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_weak_revalidate)(struct dentry *, unsigned int);
14 int (*d_hash)(const struct dentry *, struct qstr *);
15 int (*d_compare)(const struct dentry *, const struct dentry *,
16 unsigned int, const char *, const struct qstr *);
17 int (*d_delete)(struct dentry *);
18 void (*d_release)(struct dentry *);
19 void (*d_iput)(struct dentry *, struct inode *);
20 char *(*d_dname)((struct dentry *dentry, char *buffer, int buflen);
21 struct vfsmount *(*d_automount)(struct path *path);
22 int (*d_manage)(struct dentry *, bool);
25 rename_lock ->d_lock may block rcu-walk
26 d_revalidate: no no yes (ref-walk) maybe
27 d_weak_revalidate:no no yes no
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 (*rename2) (struct inode *, struct dentry *,
51 struct inode *, struct dentry *, unsigned int);
52 int (*readlink) (struct dentry *, char __user *,int);
53 const char *(*get_link) (struct dentry *, struct inode *, void **);
54 void (*truncate) (struct inode *);
55 int (*permission) (struct inode *, int, unsigned int);
56 int (*get_acl)(struct inode *, int);
57 int (*setattr) (struct dentry *, struct iattr *);
58 int (*getattr) (struct vfsmount *, struct dentry *, struct kstat *);
59 int (*setxattr) (struct dentry *, const char *,const void *,size_t,int);
60 ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t);
61 ssize_t (*listxattr) (struct dentry *, char *, size_t);
62 int (*removexattr) (struct dentry *, const char *);
63 int (*fiemap)(struct inode *, struct fiemap_extent_info *, u64 start, u64 len);
64 void (*update_time)(struct inode *, struct timespec *, int);
65 int (*atomic_open)(struct inode *, struct dentry *,
66 struct file *, unsigned open_flag,
67 umode_t create_mode, int *opened);
68 int (*tmpfile) (struct inode *, struct dentry *, umode_t);
69 int (*dentry_open)(struct dentry *, struct file *, const struct cred *);
81 rmdir: yes (both) (see below)
82 rename: yes (all) (see below)
83 rename2: yes (all) (see below)
87 permission: no (may not block if called in rcu-walk mode)
100 Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
102 cross-directory ->rename() and rename2() has (per-superblock)
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 int (*sync_fs)(struct super_block *sb, int wait);
118 int (*freeze_fs) (struct super_block *);
119 int (*unfreeze_fs) (struct super_block *);
120 int (*statfs) (struct dentry *, struct kstatfs *);
121 int (*remount_fs) (struct super_block *, int *, char *);
122 void (*umount_begin) (struct super_block *);
123 int (*show_options)(struct seq_file *, struct dentry *);
124 ssize_t (*quota_read)(struct super_block *, int, char *, size_t, loff_t);
125 ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t);
126 int (*bdev_try_to_free_page)(struct super_block*, struct page*, gfp_t);
129 All may block [not true, see below]
135 drop_inode: !!!inode->i_lock!!!
141 statfs: maybe(read) (see below)
144 show_options: no (namespace_sem)
145 quota_read: no (see below)
146 quota_write: no (see below)
147 bdev_try_to_free_page: no (see below)
149 ->statfs() has s_umount (shared) when called by ustat(2) (native or
150 compat), but that's an accident of bad API; s_umount is used to pin
151 the superblock down when we only have dev_t given us by userland to
152 identify the superblock. Everything else (statfs(), fstatfs(), etc.)
153 doesn't hold it when calling ->statfs() - superblock is pinned down
154 by resolving the pathname passed to syscall.
155 ->quota_read() and ->quota_write() functions are both guaranteed to
156 be the only ones operating on the quota file by the quota code (via
157 dqio_sem) (unless an admin really wants to screw up something and
158 writes to quota files with quotas on). For other details about locking
159 see also dquot_operations section.
160 ->bdev_try_to_free_page is called from the ->releasepage handler of
161 the block device inode. See there for more details.
163 --------------------------- file_system_type ---------------------------
165 struct dentry *(*mount) (struct file_system_type *, int,
166 const char *, void *);
167 void (*kill_sb) (struct super_block *);
173 ->mount() returns ERR_PTR or the root dentry; its superblock should be locked
175 ->kill_sb() takes a write-locked superblock, does all shutdown work on it,
176 unlocks and drops the reference.
178 --------------------------- address_space_operations --------------------------
180 int (*writepage)(struct page *page, struct writeback_control *wbc);
181 int (*readpage)(struct file *, struct page *);
182 int (*sync_page)(struct page *);
183 int (*writepages)(struct address_space *, struct writeback_control *);
184 int (*set_page_dirty)(struct page *page);
185 int (*readpages)(struct file *filp, struct address_space *mapping,
186 struct list_head *pages, unsigned nr_pages);
187 int (*write_begin)(struct file *, struct address_space *mapping,
188 loff_t pos, unsigned len, unsigned flags,
189 struct page **pagep, void **fsdata);
190 int (*write_end)(struct file *, struct address_space *mapping,
191 loff_t pos, unsigned len, unsigned copied,
192 struct page *page, void *fsdata);
193 sector_t (*bmap)(struct address_space *, sector_t);
194 void (*invalidatepage) (struct page *, unsigned int, unsigned int);
195 int (*releasepage) (struct page *, int);
196 void (*freepage)(struct page *);
197 int (*direct_IO)(struct kiocb *, struct iov_iter *iter, loff_t offset);
198 int (*migratepage)(struct address_space *, struct page *, struct page *);
199 int (*launder_page)(struct page *);
200 int (*is_partially_uptodate)(struct page *, unsigned long, unsigned long);
201 int (*error_remove_page)(struct address_space *, struct page *);
202 int (*swap_activate)(struct file *);
203 int (*swap_deactivate)(struct file *);
206 All except set_page_dirty and freepage may block
208 PageLocked(page) i_mutex
209 writepage: yes, unlocks (see below)
210 readpage: yes, unlocks
215 write_begin: locks the page yes
216 write_end: yes, unlocks yes
222 migratepage: yes (both)
224 is_partially_uptodate: yes
225 error_remove_page: yes
229 ->write_begin(), ->write_end(), ->sync_page() and ->readpage()
230 may be called from the request handler (/dev/loop).
232 ->readpage() unlocks the page, either synchronously or via I/O
235 ->readpages() populates the pagecache with the passed pages and starts
236 I/O against them. They come unlocked upon I/O completion.
238 ->writepage() is used for two purposes: for "memory cleansing" and for
239 "sync". These are quite different operations and the behaviour may differ
240 depending upon the mode.
242 If writepage is called for sync (wbc->sync_mode != WBC_SYNC_NONE) then
243 it *must* start I/O against the page, even if that would involve
244 blocking on in-progress I/O.
246 If writepage is called for memory cleansing (sync_mode ==
247 WBC_SYNC_NONE) then its role is to get as much writeout underway as
248 possible. So writepage should try to avoid blocking against
249 currently-in-progress I/O.
251 If the filesystem is not called for "sync" and it determines that it
252 would need to block against in-progress I/O to be able to start new I/O
253 against the page the filesystem should redirty the page with
254 redirty_page_for_writepage(), then unlock the page and return zero.
255 This may also be done to avoid internal deadlocks, but rarely.
257 If the filesystem is called for sync then it must wait on any
258 in-progress I/O and then start new I/O.
260 The filesystem should unlock the page synchronously, before returning to the
261 caller, unless ->writepage() returns special WRITEPAGE_ACTIVATE
262 value. WRITEPAGE_ACTIVATE means that page cannot really be written out
263 currently, and VM should stop calling ->writepage() on this page for some
264 time. VM does this by moving page to the head of the active list, hence the
267 Unless the filesystem is going to redirty_page_for_writepage(), unlock the page
268 and return zero, writepage *must* run set_page_writeback() against the page,
269 followed by unlocking it. Once set_page_writeback() has been run against the
270 page, write I/O can be submitted and the write I/O completion handler must run
271 end_page_writeback() once the I/O is complete. If no I/O is submitted, the
272 filesystem must run end_page_writeback() against the page before returning from
275 That is: after 2.5.12, pages which are under writeout are *not* locked. Note,
276 if the filesystem needs the page to be locked during writeout, that is ok, too,
277 the page is allowed to be unlocked at any point in time between the calls to
278 set_page_writeback() and end_page_writeback().
280 Note, failure to run either redirty_page_for_writepage() or the combination of
281 set_page_writeback()/end_page_writeback() on a page submitted to writepage
282 will leave the page itself marked clean but it will be tagged as dirty in the
283 radix tree. This incoherency can lead to all sorts of hard-to-debug problems
284 in the filesystem like having dirty inodes at umount and losing written data.
286 ->sync_page() locking rules are not well-defined - usually it is called
287 with lock on page, but that is not guaranteed. Considering the currently
288 existing instances of this method ->sync_page() itself doesn't look
291 ->writepages() is used for periodic writeback and for syscall-initiated
292 sync operations. The address_space should start I/O against at least
293 *nr_to_write pages. *nr_to_write must be decremented for each page which is
294 written. The address_space implementation may write more (or less) pages
295 than *nr_to_write asks for, but it should try to be reasonably close. If
296 nr_to_write is NULL, all dirty pages must be written.
298 writepages should _only_ write pages which are present on
301 ->set_page_dirty() is called from various places in the kernel
302 when the target page is marked as needing writeback. It may be called
303 under spinlock (it cannot block) and is sometimes called with the page
306 ->bmap() is currently used by legacy ioctl() (FIBMAP) provided by some
307 filesystems and by the swapper. The latter will eventually go away. Please,
308 keep it that way and don't breed new callers.
310 ->invalidatepage() is called when the filesystem must attempt to drop
311 some or all of the buffers from the page when it is being truncated. It
312 returns zero on success. If ->invalidatepage is zero, the kernel uses
313 block_invalidatepage() instead.
315 ->releasepage() is called when the kernel is about to try to drop the
316 buffers from the page in preparation for freeing it. It returns zero to
317 indicate that the buffers are (or may be) freeable. If ->releasepage is zero,
318 the kernel assumes that the fs has no private interest in the buffers.
320 ->freepage() is called when the kernel is done dropping the page
323 ->launder_page() may be called prior to releasing a page if
324 it is still found to be dirty. It returns zero if the page was successfully
325 cleaned, or an error value if not. Note that in order to prevent the page
326 getting mapped back in and redirtied, it needs to be kept locked
327 across the entire operation.
329 ->swap_activate will be called with a non-zero argument on
330 files backing (non block device backed) swapfiles. A return value
331 of zero indicates success, in which case this file can be used for
332 backing swapspace. The swapspace operations will be proxied to the
333 address space operations.
335 ->swap_deactivate() will be called in the sys_swapoff()
336 path after ->swap_activate() returned success.
338 ----------------------- file_lock_operations ------------------------------
340 void (*fl_copy_lock)(struct file_lock *, struct file_lock *);
341 void (*fl_release_private)(struct file_lock *);
345 inode->i_lock may block
347 fl_release_private: maybe maybe[1]
349 [1]: ->fl_release_private for flock or POSIX locks is currently allowed
350 to block. Leases however can still be freed while the i_lock is held and
351 so fl_release_private called on a lease should not block.
353 ----------------------- lock_manager_operations ---------------------------
355 int (*lm_compare_owner)(struct file_lock *, struct file_lock *);
356 unsigned long (*lm_owner_key)(struct file_lock *);
357 void (*lm_notify)(struct file_lock *); /* unblock callback */
358 int (*lm_grant)(struct file_lock *, struct file_lock *, int);
359 void (*lm_break)(struct file_lock *); /* break_lease callback */
360 int (*lm_change)(struct file_lock **, int);
364 inode->i_lock blocked_lock_lock may block
365 lm_compare_owner: yes[1] maybe no
366 lm_owner_key yes[1] yes no
367 lm_notify: yes yes no
372 [1]: ->lm_compare_owner and ->lm_owner_key are generally called with
373 *an* inode->i_lock held. It may not be the i_lock of the inode
374 associated with either file_lock argument! This is the case with deadlock
375 detection, since the code has to chase down the owners of locks that may
376 be entirely unrelated to the one on which the lock is being acquired.
377 For deadlock detection however, the blocked_lock_lock is also held. The
378 fact that these locks are held ensures that the file_locks do not
379 disappear out from under you while doing the comparison or generating an
382 --------------------------- buffer_head -----------------------------------
384 void (*b_end_io)(struct buffer_head *bh, int uptodate);
387 called from interrupts. In other words, extreme care is needed here.
388 bh is locked, but that's all warranties we have here. Currently only RAID1,
389 highmem, fs/buffer.c, and fs/ntfs/aops.c are providing these. Block devices
390 call this method upon the IO completion.
392 --------------------------- block_device_operations -----------------------
394 int (*open) (struct block_device *, fmode_t);
395 int (*release) (struct gendisk *, fmode_t);
396 int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
397 int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
398 int (*direct_access) (struct block_device *, sector_t, void __pmem **,
400 int (*media_changed) (struct gendisk *);
401 void (*unlock_native_capacity) (struct gendisk *);
402 int (*revalidate_disk) (struct gendisk *);
403 int (*getgeo)(struct block_device *, struct hd_geometry *);
404 void (*swap_slot_free_notify) (struct block_device *, unsigned long);
414 unlock_native_capacity: no
417 swap_slot_free_notify: no (see below)
419 media_changed, unlock_native_capacity and revalidate_disk are called only from
422 swap_slot_free_notify is called with swap_lock and sometimes the page lock
426 --------------------------- file_operations -------------------------------
428 loff_t (*llseek) (struct file *, loff_t, int);
429 ssize_t (*read) (struct file *, char __user *, size_t, loff_t *);
430 ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *);
431 ssize_t (*read_iter) (struct kiocb *, struct iov_iter *);
432 ssize_t (*write_iter) (struct kiocb *, struct iov_iter *);
433 int (*iterate) (struct file *, struct dir_context *);
434 unsigned int (*poll) (struct file *, struct poll_table_struct *);
435 long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long);
436 long (*compat_ioctl) (struct file *, unsigned int, unsigned long);
437 int (*mmap) (struct file *, struct vm_area_struct *);
438 int (*open) (struct inode *, struct file *);
439 int (*flush) (struct file *);
440 int (*release) (struct inode *, struct file *);
441 int (*fsync) (struct file *, loff_t start, loff_t end, int datasync);
442 int (*aio_fsync) (struct kiocb *, int datasync);
443 int (*fasync) (int, struct file *, int);
444 int (*lock) (struct file *, int, struct file_lock *);
445 ssize_t (*readv) (struct file *, const struct iovec *, unsigned long,
447 ssize_t (*writev) (struct file *, const struct iovec *, unsigned long,
449 ssize_t (*sendfile) (struct file *, loff_t *, size_t, read_actor_t,
451 ssize_t (*sendpage) (struct file *, struct page *, int, size_t,
453 unsigned long (*get_unmapped_area)(struct file *, unsigned long,
454 unsigned long, unsigned long, unsigned long);
455 int (*check_flags)(int);
456 int (*flock) (struct file *, int, struct file_lock *);
457 ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *,
458 size_t, unsigned int);
459 ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *,
460 size_t, unsigned int);
461 int (*setlease)(struct file *, long, struct file_lock **, void **);
462 long (*fallocate)(struct file *, int, loff_t, loff_t);
468 ->llseek() locking has moved from llseek to the individual llseek
469 implementations. If your fs is not using generic_file_llseek, you
470 need to acquire and release the appropriate locks in your ->llseek().
471 For many filesystems, it is probably safe to acquire the inode
472 mutex or just to use i_size_read() instead.
473 Note: this does not protect the file->f_pos against concurrent modifications
474 since this is something the userspace has to take care about.
476 ->fasync() is responsible for maintaining the FASYNC bit in filp->f_flags.
477 Most instances call fasync_helper(), which does that maintenance, so it's
478 not normally something one needs to worry about. Return values > 0 will be
479 mapped to zero in the VFS layer.
481 ->readdir() and ->ioctl() on directories must be changed. Ideally we would
482 move ->readdir() to inode_operations and use a separate method for directory
483 ->ioctl() or kill the latter completely. One of the problems is that for
484 anything that resembles union-mount we won't have a struct file for all
485 components. And there are other reasons why the current interface is a mess...
487 ->read on directories probably must go away - we should just enforce -EISDIR
488 in sys_read() and friends.
490 ->setlease operations should call generic_setlease() before or after setting
491 the lease within the individual filesystem to record the result of the
494 --------------------------- dquot_operations -------------------------------
496 int (*write_dquot) (struct dquot *);
497 int (*acquire_dquot) (struct dquot *);
498 int (*release_dquot) (struct dquot *);
499 int (*mark_dirty) (struct dquot *);
500 int (*write_info) (struct super_block *, int);
502 These operations are intended to be more or less wrapping functions that ensure
503 a proper locking wrt the filesystem and call the generic quota operations.
505 What filesystem should expect from the generic quota functions:
507 FS recursion Held locks when called
508 write_dquot: yes dqonoff_sem or dqptr_sem
509 acquire_dquot: yes dqonoff_sem or dqptr_sem
510 release_dquot: yes dqonoff_sem or dqptr_sem
512 write_info: yes dqonoff_sem
514 FS recursion means calling ->quota_read() and ->quota_write() from superblock
517 More details about quota locking can be found in fs/dquot.c.
519 --------------------------- vm_operations_struct -----------------------------
521 void (*open)(struct vm_area_struct*);
522 void (*close)(struct vm_area_struct*);
523 int (*fault)(struct vm_area_struct*, struct vm_fault *);
524 int (*page_mkwrite)(struct vm_area_struct *, struct vm_fault *);
525 int (*pfn_mkwrite)(struct vm_area_struct *, struct vm_fault *);
526 int (*access)(struct vm_area_struct *, unsigned long, void*, int, int);
529 mmap_sem PageLocked(page)
532 fault: yes can return with page locked
534 page_mkwrite: yes can return with page locked
538 ->fault() is called when a previously not present pte is about
539 to be faulted in. The filesystem must find and return the page associated
540 with the passed in "pgoff" in the vm_fault structure. If it is possible that
541 the page may be truncated and/or invalidated, then the filesystem must lock
542 the page, then ensure it is not already truncated (the page lock will block
543 subsequent truncate), and then return with VM_FAULT_LOCKED, and the page
544 locked. The VM will unlock the page.
546 ->map_pages() is called when VM asks to map easy accessible pages.
547 Filesystem should find and map pages associated with offsets from "pgoff"
548 till "max_pgoff". ->map_pages() is called with page table locked and must
549 not block. If it's not possible to reach a page without blocking,
550 filesystem should skip it. Filesystem should use do_set_pte() to setup
551 page table entry. Pointer to entry associated with offset "pgoff" is
552 passed in "pte" field in vm_fault structure. Pointers to entries for other
553 offsets should be calculated relative to "pte".
555 ->page_mkwrite() is called when a previously read-only pte is
556 about to become writeable. The filesystem again must ensure that there are
557 no truncate/invalidate races, and then return with the page locked. If
558 the page has been truncated, the filesystem should not look up a new page
559 like the ->fault() handler, but simply return with VM_FAULT_NOPAGE, which
560 will cause the VM to retry the fault.
562 ->pfn_mkwrite() is the same as page_mkwrite but when the pte is
563 VM_PFNMAP or VM_MIXEDMAP with a page-less entry. Expected return is
564 VM_FAULT_NOPAGE. Or one of the VM_FAULT_ERROR types. The default behavior
565 after this call is to make the pte read-write, unless pfn_mkwrite returns
568 ->access() is called when get_user_pages() fails in
569 access_process_vm(), typically used to debug a process through
570 /proc/pid/mem or ptrace. This function is needed only for
571 VM_IO | VM_PFNMAP VMAs.
573 ================================================================================
576 (if you break something or notice that it is broken and do not fix it yourself
577 - at least put it here)