From 41353fbad4f551e82c2792f7e82ac225c79cc710 Mon Sep 17 00:00:00 2001 From: Guixin Liu Date: Thu, 18 Jan 2024 20:51:45 +0800 Subject: nvmet: unify aer type enum The host and target use two definition of aer type, unify them into a single one. Signed-off-by: Guixin Liu Reviewed-by: Christoph Hellwig Reviewed-by: Sagi Grimberg Signed-off-by: Keith Busch --- include/linux/nvme.h | 6 ------ 1 file changed, 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 462c21e0e417..68eff8c86ce3 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -816,12 +816,6 @@ struct nvme_reservation_status_ext { struct nvme_registered_ctrl_ext regctl_eds[]; }; -enum nvme_async_event_type { - NVME_AER_TYPE_ERROR = 0, - NVME_AER_TYPE_SMART = 1, - NVME_AER_TYPE_NOTICE = 2, -}; - /* I/O commands */ enum nvme_opcode { -- cgit v1.2.3 From a22fe1d6dec7e98535b97249fdc95c2be79120bb Mon Sep 17 00:00:00 2001 From: Frank Li Date: Tue, 23 Jan 2024 12:28:41 -0500 Subject: dmaengine: fix is_slave_direction() return false when DMA_DEV_TO_DEV is_slave_direction() should return true when direction is DMA_DEV_TO_DEV. Fixes: 49920bc66984 ("dmaengine: add new enum dma_transfer_direction") Signed-off-by: Frank Li Link: https://lore.kernel.org/r/20240123172842.3764529-1-Frank.Li@nxp.com Signed-off-by: Vinod Koul --- include/linux/dmaengine.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index 3df70d6131c8..752dbde4cec1 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -953,7 +953,8 @@ static inline int dmaengine_slave_config(struct dma_chan *chan, static inline bool is_slave_direction(enum dma_transfer_direction direction) { - return (direction == DMA_MEM_TO_DEV) || (direction == DMA_DEV_TO_MEM); + return (direction == DMA_MEM_TO_DEV) || (direction == DMA_DEV_TO_MEM) || + (direction == DMA_DEV_TO_DEV); } static inline struct dma_async_tx_descriptor *dmaengine_prep_slave_single( -- cgit v1.2.3 From 5a287d3d2b9de2b3e747132c615599907ba5c3c1 Mon Sep 17 00:00:00 2001 From: Ondrej Mosnacek Date: Fri, 26 Jan 2024 19:45:31 +0100 Subject: lsm: fix default return value of the socket_getpeersec_*() hooks For these hooks the true "neutral" value is -EOPNOTSUPP, which is currently what is returned when no LSM provides this hook and what LSMs return when there is no security context set on the socket. Correct the value in and adjust the dispatch functions in security/security.c to avoid issues when the BPF LSM is enabled. Cc: stable@vger.kernel.org Fixes: 98e828a0650f ("security: Refactor declaration of LSM hooks") Signed-off-by: Ondrej Mosnacek [PM: subject line tweak] Signed-off-by: Paul Moore --- include/linux/lsm_hook_defs.h | 4 ++-- security/security.c | 31 +++++++++++++++++++++++++++---- 2 files changed, 29 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/lsm_hook_defs.h b/include/linux/lsm_hook_defs.h index 185924c56378..76458b6d53da 100644 --- a/include/linux/lsm_hook_defs.h +++ b/include/linux/lsm_hook_defs.h @@ -315,9 +315,9 @@ LSM_HOOK(int, 0, socket_getsockopt, struct socket *sock, int level, int optname) LSM_HOOK(int, 0, socket_setsockopt, struct socket *sock, int level, int optname) LSM_HOOK(int, 0, socket_shutdown, struct socket *sock, int how) LSM_HOOK(int, 0, socket_sock_rcv_skb, struct sock *sk, struct sk_buff *skb) -LSM_HOOK(int, 0, socket_getpeersec_stream, struct socket *sock, +LSM_HOOK(int, -ENOPROTOOPT, socket_getpeersec_stream, struct socket *sock, sockptr_t optval, sockptr_t optlen, unsigned int len) -LSM_HOOK(int, 0, socket_getpeersec_dgram, struct socket *sock, +LSM_HOOK(int, -ENOPROTOOPT, socket_getpeersec_dgram, struct socket *sock, struct sk_buff *skb, u32 *secid) LSM_HOOK(int, 0, sk_alloc_security, struct sock *sk, int family, gfp_t priority) LSM_HOOK(void, LSM_RET_VOID, sk_free_security, struct sock *sk) diff --git a/security/security.c b/security/security.c index 6196ccaba433..3aaad75c9ce8 100644 --- a/security/security.c +++ b/security/security.c @@ -4624,8 +4624,20 @@ EXPORT_SYMBOL(security_sock_rcv_skb); int security_socket_getpeersec_stream(struct socket *sock, sockptr_t optval, sockptr_t optlen, unsigned int len) { - return call_int_hook(socket_getpeersec_stream, -ENOPROTOOPT, sock, - optval, optlen, len); + struct security_hook_list *hp; + int rc; + + /* + * Only one module will provide a security context. + */ + hlist_for_each_entry(hp, &security_hook_heads.socket_getpeersec_stream, + list) { + rc = hp->hook.socket_getpeersec_stream(sock, optval, optlen, + len); + if (rc != LSM_RET_DEFAULT(socket_getpeersec_stream)) + return rc; + } + return LSM_RET_DEFAULT(socket_getpeersec_stream); } /** @@ -4645,8 +4657,19 @@ int security_socket_getpeersec_stream(struct socket *sock, sockptr_t optval, int security_socket_getpeersec_dgram(struct socket *sock, struct sk_buff *skb, u32 *secid) { - return call_int_hook(socket_getpeersec_dgram, -ENOPROTOOPT, sock, - skb, secid); + struct security_hook_list *hp; + int rc; + + /* + * Only one module will provide a security context. + */ + hlist_for_each_entry(hp, &security_hook_heads.socket_getpeersec_dgram, + list) { + rc = hp->hook.socket_getpeersec_dgram(sock, skb, secid); + if (rc != LSM_RET_DEFAULT(socket_getpeersec_dgram)) + return rc; + } + return LSM_RET_DEFAULT(socket_getpeersec_dgram); } EXPORT_SYMBOL(security_socket_getpeersec_dgram); -- cgit v1.2.3 From 764ad6b02777d77dca3659ca490f0898aa593670 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Wed, 24 Jan 2024 12:26:59 +0100 Subject: HID: bpf: use __bpf_kfunc instead of noinline Follow the docs at Documentation/bpf/kfuncs.rst: - declare the function with `__bpf_kfunc` - disables missing prototype warnings, which allows to remove them from include/linux/hid-bpf.h Removing the prototypes is not an issue because we currently have to redeclare them when writing the BPF program. They will eventually be generated by bpftool directly AFAIU. Link: https://lore.kernel.org/r/20240124-b4-hid-bpf-fixes-v2-3-052520b1e5e6@kernel.org Signed-off-by: Benjamin Tissoires --- drivers/hid/bpf/hid_bpf_dispatch.c | 18 +++++++++++++----- include/linux/hid_bpf.h | 11 ----------- 2 files changed, 13 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/drivers/hid/bpf/hid_bpf_dispatch.c b/drivers/hid/bpf/hid_bpf_dispatch.c index 7903c8638e81..470ae2c29c94 100644 --- a/drivers/hid/bpf/hid_bpf_dispatch.c +++ b/drivers/hid/bpf/hid_bpf_dispatch.c @@ -143,6 +143,9 @@ u8 *call_hid_bpf_rdesc_fixup(struct hid_device *hdev, u8 *rdesc, unsigned int *s } EXPORT_SYMBOL_GPL(call_hid_bpf_rdesc_fixup); +/* Disables missing prototype warnings */ +__bpf_kfunc_start_defs(); + /** * hid_bpf_get_data - Get the kernel memory pointer associated with the context @ctx * @@ -152,7 +155,7 @@ EXPORT_SYMBOL_GPL(call_hid_bpf_rdesc_fixup); * * @returns %NULL on error, an %__u8 memory pointer on success */ -noinline __u8 * +__bpf_kfunc __u8 * hid_bpf_get_data(struct hid_bpf_ctx *ctx, unsigned int offset, const size_t rdwr_buf_size) { struct hid_bpf_ctx_kern *ctx_kern; @@ -167,6 +170,7 @@ hid_bpf_get_data(struct hid_bpf_ctx *ctx, unsigned int offset, const size_t rdwr return ctx_kern->data + offset; } +__bpf_kfunc_end_defs(); /* * The following set contains all functions we agree BPF programs @@ -274,6 +278,9 @@ static int do_hid_bpf_attach_prog(struct hid_device *hdev, int prog_fd, struct b return fd; } +/* Disables missing prototype warnings */ +__bpf_kfunc_start_defs(); + /** * hid_bpf_attach_prog - Attach the given @prog_fd to the given HID device * @@ -286,7 +293,7 @@ static int do_hid_bpf_attach_prog(struct hid_device *hdev, int prog_fd, struct b * is pinned to the BPF file system). */ /* called from syscall */ -noinline int +__bpf_kfunc int hid_bpf_attach_prog(unsigned int hid_id, int prog_fd, __u32 flags) { struct hid_device *hdev; @@ -338,7 +345,7 @@ hid_bpf_attach_prog(unsigned int hid_id, int prog_fd, __u32 flags) * * @returns A pointer to &struct hid_bpf_ctx on success, %NULL on error. */ -noinline struct hid_bpf_ctx * +__bpf_kfunc struct hid_bpf_ctx * hid_bpf_allocate_context(unsigned int hid_id) { struct hid_device *hdev; @@ -371,7 +378,7 @@ hid_bpf_allocate_context(unsigned int hid_id) * @ctx: the HID-BPF context to release * */ -noinline void +__bpf_kfunc void hid_bpf_release_context(struct hid_bpf_ctx *ctx) { struct hid_bpf_ctx_kern *ctx_kern; @@ -397,7 +404,7 @@ hid_bpf_release_context(struct hid_bpf_ctx *ctx) * * @returns %0 on success, a negative error code otherwise. */ -noinline int +__bpf_kfunc int hid_bpf_hw_request(struct hid_bpf_ctx *ctx, __u8 *buf, size_t buf__sz, enum hid_report_type rtype, enum hid_class_request reqtype) { @@ -465,6 +472,7 @@ hid_bpf_hw_request(struct hid_bpf_ctx *ctx, __u8 *buf, size_t buf__sz, kfree(dma_data); return ret; } +__bpf_kfunc_end_defs(); /* our HID-BPF entrypoints */ BTF_SET8_START(hid_bpf_fmodret_ids) diff --git a/include/linux/hid_bpf.h b/include/linux/hid_bpf.h index 840cd254172d..7118ac28d468 100644 --- a/include/linux/hid_bpf.h +++ b/include/linux/hid_bpf.h @@ -77,17 +77,6 @@ enum hid_bpf_attach_flags { int hid_bpf_device_event(struct hid_bpf_ctx *ctx); int hid_bpf_rdesc_fixup(struct hid_bpf_ctx *ctx); -/* Following functions are kfunc that we export to BPF programs */ -/* available everywhere in HID-BPF */ -__u8 *hid_bpf_get_data(struct hid_bpf_ctx *ctx, unsigned int offset, const size_t __sz); - -/* only available in syscall */ -int hid_bpf_attach_prog(unsigned int hid_id, int prog_fd, __u32 flags); -int hid_bpf_hw_request(struct hid_bpf_ctx *ctx, __u8 *buf, size_t buf__sz, - enum hid_report_type rtype, enum hid_class_request reqtype); -struct hid_bpf_ctx *hid_bpf_allocate_context(unsigned int hid_id); -void hid_bpf_release_context(struct hid_bpf_ctx *ctx); - /* * Below is HID internal */ -- cgit v1.2.3 From 1e560864159d002b453da42bd2c13a1805515a20 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 30 Jan 2024 11:02:43 +0100 Subject: PCI/ASPM: Fix deadlock when enabling ASPM A last minute revert in 6.7-final introduced a potential deadlock when enabling ASPM during probe of Qualcomm PCIe controllers as reported by lockdep: ============================================ WARNING: possible recursive locking detected 6.7.0 #40 Not tainted -------------------------------------------- kworker/u16:5/90 is trying to acquire lock: ffffacfa78ced000 (pci_bus_sem){++++}-{3:3}, at: pcie_aspm_pm_state_change+0x58/0xdc but task is already holding lock: ffffacfa78ced000 (pci_bus_sem){++++}-{3:3}, at: pci_walk_bus+0x34/0xbc other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(pci_bus_sem); lock(pci_bus_sem); *** DEADLOCK *** Call trace: print_deadlock_bug+0x25c/0x348 __lock_acquire+0x10a4/0x2064 lock_acquire+0x1e8/0x318 down_read+0x60/0x184 pcie_aspm_pm_state_change+0x58/0xdc pci_set_full_power_state+0xa8/0x114 pci_set_power_state+0xc4/0x120 qcom_pcie_enable_aspm+0x1c/0x3c [pcie_qcom] pci_walk_bus+0x64/0xbc qcom_pcie_host_post_init_2_7_0+0x28/0x34 [pcie_qcom] The deadlock can easily be reproduced on machines like the Lenovo ThinkPad X13s by adding a delay to increase the race window during asynchronous probe where another thread can take a write lock. Add a new pci_set_power_state_locked() and associated helper functions that can be called with the PCI bus semaphore held to avoid taking the read lock twice. Link: https://lore.kernel.org/r/ZZu0qx2cmn7IwTyQ@hovoldconsulting.com Link: https://lore.kernel.org/r/20240130100243.11011-1-johan+linaro@kernel.org Fixes: f93e71aea6c6 ("Revert "PCI/ASPM: Remove pcie_aspm_pm_state_change()"") Signed-off-by: Johan Hovold Signed-off-by: Bjorn Helgaas Cc: # 6.7 --- drivers/pci/bus.c | 49 +++++++++++++-------- drivers/pci/controller/dwc/pcie-qcom.c | 2 +- drivers/pci/pci.c | 78 ++++++++++++++++++++++------------ drivers/pci/pci.h | 4 +- drivers/pci/pcie/aspm.c | 13 ++++-- include/linux/pci.h | 5 +++ 6 files changed, 101 insertions(+), 50 deletions(-) (limited to 'include/linux') diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c index 9c2137dae429..826b5016a101 100644 --- a/drivers/pci/bus.c +++ b/drivers/pci/bus.c @@ -386,21 +386,8 @@ void pci_bus_add_devices(const struct pci_bus *bus) } EXPORT_SYMBOL(pci_bus_add_devices); -/** pci_walk_bus - walk devices on/under bus, calling callback. - * @top bus whose devices should be walked - * @cb callback to be called for each device found - * @userdata arbitrary pointer to be passed to callback. - * - * Walk the given bus, including any bridged devices - * on buses under this bus. Call the provided callback - * on each device found. - * - * We check the return of @cb each time. If it returns anything - * other than 0, we break out. - * - */ -void pci_walk_bus(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), - void *userdata) +static void __pci_walk_bus(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), + void *userdata, bool locked) { struct pci_dev *dev; struct pci_bus *bus; @@ -408,7 +395,8 @@ void pci_walk_bus(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), int retval; bus = top; - down_read(&pci_bus_sem); + if (!locked) + down_read(&pci_bus_sem); next = top->devices.next; for (;;) { if (next == &bus->devices) { @@ -431,10 +419,37 @@ void pci_walk_bus(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), if (retval) break; } - up_read(&pci_bus_sem); + if (!locked) + up_read(&pci_bus_sem); +} + +/** + * pci_walk_bus - walk devices on/under bus, calling callback. + * @top: bus whose devices should be walked + * @cb: callback to be called for each device found + * @userdata: arbitrary pointer to be passed to callback + * + * Walk the given bus, including any bridged devices + * on buses under this bus. Call the provided callback + * on each device found. + * + * We check the return of @cb each time. If it returns anything + * other than 0, we break out. + */ +void pci_walk_bus(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), void *userdata) +{ + __pci_walk_bus(top, cb, userdata, false); } EXPORT_SYMBOL_GPL(pci_walk_bus); +void pci_walk_bus_locked(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), void *userdata) +{ + lockdep_assert_held(&pci_bus_sem); + + __pci_walk_bus(top, cb, userdata, true); +} +EXPORT_SYMBOL_GPL(pci_walk_bus_locked); + struct pci_bus *pci_bus_get(struct pci_bus *bus) { if (bus) diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c index 10f2d0bb86be..2ce2a3bd932b 100644 --- a/drivers/pci/controller/dwc/pcie-qcom.c +++ b/drivers/pci/controller/dwc/pcie-qcom.c @@ -972,7 +972,7 @@ static int qcom_pcie_enable_aspm(struct pci_dev *pdev, void *userdata) * Downstream devices need to be in D0 state before enabling PCI PM * substates. */ - pci_set_power_state(pdev, PCI_D0); + pci_set_power_state_locked(pdev, PCI_D0); pci_enable_link_state_locked(pdev, PCIE_LINK_STATE_ALL); return 0; diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index d8f11a078924..9ab9b1008d8b 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -1354,6 +1354,7 @@ end: /** * pci_set_full_power_state - Put a PCI device into D0 and update its state * @dev: PCI device to power up + * @locked: whether pci_bus_sem is held * * Call pci_power_up() to put @dev into D0, read from its PCI_PM_CTRL register * to confirm the state change, restore its BARs if they might be lost and @@ -1363,7 +1364,7 @@ end: * to D0, it is more efficient to use pci_power_up() directly instead of this * function. */ -static int pci_set_full_power_state(struct pci_dev *dev) +static int pci_set_full_power_state(struct pci_dev *dev, bool locked) { u16 pmcsr; int ret; @@ -1399,7 +1400,7 @@ static int pci_set_full_power_state(struct pci_dev *dev) } if (dev->bus->self) - pcie_aspm_pm_state_change(dev->bus->self); + pcie_aspm_pm_state_change(dev->bus->self, locked); return 0; } @@ -1428,10 +1429,22 @@ void pci_bus_set_current_state(struct pci_bus *bus, pci_power_t state) pci_walk_bus(bus, __pci_dev_set_current_state, &state); } +static void __pci_bus_set_current_state(struct pci_bus *bus, pci_power_t state, bool locked) +{ + if (!bus) + return; + + if (locked) + pci_walk_bus_locked(bus, __pci_dev_set_current_state, &state); + else + pci_walk_bus(bus, __pci_dev_set_current_state, &state); +} + /** * pci_set_low_power_state - Put a PCI device into a low-power state. * @dev: PCI device to handle. * @state: PCI power state (D1, D2, D3hot) to put the device into. + * @locked: whether pci_bus_sem is held * * Use the device's PCI_PM_CTRL register to put it into a low-power state. * @@ -1442,7 +1455,7 @@ void pci_bus_set_current_state(struct pci_bus *bus, pci_power_t state) * 0 if device already is in the requested state. * 0 if device's power state has been successfully changed. */ -static int pci_set_low_power_state(struct pci_dev *dev, pci_power_t state) +static int pci_set_low_power_state(struct pci_dev *dev, pci_power_t state, bool locked) { u16 pmcsr; @@ -1496,29 +1509,12 @@ static int pci_set_low_power_state(struct pci_dev *dev, pci_power_t state) pci_power_name(state)); if (dev->bus->self) - pcie_aspm_pm_state_change(dev->bus->self); + pcie_aspm_pm_state_change(dev->bus->self, locked); return 0; } -/** - * pci_set_power_state - Set the power state of a PCI device - * @dev: PCI device to handle. - * @state: PCI power state (D0, D1, D2, D3hot) to put the device into. - * - * Transition a device to a new power state, using the platform firmware and/or - * the device's PCI PM registers. - * - * RETURN VALUE: - * -EINVAL if the requested state is invalid. - * -EIO if device does not support PCI PM or its PM capabilities register has a - * wrong version, or device doesn't support the requested state. - * 0 if the transition is to D1 or D2 but D1 and D2 are not supported. - * 0 if device already is in the requested state. - * 0 if the transition is to D3 but D3 is not supported. - * 0 if device's power state has been successfully changed. - */ -int pci_set_power_state(struct pci_dev *dev, pci_power_t state) +static int __pci_set_power_state(struct pci_dev *dev, pci_power_t state, bool locked) { int error; @@ -1542,7 +1538,7 @@ int pci_set_power_state(struct pci_dev *dev, pci_power_t state) return 0; if (state == PCI_D0) - return pci_set_full_power_state(dev); + return pci_set_full_power_state(dev, locked); /* * This device is quirked not to be put into D3, so don't put it in @@ -1556,16 +1552,16 @@ int pci_set_power_state(struct pci_dev *dev, pci_power_t state) * To put the device in D3cold, put it into D3hot in the native * way, then put it into D3cold using platform ops. */ - error = pci_set_low_power_state(dev, PCI_D3hot); + error = pci_set_low_power_state(dev, PCI_D3hot, locked); if (pci_platform_power_transition(dev, PCI_D3cold)) return error; /* Powering off a bridge may power off the whole hierarchy */ if (dev->current_state == PCI_D3cold) - pci_bus_set_current_state(dev->subordinate, PCI_D3cold); + __pci_bus_set_current_state(dev->subordinate, PCI_D3cold, locked); } else { - error = pci_set_low_power_state(dev, state); + error = pci_set_low_power_state(dev, state, locked); if (pci_platform_power_transition(dev, state)) return error; @@ -1573,8 +1569,38 @@ int pci_set_power_state(struct pci_dev *dev, pci_power_t state) return 0; } + +/** + * pci_set_power_state - Set the power state of a PCI device + * @dev: PCI device to handle. + * @state: PCI power state (D0, D1, D2, D3hot) to put the device into. + * + * Transition a device to a new power state, using the platform firmware and/or + * the device's PCI PM registers. + * + * RETURN VALUE: + * -EINVAL if the requested state is invalid. + * -EIO if device does not support PCI PM or its PM capabilities register has a + * wrong version, or device doesn't support the requested state. + * 0 if the transition is to D1 or D2 but D1 and D2 are not supported. + * 0 if device already is in the requested state. + * 0 if the transition is to D3 but D3 is not supported. + * 0 if device's power state has been successfully changed. + */ +int pci_set_power_state(struct pci_dev *dev, pci_power_t state) +{ + return __pci_set_power_state(dev, state, false); +} EXPORT_SYMBOL(pci_set_power_state); +int pci_set_power_state_locked(struct pci_dev *dev, pci_power_t state) +{ + lockdep_assert_held(&pci_bus_sem); + + return __pci_set_power_state(dev, state, true); +} +EXPORT_SYMBOL(pci_set_power_state_locked); + #define PCI_EXP_SAVE_REGS 7 static struct pci_cap_saved_state *_pci_find_saved_cap(struct pci_dev *pci_dev, diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index 2336a8d1edab..e9750b1b19ba 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -571,12 +571,12 @@ int pcie_retrain_link(struct pci_dev *pdev, bool use_lt); #ifdef CONFIG_PCIEASPM void pcie_aspm_init_link_state(struct pci_dev *pdev); void pcie_aspm_exit_link_state(struct pci_dev *pdev); -void pcie_aspm_pm_state_change(struct pci_dev *pdev); +void pcie_aspm_pm_state_change(struct pci_dev *pdev, bool locked); void pcie_aspm_powersave_config_link(struct pci_dev *pdev); #else static inline void pcie_aspm_init_link_state(struct pci_dev *pdev) { } static inline void pcie_aspm_exit_link_state(struct pci_dev *pdev) { } -static inline void pcie_aspm_pm_state_change(struct pci_dev *pdev) { } +static inline void pcie_aspm_pm_state_change(struct pci_dev *pdev, bool locked) { } static inline void pcie_aspm_powersave_config_link(struct pci_dev *pdev) { } #endif diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c index 5a0066ecc3c5..bc0bd86695ec 100644 --- a/drivers/pci/pcie/aspm.c +++ b/drivers/pci/pcie/aspm.c @@ -1003,8 +1003,11 @@ void pcie_aspm_exit_link_state(struct pci_dev *pdev) up_read(&pci_bus_sem); } -/* @pdev: the root port or switch downstream port */ -void pcie_aspm_pm_state_change(struct pci_dev *pdev) +/* + * @pdev: the root port or switch downstream port + * @locked: whether pci_bus_sem is held + */ +void pcie_aspm_pm_state_change(struct pci_dev *pdev, bool locked) { struct pcie_link_state *link = pdev->link_state; @@ -1014,12 +1017,14 @@ void pcie_aspm_pm_state_change(struct pci_dev *pdev) * Devices changed PM state, we should recheck if latency * meets all functions' requirement */ - down_read(&pci_bus_sem); + if (!locked) + down_read(&pci_bus_sem); mutex_lock(&aspm_lock); pcie_update_aspm_capable(link->root); pcie_config_aspm_path(link); mutex_unlock(&aspm_lock); - up_read(&pci_bus_sem); + if (!locked) + up_read(&pci_bus_sem); } void pcie_aspm_powersave_config_link(struct pci_dev *pdev) diff --git a/include/linux/pci.h b/include/linux/pci.h index add9368e6314..7ab0d13672da 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1422,6 +1422,7 @@ int pci_load_and_free_saved_state(struct pci_dev *dev, struct pci_saved_state **state); int pci_platform_power_transition(struct pci_dev *dev, pci_power_t state); int pci_set_power_state(struct pci_dev *dev, pci_power_t state); +int pci_set_power_state_locked(struct pci_dev *dev, pci_power_t state); pci_power_t pci_choose_state(struct pci_dev *dev, pm_message_t state); bool pci_pme_capable(struct pci_dev *dev, pci_power_t state); void pci_pme_active(struct pci_dev *dev, bool enable); @@ -1625,6 +1626,8 @@ int pci_scan_bridge(struct pci_bus *bus, struct pci_dev *dev, int max, void pci_walk_bus(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), void *userdata); +void pci_walk_bus_locked(struct pci_bus *top, int (*cb)(struct pci_dev *, void *), + void *userdata); int pci_cfg_space_size(struct pci_dev *dev); unsigned char pci_bus_max_busnr(struct pci_bus *bus); void pci_setup_bridge(struct pci_bus *bus); @@ -2025,6 +2028,8 @@ static inline int pci_save_state(struct pci_dev *dev) { return 0; } static inline void pci_restore_state(struct pci_dev *dev) { } static inline int pci_set_power_state(struct pci_dev *dev, pci_power_t state) { return 0; } +static inline int pci_set_power_state_locked(struct pci_dev *dev, pci_power_t state) +{ return 0; } static inline int pci_wake_from_d3(struct pci_dev *dev, bool enable) { return 0; } static inline pci_power_t pci_choose_state(struct pci_dev *dev, -- cgit v1.2.3 From 97f7cf1cd80eeed3b7c808b7c12463295c751001 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Mon, 29 Jan 2024 10:57:01 +0100 Subject: netfilter: ipset: fix performance regression in swap operation The patch "netfilter: ipset: fix race condition between swap/destroy and kernel side add/del/test", commit 28628fa9 fixes a race condition. But the synchronize_rcu() added to the swap function unnecessarily slows it down: it can safely be moved to destroy and use call_rcu() instead. Eric Dumazet pointed out that simply calling the destroy functions as rcu callback does not work: sets with timeout use garbage collectors which need cancelling at destroy which can wait. Therefore the destroy functions are split into two: cancelling garbage collectors safely at executing the command received by netlink and moving the remaining part only into the rcu callback. Link: https://lore.kernel.org/lkml/C0829B10-EAA6-4809-874E-E1E9C05A8D84@automattic.com/ Fixes: 28628fa952fe ("netfilter: ipset: fix race condition between swap/destroy and kernel side add/del/test") Reported-by: Ale Crismani Reported-by: David Wang <00107082@163.com> Tested-by: David Wang <00107082@163.com> Signed-off-by: Jozsef Kadlecsik Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/ipset/ip_set.h | 4 ++++ net/netfilter/ipset/ip_set_bitmap_gen.h | 14 ++++++++++--- net/netfilter/ipset/ip_set_core.c | 37 +++++++++++++++++++++++++-------- net/netfilter/ipset/ip_set_hash_gen.h | 15 ++++++++++--- net/netfilter/ipset/ip_set_list_set.c | 13 +++++++++--- 5 files changed, 65 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/include/linux/netfilter/ipset/ip_set.h b/include/linux/netfilter/ipset/ip_set.h index e8c350a3ade1..e9f4f845d760 100644 --- a/include/linux/netfilter/ipset/ip_set.h +++ b/include/linux/netfilter/ipset/ip_set.h @@ -186,6 +186,8 @@ struct ip_set_type_variant { /* Return true if "b" set is the same as "a" * according to the create set parameters */ bool (*same_set)(const struct ip_set *a, const struct ip_set *b); + /* Cancel ongoing garbage collectors before destroying the set*/ + void (*cancel_gc)(struct ip_set *set); /* Region-locking is used */ bool region_lock; }; @@ -242,6 +244,8 @@ extern void ip_set_type_unregister(struct ip_set_type *set_type); /* A generic IP set */ struct ip_set { + /* For call_cru in destroy */ + struct rcu_head rcu; /* The name of the set */ char name[IPSET_MAXNAMELEN]; /* Lock protecting the set data */ diff --git a/net/netfilter/ipset/ip_set_bitmap_gen.h b/net/netfilter/ipset/ip_set_bitmap_gen.h index 21f7860e8fa1..cb48a2b9cb9f 100644 --- a/net/netfilter/ipset/ip_set_bitmap_gen.h +++ b/net/netfilter/ipset/ip_set_bitmap_gen.h @@ -30,6 +30,7 @@ #define mtype_del IPSET_TOKEN(MTYPE, _del) #define mtype_list IPSET_TOKEN(MTYPE, _list) #define mtype_gc IPSET_TOKEN(MTYPE, _gc) +#define mtype_cancel_gc IPSET_TOKEN(MTYPE, _cancel_gc) #define mtype MTYPE #define get_ext(set, map, id) ((map)->extensions + ((set)->dsize * (id))) @@ -59,9 +60,6 @@ mtype_destroy(struct ip_set *set) { struct mtype *map = set->data; - if (SET_WITH_TIMEOUT(set)) - del_timer_sync(&map->gc); - if (set->dsize && set->extensions & IPSET_EXT_DESTROY) mtype_ext_cleanup(set); ip_set_free(map->members); @@ -290,6 +288,15 @@ mtype_gc(struct timer_list *t) add_timer(&map->gc); } +static void +mtype_cancel_gc(struct ip_set *set) +{ + struct mtype *map = set->data; + + if (SET_WITH_TIMEOUT(set)) + del_timer_sync(&map->gc); +} + static const struct ip_set_type_variant mtype = { .kadt = mtype_kadt, .uadt = mtype_uadt, @@ -303,6 +310,7 @@ static const struct ip_set_type_variant mtype = { .head = mtype_head, .list = mtype_list, .same_set = mtype_same_set, + .cancel_gc = mtype_cancel_gc, }; #endif /* __IP_SET_BITMAP_IP_GEN_H */ diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index 4c133e06be1d..bcaad9c009fe 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -1182,6 +1182,14 @@ ip_set_destroy_set(struct ip_set *set) kfree(set); } +static void +ip_set_destroy_set_rcu(struct rcu_head *head) +{ + struct ip_set *set = container_of(head, struct ip_set, rcu); + + ip_set_destroy_set(set); +} + static int ip_set_destroy(struct sk_buff *skb, const struct nfnl_info *info, const struct nlattr * const attr[]) { @@ -1193,8 +1201,6 @@ static int ip_set_destroy(struct sk_buff *skb, const struct nfnl_info *info, if (unlikely(protocol_min_failed(attr))) return -IPSET_ERR_PROTOCOL; - /* Must wait for flush to be really finished in list:set */ - rcu_barrier(); /* Commands are serialized and references are * protected by the ip_set_ref_lock. @@ -1206,8 +1212,10 @@ static int ip_set_destroy(struct sk_buff *skb, const struct nfnl_info *info, * counter, so if it's already zero, we can proceed * without holding the lock. */ - read_lock_bh(&ip_set_ref_lock); if (!attr[IPSET_ATTR_SETNAME]) { + /* Must wait for flush to be really finished in list:set */ + rcu_barrier(); + read_lock_bh(&ip_set_ref_lock); for (i = 0; i < inst->ip_set_max; i++) { s = ip_set(inst, i); if (s && (s->ref || s->ref_netlink)) { @@ -1221,6 +1229,8 @@ static int ip_set_destroy(struct sk_buff *skb, const struct nfnl_info *info, s = ip_set(inst, i); if (s) { ip_set(inst, i) = NULL; + /* Must cancel garbage collectors */ + s->variant->cancel_gc(s); ip_set_destroy_set(s); } } @@ -1228,6 +1238,9 @@ static int ip_set_destroy(struct sk_buff *skb, const struct nfnl_info *info, inst->is_destroyed = false; } else { u32 flags = flag_exist(info->nlh); + u16 features = 0; + + read_lock_bh(&ip_set_ref_lock); s = find_set_and_id(inst, nla_data(attr[IPSET_ATTR_SETNAME]), &i); if (!s) { @@ -1238,10 +1251,16 @@ static int ip_set_destroy(struct sk_buff *skb, const struct nfnl_info *info, ret = -IPSET_ERR_BUSY; goto out; } + features = s->type->features; ip_set(inst, i) = NULL; read_unlock_bh(&ip_set_ref_lock); - - ip_set_destroy_set(s); + if (features & IPSET_TYPE_NAME) { + /* Must wait for flush to be really finished */ + rcu_barrier(); + } + /* Must cancel garbage collectors */ + s->variant->cancel_gc(s); + call_rcu(&s->rcu, ip_set_destroy_set_rcu); } return 0; out: @@ -1394,9 +1413,6 @@ static int ip_set_swap(struct sk_buff *skb, const struct nfnl_info *info, ip_set(inst, to_id) = from; write_unlock_bh(&ip_set_ref_lock); - /* Make sure all readers of the old set pointers are completed. */ - synchronize_rcu(); - return 0; } @@ -2409,8 +2425,11 @@ ip_set_fini(void) { nf_unregister_sockopt(&so_set); nfnetlink_subsys_unregister(&ip_set_netlink_subsys); - unregister_pernet_subsys(&ip_set_net_ops); + + /* Wait for call_rcu() in destroy */ + rcu_barrier(); + pr_debug("these are the famous last words\n"); } diff --git a/net/netfilter/ipset/ip_set_hash_gen.h b/net/netfilter/ipset/ip_set_hash_gen.h index cbf80da9a01c..1136510521a8 100644 --- a/net/netfilter/ipset/ip_set_hash_gen.h +++ b/net/netfilter/ipset/ip_set_hash_gen.h @@ -222,6 +222,7 @@ static const union nf_inet_addr zeromask = {}; #undef mtype_gc_do #undef mtype_gc #undef mtype_gc_init +#undef mtype_cancel_gc #undef mtype_variant #undef mtype_data_match @@ -266,6 +267,7 @@ static const union nf_inet_addr zeromask = {}; #define mtype_gc_do IPSET_TOKEN(MTYPE, _gc_do) #define mtype_gc IPSET_TOKEN(MTYPE, _gc) #define mtype_gc_init IPSET_TOKEN(MTYPE, _gc_init) +#define mtype_cancel_gc IPSET_TOKEN(MTYPE, _cancel_gc) #define mtype_variant IPSET_TOKEN(MTYPE, _variant) #define mtype_data_match IPSET_TOKEN(MTYPE, _data_match) @@ -450,9 +452,6 @@ mtype_destroy(struct ip_set *set) struct htype *h = set->data; struct list_head *l, *lt; - if (SET_WITH_TIMEOUT(set)) - cancel_delayed_work_sync(&h->gc.dwork); - mtype_ahash_destroy(set, ipset_dereference_nfnl(h->table), true); list_for_each_safe(l, lt, &h->ad) { list_del(l); @@ -599,6 +598,15 @@ mtype_gc_init(struct htable_gc *gc) queue_delayed_work(system_power_efficient_wq, &gc->dwork, HZ); } +static void +mtype_cancel_gc(struct ip_set *set) +{ + struct htype *h = set->data; + + if (SET_WITH_TIMEOUT(set)) + cancel_delayed_work_sync(&h->gc.dwork); +} + static int mtype_add(struct ip_set *set, void *value, const struct ip_set_ext *ext, struct ip_set_ext *mext, u32 flags); @@ -1441,6 +1449,7 @@ static const struct ip_set_type_variant mtype_variant = { .uref = mtype_uref, .resize = mtype_resize, .same_set = mtype_same_set, + .cancel_gc = mtype_cancel_gc, .region_lock = true, }; diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index e162636525cf..6c3f28bc59b3 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -426,9 +426,6 @@ list_set_destroy(struct ip_set *set) struct list_set *map = set->data; struct set_elem *e, *n; - if (SET_WITH_TIMEOUT(set)) - timer_shutdown_sync(&map->gc); - list_for_each_entry_safe(e, n, &map->members, list) { list_del(&e->list); ip_set_put_byindex(map->net, e->id); @@ -545,6 +542,15 @@ list_set_same_set(const struct ip_set *a, const struct ip_set *b) a->extensions == b->extensions; } +static void +list_set_cancel_gc(struct ip_set *set) +{ + struct list_set *map = set->data; + + if (SET_WITH_TIMEOUT(set)) + timer_shutdown_sync(&map->gc); +} + static const struct ip_set_type_variant set_variant = { .kadt = list_set_kadt, .uadt = list_set_uadt, @@ -558,6 +564,7 @@ static const struct ip_set_type_variant set_variant = { .head = list_set_head, .list = list_set_list, .same_set = list_set_same_set, + .cancel_gc = list_set_cancel_gc, }; static void -- cgit v1.2.3 From f9e9115d0c014dec3278d68823eaff159f98f4d6 Mon Sep 17 00:00:00 2001 From: Caleb Sander Date: Wed, 31 Jan 2024 09:43:13 -0700 Subject: nvme: take const cmd pointer in read-only helpers nvme_is_fabrics() and nvme_is_write() only read struct nvme_command, so take it by const pointer. This allows callers to pass a const pointer and communicates that these functions don't modify the command. Signed-off-by: Caleb Sander Reviewed-by: Chaitanya Kulkarni Reviewed-by: Christoph Hellwig Reviewed-by: Sagi Grimberg Signed-off-by: Keith Busch --- include/linux/nvme.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 68eff8c86ce3..bc605ec4a3fd 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -1812,7 +1812,7 @@ struct nvme_command { }; }; -static inline bool nvme_is_fabrics(struct nvme_command *cmd) +static inline bool nvme_is_fabrics(const struct nvme_command *cmd) { return cmd->common.opcode == nvme_fabrics_command; } @@ -1831,7 +1831,7 @@ struct nvme_error_slot { __u8 resv2[24]; }; -static inline bool nvme_is_write(struct nvme_command *cmd) +static inline bool nvme_is_write(const struct nvme_command *cmd) { /* * What a mess... -- cgit v1.2.3