From b8a7a3a6674725d7ca0ff6e322f6c1cab6e6a11d Mon Sep 17 00:00:00 2001
From: Andreas Gruenbacher <agruenba@redhat.com>
Date: Thu, 24 Mar 2016 14:38:37 +0100
Subject: posix_acl: Inode acl caching fixes

When get_acl() is called for an inode whose ACL is not cached yet, the
get_acl inode operation is called to fetch the ACL from the filesystem.
The inode operation is responsible for updating the cached acl with
set_cached_acl().  This is done without locking at the VFS level, so
another task can call set_cached_acl() or forget_cached_acl() before the
get_acl inode operation gets to calling set_cached_acl(), and then
get_acl's call to set_cached_acl() results in caching an outdate ACL.

Prevent this from happening by setting the cached ACL pointer to a
task-specific sentinel value before calling the get_acl inode operation.
Move the responsibility for updating the cached ACL from the get_acl
inode operations to get_acl().  There, only set the cached ACL if the
sentinel value hasn't changed.

The sentinel values are chosen to have odd values.  Likewise, the value
of ACL_NOT_CACHED is odd.  In contrast, ACL object pointers always have
an even value (ACLs are aligned in memory).  This allows to distinguish
uncached ACLs values from ACL objects.

In addition, switch from guarding inode->i_acl and inode->i_default_acl
upates by the inode->i_lock spinlock to using xchg() and cmpxchg().

Filesystems that do not want ACLs returned from their get_acl inode
operations to be cached must call forget_cached_acl() to prevent the VFS
from doing so.

(Patch written by Al Viro and Andreas Gruenbacher.)

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 include/linux/fs.h | 12 ++++++++++++
 1 file changed, 12 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/fs.h b/include/linux/fs.h
index 14a97194b34b..329ed372d708 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -577,6 +577,18 @@ static inline void mapping_allow_writable(struct address_space *mapping)
 struct posix_acl;
 #define ACL_NOT_CACHED ((void *)(-1))
 
+static inline struct posix_acl *
+uncached_acl_sentinel(struct task_struct *task)
+{
+	return (void *)task + 1;
+}
+
+static inline bool
+is_uncached_acl(struct posix_acl *acl)
+{
+	return (long)acl & 1;
+}
+
 #define IOP_FASTPERM	0x0001
 #define IOP_LOOKUP	0x0002
 #define IOP_NOFOLLOW	0x0004
-- 
cgit v1.2.3


From 04c57f4501909b60353031cfe5b991751d745658 Mon Sep 17 00:00:00 2001
From: Andreas Gruenbacher <agruenba@redhat.com>
Date: Thu, 24 Mar 2016 14:38:38 +0100
Subject: posix_acl: Unexport acl_by_type and make it static

acl_by_type(inode, type) returns a pointer to either inode->i_acl or
inode->i_default_acl depending on type.  This is useful in
fs/posix_acl.c, but should never have been visible outside that file.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/posix_acl.c            | 3 +--
 include/linux/posix_acl.h | 1 -
 2 files changed, 1 insertion(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/fs/posix_acl.c b/fs/posix_acl.c
index bc6736d60786..db1fb0f5d9ff 100644
--- a/fs/posix_acl.c
+++ b/fs/posix_acl.c
@@ -21,7 +21,7 @@
 #include <linux/export.h>
 #include <linux/user_namespace.h>
 
-struct posix_acl **acl_by_type(struct inode *inode, int type)
+static struct posix_acl **acl_by_type(struct inode *inode, int type)
 {
 	switch (type) {
 	case ACL_TYPE_ACCESS:
@@ -32,7 +32,6 @@ struct posix_acl **acl_by_type(struct inode *inode, int type)
 		BUG();
 	}
 }
-EXPORT_SYMBOL(acl_by_type);
 
 struct posix_acl *get_cached_acl(struct inode *inode, int type)
 {
diff --git a/include/linux/posix_acl.h b/include/linux/posix_acl.h
index 3e96a6a76103..5b5a80cc5926 100644
--- a/include/linux/posix_acl.h
+++ b/include/linux/posix_acl.h
@@ -99,7 +99,6 @@ extern int posix_acl_create(struct inode *, umode_t *, struct posix_acl **,
 extern int simple_set_acl(struct inode *, struct posix_acl *, int);
 extern int simple_acl_create(struct inode *, struct inode *);
 
-struct posix_acl **acl_by_type(struct inode *inode, int type);
 struct posix_acl *get_cached_acl(struct inode *inode, int type);
 struct posix_acl *get_cached_acl_rcu(struct inode *inode, int type);
 void set_cached_acl(struct inode *inode, int type, struct posix_acl *acl);
-- 
cgit v1.2.3


From b296821a7c42fa58baa17513b2b7b30ae66f3336 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 10 Apr 2016 20:48:24 -0400
Subject: xattr_handler: pass dentry and inode as separate arguments of ->get()

... and do not assume they are already attached to each other

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/9p/acl.c                  |  6 +++---
 fs/9p/xattr.c                |  4 ++--
 fs/btrfs/xattr.c             |  6 ++----
 fs/ext2/xattr_security.c     |  6 +++---
 fs/ext2/xattr_trusted.c      |  6 +++---
 fs/ext2/xattr_user.c         |  8 ++++----
 fs/ext4/xattr_security.c     |  6 +++---
 fs/ext4/xattr_trusted.c      |  6 +++---
 fs/ext4/xattr_user.c         |  8 ++++----
 fs/f2fs/xattr.c              | 14 ++++++--------
 fs/gfs2/xattr.c              |  6 +++---
 fs/hfsplus/xattr.c           | 10 +++++-----
 fs/hfsplus/xattr.h           |  2 +-
 fs/hfsplus/xattr_security.c  |  6 +++---
 fs/hfsplus/xattr_trusted.c   |  6 +++---
 fs/hfsplus/xattr_user.c      |  6 +++---
 fs/jffs2/security.c          |  6 +++---
 fs/jffs2/xattr_trusted.c     |  6 +++---
 fs/jffs2/xattr_user.c        |  6 +++---
 fs/nfs/nfs4proc.c            | 12 ++++++------
 fs/ocfs2/xattr.c             | 20 ++++++++++----------
 fs/orangefs/xattr.c          | 10 ++++++----
 fs/posix_acl.c               | 10 +++++-----
 fs/reiserfs/xattr_security.c |  9 ++++-----
 fs/reiserfs/xattr_trusted.c  |  9 ++++-----
 fs/reiserfs/xattr_user.c     |  9 ++++-----
 fs/squashfs/xattr.c          |  6 ++++--
 fs/xattr.c                   |  3 ++-
 fs/xfs/xfs_xattr.c           |  6 +++---
 include/linux/xattr.h        |  3 ++-
 mm/shmem.c                   |  6 +++---
 31 files changed, 113 insertions(+), 114 deletions(-)

(limited to 'include/linux')

diff --git a/fs/9p/acl.c b/fs/9p/acl.c
index 2d94e94b6b59..eb3589edf485 100644
--- a/fs/9p/acl.c
+++ b/fs/9p/acl.c
@@ -213,8 +213,8 @@ int v9fs_acl_mode(struct inode *dir, umode_t *modep,
 }
 
 static int v9fs_xattr_get_acl(const struct xattr_handler *handler,
-			      struct dentry *dentry, const char *name,
-			      void *buffer, size_t size)
+			      struct dentry *dentry, struct inode *inode,
+			      const char *name, void *buffer, size_t size)
 {
 	struct v9fs_session_info *v9ses;
 	struct posix_acl *acl;
@@ -227,7 +227,7 @@ static int v9fs_xattr_get_acl(const struct xattr_handler *handler,
 	if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT)
 		return v9fs_xattr_get(dentry, handler->name, buffer, size);
 
-	acl = v9fs_get_cached_acl(d_inode(dentry), handler->flags);
+	acl = v9fs_get_cached_acl(inode, handler->flags);
 	if (IS_ERR(acl))
 		return PTR_ERR(acl);
 	if (acl == NULL)
diff --git a/fs/9p/xattr.c b/fs/9p/xattr.c
index 9dd9b47a6c1a..18c62bae9591 100644
--- a/fs/9p/xattr.c
+++ b/fs/9p/xattr.c
@@ -138,8 +138,8 @@ ssize_t v9fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size)
 }
 
 static int v9fs_xattr_handler_get(const struct xattr_handler *handler,
-				  struct dentry *dentry, const char *name,
-				  void *buffer, size_t size)
+				  struct dentry *dentry, struct inode *inode,
+				  const char *name, void *buffer, size_t size)
 {
 	const char *full_name = xattr_full_name(handler, name);
 
diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c
index 145d2b89e62d..03224b00ea70 100644
--- a/fs/btrfs/xattr.c
+++ b/fs/btrfs/xattr.c
@@ -369,11 +369,9 @@ err:
 }
 
 static int btrfs_xattr_handler_get(const struct xattr_handler *handler,
-				   struct dentry *dentry, const char *name,
-				   void *buffer, size_t size)
+				   struct dentry *unused, struct inode *inode,
+				   const char *name, void *buffer, size_t size)
 {
-	struct inode *inode = d_inode(dentry);
-
 	name = xattr_full_name(handler, name);
 	return __btrfs_getxattr(inode, name, buffer, size);
 }
diff --git a/fs/ext2/xattr_security.c b/fs/ext2/xattr_security.c
index ba97f243b050..7fd3b867ce65 100644
--- a/fs/ext2/xattr_security.c
+++ b/fs/ext2/xattr_security.c
@@ -9,10 +9,10 @@
 
 static int
 ext2_xattr_security_get(const struct xattr_handler *handler,
-			struct dentry *dentry, const char *name,
-			void *buffer, size_t size)
+			struct dentry *unused, struct inode *inode,
+			const char *name, void *buffer, size_t size)
 {
-	return ext2_xattr_get(d_inode(dentry), EXT2_XATTR_INDEX_SECURITY, name,
+	return ext2_xattr_get(inode, EXT2_XATTR_INDEX_SECURITY, name,
 			      buffer, size);
 }
 
diff --git a/fs/ext2/xattr_trusted.c b/fs/ext2/xattr_trusted.c
index 2c94d1930626..0f85705ff519 100644
--- a/fs/ext2/xattr_trusted.c
+++ b/fs/ext2/xattr_trusted.c
@@ -16,10 +16,10 @@ ext2_xattr_trusted_list(struct dentry *dentry)
 
 static int
 ext2_xattr_trusted_get(const struct xattr_handler *handler,
-		       struct dentry *dentry, const char *name,
-		       void *buffer, size_t size)
+		       struct dentry *unused, struct inode *inode,
+		       const char *name, void *buffer, size_t size)
 {
-	return ext2_xattr_get(d_inode(dentry), EXT2_XATTR_INDEX_TRUSTED, name,
+	return ext2_xattr_get(inode, EXT2_XATTR_INDEX_TRUSTED, name,
 			      buffer, size);
 }
 
diff --git a/fs/ext2/xattr_user.c b/fs/ext2/xattr_user.c
index 72a2a96d677f..1fafd27037cc 100644
--- a/fs/ext2/xattr_user.c
+++ b/fs/ext2/xattr_user.c
@@ -18,12 +18,12 @@ ext2_xattr_user_list(struct dentry *dentry)
 
 static int
 ext2_xattr_user_get(const struct xattr_handler *handler,
-		    struct dentry *dentry, const char *name,
-		    void *buffer, size_t size)
+		    struct dentry *unused, struct inode *inode,
+		    const char *name, void *buffer, size_t size)
 {
-	if (!test_opt(dentry->d_sb, XATTR_USER))
+	if (!test_opt(inode->i_sb, XATTR_USER))
 		return -EOPNOTSUPP;
-	return ext2_xattr_get(d_inode(dentry), EXT2_XATTR_INDEX_USER,
+	return ext2_xattr_get(inode, EXT2_XATTR_INDEX_USER,
 			      name, buffer, size);
 }
 
diff --git a/fs/ext4/xattr_security.c b/fs/ext4/xattr_security.c
index 3e81bdca071a..123a7d010efe 100644
--- a/fs/ext4/xattr_security.c
+++ b/fs/ext4/xattr_security.c
@@ -13,10 +13,10 @@
 
 static int
 ext4_xattr_security_get(const struct xattr_handler *handler,
-			struct dentry *dentry, const char *name,
-			void *buffer, size_t size)
+			struct dentry *unused, struct inode *inode,
+			const char *name, void *buffer, size_t size)
 {
-	return ext4_xattr_get(d_inode(dentry), EXT4_XATTR_INDEX_SECURITY,
+	return ext4_xattr_get(inode, EXT4_XATTR_INDEX_SECURITY,
 			      name, buffer, size);
 }
 
diff --git a/fs/ext4/xattr_trusted.c b/fs/ext4/xattr_trusted.c
index 2a3c6f9b8cb8..60652fa24cbc 100644
--- a/fs/ext4/xattr_trusted.c
+++ b/fs/ext4/xattr_trusted.c
@@ -20,10 +20,10 @@ ext4_xattr_trusted_list(struct dentry *dentry)
 
 static int
 ext4_xattr_trusted_get(const struct xattr_handler *handler,
-		       struct dentry *dentry, const char *name, void *buffer,
-		       size_t size)
+		       struct dentry *unused, struct inode *inode,
+		       const char *name, void *buffer, size_t size)
 {
-	return ext4_xattr_get(d_inode(dentry), EXT4_XATTR_INDEX_TRUSTED,
+	return ext4_xattr_get(inode, EXT4_XATTR_INDEX_TRUSTED,
 			      name, buffer, size);
 }
 
diff --git a/fs/ext4/xattr_user.c b/fs/ext4/xattr_user.c
index d152f431e432..17a446ffecd3 100644
--- a/fs/ext4/xattr_user.c
+++ b/fs/ext4/xattr_user.c
@@ -19,12 +19,12 @@ ext4_xattr_user_list(struct dentry *dentry)
 
 static int
 ext4_xattr_user_get(const struct xattr_handler *handler,
-		    struct dentry *dentry, const char *name,
-		    void *buffer, size_t size)
+		    struct dentry *unused, struct inode *inode,
+		    const char *name, void *buffer, size_t size)
 {
-	if (!test_opt(dentry->d_sb, XATTR_USER))
+	if (!test_opt(inode->i_sb, XATTR_USER))
 		return -EOPNOTSUPP;
-	return ext4_xattr_get(d_inode(dentry), EXT4_XATTR_INDEX_USER,
+	return ext4_xattr_get(inode, EXT4_XATTR_INDEX_USER,
 			      name, buffer, size);
 }
 
diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c
index 06a72dc0191a..17fd2b1a6848 100644
--- a/fs/f2fs/xattr.c
+++ b/fs/f2fs/xattr.c
@@ -26,10 +26,10 @@
 #include "xattr.h"
 
 static int f2fs_xattr_generic_get(const struct xattr_handler *handler,
-		struct dentry *dentry, const char *name, void *buffer,
-		size_t size)
+		struct dentry *unused, struct inode *inode,
+		const char *name, void *buffer, size_t size)
 {
-	struct f2fs_sb_info *sbi = F2FS_SB(dentry->d_sb);
+	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 
 	switch (handler->flags) {
 	case F2FS_XATTR_INDEX_USER:
@@ -45,7 +45,7 @@ static int f2fs_xattr_generic_get(const struct xattr_handler *handler,
 	default:
 		return -EINVAL;
 	}
-	return f2fs_getxattr(d_inode(dentry), handler->flags, name,
+	return f2fs_getxattr(inode, handler->flags, name,
 			     buffer, size, NULL);
 }
 
@@ -86,11 +86,9 @@ static bool f2fs_xattr_trusted_list(struct dentry *dentry)
 }
 
 static int f2fs_xattr_advise_get(const struct xattr_handler *handler,
-		struct dentry *dentry, const char *name, void *buffer,
-		size_t size)
+		struct dentry *unused, struct inode *inode,
+		const char *name, void *buffer, size_t size)
 {
-	struct inode *inode = d_inode(dentry);
-
 	if (buffer)
 		*((char *)buffer) = F2FS_I(inode)->i_advise;
 	return sizeof(char);
diff --git a/fs/gfs2/xattr.c b/fs/gfs2/xattr.c
index e8dfb4740c04..619886ba6e78 100644
--- a/fs/gfs2/xattr.c
+++ b/fs/gfs2/xattr.c
@@ -584,10 +584,10 @@ out:
  * Returns: actual size of data on success, -errno on error
  */
 static int gfs2_xattr_get(const struct xattr_handler *handler,
-			  struct dentry *dentry, const char *name,
-			  void *buffer, size_t size)
+			  struct dentry *unused, struct inode *inode,
+			  const char *name, void *buffer, size_t size)
 {
-	struct gfs2_inode *ip = GFS2_I(d_inode(dentry));
+	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_ea_location el;
 	int type = handler->flags;
 	int error;
diff --git a/fs/hfsplus/xattr.c b/fs/hfsplus/xattr.c
index ab01530b4930..45dc4ae3791a 100644
--- a/fs/hfsplus/xattr.c
+++ b/fs/hfsplus/xattr.c
@@ -579,7 +579,7 @@ failed_getxattr_init:
 	return res;
 }
 
-ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name,
+ssize_t hfsplus_getxattr(struct inode *inode, const char *name,
 			 void *value, size_t size,
 			 const char *prefix, size_t prefixlen)
 {
@@ -594,7 +594,7 @@ ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name,
 	strcpy(xattr_name, prefix);
 	strcpy(xattr_name + prefixlen, name);
 
-	res = __hfsplus_getxattr(d_inode(dentry), xattr_name, value, size);
+	res = __hfsplus_getxattr(inode, xattr_name, value, size);
 	kfree(xattr_name);
 	return res;
 
@@ -844,8 +844,8 @@ end_removexattr:
 }
 
 static int hfsplus_osx_getxattr(const struct xattr_handler *handler,
-				struct dentry *dentry, const char *name,
-				void *buffer, size_t size)
+				struct dentry *unused, struct inode *inode,
+				const char *name, void *buffer, size_t size)
 {
 	/*
 	 * Don't allow retrieving properly prefixed attributes
@@ -860,7 +860,7 @@ static int hfsplus_osx_getxattr(const struct xattr_handler *handler,
 	 * creates), so we pass the name through unmodified (after
 	 * ensuring it doesn't conflict with another namespace).
 	 */
-	return __hfsplus_getxattr(d_inode(dentry), name, buffer, size);
+	return __hfsplus_getxattr(inode, name, buffer, size);
 }
 
 static int hfsplus_osx_setxattr(const struct xattr_handler *handler,
diff --git a/fs/hfsplus/xattr.h b/fs/hfsplus/xattr.h
index f9b0955b3d28..d04ba6f58df2 100644
--- a/fs/hfsplus/xattr.h
+++ b/fs/hfsplus/xattr.h
@@ -28,7 +28,7 @@ int hfsplus_setxattr(struct dentry *dentry, const char *name,
 ssize_t __hfsplus_getxattr(struct inode *inode, const char *name,
 			   void *value, size_t size);
 
-ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name,
+ssize_t hfsplus_getxattr(struct inode *inode, const char *name,
 			 void *value, size_t size,
 			 const char *prefix, size_t prefixlen);
 
diff --git a/fs/hfsplus/xattr_security.c b/fs/hfsplus/xattr_security.c
index 72a68a3a0c99..ae2ca8c2e335 100644
--- a/fs/hfsplus/xattr_security.c
+++ b/fs/hfsplus/xattr_security.c
@@ -14,10 +14,10 @@
 #include "acl.h"
 
 static int hfsplus_security_getxattr(const struct xattr_handler *handler,
-				     struct dentry *dentry, const char *name,
-				     void *buffer, size_t size)
+				     struct dentry *unused, struct inode *inode,
+				     const char *name, void *buffer, size_t size)
 {
-	return hfsplus_getxattr(dentry, name, buffer, size,
+	return hfsplus_getxattr(inode, name, buffer, size,
 				XATTR_SECURITY_PREFIX,
 				XATTR_SECURITY_PREFIX_LEN);
 }
diff --git a/fs/hfsplus/xattr_trusted.c b/fs/hfsplus/xattr_trusted.c
index 95a7704c7abb..eae2947060aa 100644
--- a/fs/hfsplus/xattr_trusted.c
+++ b/fs/hfsplus/xattr_trusted.c
@@ -12,10 +12,10 @@
 #include "xattr.h"
 
 static int hfsplus_trusted_getxattr(const struct xattr_handler *handler,
-				    struct dentry *dentry, const char *name,
-				    void *buffer, size_t size)
+				    struct dentry *unused, struct inode *inode,
+				    const char *name, void *buffer, size_t size)
 {
-	return hfsplus_getxattr(dentry, name, buffer, size,
+	return hfsplus_getxattr(inode, name, buffer, size,
 				XATTR_TRUSTED_PREFIX,
 				XATTR_TRUSTED_PREFIX_LEN);
 }
diff --git a/fs/hfsplus/xattr_user.c b/fs/hfsplus/xattr_user.c
index 6fc269baf959..3c9eec3e4c7b 100644
--- a/fs/hfsplus/xattr_user.c
+++ b/fs/hfsplus/xattr_user.c
@@ -12,11 +12,11 @@
 #include "xattr.h"
 
 static int hfsplus_user_getxattr(const struct xattr_handler *handler,
-				 struct dentry *dentry, const char *name,
-				 void *buffer, size_t size)
+				 struct dentry *unused, struct inode *inode,
+				 const char *name, void *buffer, size_t size)
 {
 
-	return hfsplus_getxattr(dentry, name, buffer, size,
+	return hfsplus_getxattr(inode, name, buffer, size,
 				XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN);
 }
 
diff --git a/fs/jffs2/security.c b/fs/jffs2/security.c
index 7a28facd7175..3ed9a4b49778 100644
--- a/fs/jffs2/security.c
+++ b/fs/jffs2/security.c
@@ -49,10 +49,10 @@ int jffs2_init_security(struct inode *inode, struct inode *dir,
 
 /* ---- XATTR Handler for "security.*" ----------------- */
 static int jffs2_security_getxattr(const struct xattr_handler *handler,
-				   struct dentry *dentry, const char *name,
-				   void *buffer, size_t size)
+				   struct dentry *unused, struct inode *inode,
+				   const char *name, void *buffer, size_t size)
 {
-	return do_jffs2_getxattr(d_inode(dentry), JFFS2_XPREFIX_SECURITY,
+	return do_jffs2_getxattr(inode, JFFS2_XPREFIX_SECURITY,
 				 name, buffer, size);
 }
 
diff --git a/fs/jffs2/xattr_trusted.c b/fs/jffs2/xattr_trusted.c
index b2555ef07a12..4ebecff1d922 100644
--- a/fs/jffs2/xattr_trusted.c
+++ b/fs/jffs2/xattr_trusted.c
@@ -17,10 +17,10 @@
 #include "nodelist.h"
 
 static int jffs2_trusted_getxattr(const struct xattr_handler *handler,
-				  struct dentry *dentry, const char *name,
-				  void *buffer, size_t size)
+				  struct dentry *unused, struct inode *inode,
+				  const char *name, void *buffer, size_t size)
 {
-	return do_jffs2_getxattr(d_inode(dentry), JFFS2_XPREFIX_TRUSTED,
+	return do_jffs2_getxattr(inode, JFFS2_XPREFIX_TRUSTED,
 				 name, buffer, size);
 }
 
diff --git a/fs/jffs2/xattr_user.c b/fs/jffs2/xattr_user.c
index 539bd630b5e4..bce249e1b277 100644
--- a/fs/jffs2/xattr_user.c
+++ b/fs/jffs2/xattr_user.c
@@ -17,10 +17,10 @@
 #include "nodelist.h"
 
 static int jffs2_user_getxattr(const struct xattr_handler *handler,
-			       struct dentry *dentry, const char *name,
-			       void *buffer, size_t size)
+			       struct dentry *unused, struct inode *inode,
+			       const char *name, void *buffer, size_t size)
 {
-	return do_jffs2_getxattr(d_inode(dentry), JFFS2_XPREFIX_USER,
+	return do_jffs2_getxattr(inode, JFFS2_XPREFIX_USER,
 				 name, buffer, size);
 }
 
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 327b8c34d360..7a7ac1dafa02 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -6263,10 +6263,10 @@ static int nfs4_xattr_set_nfs4_acl(const struct xattr_handler *handler,
 }
 
 static int nfs4_xattr_get_nfs4_acl(const struct xattr_handler *handler,
-				   struct dentry *dentry, const char *key,
-				   void *buf, size_t buflen)
+				   struct dentry *unused, struct inode *inode,
+				   const char *key, void *buf, size_t buflen)
 {
-	return nfs4_proc_get_acl(d_inode(dentry), buf, buflen);
+	return nfs4_proc_get_acl(inode, buf, buflen);
 }
 
 static bool nfs4_xattr_list_nfs4_acl(struct dentry *dentry)
@@ -6288,11 +6288,11 @@ static int nfs4_xattr_set_nfs4_label(const struct xattr_handler *handler,
 }
 
 static int nfs4_xattr_get_nfs4_label(const struct xattr_handler *handler,
-				     struct dentry *dentry, const char *key,
-				     void *buf, size_t buflen)
+				     struct dentry *unused, struct inode *inode,
+				     const char *key, void *buf, size_t buflen)
 {
 	if (security_ismaclabel(key))
-		return nfs4_get_security_label(d_inode(dentry), buf, buflen);
+		return nfs4_get_security_label(inode, buf, buflen);
 	return -EOPNOTSUPP;
 }
 
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index 7d3d979f57d9..72eef4cfe8a9 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -7250,10 +7250,10 @@ leave:
  * 'security' attributes support
  */
 static int ocfs2_xattr_security_get(const struct xattr_handler *handler,
-				    struct dentry *dentry, const char *name,
-				    void *buffer, size_t size)
+				    struct dentry *unused, struct inode *inode,
+				    const char *name, void *buffer, size_t size)
 {
-	return ocfs2_xattr_get(d_inode(dentry), OCFS2_XATTR_INDEX_SECURITY,
+	return ocfs2_xattr_get(inode, OCFS2_XATTR_INDEX_SECURITY,
 			       name, buffer, size);
 }
 
@@ -7321,10 +7321,10 @@ const struct xattr_handler ocfs2_xattr_security_handler = {
  * 'trusted' attributes support
  */
 static int ocfs2_xattr_trusted_get(const struct xattr_handler *handler,
-				   struct dentry *dentry, const char *name,
-				   void *buffer, size_t size)
+				   struct dentry *unused, struct inode *inode,
+				   const char *name, void *buffer, size_t size)
 {
-	return ocfs2_xattr_get(d_inode(dentry), OCFS2_XATTR_INDEX_TRUSTED,
+	return ocfs2_xattr_get(inode, OCFS2_XATTR_INDEX_TRUSTED,
 			       name, buffer, size);
 }
 
@@ -7346,14 +7346,14 @@ const struct xattr_handler ocfs2_xattr_trusted_handler = {
  * 'user' attributes support
  */
 static int ocfs2_xattr_user_get(const struct xattr_handler *handler,
-				struct dentry *dentry, const char *name,
-				void *buffer, size_t size)
+				struct dentry *unusde, struct inode *inode,
+				const char *name, void *buffer, size_t size)
 {
-	struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 
 	if (osb->s_mount_opt & OCFS2_MOUNT_NOUSERXATTR)
 		return -EOPNOTSUPP;
-	return ocfs2_xattr_get(d_inode(dentry), OCFS2_XATTR_INDEX_USER, name,
+	return ocfs2_xattr_get(inode, OCFS2_XATTR_INDEX_USER, name,
 			       buffer, size);
 }
 
diff --git a/fs/orangefs/xattr.c b/fs/orangefs/xattr.c
index ef5da7538cd5..6a4c0f7ce5c1 100644
--- a/fs/orangefs/xattr.c
+++ b/fs/orangefs/xattr.c
@@ -478,12 +478,13 @@ static int orangefs_xattr_set_default(const struct xattr_handler *handler,
 }
 
 static int orangefs_xattr_get_default(const struct xattr_handler *handler,
-				      struct dentry *dentry,
+				      struct dentry *unused,
+				      struct inode *inode,
 				      const char *name,
 				      void *buffer,
 				      size_t size)
 {
-	return orangefs_inode_getxattr(dentry->d_inode,
+	return orangefs_inode_getxattr(inode,
 				    ORANGEFS_XATTR_NAME_DEFAULT_PREFIX,
 				    name,
 				    buffer,
@@ -507,12 +508,13 @@ static int orangefs_xattr_set_trusted(const struct xattr_handler *handler,
 }
 
 static int orangefs_xattr_get_trusted(const struct xattr_handler *handler,
-				      struct dentry *dentry,
+				      struct dentry *unused,
+				      struct inode *inode,
 				      const char *name,
 				      void *buffer,
 				      size_t size)
 {
-	return orangefs_inode_getxattr(dentry->d_inode,
+	return orangefs_inode_getxattr(inode,
 				    ORANGEFS_XATTR_NAME_TRUSTED_PREFIX,
 				    name,
 				    buffer,
diff --git a/fs/posix_acl.c b/fs/posix_acl.c
index db1fb0f5d9ff..2c60f17e7d92 100644
--- a/fs/posix_acl.c
+++ b/fs/posix_acl.c
@@ -797,18 +797,18 @@ EXPORT_SYMBOL (posix_acl_to_xattr);
 
 static int
 posix_acl_xattr_get(const struct xattr_handler *handler,
-		    struct dentry *dentry, const char *name,
-		    void *value, size_t size)
+		    struct dentry *unused, struct inode *inode,
+		    const char *name, void *value, size_t size)
 {
 	struct posix_acl *acl;
 	int error;
 
-	if (!IS_POSIXACL(d_backing_inode(dentry)))
+	if (!IS_POSIXACL(inode))
 		return -EOPNOTSUPP;
-	if (d_is_symlink(dentry))
+	if (S_ISLNK(inode->i_mode))
 		return -EOPNOTSUPP;
 
-	acl = get_acl(d_backing_inode(dentry), handler->flags);
+	acl = get_acl(inode, handler->flags);
 	if (IS_ERR(acl))
 		return PTR_ERR(acl);
 	if (acl == NULL)
diff --git a/fs/reiserfs/xattr_security.c b/fs/reiserfs/xattr_security.c
index ac7e104ada6b..86aeb9dd805a 100644
--- a/fs/reiserfs/xattr_security.c
+++ b/fs/reiserfs/xattr_security.c
@@ -9,14 +9,13 @@
 #include <linux/uaccess.h>
 
 static int
-security_get(const struct xattr_handler *handler, struct dentry *dentry,
-	     const char *name, void *buffer, size_t size)
+security_get(const struct xattr_handler *handler, struct dentry *unused,
+	     struct inode *inode, const char *name, void *buffer, size_t size)
 {
-	if (IS_PRIVATE(d_inode(dentry)))
+	if (IS_PRIVATE(inode))
 		return -EPERM;
 
-	return reiserfs_xattr_get(d_inode(dentry),
-				  xattr_full_name(handler, name),
+	return reiserfs_xattr_get(inode, xattr_full_name(handler, name),
 				  buffer, size);
 }
 
diff --git a/fs/reiserfs/xattr_trusted.c b/fs/reiserfs/xattr_trusted.c
index cc248a581b60..31837f031f59 100644
--- a/fs/reiserfs/xattr_trusted.c
+++ b/fs/reiserfs/xattr_trusted.c
@@ -8,14 +8,13 @@
 #include <linux/uaccess.h>
 
 static int
-trusted_get(const struct xattr_handler *handler, struct dentry *dentry,
-	    const char *name, void *buffer, size_t size)
+trusted_get(const struct xattr_handler *handler, struct dentry *unused,
+	    struct inode *inode, const char *name, void *buffer, size_t size)
 {
-	if (!capable(CAP_SYS_ADMIN) || IS_PRIVATE(d_inode(dentry)))
+	if (!capable(CAP_SYS_ADMIN) || IS_PRIVATE(inode))
 		return -EPERM;
 
-	return reiserfs_xattr_get(d_inode(dentry),
-				  xattr_full_name(handler, name),
+	return reiserfs_xattr_get(inode, xattr_full_name(handler, name),
 				  buffer, size);
 }
 
diff --git a/fs/reiserfs/xattr_user.c b/fs/reiserfs/xattr_user.c
index caad583086af..f7c39731684b 100644
--- a/fs/reiserfs/xattr_user.c
+++ b/fs/reiserfs/xattr_user.c
@@ -7,13 +7,12 @@
 #include <linux/uaccess.h>
 
 static int
-user_get(const struct xattr_handler *handler, struct dentry *dentry,
-	 const char *name, void *buffer, size_t size)
+user_get(const struct xattr_handler *handler, struct dentry *unused,
+	 struct inode *inode, const char *name, void *buffer, size_t size)
 {
-	if (!reiserfs_xattrs_user(dentry->d_sb))
+	if (!reiserfs_xattrs_user(inode->i_sb))
 		return -EOPNOTSUPP;
-	return reiserfs_xattr_get(d_inode(dentry),
-				  xattr_full_name(handler, name),
+	return reiserfs_xattr_get(inode, xattr_full_name(handler, name),
 				  buffer, size);
 }
 
diff --git a/fs/squashfs/xattr.c b/fs/squashfs/xattr.c
index 1e9de96288d8..1548b3784548 100644
--- a/fs/squashfs/xattr.c
+++ b/fs/squashfs/xattr.c
@@ -214,10 +214,12 @@ failed:
 
 
 static int squashfs_xattr_handler_get(const struct xattr_handler *handler,
-				      struct dentry *d, const char *name,
+				      struct dentry *unused,
+				      struct inode *inode,
+				      const char *name,
 				      void *buffer, size_t size)
 {
-	return squashfs_xattr_get(d_inode(d), handler->flags, name,
+	return squashfs_xattr_get(inode, handler->flags, name,
 		buffer, size);
 }
 
diff --git a/fs/xattr.c b/fs/xattr.c
index 4861322e28e8..461ba45b7da9 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c
@@ -698,7 +698,8 @@ generic_getxattr(struct dentry *dentry, const char *name, void *buffer, size_t s
 	handler = xattr_resolve_name(dentry->d_sb->s_xattr, &name);
 	if (IS_ERR(handler))
 		return PTR_ERR(handler);
-	return handler->get(handler, dentry, name, buffer, size);
+	return handler->get(handler, dentry, d_inode(dentry),
+			    name, buffer, size);
 }
 
 /*
diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c
index 110f1d7d86b0..d111f691f313 100644
--- a/fs/xfs/xfs_xattr.c
+++ b/fs/xfs/xfs_xattr.c
@@ -32,11 +32,11 @@
 
 
 static int
-xfs_xattr_get(const struct xattr_handler *handler, struct dentry *dentry,
-		const char *name, void *value, size_t size)
+xfs_xattr_get(const struct xattr_handler *handler, struct dentry *unused,
+		struct inode *inode, const char *name, void *value, size_t size)
 {
 	int xflags = handler->flags;
-	struct xfs_inode *ip = XFS_I(d_inode(dentry));
+	struct xfs_inode *ip = XFS_I(inode);
 	int error, asize = size;
 
 	/* Convert Linux syscall to XFS internal ATTR flags */
diff --git a/include/linux/xattr.h b/include/linux/xattr.h
index 4457541de3c9..c11c022298b9 100644
--- a/include/linux/xattr.h
+++ b/include/linux/xattr.h
@@ -30,7 +30,8 @@ struct xattr_handler {
 	int flags;      /* fs private flags */
 	bool (*list)(struct dentry *dentry);
 	int (*get)(const struct xattr_handler *, struct dentry *dentry,
-		   const char *name, void *buffer, size_t size);
+		   struct inode *inode, const char *name, void *buffer,
+		   size_t size);
 	int (*set)(const struct xattr_handler *, struct dentry *dentry,
 		   const char *name, const void *buffer, size_t size,
 		   int flags);
diff --git a/mm/shmem.c b/mm/shmem.c
index 9428c51ab2d6..00d5d025eece 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2646,10 +2646,10 @@ static int shmem_initxattrs(struct inode *inode,
 }
 
 static int shmem_xattr_handler_get(const struct xattr_handler *handler,
-				   struct dentry *dentry, const char *name,
-				   void *buffer, size_t size)
+				   struct dentry *unused, struct inode *inode,
+				   const char *name, void *buffer, size_t size)
 {
-	struct shmem_inode_info *info = SHMEM_I(d_inode(dentry));
+	struct shmem_inode_info *info = SHMEM_I(inode);
 
 	name = xattr_full_name(handler, name);
 	return simple_xattr_get(&info->xattrs, name, buffer, size);
-- 
cgit v1.2.3


From ce23e640133484eebc20ca7b7668388213e11327 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 11 Apr 2016 00:48:00 -0400
Subject: ->getxattr(): pass dentry and inode as separate arguments

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 Documentation/filesystems/porting                  |  6 ++++++
 .../staging/lustre/lustre/llite/llite_internal.h   |  4 ++--
 drivers/staging/lustre/lustre/llite/xattr.c        |  6 ++----
 fs/bad_inode.c                                     |  4 ++--
 fs/ceph/super.h                                    |  2 +-
 fs/ceph/xattr.c                                    |  8 ++++----
 fs/cifs/cifsfs.h                                   |  2 +-
 fs/cifs/xattr.c                                    |  6 +++---
 fs/ecryptfs/crypto.c                               |  5 ++++-
 fs/ecryptfs/ecryptfs_kernel.h                      |  4 ++--
 fs/ecryptfs/inode.c                                | 23 +++++++++++-----------
 fs/ecryptfs/mmap.c                                 |  3 ++-
 fs/fuse/dir.c                                      |  5 ++---
 fs/gfs2/inode.c                                    |  9 ++++-----
 fs/hfs/attr.c                                      |  5 ++---
 fs/hfs/hfs_fs.h                                    |  4 ++--
 fs/jfs/jfs_xattr.h                                 |  2 +-
 fs/jfs/xattr.c                                     |  8 ++++----
 fs/kernfs/inode.c                                  |  6 +++---
 fs/kernfs/kernfs-internal.h                        |  4 ++--
 fs/libfs.c                                         |  4 ++--
 fs/overlayfs/inode.c                               |  4 ++--
 fs/overlayfs/overlayfs.h                           |  4 ++--
 fs/overlayfs/super.c                               |  2 +-
 fs/ubifs/ubifs.h                                   |  4 ++--
 fs/ubifs/xattr.c                                   |  6 +++---
 fs/xattr.c                                         | 11 ++++++-----
 include/linux/fs.h                                 |  3 ++-
 include/linux/xattr.h                              |  2 +-
 net/socket.c                                       |  2 +-
 security/commoncap.c                               |  6 +++---
 security/integrity/evm/evm_main.c                  |  2 +-
 security/selinux/hooks.c                           |  9 +++++----
 security/smack/smack_lsm.c                         |  4 ++--
 34 files changed, 94 insertions(+), 85 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/filesystems/porting b/Documentation/filesystems/porting
index f1b87d8aa2da..57bb3754a027 100644
--- a/Documentation/filesystems/porting
+++ b/Documentation/filesystems/porting
@@ -525,3 +525,9 @@ in your dentry operations instead.
 	set_delayed_call() where it used to set *cookie.
 	->put_link() is gone - just give the destructor to set_delayed_call()
 	in ->get_link().
+--
+[mandatory]
+	->getxattr() and xattr_handler.get() get dentry and inode passed separately.
+	dentry might be yet to be attached to inode, so do _not_ use its ->d_inode
+	in the instances.  Rationale: !@#!@# security_d_instantiate() needs to be
+	called before we attach dentry to inode.
diff --git a/drivers/staging/lustre/lustre/llite/llite_internal.h b/drivers/staging/lustre/lustre/llite/llite_internal.h
index 3e1572cb457b..d28efd27af57 100644
--- a/drivers/staging/lustre/lustre/llite/llite_internal.h
+++ b/drivers/staging/lustre/lustre/llite/llite_internal.h
@@ -1042,8 +1042,8 @@ static inline __u64 ll_file_maxbytes(struct inode *inode)
 /* llite/xattr.c */
 int ll_setxattr(struct dentry *dentry, const char *name,
 		const void *value, size_t size, int flags);
-ssize_t ll_getxattr(struct dentry *dentry, const char *name,
-		    void *buffer, size_t size);
+ssize_t ll_getxattr(struct dentry *dentry, struct inode *inode,
+		    const char *name, void *buffer, size_t size);
 ssize_t ll_listxattr(struct dentry *dentry, char *buffer, size_t size);
 int ll_removexattr(struct dentry *dentry, const char *name);
 
diff --git a/drivers/staging/lustre/lustre/llite/xattr.c b/drivers/staging/lustre/lustre/llite/xattr.c
index b68dcc921ca2..c671f221c28c 100644
--- a/drivers/staging/lustre/lustre/llite/xattr.c
+++ b/drivers/staging/lustre/lustre/llite/xattr.c
@@ -451,11 +451,9 @@ out:
 	return rc;
 }
 
-ssize_t ll_getxattr(struct dentry *dentry, const char *name,
-		    void *buffer, size_t size)
+ssize_t ll_getxattr(struct dentry *dentry, struct inode *inode,
+		    const char *name, void *buffer, size_t size)
 {
-	struct inode *inode = d_inode(dentry);
-
 	LASSERT(inode);
 	LASSERT(name);
 
diff --git a/fs/bad_inode.c b/fs/bad_inode.c
index 103f5d7c3083..72e35b721608 100644
--- a/fs/bad_inode.c
+++ b/fs/bad_inode.c
@@ -106,8 +106,8 @@ static int bad_inode_setxattr(struct dentry *dentry, const char *name,
 	return -EIO;
 }
 
-static ssize_t bad_inode_getxattr(struct dentry *dentry, const char *name,
-			void *buffer, size_t size)
+static ssize_t bad_inode_getxattr(struct dentry *dentry, struct inode *inode,
+			const char *name, void *buffer, size_t size)
 {
 	return -EIO;
 }
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index e705c4d612d7..beb893bb234f 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -795,7 +795,7 @@ extern int ceph_setxattr(struct dentry *, const char *, const void *,
 int __ceph_setxattr(struct dentry *, const char *, const void *, size_t, int);
 ssize_t __ceph_getxattr(struct inode *, const char *, void *, size_t);
 int __ceph_removexattr(struct dentry *, const char *);
-extern ssize_t ceph_getxattr(struct dentry *, const char *, void *, size_t);
+extern ssize_t ceph_getxattr(struct dentry *, struct inode *, const char *, void *, size_t);
 extern ssize_t ceph_listxattr(struct dentry *, char *, size_t);
 extern int ceph_removexattr(struct dentry *, const char *);
 extern void __ceph_build_xattrs_blob(struct ceph_inode_info *ci);
diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c
index 9410abdef3ce..c6e917d360f7 100644
--- a/fs/ceph/xattr.c
+++ b/fs/ceph/xattr.c
@@ -804,13 +804,13 @@ out:
 	return err;
 }
 
-ssize_t ceph_getxattr(struct dentry *dentry, const char *name, void *value,
-		      size_t size)
+ssize_t ceph_getxattr(struct dentry *dentry, struct inode *inode,
+		      const char *name, void *value, size_t size)
 {
 	if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
-		return generic_getxattr(dentry, name, value, size);
+		return generic_getxattr(dentry, inode, name, value, size);
 
-	return __ceph_getxattr(d_inode(dentry), name, value, size);
+	return __ceph_getxattr(inode, name, value, size);
 }
 
 ssize_t ceph_listxattr(struct dentry *dentry, char *names, size_t size)
diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
index 83aac8ba50b0..c89ecd7a5c39 100644
--- a/fs/cifs/cifsfs.h
+++ b/fs/cifs/cifsfs.h
@@ -123,7 +123,7 @@ extern int cifs_symlink(struct inode *inode, struct dentry *direntry,
 extern int	cifs_removexattr(struct dentry *, const char *);
 extern int	cifs_setxattr(struct dentry *, const char *, const void *,
 			size_t, int);
-extern ssize_t	cifs_getxattr(struct dentry *, const char *, void *, size_t);
+extern ssize_t	cifs_getxattr(struct dentry *, struct inode *, const char *, void *, size_t);
 extern ssize_t	cifs_listxattr(struct dentry *, char *, size_t);
 extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg);
 #ifdef CONFIG_CIFS_NFSD_EXPORT
diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c
index 159547c8a40b..5d57c85703a9 100644
--- a/fs/cifs/xattr.c
+++ b/fs/cifs/xattr.c
@@ -213,8 +213,8 @@ set_ea_exit:
 	return rc;
 }
 
-ssize_t cifs_getxattr(struct dentry *direntry, const char *ea_name,
-	void *ea_value, size_t buf_size)
+ssize_t cifs_getxattr(struct dentry *direntry, struct inode *inode,
+	const char *ea_name, void *ea_value, size_t buf_size)
 {
 	ssize_t rc = -EOPNOTSUPP;
 #ifdef CONFIG_CIFS_XATTR
@@ -296,7 +296,7 @@ ssize_t cifs_getxattr(struct dentry *direntry, const char *ea_name,
 				goto get_ea_exit; /* rc already EOPNOTSUPP */
 
 			pacl = pTcon->ses->server->ops->get_acl(cifs_sb,
-					d_inode(direntry), full_path, &acllen);
+					inode, full_path, &acllen);
 			if (IS_ERR(pacl)) {
 				rc = PTR_ERR(pacl);
 				cifs_dbg(VFS, "%s: error %zd getting sec desc\n",
diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c
index 64026e53722a..543a146ee019 100644
--- a/fs/ecryptfs/crypto.c
+++ b/fs/ecryptfs/crypto.c
@@ -1369,7 +1369,9 @@ int ecryptfs_read_xattr_region(char *page_virt, struct inode *ecryptfs_inode)
 	ssize_t size;
 	int rc = 0;
 
-	size = ecryptfs_getxattr_lower(lower_dentry, ECRYPTFS_XATTR_NAME,
+	size = ecryptfs_getxattr_lower(lower_dentry,
+				       ecryptfs_inode_to_lower(ecryptfs_inode),
+				       ECRYPTFS_XATTR_NAME,
 				       page_virt, ECRYPTFS_DEFAULT_EXTENT_SIZE);
 	if (size < 0) {
 		if (unlikely(ecryptfs_verbosity > 0))
@@ -1391,6 +1393,7 @@ int ecryptfs_read_and_validate_xattr_region(struct dentry *dentry,
 	int rc;
 
 	rc = ecryptfs_getxattr_lower(ecryptfs_dentry_to_lower(dentry),
+				     ecryptfs_inode_to_lower(inode),
 				     ECRYPTFS_XATTR_NAME, file_size,
 				     ECRYPTFS_SIZE_AND_MARKER_BYTES);
 	if (rc < ECRYPTFS_SIZE_AND_MARKER_BYTES)
diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h
index d123fbaa28e0..6ff907f73331 100644
--- a/fs/ecryptfs/ecryptfs_kernel.h
+++ b/fs/ecryptfs/ecryptfs_kernel.h
@@ -607,8 +607,8 @@ ecryptfs_parse_packet_set(struct ecryptfs_crypt_stat *crypt_stat,
 			  unsigned char *src, struct dentry *ecryptfs_dentry);
 int ecryptfs_truncate(struct dentry *dentry, loff_t new_length);
 ssize_t
-ecryptfs_getxattr_lower(struct dentry *lower_dentry, const char *name,
-			void *value, size_t size);
+ecryptfs_getxattr_lower(struct dentry *lower_dentry, struct inode *lower_inode,
+			const char *name, void *value, size_t size);
 int
 ecryptfs_setxattr(struct dentry *dentry, const char *name, const void *value,
 		  size_t size, int flags);
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 121114e9a464..1ac631cd9d84 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -1033,29 +1033,30 @@ out:
 }
 
 ssize_t
-ecryptfs_getxattr_lower(struct dentry *lower_dentry, const char *name,
-			void *value, size_t size)
+ecryptfs_getxattr_lower(struct dentry *lower_dentry, struct inode *lower_inode,
+			const char *name, void *value, size_t size)
 {
 	int rc = 0;
 
-	if (!d_inode(lower_dentry)->i_op->getxattr) {
+	if (!lower_inode->i_op->getxattr) {
 		rc = -EOPNOTSUPP;
 		goto out;
 	}
-	inode_lock(d_inode(lower_dentry));
-	rc = d_inode(lower_dentry)->i_op->getxattr(lower_dentry, name, value,
-						   size);
-	inode_unlock(d_inode(lower_dentry));
+	inode_lock(lower_inode);
+	rc = lower_inode->i_op->getxattr(lower_dentry, lower_inode,
+					 name, value, size);
+	inode_unlock(lower_inode);
 out:
 	return rc;
 }
 
 static ssize_t
-ecryptfs_getxattr(struct dentry *dentry, const char *name, void *value,
-		  size_t size)
+ecryptfs_getxattr(struct dentry *dentry, struct inode *inode,
+		  const char *name, void *value, size_t size)
 {
-	return ecryptfs_getxattr_lower(ecryptfs_dentry_to_lower(dentry), name,
-				       value, size);
+	return ecryptfs_getxattr_lower(ecryptfs_dentry_to_lower(dentry),
+				       ecryptfs_inode_to_lower(inode),
+				       name, value, size);
 }
 
 static ssize_t
diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c
index 1f5865263b3e..39e4381d3a65 100644
--- a/fs/ecryptfs/mmap.c
+++ b/fs/ecryptfs/mmap.c
@@ -436,7 +436,8 @@ static int ecryptfs_write_inode_size_to_xattr(struct inode *ecryptfs_inode)
 		goto out;
 	}
 	inode_lock(lower_inode);
-	size = lower_inode->i_op->getxattr(lower_dentry, ECRYPTFS_XATTR_NAME,
+	size = lower_inode->i_op->getxattr(lower_dentry, lower_inode,
+					   ECRYPTFS_XATTR_NAME,
 					   xattr_virt, PAGE_CACHE_SIZE);
 	if (size < 0)
 		size = 8;
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 4b855b65d457..b618527c05c6 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -1759,10 +1759,9 @@ static int fuse_setxattr(struct dentry *entry, const char *name,
 	return err;
 }
 
-static ssize_t fuse_getxattr(struct dentry *entry, const char *name,
-			     void *value, size_t size)
+static ssize_t fuse_getxattr(struct dentry *entry, struct inode *inode,
+			     const char *name, void *value, size_t size)
 {
-	struct inode *inode = d_inode(entry);
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	FUSE_ARGS(args);
 	struct fuse_getxattr_in inarg;
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index bb30f9a72c65..45f516cada78 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -1968,22 +1968,21 @@ static int gfs2_setxattr(struct dentry *dentry, const char *name,
 	return ret;
 }
 
-static ssize_t gfs2_getxattr(struct dentry *dentry, const char *name,
-			     void *data, size_t size)
+static ssize_t gfs2_getxattr(struct dentry *dentry, struct inode *inode,
+			     const char *name, void *data, size_t size)
 {
-	struct inode *inode = d_inode(dentry);
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_holder gh;
 	int ret;
 
 	/* For selinux during lookup */
 	if (gfs2_glock_is_locked_by_me(ip->i_gl))
-		return generic_getxattr(dentry, name, data, size);
+		return generic_getxattr(dentry, inode, name, data, size);
 
 	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh);
 	ret = gfs2_glock_nq(&gh);
 	if (ret == 0) {
-		ret = generic_getxattr(dentry, name, data, size);
+		ret = generic_getxattr(dentry, inode, name, data, size);
 		gfs2_glock_dq(&gh);
 	}
 	gfs2_holder_uninit(&gh);
diff --git a/fs/hfs/attr.c b/fs/hfs/attr.c
index 8d931b157bbe..064f92f17efc 100644
--- a/fs/hfs/attr.c
+++ b/fs/hfs/attr.c
@@ -56,10 +56,9 @@ out:
 	return res;
 }
 
-ssize_t hfs_getxattr(struct dentry *dentry, const char *name,
-			 void *value, size_t size)
+ssize_t hfs_getxattr(struct dentry *unused, struct inode *inode,
+		     const char *name, void *value, size_t size)
 {
-	struct inode *inode = d_inode(dentry);
 	struct hfs_find_data fd;
 	hfs_cat_rec rec;
 	struct hfs_cat_file *file;
diff --git a/fs/hfs/hfs_fs.h b/fs/hfs/hfs_fs.h
index 1f1c7dcbcc2f..79daa097929a 100644
--- a/fs/hfs/hfs_fs.h
+++ b/fs/hfs/hfs_fs.h
@@ -213,8 +213,8 @@ extern void hfs_delete_inode(struct inode *);
 /* attr.c */
 extern int hfs_setxattr(struct dentry *dentry, const char *name,
 			const void *value, size_t size, int flags);
-extern ssize_t hfs_getxattr(struct dentry *dentry, const char *name,
-			    void *value, size_t size);
+extern ssize_t hfs_getxattr(struct dentry *dentry, struct inode *inode,
+			    const char *name, void *value, size_t size);
 extern ssize_t hfs_listxattr(struct dentry *dentry, char *buffer, size_t size);
 
 /* mdb.c */
diff --git a/fs/jfs/jfs_xattr.h b/fs/jfs/jfs_xattr.h
index e8d717dabca3..e69e14f3777b 100644
--- a/fs/jfs/jfs_xattr.h
+++ b/fs/jfs/jfs_xattr.h
@@ -57,7 +57,7 @@ extern int __jfs_setxattr(tid_t, struct inode *, const char *, const void *,
 extern int jfs_setxattr(struct dentry *, const char *, const void *, size_t,
 			int);
 extern ssize_t __jfs_getxattr(struct inode *, const char *, void *, size_t);
-extern ssize_t jfs_getxattr(struct dentry *, const char *, void *, size_t);
+extern ssize_t jfs_getxattr(struct dentry *, struct inode *, const char *, void *, size_t);
 extern ssize_t jfs_listxattr(struct dentry *, char *, size_t);
 extern int jfs_removexattr(struct dentry *, const char *);
 
diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c
index 48b15a6e5558..5becc6a3ff8c 100644
--- a/fs/jfs/xattr.c
+++ b/fs/jfs/xattr.c
@@ -933,8 +933,8 @@ ssize_t __jfs_getxattr(struct inode *inode, const char *name, void *data,
 	return size;
 }
 
-ssize_t jfs_getxattr(struct dentry *dentry, const char *name, void *data,
-		     size_t buf_size)
+ssize_t jfs_getxattr(struct dentry *dentry, struct inode *inode,
+		     const char *name, void *data, size_t buf_size)
 {
 	int err;
 
@@ -944,7 +944,7 @@ ssize_t jfs_getxattr(struct dentry *dentry, const char *name, void *data,
 	 * for it via sb->s_xattr.
 	 */
 	if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
-		return generic_getxattr(dentry, name, data, buf_size);
+		return generic_getxattr(dentry, inode, name, data, buf_size);
 
 	if (strncmp(name, XATTR_OS2_PREFIX, XATTR_OS2_PREFIX_LEN) == 0) {
 		/*
@@ -959,7 +959,7 @@ ssize_t jfs_getxattr(struct dentry *dentry, const char *name, void *data,
 			return -EOPNOTSUPP;
 	}
 
-	err = __jfs_getxattr(d_inode(dentry), name, data, buf_size);
+	err = __jfs_getxattr(inode, name, data, buf_size);
 
 	return err;
 }
diff --git a/fs/kernfs/inode.c b/fs/kernfs/inode.c
index 16405ae88d2d..b5247226732b 100644
--- a/fs/kernfs/inode.c
+++ b/fs/kernfs/inode.c
@@ -208,10 +208,10 @@ int kernfs_iop_removexattr(struct dentry *dentry, const char *name)
 	return simple_xattr_set(&attrs->xattrs, name, NULL, 0, XATTR_REPLACE);
 }
 
-ssize_t kernfs_iop_getxattr(struct dentry *dentry, const char *name, void *buf,
-			    size_t size)
+ssize_t kernfs_iop_getxattr(struct dentry *unused, struct inode *inode,
+			    const char *name, void *buf, size_t size)
 {
-	struct kernfs_node *kn = dentry->d_fsdata;
+	struct kernfs_node *kn = inode->i_private;
 	struct kernfs_iattrs *attrs;
 
 	attrs = kernfs_iattrs(kn);
diff --git a/fs/kernfs/kernfs-internal.h b/fs/kernfs/kernfs-internal.h
index 6762bfbd8207..45c9192c276e 100644
--- a/fs/kernfs/kernfs-internal.h
+++ b/fs/kernfs/kernfs-internal.h
@@ -84,8 +84,8 @@ int kernfs_iop_getattr(struct vfsmount *mnt, struct dentry *dentry,
 int kernfs_iop_setxattr(struct dentry *dentry, const char *name, const void *value,
 			size_t size, int flags);
 int kernfs_iop_removexattr(struct dentry *dentry, const char *name);
-ssize_t kernfs_iop_getxattr(struct dentry *dentry, const char *name, void *buf,
-			    size_t size);
+ssize_t kernfs_iop_getxattr(struct dentry *dentry, struct inode *inode,
+			    const char *name, void *buf, size_t size);
 ssize_t kernfs_iop_listxattr(struct dentry *dentry, char *buf, size_t size);
 
 /*
diff --git a/fs/libfs.c b/fs/libfs.c
index 0ca80b2af420..03332f4bdedf 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -1127,8 +1127,8 @@ static int empty_dir_setxattr(struct dentry *dentry, const char *name,
 	return -EOPNOTSUPP;
 }
 
-static ssize_t empty_dir_getxattr(struct dentry *dentry, const char *name,
-				  void *value, size_t size)
+static ssize_t empty_dir_getxattr(struct dentry *dentry, struct inode *inode,
+				  const char *name, void *value, size_t size)
 {
 	return -EOPNOTSUPP;
 }
diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c
index a4ff5d0d7db9..c7b31a03dc9c 100644
--- a/fs/overlayfs/inode.c
+++ b/fs/overlayfs/inode.c
@@ -246,8 +246,8 @@ static bool ovl_need_xattr_filter(struct dentry *dentry,
 		return false;
 }
 
-ssize_t ovl_getxattr(struct dentry *dentry, const char *name,
-		     void *value, size_t size)
+ssize_t ovl_getxattr(struct dentry *dentry, struct inode *inode,
+		     const char *name, void *value, size_t size)
 {
 	struct path realpath;
 	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index 6a7090f4a441..99ec4b035237 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -173,8 +173,8 @@ int ovl_setattr(struct dentry *dentry, struct iattr *attr);
 int ovl_permission(struct inode *inode, int mask);
 int ovl_setxattr(struct dentry *dentry, const char *name,
 		 const void *value, size_t size, int flags);
-ssize_t ovl_getxattr(struct dentry *dentry, const char *name,
-		     void *value, size_t size);
+ssize_t ovl_getxattr(struct dentry *dentry, struct inode *inode,
+		     const char *name, void *value, size_t size);
 ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size);
 int ovl_removexattr(struct dentry *dentry, const char *name);
 struct inode *ovl_d_select_inode(struct dentry *dentry, unsigned file_flags);
diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index ef64984c9bbc..14cab381cece 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -274,7 +274,7 @@ static bool ovl_is_opaquedir(struct dentry *dentry)
 	if (!S_ISDIR(inode->i_mode) || !inode->i_op->getxattr)
 		return false;
 
-	res = inode->i_op->getxattr(dentry, OVL_XATTR_OPAQUE, &val, 1);
+	res = inode->i_op->getxattr(dentry, inode, OVL_XATTR_OPAQUE, &val, 1);
 	if (res == 1 && val == 'y')
 		return true;
 
diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h
index c2a57e193a81..536fb495f2f1 100644
--- a/fs/ubifs/ubifs.h
+++ b/fs/ubifs/ubifs.h
@@ -1734,8 +1734,8 @@ int ubifs_getattr(struct vfsmount *mnt, struct dentry *dentry,
 /* xattr.c */
 int ubifs_setxattr(struct dentry *dentry, const char *name,
 		   const void *value, size_t size, int flags);
-ssize_t ubifs_getxattr(struct dentry *dentry, const char *name, void *buf,
-		       size_t size);
+ssize_t ubifs_getxattr(struct dentry *dentry, struct inode *host,
+		       const char *name, void *buf, size_t size);
 ssize_t ubifs_listxattr(struct dentry *dentry, char *buffer, size_t size);
 int ubifs_removexattr(struct dentry *dentry, const char *name);
 int ubifs_init_security(struct inode *dentry, struct inode *inode,
diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c
index b043e044121d..413d650c9476 100644
--- a/fs/ubifs/xattr.c
+++ b/fs/ubifs/xattr.c
@@ -372,10 +372,10 @@ int ubifs_setxattr(struct dentry *dentry, const char *name,
 	return setxattr(d_inode(dentry), name, value, size, flags);
 }
 
-ssize_t ubifs_getxattr(struct dentry *dentry, const char *name, void *buf,
-		       size_t size)
+ssize_t ubifs_getxattr(struct dentry *dentry, struct inode *host,
+		       const char *name, void *buf, size_t size)
 {
-	struct inode *inode, *host = d_inode(dentry);
+	struct inode *inode;
 	struct ubifs_info *c = host->i_sb->s_fs_info;
 	struct qstr nm = QSTR_INIT(name, strlen(name));
 	struct ubifs_inode *ui;
diff --git a/fs/xattr.c b/fs/xattr.c
index 461ba45b7da9..b11945e15fde 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c
@@ -192,7 +192,7 @@ vfs_getxattr_alloc(struct dentry *dentry, const char *name, char **xattr_value,
 	if (!inode->i_op->getxattr)
 		return -EOPNOTSUPP;
 
-	error = inode->i_op->getxattr(dentry, name, NULL, 0);
+	error = inode->i_op->getxattr(dentry, inode, name, NULL, 0);
 	if (error < 0)
 		return error;
 
@@ -203,7 +203,7 @@ vfs_getxattr_alloc(struct dentry *dentry, const char *name, char **xattr_value,
 		memset(value, 0, error + 1);
 	}
 
-	error = inode->i_op->getxattr(dentry, name, value, error);
+	error = inode->i_op->getxattr(dentry, inode, name, value, error);
 	*xattr_value = value;
 	return error;
 }
@@ -236,7 +236,7 @@ vfs_getxattr(struct dentry *dentry, const char *name, void *value, size_t size)
 	}
 nolsm:
 	if (inode->i_op->getxattr)
-		error = inode->i_op->getxattr(dentry, name, value, size);
+		error = inode->i_op->getxattr(dentry, inode, name, value, size);
 	else
 		error = -EOPNOTSUPP;
 
@@ -691,14 +691,15 @@ xattr_resolve_name(const struct xattr_handler **handlers, const char **name)
  * Find the handler for the prefix and dispatch its get() operation.
  */
 ssize_t
-generic_getxattr(struct dentry *dentry, const char *name, void *buffer, size_t size)
+generic_getxattr(struct dentry *dentry, struct inode *inode,
+		 const char *name, void *buffer, size_t size)
 {
 	const struct xattr_handler *handler;
 
 	handler = xattr_resolve_name(dentry->d_sb->s_xattr, &name);
 	if (IS_ERR(handler))
 		return PTR_ERR(handler);
-	return handler->get(handler, dentry, d_inode(dentry),
+	return handler->get(handler, dentry, inode,
 			    name, buffer, size);
 }
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 329ed372d708..1b5fcaeea827 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1702,7 +1702,8 @@ struct inode_operations {
 	int (*setattr) (struct dentry *, struct iattr *);
 	int (*getattr) (struct vfsmount *mnt, struct dentry *, struct kstat *);
 	int (*setxattr) (struct dentry *, const char *,const void *,size_t,int);
-	ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t);
+	ssize_t (*getxattr) (struct dentry *, struct inode *,
+			     const char *, void *, size_t);
 	ssize_t (*listxattr) (struct dentry *, char *, size_t);
 	int (*removexattr) (struct dentry *, const char *);
 	int (*fiemap)(struct inode *, struct fiemap_extent_info *, u64 start,
diff --git a/include/linux/xattr.h b/include/linux/xattr.h
index c11c022298b9..1cc4c578deb9 100644
--- a/include/linux/xattr.h
+++ b/include/linux/xattr.h
@@ -52,7 +52,7 @@ int __vfs_setxattr_noperm(struct dentry *, const char *, const void *, size_t, i
 int vfs_setxattr(struct dentry *, const char *, const void *, size_t, int);
 int vfs_removexattr(struct dentry *, const char *);
 
-ssize_t generic_getxattr(struct dentry *dentry, const char *name, void *buffer, size_t size);
+ssize_t generic_getxattr(struct dentry *dentry, struct inode *inode, const char *name, void *buffer, size_t size);
 ssize_t generic_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size);
 int generic_setxattr(struct dentry *dentry, const char *name, const void *value, size_t size, int flags);
 int generic_removexattr(struct dentry *dentry, const char *name);
diff --git a/net/socket.c b/net/socket.c
index 5f77a8e93830..35e4523edada 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -466,7 +466,7 @@ static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed)
 #define XATTR_SOCKPROTONAME_SUFFIX "sockprotoname"
 #define XATTR_NAME_SOCKPROTONAME (XATTR_SYSTEM_PREFIX XATTR_SOCKPROTONAME_SUFFIX)
 #define XATTR_NAME_SOCKPROTONAME_LEN (sizeof(XATTR_NAME_SOCKPROTONAME)-1)
-static ssize_t sockfs_getxattr(struct dentry *dentry,
+static ssize_t sockfs_getxattr(struct dentry *dentry, struct inode *inode,
 			       const char *name, void *value, size_t size)
 {
 	const char *proto_name;
diff --git a/security/commoncap.c b/security/commoncap.c
index 48071ed7c445..a042077312a5 100644
--- a/security/commoncap.c
+++ b/security/commoncap.c
@@ -313,7 +313,7 @@ int cap_inode_need_killpriv(struct dentry *dentry)
 	if (!inode->i_op->getxattr)
 	       return 0;
 
-	error = inode->i_op->getxattr(dentry, XATTR_NAME_CAPS, NULL, 0);
+	error = inode->i_op->getxattr(dentry, inode, XATTR_NAME_CAPS, NULL, 0);
 	if (error <= 0)
 		return 0;
 	return 1;
@@ -397,8 +397,8 @@ int get_vfs_caps_from_disk(const struct dentry *dentry, struct cpu_vfs_cap_data
 	if (!inode || !inode->i_op->getxattr)
 		return -ENODATA;
 
-	size = inode->i_op->getxattr((struct dentry *)dentry, XATTR_NAME_CAPS, &caps,
-				   XATTR_CAPS_SZ);
+	size = inode->i_op->getxattr((struct dentry *)dentry, inode,
+				     XATTR_NAME_CAPS, &caps, XATTR_CAPS_SZ);
 	if (size == -ENODATA || size == -EOPNOTSUPP)
 		/* no data, that's ok */
 		return -ENODATA;
diff --git a/security/integrity/evm/evm_main.c b/security/integrity/evm/evm_main.c
index 84c6d11fc096..b9e26288d30c 100644
--- a/security/integrity/evm/evm_main.c
+++ b/security/integrity/evm/evm_main.c
@@ -82,7 +82,7 @@ static int evm_find_protected_xattrs(struct dentry *dentry)
 		return -EOPNOTSUPP;
 
 	for (xattr = evm_config_xattrnames; *xattr != NULL; xattr++) {
-		error = inode->i_op->getxattr(dentry, *xattr, NULL, 0);
+		error = inode->i_op->getxattr(dentry, inode, *xattr, NULL, 0);
 		if (error < 0) {
 			if (error == -ENODATA)
 				continue;
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 889cd59ca5a7..469f5c75bd4b 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -506,7 +506,8 @@ static int sb_finish_set_opts(struct super_block *sb)
 			rc = -EOPNOTSUPP;
 			goto out;
 		}
-		rc = root_inode->i_op->getxattr(root, XATTR_NAME_SELINUX, NULL, 0);
+		rc = root_inode->i_op->getxattr(root, root_inode,
+						XATTR_NAME_SELINUX, NULL, 0);
 		if (rc < 0 && rc != -ENODATA) {
 			if (rc == -EOPNOTSUPP)
 				printk(KERN_WARNING "SELinux: (dev %s, type "
@@ -1412,13 +1413,13 @@ static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dent
 			goto out_unlock;
 		}
 		context[len] = '\0';
-		rc = inode->i_op->getxattr(dentry, XATTR_NAME_SELINUX,
+		rc = inode->i_op->getxattr(dentry, inode, XATTR_NAME_SELINUX,
 					   context, len);
 		if (rc == -ERANGE) {
 			kfree(context);
 
 			/* Need a larger buffer.  Query for the right size. */
-			rc = inode->i_op->getxattr(dentry, XATTR_NAME_SELINUX,
+			rc = inode->i_op->getxattr(dentry, inode, XATTR_NAME_SELINUX,
 						   NULL, 0);
 			if (rc < 0) {
 				dput(dentry);
@@ -1432,7 +1433,7 @@ static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dent
 				goto out_unlock;
 			}
 			context[len] = '\0';
-			rc = inode->i_op->getxattr(dentry,
+			rc = inode->i_op->getxattr(dentry, inode,
 						   XATTR_NAME_SELINUX,
 						   context, len);
 		}
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index 50bcca26c0b7..ff2b8c3cf7a9 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -272,7 +272,7 @@ static struct smack_known *smk_fetch(const char *name, struct inode *ip,
 	if (buffer == NULL)
 		return ERR_PTR(-ENOMEM);
 
-	rc = ip->i_op->getxattr(dp, name, buffer, SMK_LONGLABEL);
+	rc = ip->i_op->getxattr(dp, ip, name, buffer, SMK_LONGLABEL);
 	if (rc < 0)
 		skp = ERR_PTR(rc);
 	else if (rc == 0)
@@ -3519,7 +3519,7 @@ static void smack_d_instantiate(struct dentry *opt_dentry, struct inode *inode)
 					TRANS_TRUE, TRANS_TRUE_SIZE,
 					0);
 			} else {
-				rc = inode->i_op->getxattr(dp,
+				rc = inode->i_op->getxattr(dp, inode,
 					XATTR_NAME_SMACKTRANSMUTE, trattr,
 					TRANS_TRUE_SIZE);
 				if (rc >= 0 && strncmp(trattr, TRANS_TRUE,
-- 
cgit v1.2.3


From 85c7f81041d57cfe9dc97f4680d5586b54534a39 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 14 Apr 2016 19:52:13 -0400
Subject: beginning of transition to parallel lookups - marking in-lookup
 dentries

marked as such when (would be) parallel lookup is about to pass them
to actual ->lookup(); unmarked when
	* __d_add() is about to make it hashed, positive or not.
	* __d_move() (from d_splice_alias(), directly or via
__d_unalias()) puts a preexisting dentry in its place
	* in caller of ->lookup() if it has escaped all of the
above.  Bug (WARN_ON, actually) if it reaches the final dput()
or d_instantiate() while still marked such.

As the result, we are guaranteed that for as long as the flag is
set, dentry will
	* remain negative unhashed with positive refcount
	* never have its ->d_alias looked at
	* never have its ->d_lru looked at
	* never have its ->d_parent and ->d_name changed

Right now we have at most one such for any given parent directory.
With parallel lookups that restriction will weaken to
	* only exist when parent is locked shared
	* at most one with given (parent,name) pair (comparison of
names is according to ->d_compare())
	* only exist when there's no hashed dentry with the same
(parent,name)

Transition will take the next several commits; unfortunately, we'll
only be able to switch to rwsem at the end of this series.  The
reason for not making it a single patch is to simplify review.

New primitives: d_in_lookup() (a predicate checking if dentry is in
the in-lookup state) and d_lookup_done() (tells the system that
we are done with lookup and if it's still marked as in-lookup, it
should cease to be such).

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c            | 13 +++++++++++++
 fs/namei.c             |  4 ++++
 include/linux/dcache.h | 18 ++++++++++++++++++
 3 files changed, 35 insertions(+)

(limited to 'include/linux')

diff --git a/fs/dcache.c b/fs/dcache.c
index 20394fb6f967..0f1d93866e69 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -761,6 +761,8 @@ repeat:
 	/* Slow case: now with the dentry lock held */
 	rcu_read_unlock();
 
+	WARN_ON(d_in_lookup(dentry));
+
 	/* Unreachable? Get rid of it */
 	if (unlikely(d_unhashed(dentry)))
 		goto kill_it;
@@ -1746,6 +1748,7 @@ type_determined:
 static void __d_instantiate(struct dentry *dentry, struct inode *inode)
 {
 	unsigned add_flags = d_flags_for_inode(inode);
+	WARN_ON(d_in_lookup(dentry));
 
 	spin_lock(&dentry->d_lock);
 	hlist_add_head(&dentry->d_u.d_alias, &inode->i_dentry);
@@ -2361,12 +2364,20 @@ void d_rehash(struct dentry * entry)
 }
 EXPORT_SYMBOL(d_rehash);
 
+void __d_lookup_done(struct dentry *dentry)
+{
+	dentry->d_flags &= ~DCACHE_PAR_LOOKUP;
+	/* more stuff will land here */
+}
+EXPORT_SYMBOL(__d_lookup_done);
 
 /* inode->i_lock held if inode is non-NULL */
 
 static inline void __d_add(struct dentry *dentry, struct inode *inode)
 {
 	spin_lock(&dentry->d_lock);
+	if (unlikely(d_in_lookup(dentry)))
+		__d_lookup_done(dentry);
 	if (inode) {
 		unsigned add_flags = d_flags_for_inode(inode);
 		hlist_add_head(&dentry->d_u.d_alias, &inode->i_dentry);
@@ -2612,6 +2623,8 @@ static void __d_move(struct dentry *dentry, struct dentry *target,
 	BUG_ON(d_ancestor(target, dentry));
 
 	dentry_lock_for_move(dentry, target);
+	if (unlikely(d_in_lookup(target)))
+		__d_lookup_done(target);
 
 	write_seqcount_begin(&dentry->d_seq);
 	write_seqcount_begin_nested(&target->d_seq, DENTRY_D_LOCK_NESTED);
diff --git a/fs/namei.c b/fs/namei.c
index c275635c4b9e..26e5f84e0c36 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1634,7 +1634,11 @@ static struct dentry *lookup_slow(const struct qstr *name,
 		inode_unlock(inode);
 		return ERR_PTR(-ENOMEM);
 	}
+	spin_lock(&dentry->d_lock);
+	dentry->d_flags |= DCACHE_PAR_LOOKUP;
+	spin_unlock(&dentry->d_lock);
 	old = inode->i_op->lookup(inode, dentry, flags);
+	d_lookup_done(dentry);
 	if (unlikely(old)) {
 		dput(dentry);
 		dentry = old;
diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index 4bb4de8d95ea..9a7aa890b642 100644
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -232,6 +232,8 @@ struct dentry_operations {
 #define DCACHE_ENCRYPTED_WITH_KEY	0x04000000 /* dir is encrypted with a valid key */
 #define DCACHE_OP_REAL			0x08000000
 
+#define DCACHE_PAR_LOOKUP		0x10000000 /* being looked up (with parent locked shared) */
+
 extern seqlock_t rename_lock;
 
 /*
@@ -367,6 +369,22 @@ static inline void dont_mount(struct dentry *dentry)
 	spin_unlock(&dentry->d_lock);
 }
 
+extern void __d_lookup_done(struct dentry *);
+
+static inline int d_in_lookup(struct dentry *dentry)
+{
+	return dentry->d_flags & DCACHE_PAR_LOOKUP;
+}
+
+static inline void d_lookup_done(struct dentry *dentry)
+{
+	if (unlikely(d_in_lookup(dentry))) {
+		spin_lock(&dentry->d_lock);
+		__d_lookup_done(dentry);
+		spin_unlock(&dentry->d_lock);
+	}
+}
+
 extern void dput(struct dentry *);
 
 static inline bool d_managed(const struct dentry *dentry)
-- 
cgit v1.2.3


From 84e710da2a1dfacfc87f604869a4d22df91ce6cd Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 15 Apr 2016 00:58:55 -0400
Subject: parallel lookups machinery, part 2

We'll need to verify that there's neither a hashed nor in-lookup
dentry with desired parent/name before adding to in-lookup set.

One possible solution would be to hold the parent's ->d_lock through
both checks, but while the in-lookup set is relatively small at any
time, dcache is not.  And holding the parent's ->d_lock through
something like __d_lookup_rcu() would suck too badly.

So we leave the parent's ->d_lock alone, which means that we watch
out for the following scenario:
	* we verify that there's no hashed match
	* existing in-lookup match gets hashed by another process
	* we verify that there's no in-lookup matches and decide
that everything's fine.

Solution: per-directory kinda-sorta seqlock, bumped around the times
we hash something that used to be in-lookup or move (and hash)
something in place of in-lookup.  Then the above would turn into
	* read the counter
	* do dcache lookup
	* if no matches found, check for in-lookup matches
	* if there had been none of those either, check if the
counter has changed; repeat if it has.

The "kinda-sorta" part is due to the fact that we don't have much spare
space in inode.  There is a spare word (shared with i_bdev/i_cdev/i_pipe),
so the counter part is not a problem, but spinlock is a different story.

We could use the parent's ->d_lock, and it would be less painful in
terms of contention, for __d_add() it would be rather inconvenient to
grab; we could do that (using lock_parent()), but...

Fortunately, we can get serialization on the counter itself, and it
might be a good idea in general; we can use cmpxchg() in a loop to
get from even to odd and smp_store_release() from odd to even.

This commit adds the counter and updating logics; the readers will be
added in the next commit.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 Documentation/filesystems/porting |  8 ++++++++
 fs/dcache.c                       | 34 ++++++++++++++++++++++++++++++++--
 fs/inode.c                        |  1 +
 include/linux/fs.h                |  1 +
 mm/shmem.c                        |  3 ++-
 5 files changed, 44 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/filesystems/porting b/Documentation/filesystems/porting
index 57bb3754a027..8810e2367fe6 100644
--- a/Documentation/filesystems/porting
+++ b/Documentation/filesystems/porting
@@ -531,3 +531,11 @@ in your dentry operations instead.
 	dentry might be yet to be attached to inode, so do _not_ use its ->d_inode
 	in the instances.  Rationale: !@#!@# security_d_instantiate() needs to be
 	called before we attach dentry to inode.
+--
+[mandatory]
+	symlinks are no longer the only inodes that do *not* have i_bdev/i_cdev/
+	i_pipe/i_link union zeroed out at inode eviction.  As the result, you can't
+	assume that non-NULL value in ->i_nlink at ->destroy_inode() implies that
+	it's a symlink.  Checking ->i_mode is really needed now.  In-tree we had
+	to fix shmem_destroy_callback() that used to take that kind of shortcut;
+	watch out, since that shortcut is no longer valid.
diff --git a/fs/dcache.c b/fs/dcache.c
index 0f1d93866e69..10988f7e5a23 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -2364,6 +2364,22 @@ void d_rehash(struct dentry * entry)
 }
 EXPORT_SYMBOL(d_rehash);
 
+static inline unsigned start_dir_add(struct inode *dir)
+{
+
+	for (;;) {
+		unsigned n = dir->i_dir_seq;
+		if (!(n & 1) && cmpxchg(&dir->i_dir_seq, n, n + 1) == n)
+			return n;
+		cpu_relax();
+	}
+}
+
+static inline void end_dir_add(struct inode *dir, unsigned n)
+{
+	smp_store_release(&dir->i_dir_seq, n + 2);
+}
+
 void __d_lookup_done(struct dentry *dentry)
 {
 	dentry->d_flags &= ~DCACHE_PAR_LOOKUP;
@@ -2375,9 +2391,14 @@ EXPORT_SYMBOL(__d_lookup_done);
 
 static inline void __d_add(struct dentry *dentry, struct inode *inode)
 {
+	struct inode *dir = NULL;
+	unsigned n;
 	spin_lock(&dentry->d_lock);
-	if (unlikely(d_in_lookup(dentry)))
+	if (unlikely(d_in_lookup(dentry))) {
+		dir = dentry->d_parent->d_inode;
+		n = start_dir_add(dir);
 		__d_lookup_done(dentry);
+	}
 	if (inode) {
 		unsigned add_flags = d_flags_for_inode(inode);
 		hlist_add_head(&dentry->d_u.d_alias, &inode->i_dentry);
@@ -2387,6 +2408,8 @@ static inline void __d_add(struct dentry *dentry, struct inode *inode)
 		__fsnotify_d_instantiate(dentry);
 	}
 	_d_rehash(dentry);
+	if (dir)
+		end_dir_add(dir, n);
 	spin_unlock(&dentry->d_lock);
 	if (inode)
 		spin_unlock(&inode->i_lock);
@@ -2616,6 +2639,8 @@ static void dentry_unlock_for_move(struct dentry *dentry, struct dentry *target)
 static void __d_move(struct dentry *dentry, struct dentry *target,
 		     bool exchange)
 {
+	struct inode *dir = NULL;
+	unsigned n;
 	if (!dentry->d_inode)
 		printk(KERN_WARNING "VFS: moving negative dcache entry\n");
 
@@ -2623,8 +2648,11 @@ static void __d_move(struct dentry *dentry, struct dentry *target,
 	BUG_ON(d_ancestor(target, dentry));
 
 	dentry_lock_for_move(dentry, target);
-	if (unlikely(d_in_lookup(target)))
+	if (unlikely(d_in_lookup(target))) {
+		dir = target->d_parent->d_inode;
+		n = start_dir_add(dir);
 		__d_lookup_done(target);
+	}
 
 	write_seqcount_begin(&dentry->d_seq);
 	write_seqcount_begin_nested(&target->d_seq, DENTRY_D_LOCK_NESTED);
@@ -2674,6 +2702,8 @@ static void __d_move(struct dentry *dentry, struct dentry *target,
 	write_seqcount_end(&target->d_seq);
 	write_seqcount_end(&dentry->d_seq);
 
+	if (dir)
+		end_dir_add(dir, n);
 	dentry_unlock_for_move(dentry, target);
 }
 
diff --git a/fs/inode.c b/fs/inode.c
index 4202aac99464..4b884f73214e 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -151,6 +151,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
 	inode->i_bdev = NULL;
 	inode->i_cdev = NULL;
 	inode->i_link = NULL;
+	inode->i_dir_seq = 0;
 	inode->i_rdev = 0;
 	inode->dirtied_when = 0;
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 6d0fa9174a24..00cecc5a2f75 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -684,6 +684,7 @@ struct inode {
 		struct block_device	*i_bdev;
 		struct cdev		*i_cdev;
 		char			*i_link;
+		unsigned		i_dir_seq;
 	};
 
 	__u32			i_generation;
diff --git a/mm/shmem.c b/mm/shmem.c
index 4640699b209b..e684a9140228 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -3123,7 +3123,8 @@ static struct inode *shmem_alloc_inode(struct super_block *sb)
 static void shmem_destroy_callback(struct rcu_head *head)
 {
 	struct inode *inode = container_of(head, struct inode, i_rcu);
-	kfree(inode->i_link);
+	if (S_ISLNK(inode->i_mode))
+		kfree(inode->i_link);
 	kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode));
 }
 
-- 
cgit v1.2.3


From 94bdd655caba2080ae81d83d756d325abdffcb9f Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 15 Apr 2016 02:42:04 -0400
Subject: parallel lookups machinery, part 3

We will need to be able to check if there is an in-lookup
dentry with matching parent/name.  Right now it's impossible,
but as soon as start locking directories shared such beasts
will appear.

Add a secondary hash for locating those.  Hash chains go through
the same space where d_alias will be once it's not in-lookup anymore.
Search is done under the same bitlock we use for modifications -
with the primary hash we can rely on d_rehash() into the wrong
chain being the worst that could happen, but here the pointers are
buggered once it's removed from the chain.  On the other hand,
the chains are not going to be long and normally we'll end up
adding to the chain anyway.  That allows us to avoid bothering with
->d_lock when doing the comparisons - everything is stable until
removed from chain.

New helper: d_alloc_parallel().  Right now it allocates, verifies
that no hashed and in-lookup matches exist and adds to in-lookup
hash.

Returns ERR_PTR() for error, hashed match (in the unlikely case it's
been found) or new dentry.  In-lookup matches trigger BUG() for
now; that will change in the next commit when we introduce waiting
for ongoing lookup to finish.  Note that in-lookup matches won't be
possible until we actually go for shared locking.

lookup_slow() switched to use of d_alloc_parallel().

Again, these commits are separated only for making it easier to
review.  All this machinery will start doing something useful only
when we go for shared locking; it's just that the combination is
too large for my taste.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c            | 104 +++++++++++++++++++++++++++++++++++++++++++++++++
 fs/namei.c             |  44 +++++++++------------
 include/linux/dcache.h |   2 +
 3 files changed, 125 insertions(+), 25 deletions(-)

(limited to 'include/linux')

diff --git a/fs/dcache.c b/fs/dcache.c
index 10988f7e5a23..ea2de7c19b08 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -111,6 +111,17 @@ static inline struct hlist_bl_head *d_hash(const struct dentry *parent,
 	return dentry_hashtable + hash_32(hash, d_hash_shift);
 }
 
+#define IN_LOOKUP_SHIFT 10
+static struct hlist_bl_head in_lookup_hashtable[1 << IN_LOOKUP_SHIFT];
+
+static inline struct hlist_bl_head *in_lookup_hash(const struct dentry *parent,
+					unsigned int hash)
+{
+	hash += (unsigned long) parent / L1_CACHE_BYTES;
+	return in_lookup_hashtable + hash_32(hash, IN_LOOKUP_SHIFT);
+}
+
+
 /* Statistics gathering. */
 struct dentry_stat_t dentry_stat = {
 	.age_limit = 45,
@@ -2380,9 +2391,102 @@ static inline void end_dir_add(struct inode *dir, unsigned n)
 	smp_store_release(&dir->i_dir_seq, n + 2);
 }
 
+struct dentry *d_alloc_parallel(struct dentry *parent,
+				const struct qstr *name)
+{
+	unsigned int len = name->len;
+	unsigned int hash = name->hash;
+	const unsigned char *str = name->name;
+	struct hlist_bl_head *b = in_lookup_hash(parent, hash);
+	struct hlist_bl_node *node;
+	struct dentry *new = d_alloc(parent, name);
+	struct dentry *dentry;
+	unsigned seq, r_seq, d_seq;
+
+	if (unlikely(!new))
+		return ERR_PTR(-ENOMEM);
+
+retry:
+	rcu_read_lock();
+	seq = smp_load_acquire(&parent->d_inode->i_dir_seq) & ~1;
+	r_seq = read_seqbegin(&rename_lock);
+	dentry = __d_lookup_rcu(parent, name, &d_seq);
+	if (unlikely(dentry)) {
+		if (!lockref_get_not_dead(&dentry->d_lockref)) {
+			rcu_read_unlock();
+			goto retry;
+		}
+		if (read_seqcount_retry(&dentry->d_seq, d_seq)) {
+			rcu_read_unlock();
+			dput(dentry);
+			goto retry;
+		}
+		rcu_read_unlock();
+		dput(new);
+		return dentry;
+	}
+	if (unlikely(read_seqretry(&rename_lock, r_seq))) {
+		rcu_read_unlock();
+		goto retry;
+	}
+	hlist_bl_lock(b);
+	if (unlikely(parent->d_inode->i_dir_seq != seq)) {
+		hlist_bl_unlock(b);
+		rcu_read_unlock();
+		goto retry;
+	}
+	rcu_read_unlock();
+	/*
+	 * No changes for the parent since the beginning of d_lookup().
+	 * Since all removals from the chain happen with hlist_bl_lock(),
+	 * any potential in-lookup matches are going to stay here until
+	 * we unlock the chain.  All fields are stable in everything
+	 * we encounter.
+	 */
+	hlist_bl_for_each_entry(dentry, node, b, d_u.d_in_lookup_hash) {
+		if (dentry->d_name.hash != hash)
+			continue;
+		if (dentry->d_parent != parent)
+			continue;
+		if (d_unhashed(dentry))
+			continue;
+		if (parent->d_flags & DCACHE_OP_COMPARE) {
+			int tlen = dentry->d_name.len;
+			const char *tname = dentry->d_name.name;
+			if (parent->d_op->d_compare(parent, dentry, tlen, tname, name))
+				continue;
+		} else {
+			if (dentry->d_name.len != len)
+				continue;
+			if (dentry_cmp(dentry, str, len))
+				continue;
+		}
+		dget(dentry);
+		hlist_bl_unlock(b);
+		/* impossible until we actually enable parallel lookups */
+		BUG();
+		/* and this will be "wait for it to stop being in-lookup" */
+		/* this one will be handled in the next commit */
+		dput(new);
+		return dentry;
+	}
+	/* we can't take ->d_lock here; it's OK, though. */
+	new->d_flags |= DCACHE_PAR_LOOKUP;
+	hlist_bl_add_head_rcu(&new->d_u.d_in_lookup_hash, b);
+	hlist_bl_unlock(b);
+	return new;
+}
+EXPORT_SYMBOL(d_alloc_parallel);
+
 void __d_lookup_done(struct dentry *dentry)
 {
+	struct hlist_bl_head *b = in_lookup_hash(dentry->d_parent,
+						 dentry->d_name.hash);
+	hlist_bl_lock(b);
 	dentry->d_flags &= ~DCACHE_PAR_LOOKUP;
+	__hlist_bl_del(&dentry->d_u.d_in_lookup_hash);
+	hlist_bl_unlock(b);
+	INIT_HLIST_NODE(&dentry->d_u.d_alias);
 	/* more stuff will land here */
 }
 EXPORT_SYMBOL(__d_lookup_done);
diff --git a/fs/namei.c b/fs/namei.c
index 26e5f84e0c36..aa04320e1f37 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1603,46 +1603,40 @@ static struct dentry *lookup_slow(const struct qstr *name,
 				  struct dentry *dir,
 				  unsigned int flags)
 {
-	struct dentry *dentry, *old;
+	struct dentry *dentry = ERR_PTR(-ENOENT), *old;
 	struct inode *inode = dir->d_inode;
 
 	inode_lock(inode);
 	/* Don't go there if it's already dead */
-	if (unlikely(IS_DEADDIR(inode))) {
-		inode_unlock(inode);
-		return ERR_PTR(-ENOENT);
-	}
-	dentry = d_lookup(dir, name);
-	if (unlikely(dentry)) {
+	if (unlikely(IS_DEADDIR(inode)))
+		goto out;
+again:
+	dentry = d_alloc_parallel(dir, name);
+	if (IS_ERR(dentry))
+		goto out;
+	if (unlikely(!d_in_lookup(dentry))) {
 		if ((dentry->d_flags & DCACHE_OP_REVALIDATE) &&
 		    !(flags & LOOKUP_NO_REVAL)) {
 			int error = d_revalidate(dentry, flags);
 			if (unlikely(error <= 0)) {
-				if (!error)
+				if (!error) {
 					d_invalidate(dentry);
+					dput(dentry);
+					goto again;
+				}
 				dput(dentry);
 				dentry = ERR_PTR(error);
 			}
 		}
-		if (dentry) {
-			inode_unlock(inode);
-			return dentry;
+	} else {
+		old = inode->i_op->lookup(inode, dentry, flags);
+		d_lookup_done(dentry);
+		if (unlikely(old)) {
+			dput(dentry);
+			dentry = old;
 		}
 	}
-	dentry = d_alloc(dir, name);
-	if (unlikely(!dentry)) {
-		inode_unlock(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	spin_lock(&dentry->d_lock);
-	dentry->d_flags |= DCACHE_PAR_LOOKUP;
-	spin_unlock(&dentry->d_lock);
-	old = inode->i_op->lookup(inode, dentry, flags);
-	d_lookup_done(dentry);
-	if (unlikely(old)) {
-		dput(dentry);
-		dentry = old;
-	}
+out:
 	inode_unlock(inode);
 	return dentry;
 }
diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index 9a7aa890b642..3eea562f5f27 100644
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -131,6 +131,7 @@ struct dentry {
 	 */
 	union {
 		struct hlist_node d_alias;	/* inode alias list */
+		struct hlist_bl_node d_in_lookup_hash;	/* only for in-lookup ones */
 	 	struct rcu_head d_rcu;
 	} d_u;
 };
@@ -250,6 +251,7 @@ extern void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op
 /* allocate/de-allocate */
 extern struct dentry * d_alloc(struct dentry *, const struct qstr *);
 extern struct dentry * d_alloc_pseudo(struct super_block *, const struct qstr *);
+extern struct dentry * d_alloc_parallel(struct dentry *, const struct qstr *);
 extern struct dentry * d_splice_alias(struct inode *, struct dentry *);
 extern struct dentry * d_add_ci(struct dentry *, struct inode *, struct qstr *);
 extern struct dentry * d_exact_alias(struct dentry *, struct inode *);
-- 
cgit v1.2.3


From d9171b9345261e0d941d92fdda5672b5db67f968 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 15 Apr 2016 03:33:13 -0400
Subject: parallel lookups machinery, part 4 (and last)

If we *do* run into an in-lookup match, we need to wait for it to
cease being in-lookup.  Fortunately, we do have unused space in
in-lookup dentries - d_lru is never looked at until it stops being
in-lookup.

So we can stash a pointer to wait_queue_head from stack frame of
the caller of ->lookup().  Some precautions are needed while
waiting, but it's not that hard - we do hold a reference to dentry
we are waiting for, so it can't go away.  If it's found to be
in-lookup the wait_queue_head is still alive and will remain so
at least while ->d_lock is held.  Moreover, the condition we
are waiting for becomes true at the same point where everything
on that wq gets woken up, so we can just add ourselves to the
queue once.

d_alloc_parallel() gets a pointer to wait_queue_head_t from its
caller; lookup_slow() adjusted, d_add_ci() taught to use
d_alloc_parallel() if the dentry passed to it happens to be
in-lookup one (i.e. if it's been called from the parallel lookup).

That's pretty much it - all that remains is to switch ->i_mutex
to rwsem and have lookup_slow() take it shared.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c            | 94 +++++++++++++++++++++++++++++++++++++++-----------
 fs/namei.c             |  3 +-
 include/linux/dcache.h |  8 +++--
 3 files changed, 82 insertions(+), 23 deletions(-)

(limited to 'include/linux')

diff --git a/fs/dcache.c b/fs/dcache.c
index ea2de7c19b08..59fcffcbf096 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1987,28 +1987,36 @@ EXPORT_SYMBOL(d_obtain_root);
 struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
 			struct qstr *name)
 {
-	struct dentry *found;
-	struct dentry *new;
+	struct dentry *found, *res;
 
 	/*
 	 * First check if a dentry matching the name already exists,
 	 * if not go ahead and create it now.
 	 */
 	found = d_hash_and_lookup(dentry->d_parent, name);
-	if (!found) {
-		new = d_alloc(dentry->d_parent, name);
-		if (!new) {
-			found = ERR_PTR(-ENOMEM);
-		} else {
-			found = d_splice_alias(inode, new);
-			if (found) {
-				dput(new);
-				return found;
-			}
-			return new;
+	if (found) {
+		iput(inode);
+		return found;
+	}
+	if (d_in_lookup(dentry)) {
+		found = d_alloc_parallel(dentry->d_parent, name,
+					dentry->d_wait);
+		if (IS_ERR(found) || !d_in_lookup(found)) {
+			iput(inode);
+			return found;
 		}
+	} else {
+		found = d_alloc(dentry->d_parent, name);
+		if (!found) {
+			iput(inode);
+			return ERR_PTR(-ENOMEM);
+		} 
+	}
+	res = d_splice_alias(inode, found);
+	if (res) {
+		dput(found);
+		return res;
 	}
-	iput(inode);
 	return found;
 }
 EXPORT_SYMBOL(d_add_ci);
@@ -2391,8 +2399,23 @@ static inline void end_dir_add(struct inode *dir, unsigned n)
 	smp_store_release(&dir->i_dir_seq, n + 2);
 }
 
+static void d_wait_lookup(struct dentry *dentry)
+{
+	if (d_in_lookup(dentry)) {
+		DECLARE_WAITQUEUE(wait, current);
+		add_wait_queue(dentry->d_wait, &wait);
+		do {
+			set_current_state(TASK_UNINTERRUPTIBLE);
+			spin_unlock(&dentry->d_lock);
+			schedule();
+			spin_lock(&dentry->d_lock);
+		} while (d_in_lookup(dentry));
+	}
+}
+
 struct dentry *d_alloc_parallel(struct dentry *parent,
-				const struct qstr *name)
+				const struct qstr *name,
+				wait_queue_head_t *wq)
 {
 	unsigned int len = name->len;
 	unsigned int hash = name->hash;
@@ -2463,18 +2486,47 @@ retry:
 		}
 		dget(dentry);
 		hlist_bl_unlock(b);
-		/* impossible until we actually enable parallel lookups */
-		BUG();
-		/* and this will be "wait for it to stop being in-lookup" */
-		/* this one will be handled in the next commit */
+		/* somebody is doing lookup for it right now; wait for it */
+		spin_lock(&dentry->d_lock);
+		d_wait_lookup(dentry);
+		/*
+		 * it's not in-lookup anymore; in principle we should repeat
+		 * everything from dcache lookup, but it's likely to be what
+		 * d_lookup() would've found anyway.  If it is, just return it;
+		 * otherwise we really have to repeat the whole thing.
+		 */
+		if (unlikely(dentry->d_name.hash != hash))
+			goto mismatch;
+		if (unlikely(dentry->d_parent != parent))
+			goto mismatch;
+		if (unlikely(d_unhashed(dentry)))
+			goto mismatch;
+		if (parent->d_flags & DCACHE_OP_COMPARE) {
+			int tlen = dentry->d_name.len;
+			const char *tname = dentry->d_name.name;
+			if (parent->d_op->d_compare(parent, dentry, tlen, tname, name))
+				goto mismatch;
+		} else {
+			if (unlikely(dentry->d_name.len != len))
+				goto mismatch;
+			if (unlikely(dentry_cmp(dentry, str, len)))
+				goto mismatch;
+		}
+		/* OK, it *is* a hashed match; return it */
+		spin_unlock(&dentry->d_lock);
 		dput(new);
 		return dentry;
 	}
 	/* we can't take ->d_lock here; it's OK, though. */
 	new->d_flags |= DCACHE_PAR_LOOKUP;
+	new->d_wait = wq;
 	hlist_bl_add_head_rcu(&new->d_u.d_in_lookup_hash, b);
 	hlist_bl_unlock(b);
 	return new;
+mismatch:
+	spin_unlock(&dentry->d_lock);
+	dput(dentry);
+	goto retry;
 }
 EXPORT_SYMBOL(d_alloc_parallel);
 
@@ -2485,9 +2537,11 @@ void __d_lookup_done(struct dentry *dentry)
 	hlist_bl_lock(b);
 	dentry->d_flags &= ~DCACHE_PAR_LOOKUP;
 	__hlist_bl_del(&dentry->d_u.d_in_lookup_hash);
+	wake_up_all(dentry->d_wait);
+	dentry->d_wait = NULL;
 	hlist_bl_unlock(b);
 	INIT_HLIST_NODE(&dentry->d_u.d_alias);
-	/* more stuff will land here */
+	INIT_LIST_HEAD(&dentry->d_lru);
 }
 EXPORT_SYMBOL(__d_lookup_done);
 
diff --git a/fs/namei.c b/fs/namei.c
index aa04320e1f37..7babb5e5f276 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1605,13 +1605,14 @@ static struct dentry *lookup_slow(const struct qstr *name,
 {
 	struct dentry *dentry = ERR_PTR(-ENOENT), *old;
 	struct inode *inode = dir->d_inode;
+	DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq);
 
 	inode_lock(inode);
 	/* Don't go there if it's already dead */
 	if (unlikely(IS_DEADDIR(inode)))
 		goto out;
 again:
-	dentry = d_alloc_parallel(dir, name);
+	dentry = d_alloc_parallel(dir, name, &wq);
 	if (IS_ERR(dentry))
 		goto out;
 	if (unlikely(!d_in_lookup(dentry))) {
diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index 3eea562f5f27..6b1d8bbb3496 100644
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -123,7 +123,10 @@ struct dentry {
 	unsigned long d_time;		/* used by d_revalidate */
 	void *d_fsdata;			/* fs-specific data */
 
-	struct list_head d_lru;		/* LRU list */
+	union {
+		struct list_head d_lru;		/* LRU list */
+		wait_queue_head_t *d_wait;	/* in-lookup ones only */
+	};
 	struct list_head d_child;	/* child of parent list */
 	struct list_head d_subdirs;	/* our children */
 	/*
@@ -251,7 +254,8 @@ extern void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op
 /* allocate/de-allocate */
 extern struct dentry * d_alloc(struct dentry *, const struct qstr *);
 extern struct dentry * d_alloc_pseudo(struct super_block *, const struct qstr *);
-extern struct dentry * d_alloc_parallel(struct dentry *, const struct qstr *);
+extern struct dentry * d_alloc_parallel(struct dentry *, const struct qstr *,
+					wait_queue_head_t *);
 extern struct dentry * d_splice_alias(struct inode *, struct dentry *);
 extern struct dentry * d_add_ci(struct dentry *, struct inode *, struct qstr *);
 extern struct dentry * d_exact_alias(struct dentry *, struct inode *);
-- 
cgit v1.2.3


From 9902af79c01a8e39bb99b922fa3eef6d4ea23d69 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 15 Apr 2016 15:08:36 -0400
Subject: parallel lookups: actual switch to rwsem

ta-da!

The main issue is the lack of down_write_killable(), so the places
like readdir.c switched to plain inode_lock(); once killable
variants of rwsem primitives appear, that'll be dealt with.

lockdep side also might need more work

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 Documentation/filesystems/porting | 18 ++++++++++++++++++
 fs/btrfs/ioctl.c                  | 18 +++++++++++-------
 fs/configfs/inode.c               |  2 +-
 fs/dcache.c                       |  9 +++++----
 fs/gfs2/ops_fstype.c              |  2 +-
 fs/inode.c                        | 12 ++++++------
 fs/namei.c                        |  4 ++--
 fs/ocfs2/inode.c                  |  2 +-
 fs/overlayfs/readdir.c            |  4 +++-
 fs/readdir.c                      |  7 ++++---
 include/linux/fs.h                | 27 +++++++++++++++++++++------
 11 files changed, 73 insertions(+), 32 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/filesystems/porting b/Documentation/filesystems/porting
index 8810e2367fe6..1567a53857bd 100644
--- a/Documentation/filesystems/porting
+++ b/Documentation/filesystems/porting
@@ -539,3 +539,21 @@ in your dentry operations instead.
 	it's a symlink.  Checking ->i_mode is really needed now.  In-tree we had
 	to fix shmem_destroy_callback() that used to take that kind of shortcut;
 	watch out, since that shortcut is no longer valid.
+--
+[mandatory]
+	->i_mutex is replaced with ->i_rwsem now.  inode_lock() et.al. work as
+	they used to - they just take it exclusive.  However, ->lookup() may be
+	called with parent locked shared.  Its instances must not
+		* use d_instantiate) and d_rehash() separately - use d_add() or
+		  d_splice_alias() instead.
+		* use d_rehash() alone - call d_add(new_dentry, NULL) instead.
+		* in the unlikely case when (read-only) access to filesystem
+		  data structures needs exclusion for some reason, arrange it
+		  yourself.  None of the in-tree filesystems needed that.
+		* rely on ->d_parent and ->d_name not changing after dentry has
+		  been fed to d_add() or d_splice_alias().  Again, none of the
+		  in-tree instances relied upon that.
+	We are guaranteed that lookups of the same name in the same directory
+	will not happen in parallel ("same" in the sense of your ->d_compare()).
+	Lookups on different names in the same directory can and do happen in
+	parallel now.
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 5a23806ae418..0b8ba717175b 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -837,9 +837,11 @@ static noinline int btrfs_mksubvol(struct path *parent,
 	struct dentry *dentry;
 	int error;
 
-	error = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
-	if (error == -EINTR)
-		return error;
+	inode_lock_nested(dir, I_MUTEX_PARENT);
+	// XXX: should've been
+	// mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
+	// if (error == -EINTR)
+	//	return error;
 
 	dentry = lookup_one_len(name, parent->dentry, namelen);
 	error = PTR_ERR(dentry);
@@ -2366,9 +2368,11 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 		goto out;
 
 
-	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
-	if (err == -EINTR)
-		goto out_drop_write;
+	inode_lock_nested(dir, I_MUTEX_PARENT);
+	// XXX: should've been
+	// err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
+	// if (err == -EINTR)
+	//	goto out_drop_write;
 	dentry = lookup_one_len(vol_args->name, parent, namelen);
 	if (IS_ERR(dentry)) {
 		err = PTR_ERR(dentry);
@@ -2558,7 +2562,7 @@ out_dput:
 	dput(dentry);
 out_unlock_dir:
 	inode_unlock(dir);
-out_drop_write:
+//out_drop_write:
 	mnt_drop_write_file(file);
 out:
 	kfree(vol_args);
diff --git a/fs/configfs/inode.c b/fs/configfs/inode.c
index 03d124ae27d7..0387968e6f47 100644
--- a/fs/configfs/inode.c
+++ b/fs/configfs/inode.c
@@ -156,7 +156,7 @@ static void configfs_set_inode_lock_class(struct configfs_dirent *sd,
 
 	if (depth > 0) {
 		if (depth <= ARRAY_SIZE(default_group_class)) {
-			lockdep_set_class(&inode->i_mutex,
+			lockdep_set_class(&inode->i_rwsem,
 					  &default_group_class[depth - 1]);
 		} else {
 			/*
diff --git a/fs/dcache.c b/fs/dcache.c
index 59fcffcbf096..e49ba7d1b957 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -2932,7 +2932,8 @@ struct dentry *d_ancestor(struct dentry *p1, struct dentry *p2)
 static int __d_unalias(struct inode *inode,
 		struct dentry *dentry, struct dentry *alias)
 {
-	struct mutex *m1 = NULL, *m2 = NULL;
+	struct mutex *m1 = NULL;
+	struct rw_semaphore *m2 = NULL;
 	int ret = -ESTALE;
 
 	/* If alias and dentry share a parent, then no extra locks required */
@@ -2943,15 +2944,15 @@ static int __d_unalias(struct inode *inode,
 	if (!mutex_trylock(&dentry->d_sb->s_vfs_rename_mutex))
 		goto out_err;
 	m1 = &dentry->d_sb->s_vfs_rename_mutex;
-	if (!inode_trylock(alias->d_parent->d_inode))
+	if (!inode_trylock_shared(alias->d_parent->d_inode))
 		goto out_err;
-	m2 = &alias->d_parent->d_inode->i_mutex;
+	m2 = &alias->d_parent->d_inode->i_rwsem;
 out_unalias:
 	__d_move(alias, dentry, false);
 	ret = 0;
 out_err:
 	if (m2)
-		mutex_unlock(m2);
+		up_read(m2);
 	if (m1)
 		mutex_unlock(m1);
 	return ret;
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index c09c63dcd7a2..45463600fb81 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -824,7 +824,7 @@ static int init_inodes(struct gfs2_sbd *sdp, int undo)
 	 * i_mutex on quota files is special. Since this inode is hidden system
 	 * file, we are safe to define locking ourselves.
 	 */
-	lockdep_set_class(&sdp->sd_quota_inode->i_mutex,
+	lockdep_set_class(&sdp->sd_quota_inode->i_rwsem,
 			  &gfs2_quota_imutex_key);
 
 	error = gfs2_rindex_update(sdp);
diff --git a/fs/inode.c b/fs/inode.c
index 4b884f73214e..4ccbc21b30ce 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -166,8 +166,8 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
 	spin_lock_init(&inode->i_lock);
 	lockdep_set_class(&inode->i_lock, &sb->s_type->i_lock_key);
 
-	mutex_init(&inode->i_mutex);
-	lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key);
+	init_rwsem(&inode->i_rwsem);
+	lockdep_set_class(&inode->i_rwsem, &sb->s_type->i_mutex_key);
 
 	atomic_set(&inode->i_dio_count, 0);
 
@@ -925,13 +925,13 @@ void lockdep_annotate_inode_mutex_key(struct inode *inode)
 		struct file_system_type *type = inode->i_sb->s_type;
 
 		/* Set new key only if filesystem hasn't already changed it */
-		if (lockdep_match_class(&inode->i_mutex, &type->i_mutex_key)) {
+		if (lockdep_match_class(&inode->i_rwsem, &type->i_mutex_key)) {
 			/*
 			 * ensure nobody is actually holding i_mutex
 			 */
-			mutex_destroy(&inode->i_mutex);
-			mutex_init(&inode->i_mutex);
-			lockdep_set_class(&inode->i_mutex,
+			// mutex_destroy(&inode->i_mutex);
+			init_rwsem(&inode->i_rwsem);
+			lockdep_set_class(&inode->i_rwsem,
 					  &type->i_mutex_dir_key);
 		}
 	}
diff --git a/fs/namei.c b/fs/namei.c
index 7babb5e5f276..8249852b5fc6 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1607,7 +1607,7 @@ static struct dentry *lookup_slow(const struct qstr *name,
 	struct inode *inode = dir->d_inode;
 	DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq);
 
-	inode_lock(inode);
+	inode_lock_shared(inode);
 	/* Don't go there if it's already dead */
 	if (unlikely(IS_DEADDIR(inode)))
 		goto out;
@@ -1638,7 +1638,7 @@ again:
 		}
 	}
 out:
-	inode_unlock(inode);
+	inode_unlock_shared(inode);
 	return dentry;
 }
 
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index 12f4a9e9800f..0748777f2e2a 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -262,7 +262,7 @@ static int ocfs2_init_locked_inode(struct inode *inode, void *opaque)
 	inode->i_ino = args->fi_ino;
 	OCFS2_I(inode)->ip_blkno = args->fi_blkno;
 	if (args->fi_sysfile_type != 0)
-		lockdep_set_class(&inode->i_mutex,
+		lockdep_set_class(&inode->i_rwsem,
 			&ocfs2_sysfile_lock_key[args->fi_sysfile_type]);
 	if (args->fi_sysfile_type == USER_QUOTA_SYSTEM_INODE ||
 	    args->fi_sysfile_type == GROUP_QUOTA_SYSTEM_INODE ||
diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c
index 6ec1e43a9a54..da186ee4f846 100644
--- a/fs/overlayfs/readdir.c
+++ b/fs/overlayfs/readdir.c
@@ -218,7 +218,9 @@ static int ovl_check_whiteouts(struct dentry *dir, struct ovl_readdir_data *rdd)
 	cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
 	old_cred = override_creds(override_cred);
 
-	err = mutex_lock_killable(&dir->d_inode->i_mutex);
+	inode_lock(dir->d_inode);
+	err = 0;
+	// XXX: err = mutex_lock_killable(&dir->d_inode->i_mutex);
 	if (!err) {
 		while (rdd->first_maybe_whiteout) {
 			p = rdd->first_maybe_whiteout;
diff --git a/fs/readdir.c b/fs/readdir.c
index e69ef3b79787..bf583e848a1a 100644
--- a/fs/readdir.c
+++ b/fs/readdir.c
@@ -32,9 +32,10 @@ int iterate_dir(struct file *file, struct dir_context *ctx)
 	if (res)
 		goto out;
 
-	res = mutex_lock_killable(&inode->i_mutex);
-	if (res)
-		goto out;
+	inode_lock(inode);
+	// res = mutex_lock_killable(&inode->i_mutex);
+	// if (res)
+	//	goto out;
 
 	res = -ENOENT;
 	if (!IS_DEADDIR(inode)) {
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 00cecc5a2f75..3018f31f7aa0 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -647,7 +647,7 @@ struct inode {
 
 	/* Misc */
 	unsigned long		i_state;
-	struct mutex		i_mutex;
+	struct rw_semaphore	i_rwsem;
 
 	unsigned long		dirtied_when;	/* jiffies of first dirtying */
 	unsigned long		dirtied_time_when;
@@ -734,27 +734,42 @@ enum inode_i_mutex_lock_class
 
 static inline void inode_lock(struct inode *inode)
 {
-	mutex_lock(&inode->i_mutex);
+	down_write(&inode->i_rwsem);
 }
 
 static inline void inode_unlock(struct inode *inode)
 {
-	mutex_unlock(&inode->i_mutex);
+	up_write(&inode->i_rwsem);
+}
+
+static inline void inode_lock_shared(struct inode *inode)
+{
+	down_read(&inode->i_rwsem);
+}
+
+static inline void inode_unlock_shared(struct inode *inode)
+{
+	up_read(&inode->i_rwsem);
 }
 
 static inline int inode_trylock(struct inode *inode)
 {
-	return mutex_trylock(&inode->i_mutex);
+	return down_write_trylock(&inode->i_rwsem);
+}
+
+static inline int inode_trylock_shared(struct inode *inode)
+{
+	return down_read_trylock(&inode->i_rwsem);
 }
 
 static inline int inode_is_locked(struct inode *inode)
 {
-	return mutex_is_locked(&inode->i_mutex);
+	return rwsem_is_locked(&inode->i_rwsem);
 }
 
 static inline void inode_lock_nested(struct inode *inode, unsigned subclass)
 {
-	mutex_lock_nested(&inode->i_mutex, subclass);
+	down_write_nested(&inode->i_rwsem, subclass);
 }
 
 void lock_two_nondirectories(struct inode *, struct inode*);
-- 
cgit v1.2.3


From 63b6df14134ddd048984c8afadb46e721815bfc6 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 20 Apr 2016 17:08:21 -0400
Subject: give readdir(2)/getdents(2)/etc. uniform exclusion with lseek()

same as read() on regular files has, and for the same reason.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 arch/alpha/kernel/osf_sys.c |  4 ++--
 fs/compat.c                 | 12 ++++++------
 fs/file.c                   |  5 +++++
 fs/open.c                   |  2 +-
 fs/read_write.c             | 12 ------------
 fs/readdir.c                | 12 ++++++------
 include/linux/file.h        | 13 +++++++++++++
 7 files changed, 33 insertions(+), 27 deletions(-)

(limited to 'include/linux')

diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c
index 6cc08166ff00..ffb93f499c83 100644
--- a/arch/alpha/kernel/osf_sys.c
+++ b/arch/alpha/kernel/osf_sys.c
@@ -147,7 +147,7 @@ SYSCALL_DEFINE4(osf_getdirentries, unsigned int, fd,
 		long __user *, basep)
 {
 	int error;
-	struct fd arg = fdget(fd);
+	struct fd arg = fdget_pos(fd);
 	struct osf_dirent_callback buf = {
 		.ctx.actor = osf_filldir,
 		.dirent = dirent,
@@ -164,7 +164,7 @@ SYSCALL_DEFINE4(osf_getdirentries, unsigned int, fd,
 	if (count != buf.count)
 		error = count - buf.count;
 
-	fdput(arg);
+	fdput_pos(arg);
 	return error;
 }
 
diff --git a/fs/compat.c b/fs/compat.c
index a71936a3f4cb..8754e9aa14ad 100644
--- a/fs/compat.c
+++ b/fs/compat.c
@@ -884,7 +884,7 @@ COMPAT_SYSCALL_DEFINE3(old_readdir, unsigned int, fd,
 		struct compat_old_linux_dirent __user *, dirent, unsigned int, count)
 {
 	int error;
-	struct fd f = fdget(fd);
+	struct fd f = fdget_pos(fd);
 	struct compat_readdir_callback buf = {
 		.ctx.actor = compat_fillonedir,
 		.dirent = dirent
@@ -897,7 +897,7 @@ COMPAT_SYSCALL_DEFINE3(old_readdir, unsigned int, fd,
 	if (buf.result)
 		error = buf.result;
 
-	fdput(f);
+	fdput_pos(f);
 	return error;
 }
 
@@ -975,7 +975,7 @@ COMPAT_SYSCALL_DEFINE3(getdents, unsigned int, fd,
 	if (!access_ok(VERIFY_WRITE, dirent, count))
 		return -EFAULT;
 
-	f = fdget(fd);
+	f = fdget_pos(fd);
 	if (!f.file)
 		return -EBADF;
 
@@ -989,7 +989,7 @@ COMPAT_SYSCALL_DEFINE3(getdents, unsigned int, fd,
 		else
 			error = count - buf.count;
 	}
-	fdput(f);
+	fdput_pos(f);
 	return error;
 }
 
@@ -1062,7 +1062,7 @@ COMPAT_SYSCALL_DEFINE3(getdents64, unsigned int, fd,
 	if (!access_ok(VERIFY_WRITE, dirent, count))
 		return -EFAULT;
 
-	f = fdget(fd);
+	f = fdget_pos(fd);
 	if (!f.file)
 		return -EBADF;
 
@@ -1077,7 +1077,7 @@ COMPAT_SYSCALL_DEFINE3(getdents64, unsigned int, fd,
 		else
 			error = count - buf.count;
 	}
-	fdput(f);
+	fdput_pos(f);
 	return error;
 }
 #endif /* __ARCH_WANT_COMPAT_SYS_GETDENTS64 */
diff --git a/fs/file.c b/fs/file.c
index 1fbc5c0555a9..6b1acdfe59da 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -784,6 +784,11 @@ unsigned long __fdget_pos(unsigned int fd)
 	return v;
 }
 
+void __f_unlock_pos(struct file *f)
+{
+	mutex_unlock(&f->f_pos_lock);
+}
+
 /*
  * We only lock f_pos if we have threads or if the file might be
  * shared with another process. In both cases we'll have an elevated
diff --git a/fs/open.c b/fs/open.c
index 17cb6b1dab75..938a658a5c6d 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -713,7 +713,7 @@ static int do_dentry_open(struct file *f,
 	}
 
 	/* POSIX.1-2008/SUSv4 Section XSI 2.9.7 */
-	if (S_ISREG(inode->i_mode))
+	if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode))
 		f->f_mode |= FMODE_ATOMIC_POS;
 
 	f->f_op = fops_get(inode->i_fop);
diff --git a/fs/read_write.c b/fs/read_write.c
index cf377cf9dfe3..69c7c3c2955c 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -302,18 +302,6 @@ loff_t vfs_llseek(struct file *file, loff_t offset, int whence)
 }
 EXPORT_SYMBOL(vfs_llseek);
 
-static inline struct fd fdget_pos(int fd)
-{
-	return __to_fd(__fdget_pos(fd));
-}
-
-static inline void fdput_pos(struct fd f)
-{
-	if (f.flags & FDPUT_POS_UNLOCK)
-		mutex_unlock(&f.file->f_pos_lock);
-	fdput(f);
-}
-
 SYSCALL_DEFINE3(lseek, unsigned int, fd, off_t, offset, unsigned int, whence)
 {
 	off_t retval;
diff --git a/fs/readdir.c b/fs/readdir.c
index bf583e848a1a..d7308b8f6cf7 100644
--- a/fs/readdir.c
+++ b/fs/readdir.c
@@ -112,7 +112,7 @@ SYSCALL_DEFINE3(old_readdir, unsigned int, fd,
 		struct old_linux_dirent __user *, dirent, unsigned int, count)
 {
 	int error;
-	struct fd f = fdget(fd);
+	struct fd f = fdget_pos(fd);
 	struct readdir_callback buf = {
 		.ctx.actor = fillonedir,
 		.dirent = dirent
@@ -125,7 +125,7 @@ SYSCALL_DEFINE3(old_readdir, unsigned int, fd,
 	if (buf.result)
 		error = buf.result;
 
-	fdput(f);
+	fdput_pos(f);
 	return error;
 }
 
@@ -209,7 +209,7 @@ SYSCALL_DEFINE3(getdents, unsigned int, fd,
 	if (!access_ok(VERIFY_WRITE, dirent, count))
 		return -EFAULT;
 
-	f = fdget(fd);
+	f = fdget_pos(fd);
 	if (!f.file)
 		return -EBADF;
 
@@ -223,7 +223,7 @@ SYSCALL_DEFINE3(getdents, unsigned int, fd,
 		else
 			error = count - buf.count;
 	}
-	fdput(f);
+	fdput_pos(f);
 	return error;
 }
 
@@ -290,7 +290,7 @@ SYSCALL_DEFINE3(getdents64, unsigned int, fd,
 	if (!access_ok(VERIFY_WRITE, dirent, count))
 		return -EFAULT;
 
-	f = fdget(fd);
+	f = fdget_pos(fd);
 	if (!f.file)
 		return -EBADF;
 
@@ -305,6 +305,6 @@ SYSCALL_DEFINE3(getdents64, unsigned int, fd,
 		else
 			error = count - buf.count;
 	}
-	fdput(f);
+	fdput_pos(f);
 	return error;
 }
diff --git a/include/linux/file.h b/include/linux/file.h
index f87d30882a24..7444f5feda12 100644
--- a/include/linux/file.h
+++ b/include/linux/file.h
@@ -44,6 +44,7 @@ extern struct file *fget_raw(unsigned int fd);
 extern unsigned long __fdget(unsigned int fd);
 extern unsigned long __fdget_raw(unsigned int fd);
 extern unsigned long __fdget_pos(unsigned int fd);
+extern void __f_unlock_pos(struct file *);
 
 static inline struct fd __to_fd(unsigned long v)
 {
@@ -60,6 +61,18 @@ static inline struct fd fdget_raw(unsigned int fd)
 	return __to_fd(__fdget_raw(fd));
 }
 
+static inline struct fd fdget_pos(int fd)
+{
+	return __to_fd(__fdget_pos(fd));
+}
+
+static inline void fdput_pos(struct fd f)
+{
+	if (f.flags & FDPUT_POS_UNLOCK)
+		__f_unlock_pos(f.file);
+	fdput(f);
+}
+
 extern int f_dupfd(unsigned int from, struct file *file, unsigned flags);
 extern int replace_fd(unsigned fd, struct file *file, unsigned flags);
 extern void set_close_on_exec(unsigned int fd, int flag);
-- 
cgit v1.2.3


From 6192269444ebfbfb42e23c7a6a93c76ffe4b5e51 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 20 Apr 2016 23:08:32 -0400
Subject: introduce a parallel variant of ->iterate()

New method: ->iterate_shared().  Same arguments as in ->iterate(),
called with the directory locked only shared.  Once all filesystems
switch, the old one will be gone.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 Documentation/filesystems/porting | 18 ++++++++++++++++++
 fs/coda/dir.c                     | 18 ++++++++++++------
 fs/exportfs/expfs.c               |  2 +-
 fs/readdir.c                      | 20 ++++++++++++++++----
 include/linux/fs.h                |  1 +
 5 files changed, 48 insertions(+), 11 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/filesystems/porting b/Documentation/filesystems/porting
index 1567a53857bd..12c57abdaac9 100644
--- a/Documentation/filesystems/porting
+++ b/Documentation/filesystems/porting
@@ -557,3 +557,21 @@ in your dentry operations instead.
 	will not happen in parallel ("same" in the sense of your ->d_compare()).
 	Lookups on different names in the same directory can and do happen in
 	parallel now.
+--
+[recommended]
+	->iterate_shared() is added; it's a parallel variant of ->iterate().
+	Exclusion on struct file level is still provided (as well as that
+	between it and lseek on the same struct file), but if your directory
+	has been opened several times, you can get these called in parallel.
+	Exclusion between that method and all directory-modifying ones is
+	still provided, of course.
+
+	Often enough ->iterate() can serve as ->iterate_shared() without any
+	changes - it is a read-only operation, after all.  If you have any
+	per-inode or per-dentry in-core data structures modified by ->iterate(),
+	you might need something to serialize the access to them.  If you
+	do dcache pre-seeding, you'll need to switch to d_alloc_parallel() for
+	that; look for in-tree examples.
+
+	Old method is only used if the new one is absent; eventually it will
+	be removed.  Switch while you still can; the old one won't stay.
diff --git a/fs/coda/dir.c b/fs/coda/dir.c
index 42e731b8c80a..6fb8672c0892 100644
--- a/fs/coda/dir.c
+++ b/fs/coda/dir.c
@@ -424,16 +424,22 @@ static int coda_readdir(struct file *coda_file, struct dir_context *ctx)
 	BUG_ON(!cfi || cfi->cfi_magic != CODA_MAGIC);
 	host_file = cfi->cfi_container;
 
-	if (host_file->f_op->iterate) {
+	if (host_file->f_op->iterate || host_file->f_op->iterate_shared) {
 		struct inode *host_inode = file_inode(host_file);
-
-		inode_lock(host_inode);
 		ret = -ENOENT;
 		if (!IS_DEADDIR(host_inode)) {
-			ret = host_file->f_op->iterate(host_file, ctx);
-			file_accessed(host_file);
+			if (host_file->f_op->iterate_shared) {
+				inode_lock_shared(host_inode);
+				ret = host_file->f_op->iterate_shared(host_file, ctx);
+				file_accessed(host_file);
+				inode_unlock_shared(host_inode);
+			} else {
+				inode_lock(host_inode);
+				ret = host_file->f_op->iterate(host_file, ctx);
+				file_accessed(host_file);
+				inode_unlock(host_inode);
+			}
 		}
-		inode_unlock(host_inode);
 		return ret;
 	}
 	/* Venus: we must read Venus dirents from a file */
diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c
index 402c5caab5ca..207ba8d627ca 100644
--- a/fs/exportfs/expfs.c
+++ b/fs/exportfs/expfs.c
@@ -312,7 +312,7 @@ static int get_name(const struct path *path, char *name, struct dentry *child)
 		goto out;
 
 	error = -EINVAL;
-	if (!file->f_op->iterate)
+	if (!file->f_op->iterate && !file->f_op->iterate_shared)
 		goto out_close;
 
 	buffer.sequence = 0;
diff --git a/fs/readdir.c b/fs/readdir.c
index d7308b8f6cf7..a86c6c04b9bc 100644
--- a/fs/readdir.c
+++ b/fs/readdir.c
@@ -24,15 +24,21 @@
 int iterate_dir(struct file *file, struct dir_context *ctx)
 {
 	struct inode *inode = file_inode(file);
+	bool shared = false;
 	int res = -ENOTDIR;
-	if (!file->f_op->iterate)
+	if (file->f_op->iterate_shared)
+		shared = true;
+	else if (!file->f_op->iterate)
 		goto out;
 
 	res = security_file_permission(file, MAY_READ);
 	if (res)
 		goto out;
 
-	inode_lock(inode);
+	if (shared)
+		inode_lock_shared(inode);
+	else
+		inode_lock(inode);
 	// res = mutex_lock_killable(&inode->i_mutex);
 	// if (res)
 	//	goto out;
@@ -40,12 +46,18 @@ int iterate_dir(struct file *file, struct dir_context *ctx)
 	res = -ENOENT;
 	if (!IS_DEADDIR(inode)) {
 		ctx->pos = file->f_pos;
-		res = file->f_op->iterate(file, ctx);
+		if (shared)
+			res = file->f_op->iterate_shared(file, ctx);
+		else
+			res = file->f_op->iterate(file, ctx);
 		file->f_pos = ctx->pos;
 		fsnotify_access(file);
 		file_accessed(file);
 	}
-	inode_unlock(inode);
+	if (shared)
+		inode_unlock_shared(inode);
+	else
+		inode_unlock(inode);
 out:
 	return res;
 }
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 3018f31f7aa0..3dc0258a2b64 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1674,6 +1674,7 @@ struct file_operations {
 	ssize_t (*read_iter) (struct kiocb *, struct iov_iter *);
 	ssize_t (*write_iter) (struct kiocb *, struct iov_iter *);
 	int (*iterate) (struct file *, struct dir_context *);
+	int (*iterate_shared) (struct file *, struct dir_context *);
 	unsigned int (*poll) (struct file *, struct poll_table_struct *);
 	long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long);
 	long (*compat_ioctl) (struct file *, unsigned int, unsigned long);
-- 
cgit v1.2.3


From 884be175351e73c515303118150f195dd611787c Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 28 Apr 2016 23:56:31 -0400
Subject: nfs: per-name sillyunlink exclusion

use d_alloc_parallel() for sillyunlink/lookup exclusion and
explicit rwsem (nfs_rmdir() being a writer and nfs_call_unlink() -
a reader) for rmdir/sillyunlink one.

That ought to make lookup/readdir/!O_CREAT atomic_open really
parallel on NFS.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/dir.c            |   9 +--
 fs/nfs/inode.c          |   4 +-
 fs/nfs/nfs4proc.c       |   2 +-
 fs/nfs/nfstrace.h       |   2 +-
 fs/nfs/unlink.c         | 193 +++++++++++++-----------------------------------
 include/linux/nfs_fs.h  |  11 +--
 include/linux/nfs_xdr.h |   4 +-
 7 files changed, 61 insertions(+), 164 deletions(-)

(limited to 'include/linux')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index cd3c754e3f89..aaf7bd0cbae2 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -909,7 +909,6 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx)
 	desc->decode = NFS_PROTO(inode)->decode_dirent;
 	desc->plus = nfs_use_readdirplus(inode, ctx) ? 1 : 0;
 
-	nfs_block_sillyrename(dentry);
 	if (ctx->pos == 0 || nfs_dir_mapping_need_revalidate(inode))
 		res = nfs_revalidate_mapping(inode, file->f_mapping);
 	if (res < 0)
@@ -945,7 +944,6 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx)
 			break;
 	} while (!desc->eof);
 out:
-	nfs_unblock_sillyrename(dentry);
 	if (res > 0)
 		res = 0;
 	dfprintk(FILE, "NFS: readdir(%pD2) returns %d\n", file, res);
@@ -1398,7 +1396,6 @@ struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, unsigned in
 	parent = dentry->d_parent;
 	/* Protect against concurrent sillydeletes */
 	trace_nfs_lookup_enter(dir, dentry, flags);
-	nfs_block_sillyrename(parent);
 	error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, fhandle, fattr, label);
 	if (error == -ENOENT)
 		goto no_entry;
@@ -1423,7 +1420,6 @@ no_entry:
 	}
 	nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
 out_unblock_sillyrename:
-	nfs_unblock_sillyrename(parent);
 	trace_nfs_lookup_exit(dir, dentry, flags, error);
 	nfs4_label_free(label);
 out:
@@ -1535,9 +1531,7 @@ int nfs_atomic_open(struct inode *dir, struct dentry *dentry,
 		goto out;
 
 	trace_nfs_atomic_open_enter(dir, ctx, open_flags);
-	nfs_block_sillyrename(dentry->d_parent);
 	inode = NFS_PROTO(dir)->open_context(dir, ctx, open_flags, &attr, opened);
-	nfs_unblock_sillyrename(dentry->d_parent);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		trace_nfs_atomic_open_exit(dir, ctx, open_flags, err);
@@ -1781,7 +1775,7 @@ int nfs_rmdir(struct inode *dir, struct dentry *dentry)
 
 	trace_nfs_rmdir_enter(dir, dentry);
 	if (d_really_is_positive(dentry)) {
-		nfs_wait_on_sillyrename(dentry);
+		down_write(&NFS_I(d_inode(dentry))->rmdir_sem);
 		error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name);
 		/* Ensure the VFS deletes this inode */
 		switch (error) {
@@ -1791,6 +1785,7 @@ int nfs_rmdir(struct inode *dir, struct dentry *dentry)
 		case -ENOENT:
 			nfs_dentry_handle_enoent(dentry);
 		}
+		up_write(&NFS_I(d_inode(dentry))->rmdir_sem);
 	} else
 		error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name);
 	trace_nfs_rmdir_exit(dir, dentry, error);
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 738c84a42eb0..52e7d6869e3b 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1958,9 +1958,7 @@ static void init_once(void *foo)
 	nfsi->nrequests = 0;
 	nfsi->commit_info.ncommit = 0;
 	atomic_set(&nfsi->commit_info.rpcs_out, 0);
-	atomic_set(&nfsi->silly_count, 1);
-	INIT_HLIST_HEAD(&nfsi->silly_list);
-	init_waitqueue_head(&nfsi->waitqueue);
+	init_rwsem(&nfsi->rmdir_sem);
 	nfs4_init_once(nfsi);
 }
 
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 7a7ac1dafa02..084e8570da18 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3777,7 +3777,7 @@ static void nfs4_proc_unlink_setup(struct rpc_message *msg, struct inode *dir)
 
 static void nfs4_proc_unlink_rpc_prepare(struct rpc_task *task, struct nfs_unlinkdata *data)
 {
-	nfs4_setup_sequence(NFS_SERVER(data->dir),
+	nfs4_setup_sequence(NFS_SB(data->dentry->d_sb),
 			&data->args.seq_args,
 			&data->res.seq_res,
 			task);
diff --git a/fs/nfs/nfstrace.h b/fs/nfs/nfstrace.h
index 9f80a086b612..0b9e5cc9a747 100644
--- a/fs/nfs/nfstrace.h
+++ b/fs/nfs/nfstrace.h
@@ -702,7 +702,7 @@ TRACE_EVENT(nfs_sillyrename_unlink,
 		),
 
 		TP_fast_assign(
-			struct inode *dir = data->dir;
+			struct inode *dir = d_inode(data->dentry->d_parent);
 			size_t len = data->args.name.len;
 			__entry->dev = dir->i_sb->s_dev;
 			__entry->dir = NFS_FILEID(dir);
diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c
index 7d6deaccf89b..1868246f56e6 100644
--- a/fs/nfs/unlink.c
+++ b/fs/nfs/unlink.c
@@ -30,45 +30,11 @@
 static void
 nfs_free_unlinkdata(struct nfs_unlinkdata *data)
 {
-	iput(data->dir);
 	put_rpccred(data->cred);
 	kfree(data->args.name.name);
 	kfree(data);
 }
 
-#define NAME_ALLOC_LEN(len)	((len+16) & ~15)
-/**
- * nfs_copy_dname - copy dentry name to data structure
- * @dentry: pointer to dentry
- * @data: nfs_unlinkdata
- */
-static int nfs_copy_dname(struct dentry *dentry, struct nfs_unlinkdata *data)
-{
-	char		*str;
-	int		len = dentry->d_name.len;
-
-	str = kmemdup(dentry->d_name.name, NAME_ALLOC_LEN(len), GFP_KERNEL);
-	if (!str)
-		return -ENOMEM;
-	data->args.name.len = len;
-	data->args.name.name = str;
-	return 0;
-}
-
-static void nfs_free_dname(struct nfs_unlinkdata *data)
-{
-	kfree(data->args.name.name);
-	data->args.name.name = NULL;
-	data->args.name.len = 0;
-}
-
-static void nfs_dec_sillycount(struct inode *dir)
-{
-	struct nfs_inode *nfsi = NFS_I(dir);
-	if (atomic_dec_return(&nfsi->silly_count) == 1)
-		wake_up(&nfsi->waitqueue);
-}
-
 /**
  * nfs_async_unlink_done - Sillydelete post-processing
  * @task: rpc_task of the sillydelete
@@ -78,7 +44,7 @@ static void nfs_dec_sillycount(struct inode *dir)
 static void nfs_async_unlink_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs_unlinkdata *data = calldata;
-	struct inode *dir = data->dir;
+	struct inode *dir = d_inode(data->dentry->d_parent);
 
 	trace_nfs_sillyrename_unlink(data, task->tk_status);
 	if (!NFS_PROTO(dir)->unlink_done(task, dir))
@@ -95,17 +61,21 @@ static void nfs_async_unlink_done(struct rpc_task *task, void *calldata)
 static void nfs_async_unlink_release(void *calldata)
 {
 	struct nfs_unlinkdata	*data = calldata;
-	struct super_block *sb = data->dir->i_sb;
+	struct dentry *dentry = data->dentry;
+	struct super_block *sb = dentry->d_sb;
 
-	nfs_dec_sillycount(data->dir);
+	up_read_non_owner(&NFS_I(d_inode(dentry->d_parent))->rmdir_sem);
+	d_lookup_done(dentry);
 	nfs_free_unlinkdata(data);
+	dput(dentry);
 	nfs_sb_deactive(sb);
 }
 
 static void nfs_unlink_prepare(struct rpc_task *task, void *calldata)
 {
 	struct nfs_unlinkdata *data = calldata;
-	NFS_PROTO(data->dir)->unlink_rpc_prepare(task, data);
+	struct inode *dir = d_inode(data->dentry->d_parent);
+	NFS_PROTO(dir)->unlink_rpc_prepare(task, data);
 }
 
 static const struct rpc_call_ops nfs_unlink_ops = {
@@ -114,7 +84,7 @@ static const struct rpc_call_ops nfs_unlink_ops = {
 	.rpc_call_prepare = nfs_unlink_prepare,
 };
 
-static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct nfs_unlinkdata *data)
+static void nfs_do_call_unlink(struct nfs_unlinkdata *data)
 {
 	struct rpc_message msg = {
 		.rpc_argp = &data->args,
@@ -129,10 +99,31 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 		.flags = RPC_TASK_ASYNC,
 	};
 	struct rpc_task *task;
+	struct inode *dir = d_inode(data->dentry->d_parent);
+	nfs_sb_active(dir->i_sb);
+	data->args.fh = NFS_FH(dir);
+	nfs_fattr_init(data->res.dir_attr);
+
+	NFS_PROTO(dir)->unlink_setup(&msg, dir);
+
+	task_setup_data.rpc_client = NFS_CLIENT(dir);
+	task = rpc_run_task(&task_setup_data);
+	if (!IS_ERR(task))
+		rpc_put_task_async(task);
+}
+
+static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data)
+{
+	struct inode *dir = d_inode(dentry->d_parent);
 	struct dentry *alias;
 
-	alias = d_lookup(parent, &data->args.name);
-	if (alias != NULL) {
+	down_read_non_owner(&NFS_I(dir)->rmdir_sem);
+	alias = d_alloc_parallel(dentry->d_parent, &data->args.name, &data->wq);
+	if (IS_ERR(alias)) {
+		up_read_non_owner(&NFS_I(dir)->rmdir_sem);
+		return 0;
+	}
+	if (!d_in_lookup(alias)) {
 		int ret;
 		void *devname_garbage = NULL;
 
@@ -140,10 +131,8 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 		 * Hey, we raced with lookup... See if we need to transfer
 		 * the sillyrename information to the aliased dentry.
 		 */
-		nfs_free_dname(data);
-		ret = nfs_copy_dname(alias, data);
 		spin_lock(&alias->d_lock);
-		if (ret == 0 && d_really_is_positive(alias) &&
+		if (d_really_is_positive(alias) &&
 		    !(alias->d_flags & DCACHE_NFSFS_RENAMED)) {
 			devname_garbage = alias->d_fsdata;
 			alias->d_fsdata = data;
@@ -152,8 +141,8 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 		} else
 			ret = 0;
 		spin_unlock(&alias->d_lock);
-		nfs_dec_sillycount(dir);
 		dput(alias);
+		up_read_non_owner(&NFS_I(dir)->rmdir_sem);
 		/*
 		 * If we'd displaced old cached devname, free it.  At that
 		 * point dentry is definitely not a root, so we won't need
@@ -162,95 +151,18 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 		kfree(devname_garbage);
 		return ret;
 	}
-	data->dir = igrab(dir);
-	if (!data->dir) {
-		nfs_dec_sillycount(dir);
-		return 0;
-	}
-	nfs_sb_active(dir->i_sb);
-	data->args.fh = NFS_FH(dir);
-	nfs_fattr_init(data->res.dir_attr);
-
-	NFS_PROTO(dir)->unlink_setup(&msg, dir);
-
-	task_setup_data.rpc_client = NFS_CLIENT(dir);
-	task = rpc_run_task(&task_setup_data);
-	if (!IS_ERR(task))
-		rpc_put_task_async(task);
+	data->dentry = alias;
+	nfs_do_call_unlink(data);
 	return 1;
 }
 
-static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data)
-{
-	struct dentry *parent;
-	struct inode *dir;
-	int ret = 0;
-
-
-	parent = dget_parent(dentry);
-	if (parent == NULL)
-		goto out_free;
-	dir = d_inode(parent);
-	/* Non-exclusive lock protects against concurrent lookup() calls */
-	spin_lock(&dir->i_lock);
-	if (atomic_inc_not_zero(&NFS_I(dir)->silly_count) == 0) {
-		/* Deferred delete */
-		hlist_add_head(&data->list, &NFS_I(dir)->silly_list);
-		spin_unlock(&dir->i_lock);
-		ret = 1;
-		goto out_dput;
-	}
-	spin_unlock(&dir->i_lock);
-	ret = nfs_do_call_unlink(parent, dir, data);
-out_dput:
-	dput(parent);
-out_free:
-	return ret;
-}
-
-void nfs_wait_on_sillyrename(struct dentry *dentry)
-{
-	struct nfs_inode *nfsi = NFS_I(d_inode(dentry));
-
-	wait_event(nfsi->waitqueue, atomic_read(&nfsi->silly_count) <= 1);
-}
-
-void nfs_block_sillyrename(struct dentry *dentry)
-{
-	struct nfs_inode *nfsi = NFS_I(d_inode(dentry));
-
-	wait_event(nfsi->waitqueue, atomic_cmpxchg(&nfsi->silly_count, 1, 0) == 1);
-}
-
-void nfs_unblock_sillyrename(struct dentry *dentry)
-{
-	struct inode *dir = d_inode(dentry);
-	struct nfs_inode *nfsi = NFS_I(dir);
-	struct nfs_unlinkdata *data;
-
-	atomic_inc(&nfsi->silly_count);
-	wake_up(&nfsi->waitqueue);
-	spin_lock(&dir->i_lock);
-	while (!hlist_empty(&nfsi->silly_list)) {
-		if (!atomic_inc_not_zero(&nfsi->silly_count))
-			break;
-		data = hlist_entry(nfsi->silly_list.first, struct nfs_unlinkdata, list);
-		hlist_del(&data->list);
-		spin_unlock(&dir->i_lock);
-		if (nfs_do_call_unlink(dentry, dir, data) == 0)
-			nfs_free_unlinkdata(data);
-		spin_lock(&dir->i_lock);
-	}
-	spin_unlock(&dir->i_lock);
-}
-
 /**
  * nfs_async_unlink - asynchronous unlinking of a file
  * @dir: parent directory of dentry
  * @dentry: dentry to unlink
  */
 static int
-nfs_async_unlink(struct inode *dir, struct dentry *dentry)
+nfs_async_unlink(struct dentry *dentry, struct qstr *name)
 {
 	struct nfs_unlinkdata *data;
 	int status = -ENOMEM;
@@ -259,13 +171,18 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry)
 	data = kzalloc(sizeof(*data), GFP_KERNEL);
 	if (data == NULL)
 		goto out;
+	data->args.name.name = kstrdup(name->name, GFP_KERNEL);
+	if (!data->args.name.name)
+		goto out_free;
+	data->args.name.len = name->len;
 
 	data->cred = rpc_lookup_cred();
 	if (IS_ERR(data->cred)) {
 		status = PTR_ERR(data->cred);
-		goto out_free;
+		goto out_free_name;
 	}
 	data->res.dir_attr = &data->dir_attr;
+	init_waitqueue_head(&data->wq);
 
 	status = -EBUSY;
 	spin_lock(&dentry->d_lock);
@@ -285,6 +202,8 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry)
 out_unlock:
 	spin_unlock(&dentry->d_lock);
 	put_rpccred(data->cred);
+out_free_name:
+	kfree(data->args.name.name);
 out_free:
 	kfree(data);
 out:
@@ -303,17 +222,15 @@ out:
 void
 nfs_complete_unlink(struct dentry *dentry, struct inode *inode)
 {
-	struct nfs_unlinkdata	*data = NULL;
+	struct nfs_unlinkdata	*data;
 
 	spin_lock(&dentry->d_lock);
-	if (dentry->d_flags & DCACHE_NFSFS_RENAMED) {
-		dentry->d_flags &= ~DCACHE_NFSFS_RENAMED;
-		data = dentry->d_fsdata;
-		dentry->d_fsdata = NULL;
-	}
+	dentry->d_flags &= ~DCACHE_NFSFS_RENAMED;
+	data = dentry->d_fsdata;
+	dentry->d_fsdata = NULL;
 	spin_unlock(&dentry->d_lock);
 
-	if (data != NULL && (NFS_STALE(inode) || !nfs_call_unlink(dentry, data)))
+	if (NFS_STALE(inode) || !nfs_call_unlink(dentry, data))
 		nfs_free_unlinkdata(data);
 }
 
@@ -560,18 +477,10 @@ nfs_sillyrename(struct inode *dir, struct dentry *dentry)
 	/* queue unlink first. Can't do this from rpc_release as it
 	 * has to allocate memory
 	 */
-	error = nfs_async_unlink(dir, dentry);
+	error = nfs_async_unlink(dentry, &sdentry->d_name);
 	if (error)
 		goto out_dput;
 
-	/* populate unlinkdata with the right dname */
-	error = nfs_copy_dname(sdentry,
-				(struct nfs_unlinkdata *)dentry->d_fsdata);
-	if (error) {
-		nfs_cancel_async_unlink(dentry);
-		goto out_dput;
-	}
-
 	/* run the rename task, undo unlink if it fails */
 	task = nfs_async_rename(dir, dir, dentry, sdentry,
 					nfs_complete_sillyrename);
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 67300f8e5f2f..fa167f25465d 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -163,11 +163,9 @@ struct nfs_inode {
 	/* Open contexts for shared mmap writes */
 	struct list_head	open_files;
 
-	/* Number of in-flight sillydelete RPC calls */
-	atomic_t		silly_count;
-	/* List of deferred sillydelete requests */
-	struct hlist_head	silly_list;
-	wait_queue_head_t	waitqueue;
+	/* Readers: in-flight sillydelete RPC calls */
+	/* Writers: rmdir */
+	struct rw_semaphore	rmdir_sem;
 
 #if IS_ENABLED(CONFIG_NFS_V4)
 	struct nfs4_cached_acl	*nfs4_acl;
@@ -492,9 +490,6 @@ extern void nfs_release_automount_timer(void);
  * linux/fs/nfs/unlink.c
  */
 extern void nfs_complete_unlink(struct dentry *dentry, struct inode *);
-extern void nfs_wait_on_sillyrename(struct dentry *dentry);
-extern void nfs_block_sillyrename(struct dentry *dentry);
-extern void nfs_unblock_sillyrename(struct dentry *dentry);
 
 /*
  * linux/fs/nfs/write.c
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index d320906cf13e..ee8491dadbf3 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -1468,10 +1468,10 @@ struct nfs_pgio_completion_ops {
 };
 
 struct nfs_unlinkdata {
-	struct hlist_node list;
 	struct nfs_removeargs args;
 	struct nfs_removeres res;
-	struct inode *dir;
+	struct dentry *dentry;
+	wait_queue_head_t wq;
 	struct rpc_cred	*cred;
 	struct nfs_fattr dir_attr;
 	long timeout;
-- 
cgit v1.2.3


From ae05327a00fd47c34dfe25294b359a3f3fef96e8 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 12 May 2016 20:36:01 -0400
Subject: ext4: switch to ->iterate_shared()

Note that we need relax_dir() equivalent for directories
locked shared.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ext4/dir.c      | 4 ++--
 include/linux/fs.h | 7 +++++++
 2 files changed, 9 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
index 561d7308b393..5d00bf060254 100644
--- a/fs/ext4/dir.c
+++ b/fs/ext4/dir.c
@@ -266,7 +266,7 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx)
 			ctx->pos += ext4_rec_len_from_disk(de->rec_len,
 						sb->s_blocksize);
 		}
-		if ((ctx->pos < inode->i_size) && !dir_relax(inode))
+		if ((ctx->pos < inode->i_size) && !dir_relax_shared(inode))
 			goto done;
 		brelse(bh);
 		bh = NULL;
@@ -644,7 +644,7 @@ int ext4_check_all_de(struct inode *dir, struct buffer_head *bh, void *buf,
 const struct file_operations ext4_dir_operations = {
 	.llseek		= ext4_dir_llseek,
 	.read		= generic_read_dir,
-	.iterate	= ext4_readdir,
+	.iterate_shared	= ext4_readdir,
 	.unlocked_ioctl = ext4_ioctl,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= ext4_compat_ioctl,
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 3dc0258a2b64..e87245ac6941 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -3134,6 +3134,13 @@ static inline bool dir_relax(struct inode *inode)
 	return !IS_DEADDIR(inode);
 }
 
+static inline bool dir_relax_shared(struct inode *inode)
+{
+	inode_unlock_shared(inode);
+	inode_lock_shared(inode);
+	return !IS_DEADDIR(inode);
+}
+
 extern bool path_noexec(const struct path *path);
 extern void inode_nohighmem(struct inode *inode);
 
-- 
cgit v1.2.3