From 60545d0d4610b02e55f65d141c95b18ccf855b6e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 7 Jun 2013 01:20:27 -0400 Subject: [O_TMPFILE] it's still short a few helpers, but infrastructure should be OK now... Signed-off-by: Al Viro --- include/linux/dcache.h | 2 ++ include/linux/fs.h | 1 + include/uapi/asm-generic/fcntl.h | 4 ++++ 3 files changed, 7 insertions(+) (limited to 'include') diff --git a/include/linux/dcache.h b/include/linux/dcache.h index 1a6bb81f0fe5..86da7595ba31 100644 --- a/include/linux/dcache.h +++ b/include/linux/dcache.h @@ -246,6 +246,8 @@ extern struct dentry * d_make_root(struct inode *); /* - the ramfs-type tree */ extern void d_genocide(struct dentry *); +extern void d_tmpfile(struct dentry *, struct inode *); + extern struct dentry *d_find_alias(struct inode *); extern void d_prune_aliases(struct inode *); diff --git a/include/linux/fs.h b/include/linux/fs.h index 7c30e3a62baf..dd6615f0fd13 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1580,6 +1580,7 @@ struct inode_operations { int (*atomic_open)(struct inode *, struct dentry *, struct file *, unsigned open_flag, umode_t create_mode, int *opened); + int (*tmpfile) (struct inode *, struct dentry *, umode_t); } ____cacheline_aligned; ssize_t rw_copy_check_uvector(int type, const struct iovec __user * uvector, diff --git a/include/uapi/asm-generic/fcntl.h b/include/uapi/asm-generic/fcntl.h index a48937d4a5ea..06632beaa6d5 100644 --- a/include/uapi/asm-generic/fcntl.h +++ b/include/uapi/asm-generic/fcntl.h @@ -84,6 +84,10 @@ #define O_PATH 010000000 #endif +#ifndef O_TMPFILE +#define O_TMPFILE 020000000 +#endif + #ifndef O_NDELAY #define O_NDELAY O_NONBLOCK #endif -- cgit v1.2.3 From f4e0c30c191f87851c4a53454abb55ee276f4a7e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 11 Jun 2013 08:34:36 +0400 Subject: allow the temp files created by open() to be linked to O_TMPFILE | O_CREAT => linkat() with AT_SYMLINK_FOLLOW and /proc/self/fd/ as oldpath (i.e. flink()) will create a link O_TMPFILE | O_CREAT | O_EXCL => ENOENT on attempt to link those guys Signed-off-by: Al Viro --- fs/inode.c | 4 +++- fs/namei.c | 16 ++++++++++++++-- include/linux/fs.h | 1 + 3 files changed, 18 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/inode.c b/fs/inode.c index 00d5fc3b86e1..d6dfb09c8280 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -333,8 +333,10 @@ EXPORT_SYMBOL(set_nlink); */ void inc_nlink(struct inode *inode) { - if (WARN_ON(inode->i_nlink == 0)) + if (unlikely(inode->i_nlink == 0)) { + WARN_ON(!(inode->i_state & I_LINKABLE)); atomic_long_dec(&inode->i_sb->s_remove_count); + } inode->__i_nlink++; } diff --git a/fs/namei.c b/fs/namei.c index 778e253e3d48..66998b06d822 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2948,8 +2948,14 @@ static int do_tmpfile(int dfd, struct filename *pathname, if (error) goto out2; error = open_check_o_direct(file); - if (error) + if (error) { fput(file); + } else if (!(op->open_flag & O_EXCL)) { + struct inode *inode = file_inode(file); + spin_lock(&inode->i_lock); + inode->i_state |= I_LINKABLE; + spin_unlock(&inode->i_lock); + } out2: mnt_drop_write(nd->path.mnt); out: @@ -3628,12 +3634,18 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de mutex_lock(&inode->i_mutex); /* Make sure we don't allow creating hardlink to an unlinked file */ - if (inode->i_nlink == 0) + if (inode->i_nlink == 0 && !(inode->i_state & I_LINKABLE)) error = -ENOENT; else if (max_links && inode->i_nlink >= max_links) error = -EMLINK; else error = dir->i_op->link(old_dentry, dir, new_dentry); + + if (!error && (inode->i_state & I_LINKABLE)) { + spin_lock(&inode->i_lock); + inode->i_state &= ~I_LINKABLE; + spin_unlock(&inode->i_lock); + } mutex_unlock(&inode->i_mutex); if (!error) fsnotify_link(dir, inode, new_dentry); diff --git a/include/linux/fs.h b/include/linux/fs.h index dd6615f0fd13..ab11c44b0697 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1744,6 +1744,7 @@ struct super_operations { #define I_REFERENCED (1 << 8) #define __I_DIO_WAKEUP 9 #define I_DIO_WAKEUP (1 << I_DIO_WAKEUP) +#define I_LINKABLE (1 << 10) #define I_DIRTY (I_DIRTY_SYNC | I_DIRTY_DATASYNC | I_DIRTY_PAGES) -- cgit v1.2.3 From c77cecee52e9b599da1f8ffd9170d4374c99a345 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 13 Jun 2013 23:37:49 +0100 Subject: Replace a bunch of file->dentry->d_inode refs with file_inode() Replace a bunch of file->dentry->d_inode refs with file_inode(). In __fput(), use file->f_inode instead so as not to be affected by any tricks that file_inode() might grow. Signed-off-by: David Howells Signed-off-by: Al Viro --- fs/file_table.c | 2 +- include/linux/fsnotify.h | 8 ++++---- security/integrity/ima/ima_main.c | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/fs/file_table.c b/fs/file_table.c index 485dc0eddd67..08e719b884ca 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -227,7 +227,7 @@ static void __fput(struct file *file) { struct dentry *dentry = file->f_path.dentry; struct vfsmount *mnt = file->f_path.mnt; - struct inode *inode = dentry->d_inode; + struct inode *inode = file->f_inode; might_sleep(); diff --git a/include/linux/fsnotify.h b/include/linux/fsnotify.h index a78680a92dba..1c804b057fb1 100644 --- a/include/linux/fsnotify.h +++ b/include/linux/fsnotify.h @@ -38,7 +38,7 @@ static inline int fsnotify_parent(struct path *path, struct dentry *dentry, __u3 static inline int fsnotify_perm(struct file *file, int mask) { struct path *path = &file->f_path; - struct inode *inode = path->dentry->d_inode; + struct inode *inode = file_inode(file); __u32 fsnotify_mask = 0; int ret; @@ -192,7 +192,7 @@ static inline void fsnotify_mkdir(struct inode *inode, struct dentry *dentry) static inline void fsnotify_access(struct file *file) { struct path *path = &file->f_path; - struct inode *inode = path->dentry->d_inode; + struct inode *inode = file_inode(file); __u32 mask = FS_ACCESS; if (S_ISDIR(inode->i_mode)) @@ -210,7 +210,7 @@ static inline void fsnotify_access(struct file *file) static inline void fsnotify_modify(struct file *file) { struct path *path = &file->f_path; - struct inode *inode = path->dentry->d_inode; + struct inode *inode = file_inode(file); __u32 mask = FS_MODIFY; if (S_ISDIR(inode->i_mode)) @@ -228,7 +228,7 @@ static inline void fsnotify_modify(struct file *file) static inline void fsnotify_open(struct file *file) { struct path *path = &file->f_path; - struct inode *inode = path->dentry->d_inode; + struct inode *inode = file_inode(file); __u32 mask = FS_OPEN; if (S_ISDIR(inode->i_mode)) diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c index 6c491a63128e..e9508d5bbfcf 100644 --- a/security/integrity/ima/ima_main.c +++ b/security/integrity/ima/ima_main.c @@ -57,7 +57,7 @@ __setup("ima_hash=", hash_setup); static void ima_rdwr_violation_check(struct file *file) { struct dentry *dentry = file->f_path.dentry; - struct inode *inode = dentry->d_inode; + struct inode *inode = file_inode(file); fmode_t mode = file->f_mode; int must_measure; bool send_tomtou = false, send_writers = false; -- cgit v1.2.3 From 0b3fca1fd1499f0f5a7486d494f96538f2b7e5b9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 15 Jun 2013 11:37:47 +0400 Subject: kill find_inode_number() the only remaining caller (in ncpfs) is guaranteed to return 0 - we only hit it if we'd just checked that there's no dentry with such name. Signed-off-by: Al Viro --- fs/dcache.c | 29 ----------------------------- fs/ncpfs/dir.c | 2 -- include/linux/fs.h | 1 - 3 files changed, 32 deletions(-) (limited to 'include') diff --git a/fs/dcache.c b/fs/dcache.c index b7f049c31526..b692c7e097c5 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -2984,35 +2984,6 @@ void d_tmpfile(struct dentry *dentry, struct inode *inode) } EXPORT_SYMBOL(d_tmpfile); -/** - * find_inode_number - check for dentry with name - * @dir: directory to check - * @name: Name to find. - * - * Check whether a dentry already exists for the given name, - * and return the inode number if it has an inode. Otherwise - * 0 is returned. - * - * This routine is used to post-process directory listings for - * filesystems using synthetic inode numbers, and is necessary - * to keep getcwd() working. - */ - -ino_t find_inode_number(struct dentry *dir, struct qstr *name) -{ - struct dentry * dentry; - ino_t ino = 0; - - dentry = d_hash_and_lookup(dir, name); - if (!IS_ERR_OR_NULL(dentry)) { - if (dentry->d_inode) - ino = dentry->d_inode->i_ino; - dput(dentry); - } - return ino; -} -EXPORT_SYMBOL(find_inode_number); - static __initdata unsigned long dhash_entries; static int __init set_dhash_entries(char *str) { diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index e5d488530580..3bc105d36f10 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -659,8 +659,6 @@ end_advance: if (!valid) ctl.valid = 0; if (!ctl.filled && (ctl.fpos == ctx->pos)) { - if (!ino) - ino = find_inode_number(dentry, &qname); if (!ino) ino = iunique(dir->i_sb, 2); ctl.filled = !dir_emit(ctx, qname.name, qname.len, diff --git a/include/linux/fs.h b/include/linux/fs.h index ab11c44b0697..1db01c13ddce 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2311,7 +2311,6 @@ extern struct file * open_exec(const char *); /* fs/dcache.c -- generic fs support functions */ extern int is_subdir(struct dentry *, struct dentry *); extern int path_is_under(struct path *, struct path *); -extern ino_t find_inode_number(struct dentry *, struct qstr *); #include -- cgit v1.2.3 From 1bf9d14dff4a2c4de6152c6f751bdaf6896b68bb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 16 Jun 2013 20:27:42 +0400 Subject: new helper: fixed_size_llseek() Signed-off-by: Al Viro --- fs/read_write.c | 20 ++++++++++++++++++++ include/linux/fs.h | 2 ++ 2 files changed, 22 insertions(+) (limited to 'include') diff --git a/fs/read_write.c b/fs/read_write.c index f646c8b565b9..782dfc3acebc 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -144,6 +144,26 @@ loff_t generic_file_llseek(struct file *file, loff_t offset, int whence) } EXPORT_SYMBOL(generic_file_llseek); +/** + * fixed_size_llseek - llseek implementation for fixed-sized devices + * @file: file structure to seek on + * @offset: file offset to seek to + * @whence: type of seek + * @size: size of the file + * + */ +loff_t fixed_size_llseek(struct file *file, loff_t offset, int whence, loff_t size) +{ + switch (whence) { + case SEEK_SET: case SEEK_CUR: case SEEK_END: + return generic_file_llseek_size(file, offset, whence, + size, size); + default: + return -EINVAL; + } +} +EXPORT_SYMBOL(fixed_size_llseek); + /** * noop_llseek - No Operation Performed llseek implementation * @file: file structure to seek on diff --git a/include/linux/fs.h b/include/linux/fs.h index 1db01c13ddce..803b7fa2520a 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2428,6 +2428,8 @@ extern loff_t no_llseek(struct file *file, loff_t offset, int whence); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); extern loff_t generic_file_llseek_size(struct file *file, loff_t offset, int whence, loff_t maxsize, loff_t eof); +extern loff_t fixed_size_llseek(struct file *file, loff_t offset, + int whence, loff_t size); extern int generic_file_open(struct inode * inode, struct file * filp); extern int nonseekable_open(struct inode * inode, struct file * filp); -- cgit v1.2.3 From 68d70d03f8f5bd10a0e7337210b13f536fd4aeb9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 19 Jun 2013 15:26:04 +0400 Subject: constify rw_verify_area() Signed-off-by: Al Viro --- fs/aio.c | 2 ++ fs/internal.h | 1 + fs/read_write.c | 2 +- include/linux/fs.h | 1 - 4 files changed, 4 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/fs/aio.c b/fs/aio.c index 2bbcacf74d0c..a8ecc8313fb0 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -39,6 +39,8 @@ #include #include +#include "internal.h" + #define AIO_RING_MAGIC 0xa10a10a1 #define AIO_RING_COMPAT_FEATURES 1 #define AIO_RING_INCOMPAT_FEATURES 0 diff --git a/fs/internal.h b/fs/internal.h index f6ad34362823..7c5f01cf619d 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -131,6 +131,7 @@ extern struct dentry *__d_alloc(struct super_block *, const struct qstr *); * read_write.c */ extern ssize_t __kernel_write(struct file *, const char *, size_t, loff_t *); +extern int rw_verify_area(int, struct file *, const loff_t *, size_t); /* * splice.c diff --git a/fs/read_write.c b/fs/read_write.c index 782dfc3acebc..fd72b592aa1b 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -316,7 +316,7 @@ out_putf: * them to something that fits in "int" so that others * won't have to do range checks all the time. */ -int rw_verify_area(int read_write, struct file *file, loff_t *ppos, size_t count) +int rw_verify_area(int read_write, struct file *file, const loff_t *ppos, size_t count) { struct inode *inode; loff_t pos; diff --git a/include/linux/fs.h b/include/linux/fs.h index 803b7fa2520a..68f10204ab29 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1898,7 +1898,6 @@ extern int current_umask(void); extern struct kobject *fs_kobj; #define MAX_RW_COUNT (INT_MAX & PAGE_CACHE_MASK) -extern int rw_verify_area(int, struct file *, loff_t *, size_t); #define FLOCK_VERIFY_READ 1 #define FLOCK_VERIFY_WRITE 2 -- cgit v1.2.3 From da53be12bbb4fabbe2e9f6f908de0cf478b5161d Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 21 May 2013 15:22:44 -0700 Subject: Don't pass inode to ->d_hash() and ->d_compare() Instances either don't look at it at all (the majority of cases) or only want it to find the superblock (which can be had as dentry->d_sb). A few cases that want more are actually safe with dentry->d_inode - the only precaution needed is the check that it hadn't been replaced with NULL by rmdir() or by overwriting rename(), which case should be simply treated as cache miss. Signed-off-by: Linus Torvalds Signed-off-by: Al Viro --- Documentation/filesystems/Locking | 6 ++--- Documentation/filesystems/vfs.txt | 19 +++++++--------- fs/adfs/dir.c | 6 ++--- fs/affs/namei.c | 26 +++++++-------------- fs/cifs/dir.c | 9 +++----- fs/dcache.c | 27 ++++++++-------------- fs/efivarfs/super.c | 9 ++++---- fs/fat/namei_msdos.c | 6 ++--- fs/fat/namei_vfat.c | 12 ++++------ fs/gfs2/dentry.c | 3 +-- fs/hfs/hfs_fs.h | 7 ++---- fs/hfs/string.c | 6 ++--- fs/hfsplus/hfsplus_fs.h | 7 ++---- fs/hfsplus/unicode.c | 7 ++---- fs/hpfs/dentry.c | 7 ++---- fs/isofs/inode.c | 48 +++++++++++++-------------------------- fs/isofs/namei.c | 3 +-- fs/jfs/namei.c | 7 ++---- fs/namei.c | 7 +++--- fs/ncpfs/dir.c | 32 +++++++++++++++++++------- fs/proc/proc_sysctl.c | 7 +++--- fs/sysv/namei.c | 3 +-- include/linux/dcache.h | 9 +++----- 23 files changed, 108 insertions(+), 165 deletions(-) (limited to 'include') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index bdd82b2339d9..f94a362f408e 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -11,10 +11,8 @@ be able to use diff(1). prototypes: int (*d_revalidate)(struct dentry *, unsigned int); int (*d_weak_revalidate)(struct dentry *, unsigned int); - int (*d_hash)(const struct dentry *, const struct inode *, - struct qstr *); - int (*d_compare)(const struct dentry *, const struct inode *, - const struct dentry *, const struct inode *, + int (*d_hash)(const struct dentry *, struct qstr *); + int (*d_compare)(const struct dentry *, const struct dentry *, unsigned int, const char *, const struct qstr *); int (*d_delete)(struct dentry *); void (*d_release)(struct dentry *); diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt index 4a35f6614a66..51ba44e3fc40 100644 --- a/Documentation/filesystems/vfs.txt +++ b/Documentation/filesystems/vfs.txt @@ -901,10 +901,8 @@ defined: struct dentry_operations { int (*d_revalidate)(struct dentry *, unsigned int); int (*d_weak_revalidate)(struct dentry *, unsigned int); - int (*d_hash)(const struct dentry *, const struct inode *, - struct qstr *); - int (*d_compare)(const struct dentry *, const struct inode *, - const struct dentry *, const struct inode *, + int (*d_hash)(const struct dentry *, struct qstr *); + int (*d_compare)(const struct dentry *, const struct dentry *, unsigned int, const char *, const struct qstr *); int (*d_delete)(const struct dentry *); void (*d_release)(struct dentry *); @@ -949,25 +947,24 @@ struct dentry_operations { d_hash: called when the VFS adds a dentry to the hash table. The first dentry passed to d_hash is the parent directory that the name is - to be hashed into. The inode is the dentry's inode. + to be hashed into. Same locking and synchronisation rules as d_compare regarding what is safe to dereference etc. d_compare: called to compare a dentry name with a given name. The first dentry is the parent of the dentry to be compared, the second is - the parent's inode, then the dentry and inode (may be NULL) of the - child dentry. len and name string are properties of the dentry to be - compared. qstr is the name to compare it with. + the child dentry. len and name string are properties of the dentry + to be compared. qstr is the name to compare it with. Must be constant and idempotent, and should not take locks if - possible, and should not or store into the dentry or inodes. - Should not dereference pointers outside the dentry or inodes without + possible, and should not or store into the dentry. + Should not dereference pointers outside the dentry without lots of care (eg. d_parent, d_inode, d_name should not be used). However, our vfsmount is pinned, and RCU held, so the dentries and inodes won't disappear, neither will our sb or filesystem module. - ->i_sb and ->d_sb may be used. + ->d_sb may be used. It is a tricky calling convention because it needs to be called under "rcu-walk", ie. without any locks or references on things. diff --git a/fs/adfs/dir.c b/fs/adfs/dir.c index ade28bb058e3..0d138c0de293 100644 --- a/fs/adfs/dir.c +++ b/fs/adfs/dir.c @@ -191,8 +191,7 @@ const struct file_operations adfs_dir_operations = { }; static int -adfs_hash(const struct dentry *parent, const struct inode *inode, - struct qstr *qstr) +adfs_hash(const struct dentry *parent, struct qstr *qstr) { const unsigned int name_len = ADFS_SB(parent->d_sb)->s_namelen; const unsigned char *name; @@ -228,8 +227,7 @@ adfs_hash(const struct dentry *parent, const struct inode *inode, * requirements of the underlying filesystem. */ static int -adfs_compare(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +adfs_compare(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { int i; diff --git a/fs/affs/namei.c b/fs/affs/namei.c index ff65884a7839..c36cbb4537a2 100644 --- a/fs/affs/namei.c +++ b/fs/affs/namei.c @@ -13,18 +13,12 @@ typedef int (*toupper_t)(int); static int affs_toupper(int ch); -static int affs_hash_dentry(const struct dentry *, - const struct inode *, struct qstr *); -static int affs_compare_dentry(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int affs_hash_dentry(const struct dentry *, struct qstr *); +static int affs_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); static int affs_intl_toupper(int ch); -static int affs_intl_hash_dentry(const struct dentry *, - const struct inode *, struct qstr *); -static int affs_intl_compare_dentry(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int affs_intl_hash_dentry(const struct dentry *, struct qstr *); +static int affs_intl_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); const struct dentry_operations affs_dentry_operations = { @@ -86,14 +80,12 @@ __affs_hash_dentry(struct qstr *qstr, toupper_t toupper) } static int -affs_hash_dentry(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +affs_hash_dentry(const struct dentry *dentry, struct qstr *qstr) { return __affs_hash_dentry(qstr, affs_toupper); } static int -affs_intl_hash_dentry(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +affs_intl_hash_dentry(const struct dentry *dentry, struct qstr *qstr) { return __affs_hash_dentry(qstr, affs_intl_toupper); } @@ -131,15 +123,13 @@ static inline int __affs_compare_dentry(unsigned int len, } static int -affs_compare_dentry(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +affs_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { return __affs_compare_dentry(len, str, name, affs_toupper); } static int -affs_intl_compare_dentry(const struct dentry *parent,const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +affs_intl_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { return __affs_compare_dentry(len, str, name, affs_intl_toupper); diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 5699b5036ed8..5175aebf6737 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -822,8 +822,7 @@ const struct dentry_operations cifs_dentry_ops = { /* d_delete: cifs_d_delete, */ /* not needed except for debugging */ }; -static int cifs_ci_hash(const struct dentry *dentry, const struct inode *inode, - struct qstr *q) +static int cifs_ci_hash(const struct dentry *dentry, struct qstr *q) { struct nls_table *codepage = CIFS_SB(dentry->d_sb)->local_nls; unsigned long hash; @@ -838,12 +837,10 @@ static int cifs_ci_hash(const struct dentry *dentry, const struct inode *inode, return 0; } -static int cifs_ci_compare(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int cifs_ci_compare(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { - struct nls_table *codepage = CIFS_SB(pinode->i_sb)->local_nls; + struct nls_table *codepage = CIFS_SB(parent->d_sb)->local_nls; if ((name->len == len) && (nls_strnicmp(codepage, name->name, str, len) == 0)) diff --git a/fs/dcache.c b/fs/dcache.c index b692c7e097c5..3199fe6863a8 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1723,7 +1723,7 @@ EXPORT_SYMBOL(d_add_ci); * Do the slow-case of the dentry name compare. * * Unlike the dentry_cmp() function, we need to atomically - * load the name, length and inode information, so that the + * load the name and length information, so that the * filesystem can rely on them, and can use the 'name' and * 'len' information without worrying about walking off the * end of memory etc. @@ -1741,22 +1741,18 @@ enum slow_d_compare { static noinline enum slow_d_compare slow_dentry_cmp( const struct dentry *parent, - struct inode *inode, struct dentry *dentry, unsigned int seq, const struct qstr *name) { int tlen = dentry->d_name.len; const char *tname = dentry->d_name.name; - struct inode *i = dentry->d_inode; if (read_seqcount_retry(&dentry->d_seq, seq)) { cpu_relax(); return D_COMP_SEQRETRY; } - if (parent->d_op->d_compare(parent, inode, - dentry, i, - tlen, tname, name)) + if (parent->d_op->d_compare(parent, dentry, tlen, tname, name)) return D_COMP_NOMATCH; return D_COMP_OK; } @@ -1766,7 +1762,6 @@ static noinline enum slow_d_compare slow_dentry_cmp( * @parent: parent dentry * @name: qstr of name we wish to find * @seqp: returns d_seq value at the point where the dentry was found - * @inode: returns dentry->d_inode when the inode was found valid. * Returns: dentry, or NULL * * __d_lookup_rcu is the dcache lookup function for rcu-walk name @@ -1793,7 +1788,7 @@ static noinline enum slow_d_compare slow_dentry_cmp( */ struct dentry *__d_lookup_rcu(const struct dentry *parent, const struct qstr *name, - unsigned *seqp, struct inode *inode) + unsigned *seqp) { u64 hashlen = name->hash_len; const unsigned char *str = name->name; @@ -1827,11 +1822,10 @@ struct dentry *__d_lookup_rcu(const struct dentry *parent, seqretry: /* * The dentry sequence count protects us from concurrent - * renames, and thus protects inode, parent and name fields. + * renames, and thus protects parent and name fields. * * The caller must perform a seqcount check in order - * to do anything useful with the returned dentry, - * including using the 'd_inode' pointer. + * to do anything useful with the returned dentry. * * NOTE! We do a "raw" seqcount_begin here. That means that * we don't wait for the sequence count to stabilize if it @@ -1845,12 +1839,12 @@ seqretry: continue; if (d_unhashed(dentry)) continue; - *seqp = seq; if (unlikely(parent->d_flags & DCACHE_OP_COMPARE)) { if (dentry->d_name.hash != hashlen_hash(hashlen)) continue; - switch (slow_dentry_cmp(parent, inode, dentry, seq, name)) { + *seqp = seq; + switch (slow_dentry_cmp(parent, dentry, seq, name)) { case D_COMP_OK: return dentry; case D_COMP_NOMATCH: @@ -1862,6 +1856,7 @@ seqretry: if (dentry->d_name.hash_len != hashlen) continue; + *seqp = seq; if (!dentry_cmp(dentry, str, hashlen_len(hashlen))) return dentry; } @@ -1959,9 +1954,7 @@ struct dentry *__d_lookup(const struct dentry *parent, const struct qstr *name) 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, parent->d_inode, - dentry, dentry->d_inode, - tlen, tname, name)) + if (parent->d_op->d_compare(parent, dentry, tlen, tname, name)) goto next; } else { if (dentry->d_name.len != len) @@ -1998,7 +1991,7 @@ struct dentry *d_hash_and_lookup(struct dentry *dir, struct qstr *name) */ name->hash = full_name_hash(name->name, name->len); if (dir->d_flags & DCACHE_OP_HASH) { - int err = dir->d_op->d_hash(dir, dir->d_inode, name); + int err = dir->d_op->d_hash(dir, name); if (unlikely(err < 0)) return ERR_PTR(err); } diff --git a/fs/efivarfs/super.c b/fs/efivarfs/super.c index 141aee31884f..a8766b880c07 100644 --- a/fs/efivarfs/super.c +++ b/fs/efivarfs/super.c @@ -45,8 +45,8 @@ static struct super_block *efivarfs_sb; * So we need to perform a case-sensitive match on part 1 and a * case-insensitive match on part 2. */ -static int efivarfs_d_compare(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int efivarfs_d_compare(const struct dentry *parent, + const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { @@ -63,8 +63,7 @@ static int efivarfs_d_compare(const struct dentry *parent, const struct inode *p return strncasecmp(name->name + guid, str + guid, EFI_VARIABLE_GUID_LEN); } -static int efivarfs_d_hash(const struct dentry *dentry, - const struct inode *inode, struct qstr *qstr) +static int efivarfs_d_hash(const struct dentry *dentry, struct qstr *qstr) { unsigned long hash = init_name_hash(); const unsigned char *s = qstr->name; @@ -108,7 +107,7 @@ static struct dentry *efivarfs_alloc_dentry(struct dentry *parent, char *name) q.name = name; q.len = strlen(name); - err = efivarfs_d_hash(NULL, NULL, &q); + err = efivarfs_d_hash(NULL, &q); if (err) return ERR_PTR(err); diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c index 081b759cff83..a783b0e1272a 100644 --- a/fs/fat/namei_msdos.c +++ b/fs/fat/namei_msdos.c @@ -148,8 +148,7 @@ static int msdos_find(struct inode *dir, const unsigned char *name, int len, * that the existing dentry can be used. The msdos fs routines will * return ENOENT or EINVAL as appropriate. */ -static int msdos_hash(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +static int msdos_hash(const struct dentry *dentry, struct qstr *qstr) { struct fat_mount_options *options = &MSDOS_SB(dentry->d_sb)->options; unsigned char msdos_name[MSDOS_NAME]; @@ -165,8 +164,7 @@ static int msdos_hash(const struct dentry *dentry, const struct inode *inode, * Compare two msdos names. If either of the names are invalid, * we fall back to doing the standard name comparison. */ -static int msdos_cmp(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int msdos_cmp(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { struct fat_mount_options *options = &MSDOS_SB(parent->d_sb)->options; diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index 2da952036a3d..6df8d3d885e5 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -107,8 +107,7 @@ static unsigned int vfat_striptail_len(const struct qstr *qstr) * that the existing dentry can be used. The vfat fs routines will * return ENOENT or EINVAL as appropriate. */ -static int vfat_hash(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +static int vfat_hash(const struct dentry *dentry, struct qstr *qstr) { qstr->hash = full_name_hash(qstr->name, vfat_striptail_len(qstr)); return 0; @@ -120,8 +119,7 @@ static int vfat_hash(const struct dentry *dentry, const struct inode *inode, * that the existing dentry can be used. The vfat fs routines will * return ENOENT or EINVAL as appropriate. */ -static int vfat_hashi(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +static int vfat_hashi(const struct dentry *dentry, struct qstr *qstr) { struct nls_table *t = MSDOS_SB(dentry->d_sb)->nls_io; const unsigned char *name; @@ -142,8 +140,7 @@ static int vfat_hashi(const struct dentry *dentry, const struct inode *inode, /* * Case insensitive compare of two vfat names. */ -static int vfat_cmpi(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int vfat_cmpi(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { struct nls_table *t = MSDOS_SB(parent->d_sb)->nls_io; @@ -162,8 +159,7 @@ static int vfat_cmpi(const struct dentry *parent, const struct inode *pinode, /* * Case sensitive compare of two vfat names. */ -static int vfat_cmp(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int vfat_cmp(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { unsigned int alen, blen; diff --git a/fs/gfs2/dentry.c b/fs/gfs2/dentry.c index 4fddb3c22d25..f2448ab2aac5 100644 --- a/fs/gfs2/dentry.c +++ b/fs/gfs2/dentry.c @@ -109,8 +109,7 @@ fail: return 0; } -static int gfs2_dhash(const struct dentry *dentry, const struct inode *inode, - struct qstr *str) +static int gfs2_dhash(const struct dentry *dentry, struct qstr *str) { str->hash = gfs2_disk_hash(str->name, str->len); return 0; diff --git a/fs/hfs/hfs_fs.h b/fs/hfs/hfs_fs.h index a73b11839a41..0524cda47a6e 100644 --- a/fs/hfs/hfs_fs.h +++ b/fs/hfs/hfs_fs.h @@ -229,13 +229,10 @@ extern int hfs_part_find(struct super_block *, sector_t *, sector_t *); /* string.c */ extern const struct dentry_operations hfs_dentry_operations; -extern int hfs_hash_dentry(const struct dentry *, const struct inode *, - struct qstr *); +extern int hfs_hash_dentry(const struct dentry *, struct qstr *); extern int hfs_strcmp(const unsigned char *, unsigned int, const unsigned char *, unsigned int); -extern int hfs_compare_dentry(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +extern int hfs_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); /* trans.c */ diff --git a/fs/hfs/string.c b/fs/hfs/string.c index 495a976a3cc9..85b610c3909f 100644 --- a/fs/hfs/string.c +++ b/fs/hfs/string.c @@ -51,8 +51,7 @@ static unsigned char caseorder[256] = { /* * Hash a string to an integer in a case-independent way */ -int hfs_hash_dentry(const struct dentry *dentry, const struct inode *inode, - struct qstr *this) +int hfs_hash_dentry(const struct dentry *dentry, struct qstr *this) { const unsigned char *name = this->name; unsigned int hash, len = this->len; @@ -93,8 +92,7 @@ int hfs_strcmp(const unsigned char *s1, unsigned int len1, * Test for equality of two strings in the HFS filename character ordering. * return 1 on failure and 0 on success */ -int hfs_compare_dentry(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +int hfs_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { const unsigned char *n1, *n2; diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h index 60b0a3388b26..ede79317cfb8 100644 --- a/fs/hfsplus/hfsplus_fs.h +++ b/fs/hfsplus/hfsplus_fs.h @@ -495,11 +495,8 @@ int hfsplus_uni2asc(struct super_block *, const struct hfsplus_unistr *, char *, int *); int hfsplus_asc2uni(struct super_block *, struct hfsplus_unistr *, int, const char *, int); -int hfsplus_hash_dentry(const struct dentry *dentry, - const struct inode *inode, struct qstr *str); -int hfsplus_compare_dentry(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +int hfsplus_hash_dentry(const struct dentry *dentry, struct qstr *str); +int hfsplus_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); /* wrapper.c */ diff --git a/fs/hfsplus/unicode.c b/fs/hfsplus/unicode.c index 2c2e47dcfdd8..e8ef121a4d8b 100644 --- a/fs/hfsplus/unicode.c +++ b/fs/hfsplus/unicode.c @@ -334,8 +334,7 @@ int hfsplus_asc2uni(struct super_block *sb, * Composed unicode characters are decomposed and case-folding is performed * if the appropriate bits are (un)set on the superblock. */ -int hfsplus_hash_dentry(const struct dentry *dentry, const struct inode *inode, - struct qstr *str) +int hfsplus_hash_dentry(const struct dentry *dentry, struct qstr *str) { struct super_block *sb = dentry->d_sb; const char *astr; @@ -386,9 +385,7 @@ int hfsplus_hash_dentry(const struct dentry *dentry, const struct inode *inode, * Composed unicode characters are decomposed and case-folding is performed * if the appropriate bits are (un)set on the superblock. */ -int hfsplus_compare_dentry(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +int hfsplus_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { struct super_block *sb = parent->d_sb; diff --git a/fs/hpfs/dentry.c b/fs/hpfs/dentry.c index 05d4816e4e77..fa27980f2229 100644 --- a/fs/hpfs/dentry.c +++ b/fs/hpfs/dentry.c @@ -12,8 +12,7 @@ * Note: the dentry argument is the parent dentry. */ -static int hpfs_hash_dentry(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +static int hpfs_hash_dentry(const struct dentry *dentry, struct qstr *qstr) { unsigned long hash; int i; @@ -35,9 +34,7 @@ static int hpfs_hash_dentry(const struct dentry *dentry, const struct inode *ino return 0; } -static int hpfs_compare_dentry(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int hpfs_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { unsigned al = len; diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index d9b8aebdeb22..c348d6d88624 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -28,31 +28,23 @@ #define BEQUIET -static int isofs_hashi(const struct dentry *parent, const struct inode *inode, - struct qstr *qstr); -static int isofs_hash(const struct dentry *parent, const struct inode *inode, - struct qstr *qstr); +static int isofs_hashi(const struct dentry *parent, struct qstr *qstr); +static int isofs_hash(const struct dentry *parent, struct qstr *qstr); static int isofs_dentry_cmpi(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, + const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); static int isofs_dentry_cmp(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, + const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); #ifdef CONFIG_JOLIET -static int isofs_hashi_ms(const struct dentry *parent, const struct inode *inode, - struct qstr *qstr); -static int isofs_hash_ms(const struct dentry *parent, const struct inode *inode, - struct qstr *qstr); +static int isofs_hashi_ms(const struct dentry *parent, struct qstr *qstr); +static int isofs_hash_ms(const struct dentry *parent, struct qstr *qstr); static int isofs_dentry_cmpi_ms(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, + const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); static int isofs_dentry_cmp_ms(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, + const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name); #endif @@ -265,30 +257,26 @@ static int isofs_dentry_cmp_common( } static int -isofs_hash(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +isofs_hash(const struct dentry *dentry, struct qstr *qstr) { return isofs_hash_common(dentry, qstr, 0); } static int -isofs_hashi(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +isofs_hashi(const struct dentry *dentry, struct qstr *qstr) { return isofs_hashi_common(dentry, qstr, 0); } static int -isofs_dentry_cmp(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +isofs_dentry_cmp(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { return isofs_dentry_cmp_common(len, str, name, 0, 0); } static int -isofs_dentry_cmpi(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +isofs_dentry_cmpi(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { return isofs_dentry_cmp_common(len, str, name, 0, 1); @@ -296,30 +284,26 @@ isofs_dentry_cmpi(const struct dentry *parent, const struct inode *pinode, #ifdef CONFIG_JOLIET static int -isofs_hash_ms(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +isofs_hash_ms(const struct dentry *dentry, struct qstr *qstr) { return isofs_hash_common(dentry, qstr, 1); } static int -isofs_hashi_ms(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +isofs_hashi_ms(const struct dentry *dentry, struct qstr *qstr) { return isofs_hashi_common(dentry, qstr, 1); } static int -isofs_dentry_cmp_ms(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +isofs_dentry_cmp_ms(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { return isofs_dentry_cmp_common(len, str, name, 1, 0); } static int -isofs_dentry_cmpi_ms(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +isofs_dentry_cmpi_ms(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { return isofs_dentry_cmp_common(len, str, name, 1, 1); diff --git a/fs/isofs/namei.c b/fs/isofs/namei.c index c167028844ed..95295640d9c8 100644 --- a/fs/isofs/namei.c +++ b/fs/isofs/namei.c @@ -37,8 +37,7 @@ isofs_cmp(struct dentry *dentry, const char *compare, int dlen) qstr.name = compare; qstr.len = dlen; - return dentry->d_op->d_compare(NULL, NULL, NULL, NULL, - dentry->d_name.len, dentry->d_name.name, &qstr); + return dentry->d_op->d_compare(NULL, NULL, dentry->d_name.len, dentry->d_name.name, &qstr); } /* diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 89186b7b9002..8b19027291d6 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -1538,8 +1538,7 @@ const struct file_operations jfs_dir_operations = { .llseek = generic_file_llseek, }; -static int jfs_ci_hash(const struct dentry *dir, const struct inode *inode, - struct qstr *this) +static int jfs_ci_hash(const struct dentry *dir, struct qstr *this) { unsigned long hash; int i; @@ -1552,9 +1551,7 @@ static int jfs_ci_hash(const struct dentry *dir, const struct inode *inode, return 0; } -static int jfs_ci_compare(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int jfs_ci_compare(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { int i, result = 1; diff --git a/fs/namei.c b/fs/namei.c index 66998b06d822..b2beee7a733f 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1352,7 +1352,7 @@ static int lookup_fast(struct nameidata *nd, */ if (nd->flags & LOOKUP_RCU) { unsigned seq; - dentry = __d_lookup_rcu(parent, &nd->last, &seq, nd->inode); + dentry = __d_lookup_rcu(parent, &nd->last, &seq); if (!dentry) goto unlazy; @@ -1787,8 +1787,7 @@ static int link_path_walk(const char *name, struct nameidata *nd) struct dentry *parent = nd->path.dentry; nd->flags &= ~LOOKUP_JUMPED; if (unlikely(parent->d_flags & DCACHE_OP_HASH)) { - err = parent->d_op->d_hash(parent, nd->inode, - &this); + err = parent->d_op->d_hash(parent, &this); if (err < 0) break; } @@ -2121,7 +2120,7 @@ struct dentry *lookup_one_len(const char *name, struct dentry *base, int len) * to use its own hash.. */ if (base->d_flags & DCACHE_OP_HASH) { - int err = base->d_op->d_hash(base, base->d_inode, &this); + int err = base->d_op->d_hash(base, &this); if (err < 0) return ERR_PTR(err); } diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index 3bc105d36f10..3be047474bfc 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -73,10 +73,8 @@ const struct inode_operations ncp_dir_inode_operations = * Dentry operations routines */ static int ncp_lookup_validate(struct dentry *, unsigned int); -static int ncp_hash_dentry(const struct dentry *, const struct inode *, - struct qstr *); -static int ncp_compare_dentry(const struct dentry *, const struct inode *, - const struct dentry *, const struct inode *, +static int ncp_hash_dentry(const struct dentry *, struct qstr *); +static int ncp_compare_dentry(const struct dentry *, const struct dentry *, unsigned int, const char *, const struct qstr *); static int ncp_delete_dentry(const struct dentry *); @@ -119,11 +117,19 @@ static inline int ncp_case_sensitive(const struct inode *i) /* * Note: leave the hash unchanged if the directory * is case-sensitive. + * + * Accessing the parent inode can be racy under RCU pathwalking. + * Use ACCESS_ONCE() to make sure we use _one_ particular inode, + * the callers will handle races. */ static int -ncp_hash_dentry(const struct dentry *dentry, const struct inode *inode, - struct qstr *this) +ncp_hash_dentry(const struct dentry *dentry, struct qstr *this) { + struct inode *inode = ACCESS_ONCE(dentry->d_inode); + + if (!inode) + return 0; + if (!ncp_case_sensitive(inode)) { struct super_block *sb = dentry->d_sb; struct nls_table *t; @@ -140,14 +146,24 @@ ncp_hash_dentry(const struct dentry *dentry, const struct inode *inode, return 0; } +/* + * Accessing the parent inode can be racy under RCU pathwalking. + * Use ACCESS_ONCE() to make sure we use _one_ particular inode, + * the callers will handle races. + */ static int -ncp_compare_dentry(const struct dentry *parent, const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +ncp_compare_dentry(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { + struct inode *pinode; + if (len != name->len) return 1; + pinode = ACCESS_ONCE(parent->d_inode); + if (!pinode) + return 1; + if (ncp_case_sensitive(pinode)) return strncmp(str, name->name, len); diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index f3a570e7c257..71290463a1d3 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -796,15 +796,16 @@ static int sysctl_is_seen(struct ctl_table_header *p) return res; } -static int proc_sys_compare(const struct dentry *parent, - const struct inode *pinode, - const struct dentry *dentry, const struct inode *inode, +static int proc_sys_compare(const struct dentry *parent, const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { struct ctl_table_header *head; + struct inode *inode; + /* Although proc doesn't have negative dentries, rcu-walk means * that inode here can be NULL */ /* AV: can it, indeed? */ + inode = ACCESS_ONCE(dentry->d_inode); if (!inode) return 1; if (name->len != len) diff --git a/fs/sysv/namei.c b/fs/sysv/namei.c index 1c0d5f264767..731b2bbcaab3 100644 --- a/fs/sysv/namei.c +++ b/fs/sysv/namei.c @@ -27,8 +27,7 @@ static int add_nondir(struct dentry *dentry, struct inode *inode) return err; } -static int sysv_hash(const struct dentry *dentry, const struct inode *inode, - struct qstr *qstr) +static int sysv_hash(const struct dentry *dentry, struct qstr *qstr) { /* Truncate the name in place, avoids having to define a compare function. */ diff --git a/include/linux/dcache.h b/include/linux/dcache.h index 86da7595ba31..f42dbe145479 100644 --- a/include/linux/dcache.h +++ b/include/linux/dcache.h @@ -146,10 +146,8 @@ enum dentry_d_lock_class struct dentry_operations { int (*d_revalidate)(struct dentry *, unsigned int); int (*d_weak_revalidate)(struct dentry *, unsigned int); - int (*d_hash)(const struct dentry *, const struct inode *, - struct qstr *); - int (*d_compare)(const struct dentry *, const struct inode *, - const struct dentry *, const struct inode *, + int (*d_hash)(const struct dentry *, struct qstr *); + int (*d_compare)(const struct dentry *, const struct dentry *, unsigned int, const char *, const struct qstr *); int (*d_delete)(const struct dentry *); void (*d_release)(struct dentry *); @@ -302,8 +300,7 @@ extern struct dentry *d_lookup(const struct dentry *, const struct qstr *); extern struct dentry *d_hash_and_lookup(struct dentry *, struct qstr *); extern struct dentry *__d_lookup(const struct dentry *, const struct qstr *); extern struct dentry *__d_lookup_rcu(const struct dentry *parent, - const struct qstr *name, - unsigned *seq, struct inode *inode); + const struct qstr *name, unsigned *seq); /** * __d_rcu_to_refcount - take a refcount on dentry if sequence check is ok -- cgit v1.2.3 From f891a29f46553a384edbaa0f6ac446b1d03bccac Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jun 2013 08:58:09 -0400 Subject: locks: drop the unused filp argument to posix_unblock_lock Signed-off-by: Jeff Layton Signed-off-by: Al Viro --- fs/lockd/svclock.c | 2 +- fs/locks.c | 4 +--- include/linux/fs.h | 5 ++--- 3 files changed, 4 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index e703318c41df..a469098682c4 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -276,7 +276,7 @@ static int nlmsvc_unlink_block(struct nlm_block *block) dprintk("lockd: unlinking block %p...\n", block); /* Remove block from list */ - status = posix_unblock_lock(block->b_file->f_file, &block->b_call->a_args.lock.fl); + status = posix_unblock_lock(&block->b_call->a_args.lock.fl); nlmsvc_remove_block(block); return status; } diff --git a/fs/locks.c b/fs/locks.c index cb424a4fed71..72fb2b722211 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -2099,13 +2099,12 @@ void locks_remove_flock(struct file *filp) /** * posix_unblock_lock - stop waiting for a file lock - * @filp: how the file was opened * @waiter: the lock which was waiting * * lockd needs to block waiting for locks. */ int -posix_unblock_lock(struct file *filp, struct file_lock *waiter) +posix_unblock_lock(struct file_lock *waiter) { int status = 0; @@ -2117,7 +2116,6 @@ posix_unblock_lock(struct file *filp, struct file_lock *waiter) unlock_flocks(); return status; } - EXPORT_SYMBOL(posix_unblock_lock); /** diff --git a/include/linux/fs.h b/include/linux/fs.h index 68f10204ab29..172303655702 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -994,7 +994,7 @@ extern void locks_release_private(struct file_lock *); extern void posix_test_lock(struct file *, struct file_lock *); extern int posix_lock_file(struct file *, struct file_lock *, struct file_lock *); extern int posix_lock_file_wait(struct file *, struct file_lock *); -extern int posix_unblock_lock(struct file *, struct file_lock *); +extern int posix_unblock_lock(struct file_lock *); extern int vfs_test_lock(struct file *, struct file_lock *); extern int vfs_lock_file(struct file *, unsigned int, struct file_lock *, struct file_lock *); extern int vfs_cancel_lock(struct file *filp, struct file_lock *fl); @@ -1084,8 +1084,7 @@ static inline int posix_lock_file_wait(struct file *filp, struct file_lock *fl) return -ENOLCK; } -static inline int posix_unblock_lock(struct file *filp, - struct file_lock *waiter) +static inline int posix_unblock_lock(struct file_lock *waiter) { return -ENOENT; } -- cgit v1.2.3 From 1a9e64a7118c5ad13dd5119da18375a5bd45b330 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jun 2013 08:58:10 -0400 Subject: cifs: use posix_unblock_lock instead of locks_delete_block commit 66189be74 (CIFS: Fix VFS lock usage for oplocked files) exported the locks_delete_block symbol. There's already an exported helper function that provides this capability however, so make cifs use that instead and turn locks_delete_block back into a static function. Note that if fl->fl_next == NULL then this lock has already been through locks_delete_block(), so we should be OK to ignore an ENOENT error here and simply not retry the lock. Cc: Pavel Shilovsky Signed-off-by: Jeff Layton Acked-by: J. Bruce Fields Signed-off-by: Al Viro --- fs/cifs/file.c | 2 +- fs/locks.c | 3 +-- include/linux/fs.h | 5 ----- 3 files changed, 2 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 48b29d24c9f4..1686e4085646 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -999,7 +999,7 @@ try_again: rc = wait_event_interruptible(flock->fl_wait, !flock->fl_next); if (!rc) goto try_again; - locks_delete_block(flock); + posix_unblock_lock(flock); } return rc; } diff --git a/fs/locks.c b/fs/locks.c index 72fb2b722211..d732e2226f17 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -496,13 +496,12 @@ static void __locks_delete_block(struct file_lock *waiter) /* */ -void locks_delete_block(struct file_lock *waiter) +static void locks_delete_block(struct file_lock *waiter) { lock_flocks(); __locks_delete_block(waiter); unlock_flocks(); } -EXPORT_SYMBOL(locks_delete_block); /* Insert waiter into blocker's block list. * We use a circular list so that processes can be easily woken up in diff --git a/include/linux/fs.h b/include/linux/fs.h index 172303655702..6cfc9a29a783 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1006,7 +1006,6 @@ extern int vfs_setlease(struct file *, long, struct file_lock **); extern int lease_modify(struct file_lock **, int); extern int lock_may_read(struct inode *, loff_t start, unsigned long count); extern int lock_may_write(struct inode *, loff_t start, unsigned long count); -extern void locks_delete_block(struct file_lock *waiter); extern void lock_flocks(void); extern void unlock_flocks(void); #else /* !CONFIG_FILE_LOCKING */ @@ -1150,10 +1149,6 @@ static inline int lock_may_write(struct inode *inode, loff_t start, return 1; } -static inline void locks_delete_block(struct file_lock *waiter) -{ -} - static inline void lock_flocks(void) { } -- cgit v1.2.3 From 1cb360125966cb6cb594e414ea80a0154617b846 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jun 2013 08:58:12 -0400 Subject: locks: comment cleanups and clarifications Signed-off-by: Jeff Layton Signed-off-by: Al Viro --- fs/locks.c | 21 +++++++++++++-------- include/linux/fs.h | 18 ++++++++++++++++++ 2 files changed, 31 insertions(+), 8 deletions(-) (limited to 'include') diff --git a/fs/locks.c b/fs/locks.c index 804bb9e01a65..ddeab49fe2be 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -518,9 +518,10 @@ static void locks_insert_block(struct file_lock *blocker, list_add(&waiter->fl_link, &blocked_list); } -/* Wake up processes blocked waiting for blocker. - * If told to wait then schedule the processes until the block list - * is empty, otherwise empty the block list ourselves. +/* + * Wake up processes blocked waiting for blocker. + * + * Must be called with the file_lock_lock held! */ static void locks_wake_up_blocks(struct file_lock *blocker) { @@ -806,6 +807,11 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str } lock_flocks(); + /* + * New lock request. Walk all POSIX locks and look for conflicts. If + * there are any, either return error or put the request on the + * blocker's list of waiters and the global blocked_list. + */ if (request->fl_type != F_UNLCK) { for_each_lock(inode, before) { fl = *before; @@ -844,7 +850,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str before = &fl->fl_next; } - /* Process locks with this owner. */ + /* Process locks with this owner. */ while ((fl = *before) && posix_same_owner(request, fl)) { /* Detect adjacent or overlapping regions (if same lock type) */ @@ -930,10 +936,9 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str } /* - * The above code only modifies existing locks in case of - * merging or replacing. If new lock(s) need to be inserted - * all modifications are done bellow this, so it's safe yet to - * bail out. + * The above code only modifies existing locks in case of merging or + * replacing. If new lock(s) need to be inserted all modifications are + * done below this, so it's safe yet to bail out. */ error = -ENOLCK; /* "no luck" */ if (right && left == right && !new_fl2) diff --git a/include/linux/fs.h b/include/linux/fs.h index 6cfc9a29a783..ed9fdaaf3223 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -926,6 +926,24 @@ int locks_in_grace(struct net *); /* that will die - we need it for nfs_lock_info */ #include +/* + * struct file_lock represents a generic "file lock". It's used to represent + * POSIX byte range locks, BSD (flock) locks, and leases. It's important to + * note that the same struct is used to represent both a request for a lock and + * the lock itself, but the same object is never used for both. + * + * FIXME: should we create a separate "struct lock_request" to help distinguish + * these two uses? + * + * The i_flock list is ordered by: + * + * 1) lock type -- FL_LEASEs first, then FL_FLOCK, and finally FL_POSIX + * 2) lock owner + * 3) lock range start + * 4) lock range end + * + * Obviously, the last two criteria only matter for POSIX locks. + */ struct file_lock { struct file_lock *fl_next; /* singly linked list for this inode */ struct list_head fl_link; /* doubly linked list of all locks */ -- cgit v1.2.3 From 1c8c601a8c0dc59fe64907dcd9d512a3d181ddc7 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jun 2013 08:58:15 -0400 Subject: locks: protect most of the file_lock handling with i_lock Having a global lock that protects all of this code is a clear scalability problem. Instead of doing that, move most of the code to be protected by the i_lock instead. The exceptions are the global lists that the ->fl_link sits on, and the ->fl_block list. ->fl_link is what connects these structures to the global lists, so we must ensure that we hold those locks when iterating over or updating these lists. Furthermore, sound deadlock detection requires that we hold the blocked_list state steady while checking for loops. We also must ensure that the search and update to the list are atomic. For the checking and insertion side of the blocked_list, push the acquisition of the global lock into __posix_lock_file and ensure that checking and update of the blocked_list is done without dropping the lock in between. On the removal side, when waking up blocked lock waiters, take the global lock before walking the blocked list and dequeue the waiters from the global list prior to removal from the fl_block list. With this, deadlock detection should be race free while we minimize excessive file_lock_lock thrashing. Finally, in order to avoid a lock inversion problem when handling /proc/locks output we must ensure that manipulations of the fl_block list are also protected by the file_lock_lock. Signed-off-by: Jeff Layton Signed-off-by: Al Viro --- Documentation/filesystems/Locking | 21 +++-- fs/afs/flock.c | 7 +- fs/ceph/locks.c | 2 +- fs/ceph/mds_client.c | 8 +- fs/cifs/cifsfs.c | 2 +- fs/cifs/file.c | 13 +-- fs/gfs2/file.c | 2 +- fs/lockd/svcsubs.c | 12 +-- fs/locks.c | 164 ++++++++++++++++++++++++-------------- fs/nfs/delegation.c | 10 +-- fs/nfs/nfs4state.c | 8 +- fs/nfsd/nfs4state.c | 8 +- include/linux/fs.h | 11 --- 13 files changed, 155 insertions(+), 113 deletions(-) (limited to 'include') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index f94a362f408e..c2963a74fbc3 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -342,7 +342,7 @@ prototypes: locking rules: - file_lock_lock may block + inode->i_lock may block fl_copy_lock: yes no fl_release_private: maybe no @@ -355,12 +355,19 @@ prototypes: int (*lm_change)(struct file_lock **, int); locking rules: - file_lock_lock may block -lm_compare_owner: yes no -lm_notify: yes no -lm_grant: no no -lm_break: yes no -lm_change yes no + + inode->i_lock file_lock_lock may block +lm_compare_owner: yes[1] maybe no +lm_notify: yes yes no +lm_grant: no no no +lm_break: yes no no +lm_change yes no no + +[1]: ->lm_compare_owner is generally called with *an* inode->i_lock held. It +may not be the i_lock of the inode for either file_lock being compared! This is +the case with deadlock detection, since the code has to chase down the owners +of locks that may be entirely unrelated to the one on which the lock is being +acquired. When doing a search for deadlocks, the file_lock_lock is also held. --------------------------- buffer_head ----------------------------------- prototypes: diff --git a/fs/afs/flock.c b/fs/afs/flock.c index 2497bf306c70..a8cf2cff836c 100644 --- a/fs/afs/flock.c +++ b/fs/afs/flock.c @@ -252,7 +252,8 @@ static void afs_defer_unlock(struct afs_vnode *vnode, struct key *key) */ static int afs_do_setlk(struct file *file, struct file_lock *fl) { - struct afs_vnode *vnode = AFS_FS_I(file->f_mapping->host); + struct inode *inode = file_inode(file); + struct afs_vnode *vnode = AFS_FS_I(inode); afs_lock_type_t type; struct key *key = file->private_data; int ret; @@ -273,7 +274,7 @@ static int afs_do_setlk(struct file *file, struct file_lock *fl) type = (fl->fl_type == F_RDLCK) ? AFS_LOCK_READ : AFS_LOCK_WRITE; - lock_flocks(); + spin_lock(&inode->i_lock); /* make sure we've got a callback on this file and that our view of the * data version is up to date */ @@ -420,7 +421,7 @@ given_lock: afs_vnode_fetch_status(vnode, NULL, key); error: - unlock_flocks(); + spin_unlock(&inode->i_lock); _leave(" = %d", ret); return ret; diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c index ebbf680378e2..690f73f42425 100644 --- a/fs/ceph/locks.c +++ b/fs/ceph/locks.c @@ -192,7 +192,7 @@ void ceph_count_locks(struct inode *inode, int *fcntl_count, int *flock_count) /** * Encode the flock and fcntl locks for the given inode into the ceph_filelock - * array. Must be called with lock_flocks() already held. + * array. Must be called with inode->i_lock already held. * If we encounter more of a specific lock type than expected, return -ENOSPC. */ int ceph_encode_locks_to_buffer(struct inode *inode, diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 4d2920304be8..74fd2898b2ab 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -2481,20 +2481,20 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap, struct ceph_filelock *flocks; encode_again: - lock_flocks(); + spin_lock(&inode->i_lock); ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks); - unlock_flocks(); + spin_unlock(&inode->i_lock); flocks = kmalloc((num_fcntl_locks+num_flock_locks) * sizeof(struct ceph_filelock), GFP_NOFS); if (!flocks) { err = -ENOMEM; goto out_free; } - lock_flocks(); + spin_lock(&inode->i_lock); err = ceph_encode_locks_to_buffer(inode, flocks, num_fcntl_locks, num_flock_locks); - unlock_flocks(); + spin_unlock(&inode->i_lock); if (err) { kfree(flocks); if (err == -ENOSPC) diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 540c1ccfcdb2..a445e71746fa 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -765,7 +765,7 @@ static loff_t cifs_llseek(struct file *file, loff_t offset, int whence) static int cifs_setlease(struct file *file, long arg, struct file_lock **lease) { - /* note that this is called by vfs setlease with lock_flocks held + /* note that this is called by vfs setlease with i_lock held to protect *lease from going away */ struct inode *inode = file_inode(file); struct cifsFileInfo *cfile = file->private_data; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 1686e4085646..0630710a9c3f 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -1092,6 +1092,7 @@ struct lock_to_push { static int cifs_push_posix_locks(struct cifsFileInfo *cfile) { + struct inode *inode = cfile->dentry->d_inode; struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); struct file_lock *flock, **before; unsigned int count = 0, i = 0; @@ -1102,12 +1103,12 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) xid = get_xid(); - lock_flocks(); - cifs_for_each_lock(cfile->dentry->d_inode, before) { + spin_lock(&inode->i_lock); + cifs_for_each_lock(inode, before) { if ((*before)->fl_flags & FL_POSIX) count++; } - unlock_flocks(); + spin_unlock(&inode->i_lock); INIT_LIST_HEAD(&locks_to_send); @@ -1126,8 +1127,8 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) } el = locks_to_send.next; - lock_flocks(); - cifs_for_each_lock(cfile->dentry->d_inode, before) { + spin_lock(&inode->i_lock); + cifs_for_each_lock(inode, before) { flock = *before; if ((flock->fl_flags & FL_POSIX) == 0) continue; @@ -1152,7 +1153,7 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) lck->offset = flock->fl_start; el = el->next; } - unlock_flocks(); + spin_unlock(&inode->i_lock); list_for_each_entry_safe(lck, tmp, &locks_to_send, llist) { int stored_rc; diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index b3333371aebb..cebfd404c1d9 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -889,7 +889,7 @@ out_uninit: * cluster; until we do, disable leases (by just returning -EINVAL), * unless the administrator has requested purely local locking. * - * Locking: called under lock_flocks + * Locking: called under i_lock * * Returns: errno */ diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index 97e87415b145..dc5c75930f0f 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -169,7 +169,7 @@ nlm_traverse_locks(struct nlm_host *host, struct nlm_file *file, again: file->f_locks = 0; - lock_flocks(); /* protects i_flock list */ + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl; fl = fl->fl_next) { if (fl->fl_lmops != &nlmsvc_lock_operations) continue; @@ -181,7 +181,7 @@ again: if (match(lockhost, host)) { struct file_lock lock = *fl; - unlock_flocks(); + spin_unlock(&inode->i_lock); lock.fl_type = F_UNLCK; lock.fl_start = 0; lock.fl_end = OFFSET_MAX; @@ -193,7 +193,7 @@ again: goto again; } } - unlock_flocks(); + spin_unlock(&inode->i_lock); return 0; } @@ -228,14 +228,14 @@ nlm_file_inuse(struct nlm_file *file) if (file->f_count || !list_empty(&file->f_blocks) || file->f_shares) return 1; - lock_flocks(); + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl; fl = fl->fl_next) { if (fl->fl_lmops == &nlmsvc_lock_operations) { - unlock_flocks(); + spin_unlock(&inode->i_lock); return 1; } } - unlock_flocks(); + spin_unlock(&inode->i_lock); file->f_locks = 0; return 0; } diff --git a/fs/locks.c b/fs/locks.c index 89d898bce166..ce302d43822b 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -153,27 +153,37 @@ int lease_break_time = 45; #define for_each_lock(inode, lockp) \ for (lockp = &inode->i_flock; *lockp != NULL; lockp = &(*lockp)->fl_next) -/* The global file_lock_list is only used for displaying /proc/locks. */ +/* + * The global file_lock_list is only used for displaying /proc/locks. Protected + * by the file_lock_lock. + */ static LIST_HEAD(file_lock_list); -/* The blocked_list is used to find POSIX lock loops for deadlock detection. */ +/* + * The blocked_list is used to find POSIX lock loops for deadlock detection. + * Protected by file_lock_lock. + */ static LIST_HEAD(blocked_list); -/* Protects the two list heads above, plus the inode->i_flock list */ +/* + * This lock protects the blocked_list, and the file_lock_list. Generally, if + * you're accessing one of those lists, you want to be holding this lock. + * + * In addition, it also protects the fl->fl_block list, and the fl->fl_next + * pointer for file_lock structures that are acting as lock requests (in + * contrast to those that are acting as records of acquired locks). + * + * Note that when we acquire this lock in order to change the above fields, + * we often hold the i_lock as well. In certain cases, when reading the fields + * protected by this lock, we can skip acquiring it iff we already hold the + * i_lock. + * + * In particular, adding an entry to the fl_block list requires that you hold + * both the i_lock and the blocked_lock_lock (acquired in that order). Deleting + * an entry from the list however only requires the file_lock_lock. + */ static DEFINE_SPINLOCK(file_lock_lock); -void lock_flocks(void) -{ - spin_lock(&file_lock_lock); -} -EXPORT_SYMBOL_GPL(lock_flocks); - -void unlock_flocks(void) -{ - spin_unlock(&file_lock_lock); -} -EXPORT_SYMBOL_GPL(unlock_flocks); - static struct kmem_cache *filelock_cache __read_mostly; static void locks_init_lock_heads(struct file_lock *fl) @@ -489,13 +499,17 @@ static int posix_same_owner(struct file_lock *fl1, struct file_lock *fl2) static inline void locks_insert_global_locks(struct file_lock *fl) { + spin_lock(&file_lock_lock); list_add_tail(&fl->fl_link, &file_lock_list); + spin_unlock(&file_lock_lock); } static inline void locks_delete_global_locks(struct file_lock *fl) { + spin_lock(&file_lock_lock); list_del_init(&fl->fl_link); + spin_unlock(&file_lock_lock); } static inline void @@ -512,6 +526,8 @@ locks_delete_global_blocked(struct file_lock *waiter) /* Remove waiter from blocker's block list. * When blocker ends up pointing to itself then the list is empty. + * + * Must be called with file_lock_lock held. */ static void __locks_delete_block(struct file_lock *waiter) { @@ -520,37 +536,47 @@ static void __locks_delete_block(struct file_lock *waiter) waiter->fl_next = NULL; } -/* - */ static void locks_delete_block(struct file_lock *waiter) { - lock_flocks(); + spin_lock(&file_lock_lock); __locks_delete_block(waiter); - unlock_flocks(); + spin_unlock(&file_lock_lock); } /* Insert waiter into blocker's block list. * We use a circular list so that processes can be easily woken up in * the order they blocked. The documentation doesn't require this but * it seems like the reasonable thing to do. + * + * Must be called with file_lock_lock held! */ -static void locks_insert_block(struct file_lock *blocker, - struct file_lock *waiter) +static void __locks_insert_block(struct file_lock *blocker, + struct file_lock *waiter) { BUG_ON(!list_empty(&waiter->fl_block)); waiter->fl_next = blocker; list_add_tail(&waiter->fl_block, &blocker->fl_block); if (IS_POSIX(blocker)) - locks_insert_global_blocked(request); + locks_insert_global_blocked(waiter); +} + +/* Must be called with i_lock held. */ +static void locks_insert_block(struct file_lock *blocker, + struct file_lock *waiter) +{ + spin_lock(&file_lock_lock); + __locks_insert_block(blocker, waiter); + spin_unlock(&file_lock_lock); } /* * Wake up processes blocked waiting for blocker. * - * Must be called with the file_lock_lock held! + * Must be called with the inode->i_lock held! */ static void locks_wake_up_blocks(struct file_lock *blocker) { + spin_lock(&file_lock_lock); while (!list_empty(&blocker->fl_block)) { struct file_lock *waiter; @@ -562,10 +588,13 @@ static void locks_wake_up_blocks(struct file_lock *blocker) else wake_up(&waiter->fl_wait); } + spin_unlock(&file_lock_lock); } /* Insert file lock fl into an inode's lock list at the position indicated * by pos. At the same time add the lock to the global file lock list. + * + * Must be called with the i_lock held! */ static void locks_insert_lock(struct file_lock **pos, struct file_lock *fl) { @@ -583,6 +612,8 @@ static void locks_insert_lock(struct file_lock **pos, struct file_lock *fl) * Wake up processes that are blocked waiting for this lock, * notify the FS that the lock has been cleared and * finally free the lock. + * + * Must be called with the i_lock held! */ static void locks_delete_lock(struct file_lock **thisfl_p) { @@ -652,8 +683,9 @@ void posix_test_lock(struct file *filp, struct file_lock *fl) { struct file_lock *cfl; + struct inode *inode = file_inode(filp); - lock_flocks(); + spin_lock(&inode->i_lock); for (cfl = file_inode(filp)->i_flock; cfl; cfl = cfl->fl_next) { if (!IS_POSIX(cfl)) continue; @@ -666,7 +698,7 @@ posix_test_lock(struct file *filp, struct file_lock *fl) fl->fl_pid = pid_vnr(cfl->fl_nspid); } else fl->fl_type = F_UNLCK; - unlock_flocks(); + spin_unlock(&inode->i_lock); return; } EXPORT_SYMBOL(posix_test_lock); @@ -710,6 +742,7 @@ static struct file_lock *what_owner_is_waiting_for(struct file_lock *block_fl) return NULL; } +/* Must be called with the file_lock_lock held! */ static int posix_locks_deadlock(struct file_lock *caller_fl, struct file_lock *block_fl) { @@ -745,7 +778,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) return -ENOMEM; } - lock_flocks(); + spin_lock(&inode->i_lock); if (request->fl_flags & FL_ACCESS) goto find_conflict; @@ -775,9 +808,9 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) * give it the opportunity to lock the file. */ if (found) { - unlock_flocks(); + spin_unlock(&inode->i_lock); cond_resched(); - lock_flocks(); + spin_lock(&inode->i_lock); } find_conflict: @@ -804,7 +837,7 @@ find_conflict: error = 0; out: - unlock_flocks(); + spin_unlock(&inode->i_lock); if (new_fl) locks_free_lock(new_fl); return error; @@ -834,7 +867,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str new_fl2 = locks_alloc_lock(); } - lock_flocks(); + spin_lock(&inode->i_lock); /* * New lock request. Walk all POSIX locks and look for conflicts. If * there are any, either return error or put the request on the @@ -852,11 +885,17 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str error = -EAGAIN; if (!(request->fl_flags & FL_SLEEP)) goto out; + /* + * Deadlock detection and insertion into the blocked + * locks list must be done while holding the same lock! + */ error = -EDEADLK; - if (posix_locks_deadlock(request, fl)) - goto out; - error = FILE_LOCK_DEFERRED; - locks_insert_block(fl, request); + spin_lock(&file_lock_lock); + if (likely(!posix_locks_deadlock(request, fl))) { + error = FILE_LOCK_DEFERRED; + __locks_insert_block(fl, request); + } + spin_unlock(&file_lock_lock); goto out; } } @@ -1006,7 +1045,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str locks_wake_up_blocks(left); } out: - unlock_flocks(); + spin_unlock(&inode->i_lock); /* * Free any unused locks. */ @@ -1081,14 +1120,14 @@ int locks_mandatory_locked(struct inode *inode) /* * Search the lock list for this inode for any POSIX locks. */ - lock_flocks(); + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!IS_POSIX(fl)) continue; if (fl->fl_owner != owner) break; } - unlock_flocks(); + spin_unlock(&inode->i_lock); return fl ? -EAGAIN : 0; } @@ -1231,7 +1270,7 @@ int __break_lease(struct inode *inode, unsigned int mode) if (IS_ERR(new_fl)) return PTR_ERR(new_fl); - lock_flocks(); + spin_lock(&inode->i_lock); time_out_leases(inode); @@ -1281,11 +1320,11 @@ restart: break_time++; } locks_insert_block(flock, new_fl); - unlock_flocks(); + spin_unlock(&inode->i_lock); error = wait_event_interruptible_timeout(new_fl->fl_wait, !new_fl->fl_next, break_time); - lock_flocks(); - __locks_delete_block(new_fl); + spin_lock(&inode->i_lock); + locks_delete_block(new_fl); if (error >= 0) { if (error == 0) time_out_leases(inode); @@ -1302,7 +1341,7 @@ restart: } out: - unlock_flocks(); + spin_unlock(&inode->i_lock); locks_free_lock(new_fl); return error; } @@ -1355,9 +1394,10 @@ EXPORT_SYMBOL(lease_get_mtime); int fcntl_getlease(struct file *filp) { struct file_lock *fl; + struct inode *inode = file_inode(filp); int type = F_UNLCK; - lock_flocks(); + spin_lock(&inode->i_lock); time_out_leases(file_inode(filp)); for (fl = file_inode(filp)->i_flock; fl && IS_LEASE(fl); fl = fl->fl_next) { @@ -1366,7 +1406,7 @@ int fcntl_getlease(struct file *filp) break; } } - unlock_flocks(); + spin_unlock(&inode->i_lock); return type; } @@ -1460,7 +1500,7 @@ static int generic_delete_lease(struct file *filp, struct file_lock **flp) * The (input) flp->fl_lmops->lm_break function is required * by break_lease(). * - * Called with file_lock_lock held. + * Called with inode->i_lock held. */ int generic_setlease(struct file *filp, long arg, struct file_lock **flp) { @@ -1529,11 +1569,12 @@ static int __vfs_setlease(struct file *filp, long arg, struct file_lock **lease) int vfs_setlease(struct file *filp, long arg, struct file_lock **lease) { + struct inode *inode = file_inode(filp); int error; - lock_flocks(); + spin_lock(&inode->i_lock); error = __vfs_setlease(filp, arg, lease); - unlock_flocks(); + spin_unlock(&inode->i_lock); return error; } @@ -1551,6 +1592,7 @@ static int do_fcntl_delete_lease(struct file *filp) static int do_fcntl_add_lease(unsigned int fd, struct file *filp, long arg) { struct file_lock *fl, *ret; + struct inode *inode = file_inode(filp); struct fasync_struct *new; int error; @@ -1564,10 +1606,10 @@ static int do_fcntl_add_lease(unsigned int fd, struct file *filp, long arg) return -ENOMEM; } ret = fl; - lock_flocks(); + spin_lock(&inode->i_lock); error = __vfs_setlease(filp, arg, &ret); if (error) { - unlock_flocks(); + spin_unlock(&inode->i_lock); locks_free_lock(fl); goto out_free_fasync; } @@ -1584,7 +1626,7 @@ static int do_fcntl_add_lease(unsigned int fd, struct file *filp, long arg) new = NULL; error = __f_setown(filp, task_pid(current), PIDTYPE_PID, 0); - unlock_flocks(); + spin_unlock(&inode->i_lock); out_free_fasync: if (new) @@ -2108,7 +2150,7 @@ void locks_remove_flock(struct file *filp) fl.fl_ops->fl_release_private(&fl); } - lock_flocks(); + spin_lock(&inode->i_lock); before = &inode->i_flock; while ((fl = *before) != NULL) { @@ -2126,7 +2168,7 @@ void locks_remove_flock(struct file *filp) } before = &fl->fl_next; } - unlock_flocks(); + spin_unlock(&inode->i_lock); } /** @@ -2140,12 +2182,12 @@ posix_unblock_lock(struct file_lock *waiter) { int status = 0; - lock_flocks(); + spin_lock(&file_lock_lock); if (waiter->fl_next) __locks_delete_block(waiter); else status = -ENOENT; - unlock_flocks(); + spin_unlock(&file_lock_lock); return status; } EXPORT_SYMBOL(posix_unblock_lock); @@ -2259,7 +2301,7 @@ static void *locks_start(struct seq_file *f, loff_t *pos) { loff_t *p = f->private; - lock_flocks(); + spin_lock(&file_lock_lock); *p = (*pos + 1); return seq_list_start(&file_lock_list, *pos); } @@ -2273,7 +2315,7 @@ static void *locks_next(struct seq_file *f, void *v, loff_t *pos) static void locks_stop(struct seq_file *f, void *v) { - unlock_flocks(); + spin_unlock(&file_lock_lock); } static const struct seq_operations locks_seq_operations = { @@ -2320,7 +2362,8 @@ int lock_may_read(struct inode *inode, loff_t start, unsigned long len) { struct file_lock *fl; int result = 1; - lock_flocks(); + + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (IS_POSIX(fl)) { if (fl->fl_type == F_RDLCK) @@ -2337,7 +2380,7 @@ int lock_may_read(struct inode *inode, loff_t start, unsigned long len) result = 0; break; } - unlock_flocks(); + spin_unlock(&inode->i_lock); return result; } @@ -2360,7 +2403,8 @@ int lock_may_write(struct inode *inode, loff_t start, unsigned long len) { struct file_lock *fl; int result = 1; - lock_flocks(); + + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (IS_POSIX(fl)) { if ((fl->fl_end < start) || (fl->fl_start > (start + len))) @@ -2375,7 +2419,7 @@ int lock_may_write(struct inode *inode, loff_t start, unsigned long len) result = 0; break; } - unlock_flocks(); + spin_unlock(&inode->i_lock); return result; } diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 57db3244f4d9..7ec4814e298d 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -73,20 +73,20 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ if (inode->i_flock == NULL) goto out; - /* Protect inode->i_flock using the file locks lock */ - lock_flocks(); + /* Protect inode->i_flock using the i_lock */ + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; if (nfs_file_open_context(fl->fl_file) != ctx) continue; - unlock_flocks(); + spin_unlock(&inode->i_lock); status = nfs4_lock_delegation_recall(fl, state, stateid); if (status < 0) goto out; - lock_flocks(); + spin_lock(&inode->i_lock); } - unlock_flocks(); + spin_unlock(&inode->i_lock); out: return status; } diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 1fab140764c4..ff10b4aa534c 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1373,13 +1373,13 @@ static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_ /* Guard against delegation returns and new lock/unlock calls */ down_write(&nfsi->rwsem); /* Protect inode->i_flock using the BKL */ - lock_flocks(); + spin_lock(&inode->i_lock); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; if (nfs_file_open_context(fl->fl_file)->state != state) continue; - unlock_flocks(); + spin_unlock(&inode->i_lock); status = ops->recover_lock(state, fl); switch (status) { case 0: @@ -1406,9 +1406,9 @@ static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_ /* kill_proc(fl->fl_pid, SIGLOST, 1); */ status = 0; } - lock_flocks(); + spin_lock(&inode->i_lock); } - unlock_flocks(); + spin_unlock(&inode->i_lock); out: up_write(&nfsi->rwsem); return status; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 316ec843dec2..f17051838b41 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2645,13 +2645,13 @@ static void nfsd_break_one_deleg(struct nfs4_delegation *dp) list_add_tail(&dp->dl_recall_lru, &nn->del_recall_lru); - /* only place dl_time is set. protected by lock_flocks*/ + /* Only place dl_time is set; protected by i_lock: */ dp->dl_time = get_seconds(); nfsd4_cb_recall(dp); } -/* Called from break_lease() with lock_flocks() held. */ +/* Called from break_lease() with i_lock held. */ static void nfsd_break_deleg_cb(struct file_lock *fl) { struct nfs4_file *fp = (struct nfs4_file *)fl->fl_owner; @@ -4520,7 +4520,7 @@ check_for_locks(struct nfs4_file *filp, struct nfs4_lockowner *lowner) struct inode *inode = filp->fi_inode; int status = 0; - lock_flocks(); + spin_lock(&inode->i_lock); for (flpp = &inode->i_flock; *flpp != NULL; flpp = &(*flpp)->fl_next) { if ((*flpp)->fl_owner == (fl_owner_t)lowner) { status = 1; @@ -4528,7 +4528,7 @@ check_for_locks(struct nfs4_file *filp, struct nfs4_lockowner *lowner) } } out: - unlock_flocks(); + spin_unlock(&inode->i_lock); return status; } diff --git a/include/linux/fs.h b/include/linux/fs.h index ed9fdaaf3223..24fe998795e1 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1024,8 +1024,6 @@ extern int vfs_setlease(struct file *, long, struct file_lock **); extern int lease_modify(struct file_lock **, int); extern int lock_may_read(struct inode *, loff_t start, unsigned long count); extern int lock_may_write(struct inode *, loff_t start, unsigned long count); -extern void lock_flocks(void); -extern void unlock_flocks(void); #else /* !CONFIG_FILE_LOCKING */ static inline int fcntl_getlk(struct file *file, struct flock __user *user) { @@ -1166,15 +1164,6 @@ static inline int lock_may_write(struct inode *inode, loff_t start, { return 1; } - -static inline void lock_flocks(void) -{ -} - -static inline void unlock_flocks(void) -{ -} - #endif /* !CONFIG_FILE_LOCKING */ -- cgit v1.2.3 From 139ca04ee572fea6c0c105e88aba3a534efcd7c4 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jun 2013 08:58:17 -0400 Subject: locks: convert fl_link to a hlist_node Testing has shown that iterating over the blocked_list for deadlock detection turns out to be a bottleneck. In order to alleviate that, begin the process of turning it into a hashtable. We start by turning the fl_link into a hlist_node and the global lists into hlists. A later patch will do the conversion of the blocked_list to a hashtable. Signed-off-by: Jeff Layton Acked-by: J. Bruce Fields Signed-off-by: Al Viro --- fs/locks.c | 24 ++++++++++++------------ include/linux/fs.h | 2 +- 2 files changed, 13 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/fs/locks.c b/fs/locks.c index 84e269fc4c69..941b7146b6be 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -157,13 +157,13 @@ int lease_break_time = 45; * The global file_lock_list is only used for displaying /proc/locks. Protected * by the file_lock_lock. */ -static LIST_HEAD(file_lock_list); +static HLIST_HEAD(file_lock_list); /* * The blocked_list is used to find POSIX lock loops for deadlock detection. * Protected by file_lock_lock. */ -static LIST_HEAD(blocked_list); +static HLIST_HEAD(blocked_list); /* * This lock protects the blocked_list, and the file_lock_list. Generally, if @@ -188,7 +188,7 @@ static struct kmem_cache *filelock_cache __read_mostly; static void locks_init_lock_heads(struct file_lock *fl) { - INIT_LIST_HEAD(&fl->fl_link); + INIT_HLIST_NODE(&fl->fl_link); INIT_LIST_HEAD(&fl->fl_block); init_waitqueue_head(&fl->fl_wait); } @@ -222,7 +222,7 @@ void locks_free_lock(struct file_lock *fl) { BUG_ON(waitqueue_active(&fl->fl_wait)); BUG_ON(!list_empty(&fl->fl_block)); - BUG_ON(!list_empty(&fl->fl_link)); + BUG_ON(!hlist_unhashed(&fl->fl_link)); locks_release_private(fl); kmem_cache_free(filelock_cache, fl); @@ -500,7 +500,7 @@ static inline void locks_insert_global_locks(struct file_lock *fl) { spin_lock(&file_lock_lock); - list_add_tail(&fl->fl_link, &file_lock_list); + hlist_add_head(&fl->fl_link, &file_lock_list); spin_unlock(&file_lock_lock); } @@ -508,20 +508,20 @@ static inline void locks_delete_global_locks(struct file_lock *fl) { spin_lock(&file_lock_lock); - list_del_init(&fl->fl_link); + hlist_del_init(&fl->fl_link); spin_unlock(&file_lock_lock); } static inline void locks_insert_global_blocked(struct file_lock *waiter) { - list_add(&waiter->fl_link, &blocked_list); + hlist_add_head(&waiter->fl_link, &blocked_list); } static inline void locks_delete_global_blocked(struct file_lock *waiter) { - list_del_init(&waiter->fl_link); + hlist_del_init(&waiter->fl_link); } /* Remove waiter from blocker's block list. @@ -748,7 +748,7 @@ static struct file_lock *what_owner_is_waiting_for(struct file_lock *block_fl) { struct file_lock *fl; - list_for_each_entry(fl, &blocked_list, fl_link) { + hlist_for_each_entry(fl, &blocked_list, fl_link) { if (posix_same_owner(fl, block_fl)) return fl->fl_next; } @@ -2300,7 +2300,7 @@ static int locks_show(struct seq_file *f, void *v) { struct file_lock *fl, *bfl; - fl = list_entry(v, struct file_lock, fl_link); + fl = hlist_entry(v, struct file_lock, fl_link); lock_get_status(f, fl, *((loff_t *)f->private), ""); @@ -2316,14 +2316,14 @@ static void *locks_start(struct seq_file *f, loff_t *pos) spin_lock(&file_lock_lock); *p = (*pos + 1); - return seq_list_start(&file_lock_list, *pos); + return seq_hlist_start(&file_lock_list, *pos); } static void *locks_next(struct seq_file *f, void *v, loff_t *pos) { loff_t *p = f->private; ++*p; - return seq_list_next(v, &file_lock_list, pos); + return seq_hlist_next(v, &file_lock_list, pos); } static void locks_stop(struct seq_file *f, void *v) diff --git a/include/linux/fs.h b/include/linux/fs.h index 24fe998795e1..fab064a3b65f 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -946,7 +946,7 @@ int locks_in_grace(struct net *); */ struct file_lock { struct file_lock *fl_next; /* singly linked list for this inode */ - struct list_head fl_link; /* doubly linked list of all locks */ + struct hlist_node fl_link; /* node in global lists */ struct list_head fl_block; /* circular list of blocked processes */ fl_owner_t fl_owner; unsigned int fl_flags; -- cgit v1.2.3 From 3999e49364193f7dbbba66e2be655fe91ba1fced Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jun 2013 08:58:19 -0400 Subject: locks: add a new "lm_owner_key" lock operation Currently, the hashing that the locking code uses to add these values to the blocked_hash is simply calculated using fl_owner field. That's valid in most cases except for server-side lockd, which validates the owner of a lock based on fl_owner and fl_pid. In the case where you have a small number of NFS clients doing a lot of locking between different processes, you could end up with all the blocked requests sitting in a very small number of hash buckets. Add a new lm_owner_key operation to the lock_manager_operations that will generate an unsigned long to use as the key in the hashtable. That function is only implemented for server-side lockd, and simply XORs the fl_owner and fl_pid. Signed-off-by: Jeff Layton Acked-by: J. Bruce Fields Signed-off-by: Al Viro --- Documentation/filesystems/Locking | 16 +++++++++++----- fs/lockd/svclock.c | 12 ++++++++++++ fs/locks.c | 12 ++++++++++-- include/linux/fs.h | 1 + 4 files changed, 34 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index c2963a74fbc3..2db7c9e492e9 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -349,6 +349,7 @@ fl_release_private: maybe no ----------------------- lock_manager_operations --------------------------- prototypes: int (*lm_compare_owner)(struct file_lock *, struct file_lock *); + unsigned long (*lm_owner_key)(struct file_lock *); void (*lm_notify)(struct file_lock *); /* unblock callback */ int (*lm_grant)(struct file_lock *, struct file_lock *, int); void (*lm_break)(struct file_lock *); /* break_lease callback */ @@ -358,16 +359,21 @@ locking rules: inode->i_lock file_lock_lock may block lm_compare_owner: yes[1] maybe no +lm_owner_key yes[1] yes no lm_notify: yes yes no lm_grant: no no no lm_break: yes no no lm_change yes no no -[1]: ->lm_compare_owner is generally called with *an* inode->i_lock held. It -may not be the i_lock of the inode for either file_lock being compared! This is -the case with deadlock detection, since the code has to chase down the owners -of locks that may be entirely unrelated to the one on which the lock is being -acquired. When doing a search for deadlocks, the file_lock_lock is also held. +[1]: ->lm_compare_owner and ->lm_owner_key are generally called with +*an* inode->i_lock held. It may not be the i_lock of the inode +associated with either file_lock argument! This is the case with deadlock +detection, since the code has to chase down the owners of locks that may +be entirely unrelated to the one on which the lock is being acquired. +For deadlock detection however, the file_lock_lock is also held. The +fact that these locks are held ensures that the file_locks do not +disappear out from under you while doing the comparison or generating an +owner key. --------------------------- buffer_head ----------------------------------- prototypes: diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index a469098682c4..067778b0ccc9 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -744,8 +744,20 @@ static int nlmsvc_same_owner(struct file_lock *fl1, struct file_lock *fl2) return fl1->fl_owner == fl2->fl_owner && fl1->fl_pid == fl2->fl_pid; } +/* + * Since NLM uses two "keys" for tracking locks, we need to hash them down + * to one for the blocked_hash. Here, we're just xor'ing the host address + * with the pid in order to create a key value for picking a hash bucket. + */ +static unsigned long +nlmsvc_owner_key(struct file_lock *fl) +{ + return (unsigned long)fl->fl_owner ^ (unsigned long)fl->fl_pid; +} + const struct lock_manager_operations nlmsvc_lock_operations = { .lm_compare_owner = nlmsvc_same_owner, + .lm_owner_key = nlmsvc_owner_key, .lm_notify = nlmsvc_notify_blocked, .lm_grant = nlmsvc_grant_deferred, }; diff --git a/fs/locks.c b/fs/locks.c index 71d847cbbb63..6242e0b1c69c 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -521,10 +521,18 @@ locks_delete_global_locks(struct file_lock *fl) spin_unlock(&file_lock_lock); } +static unsigned long +posix_owner_key(struct file_lock *fl) +{ + if (fl->fl_lmops && fl->fl_lmops->lm_owner_key) + return fl->fl_lmops->lm_owner_key(fl); + return (unsigned long)fl->fl_owner; +} + static inline void locks_insert_global_blocked(struct file_lock *waiter) { - hash_add(blocked_hash, &waiter->fl_link, (unsigned long)waiter->fl_owner); + hash_add(blocked_hash, &waiter->fl_link, posix_owner_key(waiter)); } static inline void @@ -757,7 +765,7 @@ static struct file_lock *what_owner_is_waiting_for(struct file_lock *block_fl) { struct file_lock *fl; - hash_for_each_possible(blocked_hash, fl, fl_link, (unsigned long)block_fl->fl_owner) { + hash_for_each_possible(blocked_hash, fl, fl_link, posix_owner_key(block_fl)) { if (posix_same_owner(fl, block_fl)) return fl->fl_next; } diff --git a/include/linux/fs.h b/include/linux/fs.h index fab064a3b65f..a137a73fc1fe 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -908,6 +908,7 @@ struct file_lock_operations { struct lock_manager_operations { int (*lm_compare_owner)(struct file_lock *, struct file_lock *); + unsigned long (*lm_owner_key)(struct file_lock *); void (*lm_notify)(struct file_lock *); /* unblock callback */ int (*lm_grant)(struct file_lock *, struct file_lock *, int); void (*lm_break)(struct file_lock *); -- cgit v1.2.3 From 46a1c2c7ae53de2a5676754b54a73c591a3951d2 Mon Sep 17 00:00:00 2001 From: Jie Liu Date: Tue, 25 Jun 2013 12:02:13 +0800 Subject: vfs: export lseek_execute() to modules For those file systems(btrfs/ext4/ocfs2/tmpfs) that support SEEK_DATA/SEEK_HOLE functions, we end up handling the similar matter in lseek_execute() to update the current file offset to the desired offset if it is valid, ceph also does the simliar things at ceph_llseek(). To reduce the duplications, this patch make lseek_execute() public accessible so that we can call it directly from the underlying file systems. Thanks Dave Chinner for this suggestion. [AV: call it vfs_setpos(), don't bring the removed 'inode' argument back] v2->v1: - Add kernel-doc comments for lseek_execute() - Call lseek_execute() in ceph->llseek() Signed-off-by: Jie Liu Cc: Dave Chinner Cc: Al Viro Cc: Andi Kleen Cc: Andrew Morton Cc: Christoph Hellwig Cc: Chris Mason Cc: Josef Bacik Cc: Ben Myers Cc: Ted Tso Cc: Hugh Dickins Cc: Mark Fasheh Cc: Joel Becker Cc: Sage Weil Signed-off-by: Al Viro --- fs/btrfs/file.c | 15 +-------------- fs/ceph/file.c | 11 +---------- fs/ext4/file.c | 24 ++---------------------- fs/ocfs2/file.c | 12 +----------- fs/read_write.c | 19 ++++++++++++++++--- fs/xfs/xfs_file.c | 6 ++---- include/linux/fs.h | 1 + mm/shmem.c | 5 +---- 8 files changed, 25 insertions(+), 68 deletions(-) (limited to 'include') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 4205ba752d40..89da56a58b63 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2425,20 +2425,7 @@ static loff_t btrfs_file_llseek(struct file *file, loff_t offset, int whence) } } - if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) { - offset = -EINVAL; - goto out; - } - if (offset > inode->i_sb->s_maxbytes) { - offset = -EINVAL; - goto out; - } - - /* Special lock needed here? */ - if (offset != file->f_pos) { - file->f_pos = offset; - file->f_version = 0; - } + offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes); out: mutex_unlock(&inode->i_mutex); return offset; diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 656e16907430..16c989d3e23c 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -866,16 +866,7 @@ static loff_t ceph_llseek(struct file *file, loff_t offset, int whence) break; } - if (offset < 0 || offset > inode->i_sb->s_maxbytes) { - offset = -EINVAL; - goto out; - } - - /* Special lock needed here? */ - if (offset != file->f_pos) { - file->f_pos = offset; - file->f_version = 0; - } + offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes); out: mutex_unlock(&inode->i_mutex); diff --git a/fs/ext4/file.c b/fs/ext4/file.c index b1b4d51b5d86..469361dbe619 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -494,17 +494,7 @@ static loff_t ext4_seek_data(struct file *file, loff_t offset, loff_t maxsize) if (dataoff > isize) return -ENXIO; - if (dataoff < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) - return -EINVAL; - if (dataoff > maxsize) - return -EINVAL; - - if (dataoff != file->f_pos) { - file->f_pos = dataoff; - file->f_version = 0; - } - - return dataoff; + return vfs_setpos(file, dataoff, maxsize); } /* @@ -580,17 +570,7 @@ static loff_t ext4_seek_hole(struct file *file, loff_t offset, loff_t maxsize) if (holeoff > isize) holeoff = isize; - if (holeoff < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) - return -EINVAL; - if (holeoff > maxsize) - return -EINVAL; - - if (holeoff != file->f_pos) { - file->f_pos = holeoff; - file->f_version = 0; - } - - return holeoff; + return vfs_setpos(file, holeoff, maxsize); } /* diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 8a38714f1d92..41000f223ca4 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2646,17 +2646,7 @@ static loff_t ocfs2_file_llseek(struct file *file, loff_t offset, int whence) goto out; } - if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) - ret = -EINVAL; - if (!ret && offset > inode->i_sb->s_maxbytes) - ret = -EINVAL; - if (ret) - goto out; - - if (offset != file->f_pos) { - file->f_pos = offset; - file->f_version = 0; - } + offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes); out: mutex_unlock(&inode->i_mutex); diff --git a/fs/read_write.c b/fs/read_write.c index 37d16e82b575..122a3846d9e1 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -41,7 +41,19 @@ static inline int unsigned_offsets(struct file *file) return file->f_mode & FMODE_UNSIGNED_OFFSET; } -static loff_t lseek_execute(struct file *file, loff_t offset, loff_t maxsize) +/** + * vfs_setpos - update the file offset for lseek + * @file: file structure in question + * @offset: file offset to seek to + * @maxsize: maximum file size + * + * This is a low-level filesystem helper for updating the file offset to + * the value specified by @offset if the given offset is valid and it is + * not equal to the current file offset. + * + * Return the specified offset on success and -EINVAL on invalid offset. + */ +loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize) { if (offset < 0 && !unsigned_offsets(file)) return -EINVAL; @@ -54,6 +66,7 @@ static loff_t lseek_execute(struct file *file, loff_t offset, loff_t maxsize) } return offset; } +EXPORT_SYMBOL(vfs_setpos); /** * generic_file_llseek_size - generic llseek implementation for regular files @@ -94,7 +107,7 @@ generic_file_llseek_size(struct file *file, loff_t offset, int whence, * like SEEK_SET. */ spin_lock(&file->f_lock); - offset = lseek_execute(file, file->f_pos + offset, maxsize); + offset = vfs_setpos(file, file->f_pos + offset, maxsize); spin_unlock(&file->f_lock); return offset; case SEEK_DATA: @@ -116,7 +129,7 @@ generic_file_llseek_size(struct file *file, loff_t offset, int whence, break; } - return lseek_execute(file, offset, maxsize); + return vfs_setpos(file, offset, maxsize); } EXPORT_SYMBOL(generic_file_llseek_size); diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 0ad2b95fca12..de3dc98f4e8f 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -1268,8 +1268,7 @@ xfs_seek_data( } out: - if (offset != file->f_pos) - file->f_pos = offset; + offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes); out_unlock: xfs_iunlock_map_shared(ip, lock); @@ -1377,8 +1376,7 @@ out: * situation in particular. */ offset = min_t(loff_t, offset, isize); - if (offset != file->f_pos) - file->f_pos = offset; + offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes); out_unlock: xfs_iunlock_map_shared(ip, lock); diff --git a/include/linux/fs.h b/include/linux/fs.h index a137a73fc1fe..bccb1924ec93 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2426,6 +2426,7 @@ extern void file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping); extern loff_t noop_llseek(struct file *file, loff_t offset, int whence); extern loff_t no_llseek(struct file *file, loff_t offset, int whence); +extern loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); extern loff_t generic_file_llseek_size(struct file *file, loff_t offset, int whence, loff_t maxsize, loff_t eof); diff --git a/mm/shmem.c b/mm/shmem.c index f887358dabc5..118dfa4952f4 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -1798,10 +1798,7 @@ static loff_t shmem_file_llseek(struct file *file, loff_t offset, int whence) } } - if (offset >= 0 && offset != file->f_pos) { - file->f_pos = offset; - file->f_version = 0; - } + offset = vfs_setpos(file, offset, MAX_LFS_FILESIZE); mutex_unlock(&inode->i_mutex); return offset; } -- cgit v1.2.3