Al Viro [Mon, 21 Nov 2011 19:59:34 +0000 (14:59 -0500)]
switch open and mkdir syscalls to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 21 Nov 2011 19:58:38 +0000 (14:58 -0500)]
switch ->path_mknod() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 21 Nov 2011 19:56:21 +0000 (14:56 -0500)]
switch ->path_mkdir() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sat, 27 Aug 2011 03:03:17 +0000 (23:03 -0400)]
tomoyo_mini_stat: switch to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 27 Jul 2011 18:54:36 +0000 (14:54 -0400)]
consolidate a bunch of ipcbuf.h instances
... some still remain weird :-/
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 27 Jul 2011 18:04:25 +0000 (14:04 -0400)]
auditsc: propage umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 27 Jul 2011 18:03:22 +0000 (14:03 -0400)]
switch kern_ipc_perm to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 21:04:15 +0000 (17:04 -0400)]
consolidate umode_t declarations
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 20:50:23 +0000 (16:50 -0400)]
switch spu_create(2) to use of SYSCALL_DEFINE4, make it use umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 09:26:10 +0000 (05:26 -0400)]
switch mq_open() to umode_t
Al Viro [Tue, 26 Jul 2011 08:52:22 +0000 (04:52 -0400)]
ceph: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:48:59 +0000 (04:48 -0400)]
switch hostfs_iattr to explicit unsigned short
It's shared between kernel-compiled hostfs_kern and userland-compiled
hostfs_user (it's uml stuff). Use explicit type instead of playing
silly buggers with mode_t. It's not a userland API per se; it interacts
between code compiled with types same as for host kernel and, directly
linked to it, code talking to libc. Both sides come from the same
kernel source...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:47:38 +0000 (04:47 -0400)]
switch miscdevice to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:47:14 +0000 (04:47 -0400)]
switch spufs guts to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:34:13 +0000 (04:34 -0400)]
init/initramfs.c: should use umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:31:40 +0000 (04:31 -0400)]
switch may_mknod() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:31:05 +0000 (04:31 -0400)]
should_remove_suid(): inode->i_mode is umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:30:04 +0000 (04:30 -0400)]
switch securityfs_create_file() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:25:58 +0000 (04:25 -0400)]
switch security_path_chmod() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 08:22:01 +0000 (04:22 -0400)]
switch sys_chmod()/sys_fchmod()/sys_fchmodat() to umode_t
SYSCALLx magic should take care of things, according to Linus...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:47:31 +0000 (03:47 -0400)]
sysctl: use umode_t for table permissions
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 00:49:29 +0000 (20:49 -0400)]
get rid of open-coded S_ISREG(), etc.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 18:18:20 +0000 (14:18 -0400)]
mqueue: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:21:59 +0000 (23:21 -0400)]
switch is_sxid() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:20:18 +0000 (23:20 -0400)]
switch inode_init_owner() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:30:54 +0000 (03:30 -0400)]
fs: propagate umode_t, misc bits
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:29:03 +0000 (03:29 -0400)]
autofs4: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:26:51 +0000 (03:26 -0400)]
hfsplus: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:26:05 +0000 (03:26 -0400)]
hfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:22:14 +0000 (03:22 -0400)]
cifs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:21:30 +0000 (03:21 -0400)]
fat: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:20:46 +0000 (03:20 -0400)]
ntfs: propagate umode_t
same story as with isofs and udf...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:19:52 +0000 (03:19 -0400)]
isofs: propagate umode_t
situation with mount options is the same as for udf
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:18:29 +0000 (03:18 -0400)]
udf: propagate umode_t
note re mount options: fmask and dmask are explicitly truncated to 12bit,
UDF_INVALID_MODE just needs to be guaranteed to differ from any such value.
And umask is used only in &= with umode_t, so we ignore other bits anyway.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:17:33 +0000 (03:17 -0400)]
fuse: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:16:55 +0000 (03:16 -0400)]
pohmelfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:15:03 +0000 (03:15 -0400)]
shmem, ramfs: propagate umode_t, open-coded S_ISREG
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:13:21 +0000 (03:13 -0400)]
logfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:12:59 +0000 (03:12 -0400)]
ubifs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:12:16 +0000 (03:12 -0400)]
ncpfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:17:40 +0000 (23:17 -0400)]
hugetlbfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:07:49 +0000 (03:07 -0400)]
exofs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:07:14 +0000 (03:07 -0400)]
nilfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 04:35:13 +0000 (00:35 -0400)]
sunrpc: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 07:04:30 +0000 (03:04 -0400)]
affs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 05:55:55 +0000 (01:55 -0400)]
cgroup: propagate mode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:27:00 +0000 (23:27 -0400)]
infiniband: umode_t noise, including open-coded S_ISDIR()
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:44:53 +0000 (23:44 -0400)]
sysfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:36:12 +0000 (23:36 -0400)]
usbfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 03:30:19 +0000 (23:30 -0400)]
hypfs: umode_t noise and properly spelled S_ISDIR/S_ISREG
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:55:32 +0000 (02:55 -0400)]
ocfs2: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 17:43:19 +0000 (13:43 -0400)]
dlmfs: use inode_init_owner()
don't open-code it...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:53:22 +0000 (02:53 -0400)]
9p: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 02:58:10 +0000 (22:58 -0400)]
omfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:50:53 +0000 (02:50 -0400)]
reiserfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:50:15 +0000 (02:50 -0400)]
xfs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:49:38 +0000 (02:49 -0400)]
sysv: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:49:13 +0000 (02:49 -0400)]
ufs: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:48:39 +0000 (02:48 -0400)]
minix: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:48:06 +0000 (02:48 -0400)]
ext4: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:46:57 +0000 (02:46 -0400)]
ext3: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 06:46:18 +0000 (02:46 -0400)]
ext2: propagate umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 22:59:07 +0000 (18:59 -0400)]
9p: don't bother with unixmode2p9mode() for link() and symlink()
Pass perm to v9fs_vfs_mkspecial() instead of passing mode;
calculate in caller when done for mknod(), use known value for link()
and symlink(). As the result, we avoid a bit of work *and* stop
mixing mode_t with P9_DMLINK.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 17:47:19 +0000 (13:47 -0400)]
kill ecryptfs_create_underlying_file()
it's a just a wrapper for vfs_create()
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 04:05:26 +0000 (00:05 -0400)]
configfs: convert to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 14:47:56 +0000 (10:47 -0400)]
switch devtmpfs to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 08:33:43 +0000 (04:33 -0400)]
switch debugfs to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 07:40:40 +0000 (03:40 -0400)]
switch sysfs_chmod_file() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 07:36:29 +0000 (03:36 -0400)]
switch procfs to umode_t use
both proc_dir_entry ->mode and populating functions
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 03:11:19 +0000 (23:11 -0400)]
switch ->is_visible() to returning umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 03:10:46 +0000 (23:10 -0400)]
switch sysfs attr->mode to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Sun, 24 Jul 2011 00:24:48 +0000 (20:24 -0400)]
switch device_get_devnode() and ->devnode() to umode_t *
both callers of device_get_devnode() are only interested in lower 16bits
and nobody tries to return anything wider than 16bit anyway.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 00:20:48 +0000 (20:20 -0400)]
hugetlbfs: switch to inode_init_owner()
... rather than open-coding it
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 05:52:52 +0000 (01:52 -0400)]
switch ->mknod() to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 05:42:34 +0000 (01:42 -0400)]
switch ->create() to umode_t
vfs_create() ignores everything outside of 16bit subset of its
mode argument; switching it to umode_t is obviously equivalent
and it's the only caller of the method
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 26 Jul 2011 05:41:39 +0000 (01:41 -0400)]
switch vfs_mkdir() and ->mkdir() to umode_t
vfs_mkdir() gets int, but immediately drops everything that might not
fit into umode_t and that's the only caller of ->mkdir()...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 25 Jul 2011 21:32:17 +0000 (17:32 -0400)]
switch sys_mknodat(2) to umode_t
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 4 Jan 2012 02:01:29 +0000 (21:01 -0500)]
vfs: fix the rest of sget() races
unfortunately, just checking MS_BORN after having grabbed ->s_umount in
sget() is not enough; places that pick superblock from a list and
grab s_umount shared need the same check in addition to checking for
->s_root; otherwise three-way race between failing mount, sget() and
such list-walker can leave us with list-walker coming *second*, when
temporary active ref grabbed by sget() (to be dropped when sget()
notices that original mount has failed by checking MS_BORN) has
lead to deactivate_locked_super() from failing ->mount() *not* doing
->kill_sb() and just releasing ->s_umount. Once sget() gets through
and notices that MS_BORN had never been set it will drop the active
ref and fs will be shut down and kicked out of all lists, but it's
too late for something like sync_supers().
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 3 Jan 2012 03:28:36 +0000 (22:28 -0500)]
vfs: new helper - vfs_ustat()
... and bury user_get_super()/statfs_by_dentry() - they are
purely internal now.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 9 Dec 2011 04:01:06 +0000 (23:01 -0500)]
vfs: live vfsmounts never have NULL ->mnt_sb
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 7 Dec 2011 23:21:57 +0000 (18:21 -0500)]
vfs: for usbfs, etc. internal vfsmounts ->mnt_sb->s_root == ->mnt_root
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 9 Dec 2011 01:08:42 +0000 (20:08 -0500)]
get rid of timer in kern/acct.c
... and clean it up a bit, while we are at it
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 19 Dec 2011 01:17:41 +0000 (20:17 -0500)]
vfs: pipe.c is really non-modular
... so no exitcalls there. Not much would work if pipe(2) would stop
working, after all...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Mon, 12 Dec 2011 20:51:45 +0000 (15:51 -0500)]
vfs: fix the stupidity with i_dentry in inode destructors
Seeing that just about every destructor got that INIT_LIST_HEAD() copied into
it, there is no point whatsoever keeping this INIT_LIST_HEAD in inode_init_once();
the cost of taking it into inode_init_always() will be negligible for pipes
and sockets and negative for everything else. Not to mention the removal of
boilerplate code from ->destroy_inode() instances...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 9 Dec 2011 13:06:57 +0000 (08:06 -0500)]
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>
Al Viro [Fri, 9 Dec 2011 01:18:57 +0000 (20:18 -0500)]
constify seq_file stuff
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 13 Dec 2011 04:07:05 +0000 (23:07 -0500)]
vfs: make do_kern_mount() static
the only user outside of fs/namespace.c has died
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Tue, 13 Dec 2011 03:53:00 +0000 (22:53 -0500)]
vfs: convert fs_supers to hlist
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Thu, 17 Nov 2011 02:52:06 +0000 (21:52 -0500)]
make nfs_follow_remote_path() handle ERR_PTR() passed as root_mnt
... rather than duplicating that in callers
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 25 Nov 2011 07:22:06 +0000 (02:22 -0500)]
vfs: kill ->mnt_devname use in afs printks
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 25 Nov 2011 01:14:54 +0000 (20:14 -0500)]
btrfs, nfs, apparmor: don't pull mnt_namespace.h for no reason...
it's not needed anymore; we used to, back when we had to do
mount_subtree() by hand, complete with put_mnt_ns() in it.
No more... Apparmor didn't need it since the __d_path() fix.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 25 Nov 2011 01:08:53 +0000 (20:08 -0500)]
tomoyo: stop including hell knows what
tomoyo/realpath.c needs exactly one include - that of common.h. It pulls
everything the thing needs, without doing ridiculous garbage such as trying
to include ../../fs/internal.h. If that alone doesn't scream "layering
violation", I don't know what does; and these days it's all for nothing,
since it fortunately does not use any symbols defined in there...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 25 Nov 2011 00:31:36 +0000 (19:31 -0500)]
vfs: dentry_reset_mounted() doesn't use vfsmount argument
lose it
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Thu, 17 Nov 2011 03:01:36 +0000 (22:01 -0500)]
unexport put_mnt_ns(), make create_mnt_ns() static outright
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 23 Nov 2011 17:27:01 +0000 (12:27 -0500)]
vfs: add missing parens in pnode.h macros
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Thu, 24 Nov 2011 00:34:49 +0000 (19:34 -0500)]
vfs: more mnt_parent cleanups
a) mount --move is checking that ->mnt_parent is non-NULL before
looking if that parent happens to be shared; ->mnt_parent is never
NULL and it's not even an misspelled !mnt_has_parent()
b) pivot_root open-codes is_path_reachable(), poorly.
c) so does path_is_under(), while we are at it.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Thu, 24 Nov 2011 00:26:23 +0000 (19:26 -0500)]
vfs: new internal helper: mnt_has_parent(mnt)
vfsmounts have ->mnt_parent pointing either to a different vfsmount
or to itself; it's never NULL and termination condition in loops
traversing the tree towards root is mnt == mnt->mnt_parent. At least
one place (see the next patch) is confused about what's going on;
let's add an explicit helper checking it right way and use it in
all places where we need it. Not that there had been too many,
but...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Thu, 24 Nov 2011 00:04:52 +0000 (19:04 -0500)]
vfs: kill pointless helpers in namespace.c
mnt_{inc,dec}_count() is not cleaner than doing the corresponding
mnt_add_count() directly and mnt_set_count() is not used at all.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 23 Nov 2011 17:03:18 +0000 (12:03 -0500)]
new helpers: fh_{want,drop}_write()
A bunch of places in nfsd does mnt_{want,drop}_write on vfsmount of
export of given fhandle. Switched to obvious inlined helpers...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Wed, 23 Nov 2011 16:57:51 +0000 (11:57 -0500)]
switch a bunch of places to mnt_want_write_file()
it's both faster (in case when file has been opened for write) and cleaner.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
projects / openwrt / staging / blogic.git / log