kevin/linux
1
Fork 0
Code
277,258 Commits 1 Branch 830 Tags 5.6 GiB
4ed5e82fe7
Commit Graph

268 Commits

Author SHA1 Message Date
Miklos Szeredi
4ed5e82fe7 vfs: protect remounting superblock read-only 
Currently remouting superblock read-only is racy in a major way.

With the per mount read-only infrastructure it is now possible to
prevent most races, which this patch attempts.

Before starting the remount read-only, iterate through all mounts
belonging to the superblock and if none of them have any pending
writes, set sb->s_readonly_remount.  This indicates that remount is in
progress and no further write requests are allowed.  If the remount
succeeds set MS_RDONLY and reset s_readonly_remount.

If the remounting is unsuccessful just reset s_readonly_remount.
This can result in transient EROFS errors, despite the fact the
remount failed.  Unfortunately hodling off writes is difficult as
remount itself may touch the filesystem (e.g. through load_nls())
which would deadlock.

A later patch deals with delayed writes due to nlink going to zero.

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Toshiyuki Okajima <toshi.okajima@jp.fujitsu.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-06 23:20:12 -05:00
Miklos Szeredi
39f7c4db1d vfs: keep list of mounts for each superblock 
Keep track of vfsmounts belonging to a superblock.  List is protected
by vfsmount_lock.

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Tested-by: Toshiyuki Okajima <toshi.okajima@jp.fujitsu.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-06 23:20:12 -05:00
Al Viro
34c80b1d93 vfs: switch ->show_options() to struct dentry * 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-06 23:19:54 -05:00
Al Viro
d10577a8d8 vfs: trim includes a bit 
[folded fix for missing magic.h from Tetsuo Handa]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:13 -05:00
Al Viro
be08d6d260 switch mnt_namespace ->root to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:13 -05:00
Al Viro
0226f4923f vfs: take /proc/*/mounts and friends to fs/proc_namespace.c 
rationale: that stuff is far tighter bound to fs/namespace.c than to
the guts of procfs proper.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:13 -05:00
Al Viro
3a2393d71d vfs: opencode mntget() mnt_set_mountpoint() 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:12 -05:00
Al Viro
909b0a88ef vfs: spread struct mount - remaining argument of next_mnt() 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:12 -05:00
Al Viro
c63181e6b6 vfs: move fsnotify junk to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:12 -05:00
Al Viro
52ba1621de vfs: move mnt_devname 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:11 -05:00
Al Viro
1a4eeaf2a8 vfs: move mnt_list to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:11 -05:00
Al Viro
fc7be130c7 vfs: switch pnode.h macros to struct mount * 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:11 -05:00
Al Viro
863d684f94 vfs: move the rest of int fields to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:10 -05:00
Al Viro
15169fe784 vfs: mnt_id/mnt_group_id moved 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:10 -05:00
Al Viro
143c8c91ce vfs: mnt_ns moved to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:09 -05:00
Al Viro
900148dcac vfs: spread struct mount - mntput_no_expire 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:09 -05:00
Al Viro
95bc5f25c1 vfs: spread struct mount - do_add_mount and graft_tree 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:09 -05:00
Al Viro
6776db3d32 vfs: take mnt_share/mnt_slave/mnt_slave_list and mnt_expire to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:08 -05:00
Al Viro
32301920f4 vfs: and now we can make ->mnt_master point to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:08 -05:00
Al Viro
d10e8def07 vfs: take mnt_master to struct mount 
make IS_MNT_SLAVE take struct mount * at the same time

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:08 -05:00
Al Viro
14cf1fa8f5 vfs: spread struct mount - remaining argument of mnt_set_mountpoint() 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:07 -05:00
Al Viro
a8d56d8e4f vfs: spread struct mount - propagate_mnt() 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:07 -05:00
Al Viro
6fc7871fed vfs: spread struct mount - get_dominating_id / do_make_slave 
next pile of horrors, similar to mnt_parent one; this time it's
mnt_master.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:06 -05:00
Al Viro
6b41d536f7 vfs: take mnt_child/mnt_mounts to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:06 -05:00
Al Viro
68e8a9feab vfs: all counters taken to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:06 -05:00
Al Viro
83adc75322 vfs: spread struct mount - work with counters 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:05 -05:00
Al Viro
a73324da7a vfs: move mnt_mountpoint to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:05 -05:00
Al Viro
0714a53380 vfs: now it can be done - make mnt_parent point to struct mount 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:05 -05:00
Al Viro
3376f34fff vfs: mnt_parent moved to struct mount 
the second victim...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:04 -05:00
Al Viro
643822b41e vfs: spread struct mount - is_path_reachable 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:04 -05:00
Al Viro
676da58df7 vfs: spread struct mount - mnt_has_parent 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:04 -05:00
Al Viro
1ab5973862 vfs: spread struct mount - do_umount/propagate_mount_busy 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:03 -05:00
Al Viro
44d964d609 vfs: spread struct mount mnt_set_mountpoint child argument 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:03 -05:00
Al Viro
87129cc0e3 vfs: spread struct mount - clone_mnt/copy_tree argument 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:03 -05:00
Al Viro
692afc312b vfs: spread struct mount - shrink_submounts/select_submounts 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:02 -05:00
Al Viro
761d5c38eb vfs: spread struct mount - umount_tree argument 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:02 -05:00
Al Viro
1b8e5564b9 vfs: the first spoils - mnt_hash moved 
taken out of struct vfsmount into struct mount

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:02 -05:00
Al Viro
d5e50f74dd vfs: spread struct mount to remaining users of ->mnt_hash 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:01 -05:00
Al Viro
cb338d06e9 vfs: spread struct mount - clone_mnt/copy_tree result 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:01 -05:00
Al Viro
0f0afb1dcf vfs: spread struct mount - change_mnt_propagation/set_mnt_shared 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:01 -05:00
Al Viro
b105e270b4 vfs: spread struct mount - alloc_vfsmnt/free_vfsmnt/mnt_alloc_id/mnt_free_id 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:00 -05:00
Al Viro
cbbe362cd6 vfs: spread struct mount - tree_contains_unbindable 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:00 -05:00
Al Viro
0fb54e5056 vfs: spread struct mount - attach_recursive_mnt 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:57:00 -05:00
Al Viro
4b8b21f4fe vfs: spread struct mount - mount group id handling 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:56:59 -05:00
Al Viro
4b2619a571 vfs: spread struct mount - commit_tree 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:56:59 -05:00
Al Viro
419148da6e vfs: spread struct mount - attach_mnt/detach_mnt 
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:56:59 -05:00
Al Viro
315fc83e56 vfs: spread struct mount - namespace.c internal iterators 
next_mnt() return value, first argument
skip_mnt_tree() return value and argument

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:56:58 -05:00
Al Viro
c71053659e vfs: spread struct mount - __lookup_mnt() result 
switch __lookup_mnt() to returning struct mount *; callers adjusted.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:56:58 -05:00
Al Viro
7d6fec45a5 vfs: start hiding vfsmount guts series 
Almost all fields of struct vfsmount are used only by core VFS (and
a fairly small part of it, at that).  The plan: embed struct vfsmount
into struct mount, making the latter visible only to core parts of VFS.
Then move fields from vfsmount to mount, eventually leaving only
mnt_root/mnt_sb/mnt_flags in struct vfsmount.  Filesystem code still
gets pointers to struct vfsmount and remains unchanged; all such
pointers go to struct vfsmount embedded into the instances of struct
mount allocated by fs/namespace.c.  When fs/namespace.c et.al. get
a pointer to vfsmount, they turn it into pointer to mount (using
container_of) and work with that.

This is the first part of series; struct mount is introduced,
allocation switched to using it.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:56:57 -05:00
Al Viro
2a79f17e4a vfs: mnt_drop_write_file() 
new helper (wrapper around mnt_drop_write()) to be used in pair with
mnt_want_write_file().

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
 2012-01-03 22:52:40 -05:00