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 *, unsigned int);
13 int (*d_hash)(const struct dentry *, const struct inode *,
15 int (*d_compare)(const struct dentry *, const struct inode *,
16 const struct dentry *, const struct inode *,
17 unsigned int, const char *, const struct qstr *);
18 int (*d_delete)(struct dentry *);
19 void (*d_release)(struct dentry *);
20 void (*d_iput)(struct dentry *, struct inode *);
21 char *(*d_dname)((struct dentry *dentry, char *buffer, int buflen);
22 struct vfsmount *(*d_automount)(struct path *path);
23 int (*d_manage)(struct dentry *, bool);
26 rename_lock ->d_lock may block rcu-walk
27 d_revalidate: no no yes (ref-walk) maybe
29 d_compare: yes no no maybe
30 d_delete: no yes no no
31 d_release: no no yes no
35 d_automount: no no yes no
36 d_manage: no no yes (ref-walk) maybe
38 --------------------------- inode_operations ---------------------------
40 int (*create) (struct inode *,struct dentry *,umode_t, bool);
41 struct dentry * (*lookup) (struct inode *,struct dentry *, unsigned int);
42 int (*link) (struct dentry *,struct inode *,struct dentry *);
43 int (*unlink) (struct inode *,struct dentry *);
44 int (*symlink) (struct inode *,struct dentry *,const char *);
45 int (*mkdir) (struct inode *,struct dentry *,umode_t);
46 int (*rmdir) (struct inode *,struct dentry *);
47 int (*mknod) (struct inode *,struct dentry *,umode_t,dev_t);
48 int (*rename) (struct inode *, struct dentry *,
49 struct inode *, struct dentry *);
50 int (*readlink) (struct dentry *, char __user *,int);
51 void * (*follow_link) (struct dentry *, struct nameidata *);
52 void (*put_link) (struct dentry *, struct nameidata *, void *);
53 void (*truncate) (struct inode *);
54 int (*permission) (struct inode *, int, unsigned int);
55 int (*get_acl)(struct inode *, int);
56 int (*setattr) (struct dentry *, struct iattr *);
57 int (*getattr) (struct vfsmount *, struct dentry *, struct kstat *);
58 int (*setxattr) (struct dentry *, const char *,const void *,size_t,int);
59 ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t);
60 ssize_t (*listxattr) (struct dentry *, char *, size_t);
61 int (*removexattr) (struct dentry *, const char *);
62 int (*fiemap)(struct inode *, struct fiemap_extent_info *, u64 start, u64 len);
63 void (*update_time)(struct inode *, struct timespec *, int);
64 int (*atomic_open)(struct inode *, struct dentry *,
65 struct file *, unsigned open_flag,
66 umode_t create_mode, int *opened);
78 rmdir: yes (both) (see below)
79 rename: yes (all) (see below)
83 truncate: yes (see below)
85 permission: no (may not block if called in rcu-walk mode)
96 Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
98 cross-directory ->rename() has (per-superblock) ->s_vfs_rename_sem.
99 ->truncate() is never called directly - it's a callback, not a
100 method. It's called by vmtruncate() - deprecated library function used by
101 ->setattr(). Locking information above applies to that call (i.e. is
102 inherited from ->setattr() - vmtruncate() is used when ATTR_SIZE had been
105 See Documentation/filesystems/directory-locking for more detailed discussion
106 of the locking scheme for directory operations.
108 --------------------------- super_operations ---------------------------
110 struct inode *(*alloc_inode)(struct super_block *sb);
111 void (*destroy_inode)(struct inode *);
112 void (*dirty_inode) (struct inode *, int flags);
113 int (*write_inode) (struct inode *, struct writeback_control *wbc);
114 int (*drop_inode) (struct inode *);
115 void (*evict_inode) (struct inode *);
116 void (*put_super) (struct super_block *);
117 void (*write_super) (struct super_block *);
118 int (*sync_fs)(struct super_block *sb, int wait);
119 int (*freeze_fs) (struct super_block *);
120 int (*unfreeze_fs) (struct super_block *);
121 int (*statfs) (struct dentry *, struct kstatfs *);
122 int (*remount_fs) (struct super_block *, int *, char *);
123 void (*umount_begin) (struct super_block *);
124 int (*show_options)(struct seq_file *, struct dentry *);
125 ssize_t (*quota_read)(struct super_block *, int, char *, size_t, loff_t);
126 ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t);
127 int (*bdev_try_to_free_page)(struct super_block*, struct page*, gfp_t);
130 All may block [not true, see below]
136 drop_inode: !!!inode->i_lock!!!
143 statfs: maybe(read) (see below)
146 show_options: no (namespace_sem)
147 quota_read: no (see below)
148 quota_write: no (see below)
149 bdev_try_to_free_page: no (see below)
151 ->statfs() has s_umount (shared) when called by ustat(2) (native or
152 compat), but that's an accident of bad API; s_umount is used to pin
153 the superblock down when we only have dev_t given us by userland to
154 identify the superblock. Everything else (statfs(), fstatfs(), etc.)
155 doesn't hold it when calling ->statfs() - superblock is pinned down
156 by resolving the pathname passed to syscall.
157 ->quota_read() and ->quota_write() functions are both guaranteed to
158 be the only ones operating on the quota file by the quota code (via
159 dqio_sem) (unless an admin really wants to screw up something and
160 writes to quota files with quotas on). For other details about locking
161 see also dquot_operations section.
162 ->bdev_try_to_free_page is called from the ->releasepage handler of
163 the block device inode. See there for more details.
165 --------------------------- file_system_type ---------------------------
167 int (*get_sb) (struct file_system_type *, int,
168 const char *, void *, struct vfsmount *);
169 struct dentry *(*mount) (struct file_system_type *, int,
170 const char *, void *);
171 void (*kill_sb) (struct super_block *);
177 ->mount() returns ERR_PTR or the root dentry; its superblock should be locked
179 ->kill_sb() takes a write-locked superblock, does all shutdown work on it,
180 unlocks and drops the reference.
182 --------------------------- address_space_operations --------------------------
184 int (*writepage)(struct page *page, struct writeback_control *wbc);
185 int (*readpage)(struct file *, struct page *);
186 int (*sync_page)(struct page *);
187 int (*writepages)(struct address_space *, struct writeback_control *);
188 int (*set_page_dirty)(struct page *page);
189 int (*readpages)(struct file *filp, struct address_space *mapping,
190 struct list_head *pages, unsigned nr_pages);
191 int (*write_begin)(struct file *, struct address_space *mapping,
192 loff_t pos, unsigned len, unsigned flags,
193 struct page **pagep, void **fsdata);
194 int (*write_end)(struct file *, struct address_space *mapping,
195 loff_t pos, unsigned len, unsigned copied,
196 struct page *page, void *fsdata);
197 sector_t (*bmap)(struct address_space *, sector_t);
198 int (*invalidatepage) (struct page *, unsigned long);
199 int (*releasepage) (struct page *, int);
200 void (*freepage)(struct page *);
201 int (*direct_IO)(int, struct kiocb *, const struct iovec *iov,
202 loff_t offset, unsigned long nr_segs);
203 int (*get_xip_mem)(struct address_space *, pgoff_t, int, void **,
205 int (*migratepage)(struct address_space *, struct page *, struct page *);
206 int (*launder_page)(struct page *);
207 int (*is_partially_uptodate)(struct page *, read_descriptor_t *, unsigned long);
208 int (*error_remove_page)(struct address_space *, struct page *);
209 int (*swap_activate)(struct file *);
210 int (*swap_deactivate)(struct file *);
213 All except set_page_dirty and freepage may block
215 PageLocked(page) i_mutex
216 writepage: yes, unlocks (see below)
217 readpage: yes, unlocks
222 write_begin: locks the page yes
223 write_end: yes, unlocks yes
230 migratepage: yes (both)
232 is_partially_uptodate: yes
233 error_remove_page: yes
237 ->write_begin(), ->write_end(), ->sync_page() and ->readpage()
238 may be called from the request handler (/dev/loop).
240 ->readpage() unlocks the page, either synchronously or via I/O
243 ->readpages() populates the pagecache with the passed pages and starts
244 I/O against them. They come unlocked upon I/O completion.
246 ->writepage() is used for two purposes: for "memory cleansing" and for
247 "sync". These are quite different operations and the behaviour may differ
248 depending upon the mode.
250 If writepage is called for sync (wbc->sync_mode != WBC_SYNC_NONE) then
251 it *must* start I/O against the page, even if that would involve
252 blocking on in-progress I/O.
254 If writepage is called for memory cleansing (sync_mode ==
255 WBC_SYNC_NONE) then its role is to get as much writeout underway as
256 possible. So writepage should try to avoid blocking against
257 currently-in-progress I/O.
259 If the filesystem is not called for "sync" and it determines that it
260 would need to block against in-progress I/O to be able to start new I/O
261 against the page the filesystem should redirty the page with
262 redirty_page_for_writepage(), then unlock the page and return zero.
263 This may also be done to avoid internal deadlocks, but rarely.
265 If the filesystem is called for sync then it must wait on any
266 in-progress I/O and then start new I/O.
268 The filesystem should unlock the page synchronously, before returning to the
269 caller, unless ->writepage() returns special WRITEPAGE_ACTIVATE
270 value. WRITEPAGE_ACTIVATE means that page cannot really be written out
271 currently, and VM should stop calling ->writepage() on this page for some
272 time. VM does this by moving page to the head of the active list, hence the
275 Unless the filesystem is going to redirty_page_for_writepage(), unlock the page
276 and return zero, writepage *must* run set_page_writeback() against the page,
277 followed by unlocking it. Once set_page_writeback() has been run against the
278 page, write I/O can be submitted and the write I/O completion handler must run
279 end_page_writeback() once the I/O is complete. If no I/O is submitted, the
280 filesystem must run end_page_writeback() against the page before returning from
283 That is: after 2.5.12, pages which are under writeout are *not* locked. Note,
284 if the filesystem needs the page to be locked during writeout, that is ok, too,
285 the page is allowed to be unlocked at any point in time between the calls to
286 set_page_writeback() and end_page_writeback().
288 Note, failure to run either redirty_page_for_writepage() or the combination of
289 set_page_writeback()/end_page_writeback() on a page submitted to writepage
290 will leave the page itself marked clean but it will be tagged as dirty in the
291 radix tree. This incoherency can lead to all sorts of hard-to-debug problems
292 in the filesystem like having dirty inodes at umount and losing written data.
294 ->sync_page() locking rules are not well-defined - usually it is called
295 with lock on page, but that is not guaranteed. Considering the currently
296 existing instances of this method ->sync_page() itself doesn't look
299 ->writepages() is used for periodic writeback and for syscall-initiated
300 sync operations. The address_space should start I/O against at least
301 *nr_to_write pages. *nr_to_write must be decremented for each page which is
302 written. The address_space implementation may write more (or less) pages
303 than *nr_to_write asks for, but it should try to be reasonably close. If
304 nr_to_write is NULL, all dirty pages must be written.
306 writepages should _only_ write pages which are present on
309 ->set_page_dirty() is called from various places in the kernel
310 when the target page is marked as needing writeback. It may be called
311 under spinlock (it cannot block) and is sometimes called with the page
314 ->bmap() is currently used by legacy ioctl() (FIBMAP) provided by some
315 filesystems and by the swapper. The latter will eventually go away. Please,
316 keep it that way and don't breed new callers.
318 ->invalidatepage() is called when the filesystem must attempt to drop
319 some or all of the buffers from the page when it is being truncated. It
320 returns zero on success. If ->invalidatepage is zero, the kernel uses
321 block_invalidatepage() instead.
323 ->releasepage() is called when the kernel is about to try to drop the
324 buffers from the page in preparation for freeing it. It returns zero to
325 indicate that the buffers are (or may be) freeable. If ->releasepage is zero,
326 the kernel assumes that the fs has no private interest in the buffers.
328 ->freepage() is called when the kernel is done dropping the page
331 ->launder_page() may be called prior to releasing a page if
332 it is still found to be dirty. It returns zero if the page was successfully
333 cleaned, or an error value if not. Note that in order to prevent the page
334 getting mapped back in and redirtied, it needs to be kept locked
335 across the entire operation.
337 ->swap_activate will be called with a non-zero argument on
338 files backing (non block device backed) swapfiles. A return value
339 of zero indicates success, in which case this file can be used for
340 backing swapspace. The swapspace operations will be proxied to the
341 address space operations.
343 ->swap_deactivate() will be called in the sys_swapoff()
344 path after ->swap_activate() returned success.
346 ----------------------- file_lock_operations ------------------------------
348 void (*fl_copy_lock)(struct file_lock *, struct file_lock *);
349 void (*fl_release_private)(struct file_lock *);
353 file_lock_lock may block
355 fl_release_private: maybe no
357 ----------------------- lock_manager_operations ---------------------------
359 int (*lm_compare_owner)(struct file_lock *, struct file_lock *);
360 void (*lm_notify)(struct file_lock *); /* unblock callback */
361 int (*lm_grant)(struct file_lock *, struct file_lock *, int);
362 void (*lm_break)(struct file_lock *); /* break_lease callback */
363 int (*lm_change)(struct file_lock **, int);
366 file_lock_lock may block
367 lm_compare_owner: yes no
373 --------------------------- buffer_head -----------------------------------
375 void (*b_end_io)(struct buffer_head *bh, int uptodate);
378 called from interrupts. In other words, extreme care is needed here.
379 bh is locked, but that's all warranties we have here. Currently only RAID1,
380 highmem, fs/buffer.c, and fs/ntfs/aops.c are providing these. Block devices
381 call this method upon the IO completion.
383 --------------------------- block_device_operations -----------------------
385 int (*open) (struct block_device *, fmode_t);
386 int (*release) (struct gendisk *, fmode_t);
387 int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
388 int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
389 int (*direct_access) (struct block_device *, sector_t, void **, unsigned long *);
390 int (*media_changed) (struct gendisk *);
391 void (*unlock_native_capacity) (struct gendisk *);
392 int (*revalidate_disk) (struct gendisk *);
393 int (*getgeo)(struct block_device *, struct hd_geometry *);
394 void (*swap_slot_free_notify) (struct block_device *, unsigned long);
404 unlock_native_capacity: no
407 swap_slot_free_notify: no (see below)
409 media_changed, unlock_native_capacity and revalidate_disk are called only from
412 swap_slot_free_notify is called with swap_lock and sometimes the page lock
416 --------------------------- file_operations -------------------------------
418 loff_t (*llseek) (struct file *, loff_t, int);
419 ssize_t (*read) (struct file *, char __user *, size_t, loff_t *);
420 ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *);
421 ssize_t (*aio_read) (struct kiocb *, const struct iovec *, unsigned long, loff_t);
422 ssize_t (*aio_write) (struct kiocb *, const struct iovec *, unsigned long, loff_t);
423 int (*readdir) (struct file *, void *, filldir_t);
424 unsigned int (*poll) (struct file *, struct poll_table_struct *);
425 long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long);
426 long (*compat_ioctl) (struct file *, unsigned int, unsigned long);
427 int (*mmap) (struct file *, struct vm_area_struct *);
428 int (*open) (struct inode *, struct file *);
429 int (*flush) (struct file *);
430 int (*release) (struct inode *, struct file *);
431 int (*fsync) (struct file *, loff_t start, loff_t end, int datasync);
432 int (*aio_fsync) (struct kiocb *, int datasync);
433 int (*fasync) (int, struct file *, int);
434 int (*lock) (struct file *, int, struct file_lock *);
435 ssize_t (*readv) (struct file *, const struct iovec *, unsigned long,
437 ssize_t (*writev) (struct file *, const struct iovec *, unsigned long,
439 ssize_t (*sendfile) (struct file *, loff_t *, size_t, read_actor_t,
441 ssize_t (*sendpage) (struct file *, struct page *, int, size_t,
443 unsigned long (*get_unmapped_area)(struct file *, unsigned long,
444 unsigned long, unsigned long, unsigned long);
445 int (*check_flags)(int);
446 int (*flock) (struct file *, int, struct file_lock *);
447 ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *,
448 size_t, unsigned int);
449 ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *,
450 size_t, unsigned int);
451 int (*setlease)(struct file *, long, struct file_lock **);
452 long (*fallocate)(struct file *, int, loff_t, loff_t);
456 All may block except for ->setlease.
457 No VFS locks held on entry except for ->setlease.
459 ->setlease has the file_list_lock held and must not sleep.
461 ->llseek() locking has moved from llseek to the individual llseek
462 implementations. If your fs is not using generic_file_llseek, you
463 need to acquire and release the appropriate locks in your ->llseek().
464 For many filesystems, it is probably safe to acquire the inode
465 mutex or just to use i_size_read() instead.
466 Note: this does not protect the file->f_pos against concurrent modifications
467 since this is something the userspace has to take care about.
469 ->fasync() is responsible for maintaining the FASYNC bit in filp->f_flags.
470 Most instances call fasync_helper(), which does that maintenance, so it's
471 not normally something one needs to worry about. Return values > 0 will be
472 mapped to zero in the VFS layer.
474 ->readdir() and ->ioctl() on directories must be changed. Ideally we would
475 move ->readdir() to inode_operations and use a separate method for directory
476 ->ioctl() or kill the latter completely. One of the problems is that for
477 anything that resembles union-mount we won't have a struct file for all
478 components. And there are other reasons why the current interface is a mess...
480 ->read on directories probably must go away - we should just enforce -EISDIR
481 in sys_read() and friends.
483 --------------------------- dquot_operations -------------------------------
485 int (*write_dquot) (struct dquot *);
486 int (*acquire_dquot) (struct dquot *);
487 int (*release_dquot) (struct dquot *);
488 int (*mark_dirty) (struct dquot *);
489 int (*write_info) (struct super_block *, int);
491 These operations are intended to be more or less wrapping functions that ensure
492 a proper locking wrt the filesystem and call the generic quota operations.
494 What filesystem should expect from the generic quota functions:
496 FS recursion Held locks when called
497 write_dquot: yes dqonoff_sem or dqptr_sem
498 acquire_dquot: yes dqonoff_sem or dqptr_sem
499 release_dquot: yes dqonoff_sem or dqptr_sem
501 write_info: yes dqonoff_sem
503 FS recursion means calling ->quota_read() and ->quota_write() from superblock
506 More details about quota locking can be found in fs/dquot.c.
508 --------------------------- vm_operations_struct -----------------------------
510 void (*open)(struct vm_area_struct*);
511 void (*close)(struct vm_area_struct*);
512 int (*fault)(struct vm_area_struct*, struct vm_fault *);
513 int (*page_mkwrite)(struct vm_area_struct *, struct vm_fault *);
514 int (*access)(struct vm_area_struct *, unsigned long, void*, int, int);
517 mmap_sem PageLocked(page)
520 fault: yes can return with page locked
521 page_mkwrite: yes can return with page locked
524 ->fault() is called when a previously not present pte is about
525 to be faulted in. The filesystem must find and return the page associated
526 with the passed in "pgoff" in the vm_fault structure. If it is possible that
527 the page may be truncated and/or invalidated, then the filesystem must lock
528 the page, then ensure it is not already truncated (the page lock will block
529 subsequent truncate), and then return with VM_FAULT_LOCKED, and the page
530 locked. The VM will unlock the page.
532 ->page_mkwrite() is called when a previously read-only pte is
533 about to become writeable. The filesystem again must ensure that there are
534 no truncate/invalidate races, and then return with the page locked. If
535 the page has been truncated, the filesystem should not look up a new page
536 like the ->fault() handler, but simply return with VM_FAULT_NOPAGE, which
537 will cause the VM to retry the fault.
539 ->access() is called when get_user_pages() fails in
540 acces_process_vm(), typically used to debug a process through
541 /proc/pid/mem or ptrace. This function is needed only for
542 VM_IO | VM_PFNMAP VMAs.
544 ================================================================================
547 (if you break something or notice that it is broken and do not fix it yourself
548 - at least put it here)