); void tcp_fin(struct sock *sk); diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index d6ad3b5c38e7..48ee476aa031 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4218,10 +4218,13 @@ static inline bool tcp_sequence(const struct tcp_sock *tp, u32 seq, u32 end_seq) } /* When we get a reset we do this. */ -void tcp_reset(struct sock *sk) +void tcp_reset(struct sock *sk, struct sk_buff *skb) { trace_tcp_receive_reset(sk); + if (sk_is_mptcp(sk)) + mptcp_incoming_options(sk, skb); + /* We want the right error as BSD sees it (and indeed as we do). */ switch (sk->sk_state) { case TCP_SYN_SENT: @@ -5604,7 +5607,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb, &tp->last_oow_ack_time)) tcp_send_dupack(sk, skb); } else if (tcp_reset_check(sk, skb)) { - tcp_reset(sk); + tcp_reset(sk, skb); } goto discard; } @@ -5640,7 +5643,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb, } if (rst_seq_match) - tcp_reset(sk); + tcp_reset(sk, skb); else { /* Disable TFO if RST is out-of-order * and no data has been received @@ -6077,7 +6080,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, */ if (th->rst) { - tcp_reset(sk); + tcp_reset(sk, skb); goto discard; } @@ -6519,7 +6522,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) if (TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq && after(TCP_SKB_CB(skb)->end_seq - th->fin, tp->rcv_nxt)) { NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONDATA); - tcp_reset(sk); + tcp_reset(sk, skb); return 1; } } diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index 495dda2449fe..0055ae0a3bf8 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -801,7 +801,7 @@ embryonic_reset: req->rsk_ops->send_reset(sk, skb); } else if (fastopen) { /* received a valid RST pkt */ reqsk_fastopen_remove(sk, req, true); - tcp_reset(sk); + tcp_reset(sk, skb); } if (!fastopen) { inet_csk_reqsk_queue_drop(sk, req); -- cgit v1.2.3 From 6d4634d1b09172a9f5863d8c4cec8f82fbecdf15 Mon Sep 17 00:00:00 2001 From: Cambda Zhu Date: Tue, 8 Dec 2020 17:19:10 +0800 Subject: net: Limit logical shift left of TCP probe0 timeout For each TCP zero window probe, the icsk_backoff is increased by one and its max value is tcp_retries2. If tcp_retries2 is greater than 63, the probe0 timeout shift may exceed its max bits. On x86_64/ARMv8/MIPS, the shift count would be masked to range 0 to 63. And on ARMv7 the result is zero. If the shift count is masked, only several probes will be sent with timeout shorter than TCP_RTO_MAX. But if the timeout is zero, it needs tcp_retries2 times probes to end this false timeout. Besides, bitwise shift greater than or equal to the width is an undefined behavior. This patch adds a limit to the backoff. The max value of max_when is TCP_RTO_MAX and the min value of timeout base is TCP_RTO_MIN. The limit is the backoff from TCP_RTO_MIN to TCP_RTO_MAX. Signed-off-by: Cambda Zhu Link: https://lore.kernel.org/r/20201208091910.37618-1-cambda@linux.alibaba.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index b1a05f8b35f0..78d13c88720f 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1326,7 +1326,9 @@ static inline unsigned long tcp_probe0_base(const struct sock *sk) static inline unsigned long tcp_probe0_when(const struct sock *sk, unsigned long max_when) { - u64 when = (u64)tcp_probe0_base(sk) << inet_csk(sk)->icsk_backoff; + u8 backoff = min_t(u8, ilog2(TCP_RTO_MAX / TCP_RTO_MIN) + 1, + inet_csk(sk)->icsk_backoff); + u64 when = (u64)tcp_probe0_base(sk) << backoff; return (unsigned long)min_t(u64, when, max_when); } -- cgit v1.2.3 From 0780b4145634c3e8d69905dc5d2695d1207130df Mon Sep 17 00:00:00 2001 From: Toke Høiland-Jørgensen Date: Fri, 11 Dec 2020 15:26:38 +0100 Subject: inet_ecn: Use csum16_add() helper for IP_ECN_set_* helpers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Jakub pointed out that the IP_ECN_set* helpers basically open-code csum16_add(), so let's switch them over to using the helper instead. v2: - Use __be16 for check_add stack variable in IP_ECN_set_ce() (kbot) v3: - Turns out we need __force casts to do arithmetic on __be16 types Reported-by: Jakub Kicinski Tested-by: Jonathan Morton Tested-by: Pete Heist Signed-off-by: Toke Høiland-Jørgensen Link: https://lore.kernel.org/r/20201211142638.154780-1-toke@redhat.com Signed-off-by: Jakub Kicinski --- include/net/inet_ecn.h | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_ecn.h b/include/net/inet_ecn.h index 563457fec557..ba77f47ef61e 100644 --- a/include/net/inet_ecn.h +++ b/include/net/inet_ecn.h @@ -8,6 +8,7 @@ #include #include +#include enum { INET_ECN_NOT_ECT = 0, @@ -75,8 +76,8 @@ static inline void INET_ECN_dontxmit(struct sock *sk) static inline int IP_ECN_set_ce(struct iphdr *iph) { - u32 check = (__force u32)iph->check; u32 ecn = (iph->tos + 1) & INET_ECN_MASK; + __be16 check_add; /* * After the last operation we have (in binary): @@ -93,23 +94,20 @@ static inline int IP_ECN_set_ce(struct iphdr *iph) * INET_ECN_ECT_1 => check += htons(0xFFFD) * INET_ECN_ECT_0 => check += htons(0xFFFE) */ - check += (__force u16)htons(0xFFFB) + (__force u16)htons(ecn); + check_add = (__force __be16)((__force u16)htons(0xFFFB) + + (__force u16)htons(ecn)); - iph->check = (__force __sum16)(check + (check>=0xFFFF)); + iph->check = csum16_add(iph->check, check_add); iph->tos |= INET_ECN_CE; return 1; } static inline int IP_ECN_set_ect1(struct iphdr *iph) { - u32 check = (__force u32)iph->check; - if ((iph->tos & INET_ECN_MASK) != INET_ECN_ECT_0) return 0; - check += (__force u16)htons(0x1); - - iph->check = (__force __sum16)(check + (check>=0xFFFF)); + iph->check = csum16_add(iph->check, htons(0x1)); iph->tos ^= INET_ECN_MASK; return 1; } -- cgit v1.2.3