1 The text below describes the locking rules for VFS-related methods.
2 It is (believed to be) up-to-date. *Please*, if you change anything in
3 prototypes or locking protocols - update this file. And update the relevant
4 instances in the tree, don't leave that to maintainers of filesystems/devices/
5 etc. At the very least, put the list of dubious cases in the end of this file.
6 Don't turn it into log - maintainers of out-of-the-tree code are supposed to
7 be able to use diff(1).
8 Thing currently missing here: socket operations. Alexey?
10 --------------------------- dentry_operations --------------------------
12 int (*d_revalidate)(struct dentry *, int);
13 int (*d_hash) (struct dentry *, struct qstr *);
14 int (*d_compare) (struct dentry *, struct qstr *, struct qstr *);
15 int (*d_delete)(struct dentry *);
16 void (*d_release)(struct dentry *);
17 void (*d_iput)(struct dentry *, struct inode *);
18 char *(*d_dname)((struct dentry *dentry, char *buffer, int buflen);
22 dcache_lock rename_lock ->d_lock may block
23 d_revalidate: no no no yes
25 d_compare: no yes no no
26 d_delete: yes no yes no
27 d_release: no no no yes
31 --------------------------- inode_operations ---------------------------
33 int (*create) (struct inode *,struct dentry *,int, struct nameidata *);
34 struct dentry * (*lookup) (struct inode *,struct dentry *, struct nameid
36 int (*link) (struct dentry *,struct inode *,struct dentry *);
37 int (*unlink) (struct inode *,struct dentry *);
38 int (*symlink) (struct inode *,struct dentry *,const char *);
39 int (*mkdir) (struct inode *,struct dentry *,int);
40 int (*rmdir) (struct inode *,struct dentry *);
41 int (*mknod) (struct inode *,struct dentry *,int,dev_t);
42 int (*rename) (struct inode *, struct dentry *,
43 struct inode *, struct dentry *);
44 int (*readlink) (struct dentry *, char __user *,int);
45 int (*follow_link) (struct dentry *, struct nameidata *);
46 void (*truncate) (struct inode *);
47 int (*permission) (struct inode *, int, struct nameidata *);
48 int (*setattr) (struct dentry *, struct iattr *);
49 int (*getattr) (struct vfsmount *, struct dentry *, struct kstat *);
50 int (*setxattr) (struct dentry *, const char *,const void *,size_t,int);
51 ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t);
52 ssize_t (*listxattr) (struct dentry *, char *, size_t);
53 int (*removexattr) (struct dentry *, const char *);
56 all may block, none have BKL
65 rmdir: yes (both) (see below)
66 rename: yes (all) (see below)
69 truncate: yes (see below)
77 Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
79 cross-directory ->rename() has (per-superblock) ->s_vfs_rename_sem.
80 ->truncate() is never called directly - it's a callback, not a
81 method. It's called by vmtruncate() - library function normally used by
82 ->setattr(). Locking information above applies to that call (i.e. is
83 inherited from ->setattr() - vmtruncate() is used when ATTR_SIZE had been
86 See Documentation/filesystems/directory-locking for more detailed discussion
87 of the locking scheme for directory operations.
89 --------------------------- super_operations ---------------------------
91 struct inode *(*alloc_inode)(struct super_block *sb);
92 void (*destroy_inode)(struct inode *);
93 void (*dirty_inode) (struct inode *);
94 int (*write_inode) (struct inode *, int);
95 void (*drop_inode) (struct inode *);
96 void (*delete_inode) (struct inode *);
97 void (*put_super) (struct super_block *);
98 void (*write_super) (struct super_block *);
99 int (*sync_fs)(struct super_block *sb, int wait);
100 int (*freeze_fs) (struct super_block *);
101 int (*unfreeze_fs) (struct super_block *);
102 int (*statfs) (struct dentry *, struct kstatfs *);
103 int (*remount_fs) (struct super_block *, int *, char *);
104 void (*clear_inode) (struct inode *);
105 void (*umount_begin) (struct super_block *);
106 int (*show_options)(struct seq_file *, struct vfsmount *);
107 ssize_t (*quota_read)(struct super_block *, int, char *, size_t, loff_t);
108 ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t);
116 dirty_inode: (must not sleep)
118 drop_inode: !!!inode_lock!!!
126 remount_fs: maybe (see below)
129 show_options: no (namespace_sem)
130 quota_read: no (see below)
131 quota_write: no (see below)
133 ->remount_fs() will have the s_umount exclusive lock if it's already mounted.
134 When called from get_sb_single, it does NOT have the s_umount lock.
135 ->quota_read() and ->quota_write() functions are both guaranteed to
136 be the only ones operating on the quota file by the quota code (via
137 dqio_sem) (unless an admin really wants to screw up something and
138 writes to quota files with quotas on). For other details about locking
139 see also dquot_operations section.
141 --------------------------- file_system_type ---------------------------
143 int (*get_sb) (struct file_system_type *, int,
144 const char *, void *, struct vfsmount *);
145 void (*kill_sb) (struct super_block *);
151 ->get_sb() returns error or 0 with locked superblock attached to the vfsmount
152 (exclusive on ->s_umount).
153 ->kill_sb() takes a write-locked superblock, does all shutdown work on it,
154 unlocks and drops the reference.
156 --------------------------- address_space_operations --------------------------
158 int (*writepage)(struct page *page, struct writeback_control *wbc);
159 int (*readpage)(struct file *, struct page *);
160 int (*sync_page)(struct page *);
161 int (*writepages)(struct address_space *, struct writeback_control *);
162 int (*set_page_dirty)(struct page *page);
163 int (*readpages)(struct file *filp, struct address_space *mapping,
164 struct list_head *pages, unsigned nr_pages);
165 int (*write_begin)(struct file *, struct address_space *mapping,
166 loff_t pos, unsigned len, unsigned flags,
167 struct page **pagep, void **fsdata);
168 int (*write_end)(struct file *, struct address_space *mapping,
169 loff_t pos, unsigned len, unsigned copied,
170 struct page *page, void *fsdata);
171 sector_t (*bmap)(struct address_space *, sector_t);
172 int (*invalidatepage) (struct page *, unsigned long);
173 int (*releasepage) (struct page *, int);
174 int (*direct_IO)(int, struct kiocb *, const struct iovec *iov,
175 loff_t offset, unsigned long nr_segs);
176 int (*launder_page) (struct page *);
179 All except set_page_dirty may block
181 BKL PageLocked(page) i_mutex
182 writepage: no yes, unlocks (see below)
183 readpage: no yes, unlocks
188 write_begin: no locks the page yes
189 write_end: no yes, unlocks yes
190 perform_write: no n/a yes
192 invalidatepage: no yes
197 ->write_begin(), ->write_end(), ->sync_page() and ->readpage()
198 may be called from the request handler (/dev/loop).
200 ->readpage() unlocks the page, either synchronously or via I/O
203 ->readpages() populates the pagecache with the passed pages and starts
204 I/O against them. They come unlocked upon I/O completion.
206 ->writepage() is used for two purposes: for "memory cleansing" and for
207 "sync". These are quite different operations and the behaviour may differ
208 depending upon the mode.
210 If writepage is called for sync (wbc->sync_mode != WBC_SYNC_NONE) then
211 it *must* start I/O against the page, even if that would involve
212 blocking on in-progress I/O.
214 If writepage is called for memory cleansing (sync_mode ==
215 WBC_SYNC_NONE) then its role is to get as much writeout underway as
216 possible. So writepage should try to avoid blocking against
217 currently-in-progress I/O.
219 If the filesystem is not called for "sync" and it determines that it
220 would need to block against in-progress I/O to be able to start new I/O
221 against the page the filesystem should redirty the page with
222 redirty_page_for_writepage(), then unlock the page and return zero.
223 This may also be done to avoid internal deadlocks, but rarely.
225 If the filesystem is called for sync then it must wait on any
226 in-progress I/O and then start new I/O.
228 The filesystem should unlock the page synchronously, before returning to the
229 caller, unless ->writepage() returns special WRITEPAGE_ACTIVATE
230 value. WRITEPAGE_ACTIVATE means that page cannot really be written out
231 currently, and VM should stop calling ->writepage() on this page for some
232 time. VM does this by moving page to the head of the active list, hence the
235 Unless the filesystem is going to redirty_page_for_writepage(), unlock the page
236 and return zero, writepage *must* run set_page_writeback() against the page,
237 followed by unlocking it. Once set_page_writeback() has been run against the
238 page, write I/O can be submitted and the write I/O completion handler must run
239 end_page_writeback() once the I/O is complete. If no I/O is submitted, the
240 filesystem must run end_page_writeback() against the page before returning from
243 That is: after 2.5.12, pages which are under writeout are *not* locked. Note,
244 if the filesystem needs the page to be locked during writeout, that is ok, too,
245 the page is allowed to be unlocked at any point in time between the calls to
246 set_page_writeback() and end_page_writeback().
248 Note, failure to run either redirty_page_for_writepage() or the combination of
249 set_page_writeback()/end_page_writeback() on a page submitted to writepage
250 will leave the page itself marked clean but it will be tagged as dirty in the
251 radix tree. This incoherency can lead to all sorts of hard-to-debug problems
252 in the filesystem like having dirty inodes at umount and losing written data.
254 ->sync_page() locking rules are not well-defined - usually it is called
255 with lock on page, but that is not guaranteed. Considering the currently
256 existing instances of this method ->sync_page() itself doesn't look
259 ->writepages() is used for periodic writeback and for syscall-initiated
260 sync operations. The address_space should start I/O against at least
261 *nr_to_write pages. *nr_to_write must be decremented for each page which is
262 written. The address_space implementation may write more (or less) pages
263 than *nr_to_write asks for, but it should try to be reasonably close. If
264 nr_to_write is NULL, all dirty pages must be written.
266 writepages should _only_ write pages which are present on
269 ->set_page_dirty() is called from various places in the kernel
270 when the target page is marked as needing writeback. It may be called
271 under spinlock (it cannot block) and is sometimes called with the page
274 ->bmap() is currently used by legacy ioctl() (FIBMAP) provided by some
275 filesystems and by the swapper. The latter will eventually go away. All
276 instances do not actually need the BKL. Please, keep it that way and don't
279 ->invalidatepage() is called when the filesystem must attempt to drop
280 some or all of the buffers from the page when it is being truncated. It
281 returns zero on success. If ->invalidatepage is zero, the kernel uses
282 block_invalidatepage() instead.
284 ->releasepage() is called when the kernel is about to try to drop the
285 buffers from the page in preparation for freeing it. It returns zero to
286 indicate that the buffers are (or may be) freeable. If ->releasepage is zero,
287 the kernel assumes that the fs has no private interest in the buffers.
289 ->launder_page() may be called prior to releasing a page if
290 it is still found to be dirty. It returns zero if the page was successfully
291 cleaned, or an error value if not. Note that in order to prevent the page
292 getting mapped back in and redirtied, it needs to be kept locked
293 across the entire operation.
295 Note: currently almost all instances of address_space methods are
296 using BKL for internal serialization and that's one of the worst sources
297 of contention. Normally they are calling library functions (in fs/buffer.c)
298 and pass foo_get_block() as a callback (on local block-based filesystems,
299 indeed). BKL is not needed for library stuff and is usually taken by
300 foo_get_block(). It's an overkill, since block bitmaps can be protected by
301 internal fs locking and real critical areas are much smaller than the areas
302 filesystems protect now.
304 ----------------------- file_lock_operations ------------------------------
306 void (*fl_insert)(struct file_lock *); /* lock insertion callback */
307 void (*fl_remove)(struct file_lock *); /* lock removal callback */
308 void (*fl_copy_lock)(struct file_lock *, struct file_lock *);
309 void (*fl_release_private)(struct file_lock *);
317 fl_release_private: yes yes
319 ----------------------- lock_manager_operations ---------------------------
321 int (*fl_compare_owner)(struct file_lock *, struct file_lock *);
322 void (*fl_notify)(struct file_lock *); /* unblock callback */
323 void (*fl_copy_lock)(struct file_lock *, struct file_lock *);
324 void (*fl_release_private)(struct file_lock *);
325 void (*fl_break)(struct file_lock *); /* break_lease callback */
329 fl_compare_owner: yes no
332 fl_release_private: yes yes
335 Currently only NFSD and NLM provide instances of this class. None of the
336 them block. If you have out-of-tree instances - please, show up. Locking
337 in that area will change.
338 --------------------------- buffer_head -----------------------------------
340 void (*b_end_io)(struct buffer_head *bh, int uptodate);
343 called from interrupts. In other words, extreme care is needed here.
344 bh is locked, but that's all warranties we have here. Currently only RAID1,
345 highmem, fs/buffer.c, and fs/ntfs/aops.c are providing these. Block devices
346 call this method upon the IO completion.
348 --------------------------- block_device_operations -----------------------
350 int (*open) (struct inode *, struct file *);
351 int (*release) (struct inode *, struct file *);
352 int (*ioctl) (struct inode *, struct file *, unsigned, unsigned long);
353 int (*media_changed) (struct gendisk *);
354 int (*revalidate_disk) (struct gendisk *);
362 revalidate_disk: no no
364 The last two are called only from check_disk_change().
366 --------------------------- file_operations -------------------------------
368 loff_t (*llseek) (struct file *, loff_t, int);
369 ssize_t (*read) (struct file *, char __user *, size_t, loff_t *);
370 ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *);
371 ssize_t (*aio_read) (struct kiocb *, const struct iovec *, unsigned long, loff_t);
372 ssize_t (*aio_write) (struct kiocb *, const struct iovec *, unsigned long, loff_t);
373 int (*readdir) (struct file *, void *, filldir_t);
374 unsigned int (*poll) (struct file *, struct poll_table_struct *);
375 int (*ioctl) (struct inode *, struct file *, unsigned int,
377 long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long);
378 long (*compat_ioctl) (struct file *, unsigned int, unsigned long);
379 int (*mmap) (struct file *, struct vm_area_struct *);
380 int (*open) (struct inode *, struct file *);
381 int (*flush) (struct file *);
382 int (*release) (struct inode *, struct file *);
383 int (*fsync) (struct file *, int datasync);
384 int (*aio_fsync) (struct kiocb *, int datasync);
385 int (*fasync) (int, struct file *, int);
386 int (*lock) (struct file *, int, struct file_lock *);
387 ssize_t (*readv) (struct file *, const struct iovec *, unsigned long,
389 ssize_t (*writev) (struct file *, const struct iovec *, unsigned long,
391 ssize_t (*sendfile) (struct file *, loff_t *, size_t, read_actor_t,
393 ssize_t (*sendpage) (struct file *, struct page *, int, size_t,
395 unsigned long (*get_unmapped_area)(struct file *, unsigned long,
396 unsigned long, unsigned long, unsigned long);
397 int (*check_flags)(int);
403 llseek: no (see below)
410 ioctl: yes (see below)
411 unlocked_ioctl: no (see below)
417 fsync: no (see below)
425 get_unmapped_area: no
428 ->llseek() locking has moved from llseek to the individual llseek
429 implementations. If your fs is not using generic_file_llseek, you
430 need to acquire and release the appropriate locks in your ->llseek().
431 For many filesystems, it is probably safe to acquire the inode
432 mutex or just to use i_size_read() instead.
433 Note: this does not protect the file->f_pos against concurrent modifications
434 since this is something the userspace has to take care about.
436 Note: ext2_release() was *the* source of contention on fs-intensive
437 loads and dropping BKL on ->release() helps to get rid of that (we still
438 grab BKL for cases when we close a file that had been opened r/w, but that
439 can and should be done using the internal locking with smaller critical areas).
440 Current worst offender is ext2_get_block()...
442 ->fasync() is called without BKL protection, and is responsible for
443 maintaining the FASYNC bit in filp->f_flags. Most instances call
444 fasync_helper(), which does that maintenance, so it's not normally
445 something one needs to worry about. Return values > 0 will be mapped to
446 zero in the VFS layer.
448 ->readdir() and ->ioctl() on directories must be changed. Ideally we would
449 move ->readdir() to inode_operations and use a separate method for directory
450 ->ioctl() or kill the latter completely. One of the problems is that for
451 anything that resembles union-mount we won't have a struct file for all
452 components. And there are other reasons why the current interface is a mess...
454 ->ioctl() on regular files is superceded by the ->unlocked_ioctl() that
455 doesn't take the BKL.
457 ->read on directories probably must go away - we should just enforce -EISDIR
458 in sys_read() and friends.
460 ->fsync() has i_mutex on inode.
462 --------------------------- dquot_operations -------------------------------
464 int (*write_dquot) (struct dquot *);
465 int (*acquire_dquot) (struct dquot *);
466 int (*release_dquot) (struct dquot *);
467 int (*mark_dirty) (struct dquot *);
468 int (*write_info) (struct super_block *, int);
470 These operations are intended to be more or less wrapping functions that ensure
471 a proper locking wrt the filesystem and call the generic quota operations.
473 What filesystem should expect from the generic quota functions:
475 FS recursion Held locks when called
476 write_dquot: yes dqonoff_sem or dqptr_sem
477 acquire_dquot: yes dqonoff_sem or dqptr_sem
478 release_dquot: yes dqonoff_sem or dqptr_sem
480 write_info: yes dqonoff_sem
482 FS recursion means calling ->quota_read() and ->quota_write() from superblock
485 More details about quota locking can be found in fs/dquot.c.
487 --------------------------- vm_operations_struct -----------------------------
489 void (*open)(struct vm_area_struct*);
490 void (*close)(struct vm_area_struct*);
491 int (*fault)(struct vm_area_struct*, struct vm_fault *);
492 int (*page_mkwrite)(struct vm_area_struct *, struct vm_fault *);
493 int (*access)(struct vm_area_struct *, unsigned long, void*, int, int);
496 BKL mmap_sem PageLocked(page)
499 fault: no yes can return with page locked
500 page_mkwrite: no yes can return with page locked
503 ->fault() is called when a previously not present pte is about
504 to be faulted in. The filesystem must find and return the page associated
505 with the passed in "pgoff" in the vm_fault structure. If it is possible that
506 the page may be truncated and/or invalidated, then the filesystem must lock
507 the page, then ensure it is not already truncated (the page lock will block
508 subsequent truncate), and then return with VM_FAULT_LOCKED, and the page
509 locked. The VM will unlock the page.
511 ->page_mkwrite() is called when a previously read-only pte is
512 about to become writeable. The filesystem again must ensure that there are
513 no truncate/invalidate races, and then return with the page locked. If
514 the page has been truncated, the filesystem should not look up a new page
515 like the ->fault() handler, but simply return with VM_FAULT_NOPAGE, which
516 will cause the VM to retry the fault.
518 ->access() is called when get_user_pages() fails in
519 acces_process_vm(), typically used to debug a process through
520 /proc/pid/mem or ptrace. This function is needed only for
521 VM_IO | VM_PFNMAP VMAs.
523 ================================================================================
526 (if you break something or notice that it is broken and do not fix it yourself
527 - at least put it here)
529 ipc/shm.c::shm_delete() - may need BKL.
530 ->read() and ->write() in many drivers are (probably) missing BKL.