| blob | 4471a416c27457069d722160ff533dbccc1fbf3c |
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 --------------------------
11 prototypes:
12 int (*d_revalidate)(struct dentry *, struct nameidata *);
13 int (*d_hash)(const struct dentry *, const struct inode *,
14 struct qstr *);
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);
25 locking rules:
26 rename_lock ->d_lock may block rcu-walk
27 d_revalidate: no no yes (ref-walk) maybe
28 d_hash no no no maybe
29 d_compare: yes no no maybe
30 d_delete: no yes no no
31 d_release: no no yes no
32 d_iput: no no yes no
33 d_dname: no no no no
34 d_automount: no no yes no
35 d_manage: no no yes (ref-walk) maybe
37 --------------------------- inode_operations ---------------------------
38 prototypes:
39 int (*create) (struct inode *,struct dentry *,int, struct nameidata *);
40 struct dentry * (*lookup) (struct inode *,struct dentry *, struct nameid
41 ata *);
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 *,int);
46 int (*rmdir) (struct inode *,struct dentry *);
47 int (*mknod) (struct inode *,struct dentry *,int,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 (*check_acl)(struct inode *, int, unsigned 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 void (*truncate_range)(struct inode *, loff_t, loff_t);
63 int (*fiemap)(struct inode *, struct fiemap_extent_info *, u64 start, u64 len);
65 locking rules:
66 all may block
67 i_mutex(inode)
68 lookup: yes
69 create: yes
70 link: yes (both)
71 mknod: yes
72 symlink: yes
73 mkdir: yes
74 unlink: yes (both)
75 rmdir: yes (both) (see below)
76 rename: yes (all) (see below)
77 readlink: no
78 follow_link: no
79 put_link: no
80 truncate: yes (see below)
81 setattr: yes
82 permission: no (may not block if called in rcu-walk mode)
83 check_acl: no
84 getattr: no
85 setxattr: yes
86 getxattr: no
87 listxattr: no
88 removexattr: yes
89 truncate_range: yes
90 fiemap: no
91 Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
92 victim.
93 cross-directory ->rename() has (per-superblock) ->s_vfs_rename_sem.
94 ->truncate() is never called directly - it's a callback, not a
95 method. It's called by vmtruncate() - deprecated library function used by
96 ->setattr(). Locking information above applies to that call (i.e. is
97 inherited from ->setattr() - vmtruncate() is used when ATTR_SIZE had been
98 passed).
100 See Documentation/filesystems/directory-locking for more detailed discussion
101 of the locking scheme for directory operations.
103 --------------------------- super_operations ---------------------------
104 prototypes:
105 struct inode *(*alloc_inode)(struct super_block *sb);
106 void (*destroy_inode)(struct inode *);
107 void (*dirty_inode) (struct inode *);
108 int (*write_inode) (struct inode *, struct writeback_control *wbc);
109 int (*drop_inode) (struct inode *);
110 void (*evict_inode) (struct inode *);
111 void (*put_super) (struct super_block *);
112 void (*write_super) (struct super_block *);
113 int (*sync_fs)(struct super_block *sb, int wait);
114 int (*freeze_fs) (struct super_block *);
115 int (*unfreeze_fs) (struct super_block *);
116 int (*statfs) (struct dentry *, struct kstatfs *);
117 int (*remount_fs) (struct super_block *, int *, char *);
118 void (*umount_begin) (struct super_block *);
119 int (*show_options)(struct seq_file *, struct vfsmount *);
120 ssize_t (*quota_read)(struct super_block *, int, char *, size_t, loff_t);
121 ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t);
122 int (*bdev_try_to_free_page)(struct super_block*, struct page*, gfp_t);
124 locking rules:
125 All may block [not true, see below]
126 s_umount
127 alloc_inode:
128 destroy_inode:
129 dirty_inode: (must not sleep)
130 write_inode:
131 drop_inode: !!!inode_lock!!!
132 evict_inode:
133 put_super: write
134 write_super: read
135 sync_fs: read
136 freeze_fs: read
137 unfreeze_fs: read
138 statfs: maybe(read) (see below)
139 remount_fs: write
140 umount_begin: no
141 show_options: no (namespace_sem)
142 quota_read: no (see below)
143 quota_write: no (see below)
144 bdev_try_to_free_page: no (see below)
146 ->statfs() has s_umount (shared) when called by ustat(2) (native or
147 compat), but that's an accident of bad API; s_umount is used to pin
148 the superblock down when we only have dev_t given us by userland to
149 identify the superblock. Everything else (statfs(), fstatfs(), etc.)
150 doesn't hold it when calling ->statfs() - superblock is pinned down
151 by resolving the pathname passed to syscall.
152 ->quota_read() and ->quota_write() functions are both guaranteed to
153 be the only ones operating on the quota file by the quota code (via
154 dqio_sem) (unless an admin really wants to screw up something and
155 writes to quota files with quotas on). For other details about locking
156 see also dquot_operations section.
157 ->bdev_try_to_free_page is called from the ->releasepage handler of
158 the block device inode. See there for more details.
160 --------------------------- file_system_type ---------------------------
161 prototypes:
162 int (*get_sb) (struct file_system_type *, int,
163 const char *, void *, struct vfsmount *);
164 struct dentry *(*mount) (struct file_system_type *, int,
165 const char *, void *);
166 void (*kill_sb) (struct super_block *);
167 locking rules:
168 may block
169 get_sb yes
170 mount yes
171 kill_sb yes
173 ->get_sb() returns error or 0 with locked superblock attached to the vfsmount
174 (exclusive on ->s_umount).
175 ->mount() returns ERR_PTR or the root dentry.
176 ->kill_sb() takes a write-locked superblock, does all shutdown work on it,
177 unlocks and drops the reference.
179 --------------------------- address_space_operations --------------------------
180 prototypes:
181 int (*writepage)(struct page *page, struct writeback_control *wbc);
182 int (*readpage)(struct file *, struct page *);
183 int (*sync_page)(struct page *);
184 int (*writepages)(struct address_space *, struct writeback_control *);
185 int (*set_page_dirty)(struct page *page);
186 int (*readpages)(struct file *filp, struct address_space *mapping,
187 struct list_head *pages, unsigned nr_pages);
188 int (*write_begin)(struct file *, struct address_space *mapping,
189 loff_t pos, unsigned len, unsigned flags,
190 struct page **pagep, void **fsdata);
191 int (*write_end)(struct file *, struct address_space *mapping,
192 loff_t pos, unsigned len, unsigned copied,
193 struct page *page, void *fsdata);
194 sector_t (*bmap)(struct address_space *, sector_t);
195 int (*invalidatepage) (struct page *, unsigned long);
196 int (*releasepage) (struct page *, int);
197 void (*freepage)(struct page *);
198 int (*direct_IO)(int, struct kiocb *, const struct iovec *iov,
199 loff_t offset, unsigned long nr_segs);
200 int (*get_xip_mem)(struct address_space *, pgoff_t, int, void **,
201 unsigned long *);
202 int (*migratepage)(struct address_space *, struct page *, struct page *);
203 int (*launder_page)(struct page *);
204 int (*is_partially_uptodate)(struct page *, read_descriptor_t *, unsigned long);
205 int (*error_remove_page)(struct address_space *, struct page *);
207 locking rules:
208 All except set_page_dirty and freepage may block
210 PageLocked(page) i_mutex
211 writepage: yes, unlocks (see below)
212 readpage: yes, unlocks
213 sync_page: maybe
214 writepages:
215 set_page_dirty no
216 readpages:
217 write_begin: locks the page yes
218 write_end: yes, unlocks yes
219 bmap:
220 invalidatepage: yes
221 releasepage: yes
222 freepage: yes
223 direct_IO:
224 get_xip_mem: maybe
225 migratepage: yes (both)
226 launder_page: yes
227 is_partially_uptodate: yes
228 error_remove_page: yes
230 ->write_begin(), ->write_end(), ->sync_page() and ->readpage()
231 may be called from the request handler (/dev/loop).
233 ->readpage() unlocks the page, either synchronously or via I/O
234 completion.
236 ->readpages() populates the pagecache with the passed pages and starts
237 I/O against them. They come unlocked upon I/O completion.
239 ->writepage() is used for two purposes: for "memory cleansing" and for
240 "sync". These are quite different operations and the behaviour may differ
241 depending upon the mode.
243 If writepage is called for sync (wbc->sync_mode != WBC_SYNC_NONE) then
244 it *must* start I/O against the page, even if that would involve
245 blocking on in-progress I/O.
247 If writepage is called for memory cleansing (sync_mode ==
248 WBC_SYNC_NONE) then its role is to get as much writeout underway as
249 possible. So writepage should try to avoid blocking against
250 currently-in-progress I/O.
252 If the filesystem is not called for "sync" and it determines that it
253 would need to block against in-progress I/O to be able to start new I/O
254 against the page the filesystem should redirty the page with
255 redirty_page_for_writepage(), then unlock the page and return zero.
256 This may also be done to avoid internal deadlocks, but rarely.
258 If the filesystem is called for sync then it must wait on any
259 in-progress I/O and then start new I/O.
261 The filesystem should unlock the page synchronously, before returning to the
262 caller, unless ->writepage() returns special WRITEPAGE_ACTIVATE
263 value. WRITEPAGE_ACTIVATE means that page cannot really be written out
264 currently, and VM should stop calling ->writepage() on this page for some
265 time. VM does this by moving page to the head of the active list, hence the
266 name.
268 Unless the filesystem is going to redirty_page_for_writepage(), unlock the page
269 and return zero, writepage *must* run set_page_writeback() against the page,
270 followed by unlocking it. Once set_page_writeback() has been run against the
271 page, write I/O can be submitted and the write I/O completion handler must run
272 end_page_writeback() once the I/O is complete. If no I/O is submitted, the
273 filesystem must run end_page_writeback() against the page before returning from
274 writepage.
276 That is: after 2.5.12, pages which are under writeout are *not* locked. Note,
277 if the filesystem needs the page to be locked during writeout, that is ok, too,
278 the page is allowed to be unlocked at any point in time between the calls to
279 set_page_writeback() and end_page_writeback().
281 Note, failure to run either redirty_page_for_writepage() or the combination of
282 set_page_writeback()/end_page_writeback() on a page submitted to writepage
283 will leave the page itself marked clean but it will be tagged as dirty in the
284 radix tree. This incoherency can lead to all sorts of hard-to-debug problems
285 in the filesystem like having dirty inodes at umount and losing written data.
287 ->sync_page() locking rules are not well-defined - usually it is called
288 with lock on page, but that is not guaranteed. Considering the currently
289 existing instances of this method ->sync_page() itself doesn't look
290 well-defined...
292 ->writepages() is used for periodic writeback and for syscall-initiated
293 sync operations. The address_space should start I/O against at least
294 *nr_to_write pages. *nr_to_write must be decremented for each page which is
295 written. The address_space implementation may write more (or less) pages
296 than *nr_to_write asks for, but it should try to be reasonably close. If
297 nr_to_write is NULL, all dirty pages must be written.
299 writepages should _only_ write pages which are present on
300 mapping->io_pages.
302 ->set_page_dirty() is called from various places in the kernel
303 when the target page is marked as needing writeback. It may be called
304 under spinlock (it cannot block) and is sometimes called with the page
305 not locked.
307 ->bmap() is currently used by legacy ioctl() (FIBMAP) provided by some
308 filesystems and by the swapper. The latter will eventually go away. Please,
309 keep it that way and don't breed new callers.
311 ->invalidatepage() is called when the filesystem must attempt to drop
312 some or all of the buffers from the page when it is being truncated. It
313 returns zero on success. If ->invalidatepage is zero, the kernel uses
314 block_invalidatepage() instead.
316 ->releasepage() is called when the kernel is about to try to drop the
317 buffers from the page in preparation for freeing it. It returns zero to
318 indicate that the buffers are (or may be) freeable. If ->releasepage is zero,
319 the kernel assumes that the fs has no private interest in the buffers.
321 ->freepage() is called when the kernel is done dropping the page
322 from the page cache.
324 ->launder_page() may be called prior to releasing a page if
325 it is still found to be dirty. It returns zero if the page was successfully
326 cleaned, or an error value if not. Note that in order to prevent the page
327 getting mapped back in and redirtied, it needs to be kept locked
328 across the entire operation.
330 ----------------------- file_lock_operations ------------------------------
331 prototypes:
332 void (*fl_copy_lock)(struct file_lock *, struct file_lock *);
333 void (*fl_release_private)(struct file_lock *);
336 locking rules:
337 file_lock_lock may block
338 fl_copy_lock: yes no
339 fl_release_private: maybe no
341 ----------------------- lock_manager_operations ---------------------------
342 prototypes:
343 int (*fl_compare_owner)(struct file_lock *, struct file_lock *);
344 void (*fl_notify)(struct file_lock *); /* unblock callback */
345 int (*fl_grant)(struct file_lock *, struct file_lock *, int);
346 void (*fl_release_private)(struct file_lock *);
347 void (*fl_break)(struct file_lock *); /* break_lease callback */
348 int (*fl_change)(struct file_lock **, int);
350 locking rules:
351 file_lock_lock may block
352 fl_compare_owner: yes no
353 fl_notify: yes no
354 fl_grant: no no
355 fl_release_private: maybe no
356 fl_break: yes no
357 fl_change yes no
359 --------------------------- buffer_head -----------------------------------
360 prototypes:
361 void (*b_end_io)(struct buffer_head *bh, int uptodate);
363 locking rules:
364 called from interrupts. In other words, extreme care is needed here.
365 bh is locked, but that's all warranties we have here. Currently only RAID1,
366 highmem, fs/buffer.c, and fs/ntfs/aops.c are providing these. Block devices
367 call this method upon the IO completion.
369 --------------------------- block_device_operations -----------------------
370 prototypes:
371 int (*open) (struct block_device *, fmode_t);
372 int (*release) (struct gendisk *, fmode_t);
373 int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
374 int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
375 int (*direct_access) (struct block_device *, sector_t, void **, unsigned long *);
376 int (*media_changed) (struct gendisk *);
377 void (*unlock_native_capacity) (struct gendisk *);
378 int (*revalidate_disk) (struct gendisk *);
379 int (*getgeo)(struct block_device *, struct hd_geometry *);
380 void (*swap_slot_free_notify) (struct block_device *, unsigned long);
382 locking rules:
383 bd_mutex
384 open: yes
385 release: yes
386 ioctl: no
387 compat_ioctl: no
388 direct_access: no
389 media_changed: no
390 unlock_native_capacity: no
391 revalidate_disk: no
392 getgeo: no
393 swap_slot_free_notify: no (see below)
395 media_changed, unlock_native_capacity and revalidate_disk are called only from
396 check_disk_change().
398 swap_slot_free_notify is called with swap_lock and sometimes the page lock
399 held.
402 --------------------------- file_operations -------------------------------
403 prototypes:
404 loff_t (*llseek) (struct file *, loff_t, int);
405 ssize_t (*read) (struct file *, char __user *, size_t, loff_t *);
406 ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *);
407 ssize_t (*aio_read) (struct kiocb *, const struct iovec *, unsigned long, loff_t);
408 ssize_t (*aio_write) (struct kiocb *, const struct iovec *, unsigned long, loff_t);
409 int (*readdir) (struct file *, void *, filldir_t);
410 unsigned int (*poll) (struct file *, struct poll_table_struct *);
411 long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long);
412 long (*compat_ioctl) (struct file *, unsigned int, unsigned long);
413 int (*mmap) (struct file *, struct vm_area_struct *);
414 int (*open) (struct inode *, struct file *);
415 int (*flush) (struct file *);
416 int (*release) (struct inode *, struct file *);
417 int (*fsync) (struct file *, int datasync);
418 int (*aio_fsync) (struct kiocb *, int datasync);
419 int (*fasync) (int, struct file *, int);
420 int (*lock) (struct file *, int, struct file_lock *);
421 ssize_t (*readv) (struct file *, const struct iovec *, unsigned long,
422 loff_t *);
423 ssize_t (*writev) (struct file *, const struct iovec *, unsigned long,
424 loff_t *);
425 ssize_t (*sendfile) (struct file *, loff_t *, size_t, read_actor_t,
426 void __user *);
427 ssize_t (*sendpage) (struct file *, struct page *, int, size_t,
428 loff_t *, int);
429 unsigned long (*get_unmapped_area)(struct file *, unsigned long,
430 unsigned long, unsigned long, unsigned long);
431 int (*check_flags)(int);
432 int (*flock) (struct file *, int, struct file_lock *);
433 ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *,
434 size_t, unsigned int);
435 ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *,
436 size_t, unsigned int);
437 int (*setlease)(struct file *, long, struct file_lock **);
438 long (*fallocate)(struct file *, int, loff_t, loff_t);
439 };
441 locking rules:
442 All may block except for ->setlease.
443 No VFS locks held on entry except for ->fsync and ->setlease.
445 ->fsync() has i_mutex on inode.
447 ->setlease has the file_list_lock held and must not sleep.
449 ->llseek() locking has moved from llseek to the individual llseek
450 implementations. If your fs is not using generic_file_llseek, you
451 need to acquire and release the appropriate locks in your ->llseek().
452 For many filesystems, it is probably safe to acquire the inode
453 mutex or just to use i_size_read() instead.
454 Note: this does not protect the file->f_pos against concurrent modifications
455 since this is something the userspace has to take care about.
457 ->fasync() is responsible for maintaining the FASYNC bit in filp->f_flags.
458 Most instances call fasync_helper(), which does that maintenance, so it's
459 not normally something one needs to worry about. Return values > 0 will be
460 mapped to zero in the VFS layer.
462 ->readdir() and ->ioctl() on directories must be changed. Ideally we would
463 move ->readdir() to inode_operations and use a separate method for directory
464 ->ioctl() or kill the latter completely. One of the problems is that for
465 anything that resembles union-mount we won't have a struct file for all
466 components. And there are other reasons why the current interface is a mess...
468 ->read on directories probably must go away - we should just enforce -EISDIR
469 in sys_read() and friends.
471 --------------------------- dquot_operations -------------------------------
472 prototypes:
473 int (*write_dquot) (struct dquot *);
474 int (*acquire_dquot) (struct dquot *);
475 int (*release_dquot) (struct dquot *);
476 int (*mark_dirty) (struct dquot *);
477 int (*write_info) (struct super_block *, int);
479 These operations are intended to be more or less wrapping functions that ensure
480 a proper locking wrt the filesystem and call the generic quota operations.
482 What filesystem should expect from the generic quota functions:
484 FS recursion Held locks when called
485 write_dquot: yes dqonoff_sem or dqptr_sem
486 acquire_dquot: yes dqonoff_sem or dqptr_sem
487 release_dquot: yes dqonoff_sem or dqptr_sem
488 mark_dirty: no -
489 write_info: yes dqonoff_sem
491 FS recursion means calling ->quota_read() and ->quota_write() from superblock
492 operations.
494 More details about quota locking can be found in fs/dquot.c.
496 --------------------------- vm_operations_struct -----------------------------
497 prototypes:
498 void (*open)(struct vm_area_struct*);
499 void (*close)(struct vm_area_struct*);
500 int (*fault)(struct vm_area_struct*, struct vm_fault *);
501 int (*page_mkwrite)(struct vm_area_struct *, struct vm_fault *);
502 int (*access)(struct vm_area_struct *, unsigned long, void*, int, int);
504 locking rules:
505 mmap_sem PageLocked(page)
506 open: yes
507 close: yes
508 fault: yes can return with page locked
509 page_mkwrite: yes can return with page locked
510 access: yes
512 ->fault() is called when a previously not present pte is about
513 to be faulted in. The filesystem must find and return the page associated
514 with the passed in "pgoff" in the vm_fault structure. If it is possible that
515 the page may be truncated and/or invalidated, then the filesystem must lock
516 the page, then ensure it is not already truncated (the page lock will block
517 subsequent truncate), and then return with VM_FAULT_LOCKED, and the page
518 locked. The VM will unlock the page.
520 ->page_mkwrite() is called when a previously read-only pte is
521 about to become writeable. The filesystem again must ensure that there are
522 no truncate/invalidate races, and then return with the page locked. If
523 the page has been truncated, the filesystem should not look up a new page
524 like the ->fault() handler, but simply return with VM_FAULT_NOPAGE, which
525 will cause the VM to retry the fault.
527 ->access() is called when get_user_pages() fails in
528 acces_process_vm(), typically used to debug a process through
529 /proc/pid/mem or ptrace. This function is needed only for
530 VM_IO | VM_PFNMAP VMAs.
532 ================================================================================
533 Dubious stuff
535 (if you break something or notice that it is broken and do not fix it yourself
536 - at least put it here)