Assorted work from the last cycle; among other things, there's
a filesystem API change (->evict_inode() series), a lot of truncate
rework stuff (mostly by Chris) and a bunch of fixes.
This is _not_ all; in particular, there's pending freeze/thaw
locking rework and I would really like to see inode_lock splitup
series separated and posted. Along with vfsmount_lock and files_lock
work. Nick, could you do that ASAP, please?
Please, pull from
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6.git/ for-linus
Shortlog:
Al Viro (69):
Deal with missing exports for hostfs
missing include in hppfs
Fix reiserfs_file_release()
fix leak in __logfs_create()
jffs2: don't open-code iget_failed()
get rid of file_fsync()
simplify checks for I_CLEAR/I_FREEING
unify fs/inode.c callers of clear_inode()
New method - evict_inode()
switch hugetlbfs to ->evict_inode()
generic_detach_inode() can be static now
Take ->i_bdev/->i_cdev handling out of clear_inode()
new helper: end_writeback()
switch hypfs to ->evict_inode()
simplify get_cramfs_inode()
switch procfs to ->evict_inode()
switch sysfs to ->evict_inode()
switch minix to ->evict_inode(), fix write_inode/delete_inode race
ext2: switch to dquot_free_block_nodirty()
Take dirtying the inode to callers of ext2_free_blocks()
Don't dirty the victim in ext2_xattr_delete_inode()
merge ext2 delete_inode and clear_inode, switch to ->evict_inode()
switch mqueue to ->evict_inode()
switch shmem.c to ->evice_inode()
switch sysv to ->evict_inode()
switch smbfs to evict_inode()
covert fatfs to ->evict_inode()
switch ufs to ->evict_inode()
spufs conversion to ->evict_inode()
convert ext3 to ->evict_inode()
switch bfs to ->evict_inode(), clean up
switch omfs to ->evict_inode()
switch affs to ->evict_inode()
stop icache pollution in hostfs, switch to ->evict_inode()
hostfs: don't keep a field in each inode when we are using it only in root
get rid of hostfs_read_inode()
hostfs: pass pathname to init_inode()
switch stat_file() to passing a single struct rather than fsckloads of pointers
hostfs: get rid of file_type(), fold init_inode()
hostfs: get rid of inode_dentry_name()
hostfs: sanitize symlinks
new helper: __dentry_path()
hostfs: fix races in dentry_name() and inode_name()
leak in hostfs_unlink()
try to get rid of races in hostfs open()
switch hppfs to ->evict_inode()
switch hpfs to ->evict_inode()
switch jfs to ->evict_inode()
switch ubifs to ->evict_inode()
switch udf to ->evict_inode()
switch ncpfs to ->evict_inode()
convert ocfs2 to ->evict_inode()
switch gfs2 to ->evict_inode()
convert btrfs to ->evict_inode()
convert reiserfs to ->evict_inode()
convert exofs to ->evict_inode()
convert nilfs2 to ->evict_inode()
logfs: get rid of magical inodes
convert logfs to ->evict_inode()
convert ext4 to ->evict_inode()
->delete_inode() is gone
fs/inode.c:evict() doesn't care about delete vs. non-delete paths now
fs/inode.c:clear_inode() is gone
Make ->drop_inode() just return whether inode needs to be dropped
convert remaining ->clear_inode() to ->evict_inode()
All filesystems that need invalidate_inode_buffers() are doing that explicitly
update VFS documentation for method changes.
Fix sget() race with failing mount
no need for list_for_each_entry_safe()/resetting with superblock list
Andreas Gruenbacher (2):
mbcache: Remove unused features
mbcache: fix shrinker function return value
Artem Bityutskiy (6):
AFFS: clean up dirty flag usage
AFFS: wait for sb synchronization when needed
BFS: clean up the superblock usage
btrfs: remove junk sb_dirt change
sysv: do not mark superblock dirty on mount
sysv: do not mark superblock dirty on remount
Boaz Harrosh (1):
exofs: New truncate sequence
Christoph Hellwig (17):
sort out blockdev_direct_IO variants
get rid of nobh_write_begin_newtrunc
get rid of cont_write_begin_newtrunc
clean up write_begin usage for directories in pagecache
introduce __block_write_begin
get rid of block_write_begin_newtrunc
add missing setattr methods
rename generic_setattr
default to simple_setattr
remove inode_setattr
always call inode_change_ok early in ->setattr
check ATTR_SIZE contraints in inode_change_ok
update documentation for the new truncate sequence
xfs: new truncate sequence
pass a struct path to vfs_statfs
add f_flags to struct statfs(64)
cifs: truncate fallout
Tejun Heo (1):
vfs: don't hold s_umount over close_bdev_exclusive() call
Diffstat:
Documentation/filesystems/Locking | 22 +-
Documentation/filesystems/porting | 45 +++
arch/alpha/kernel/osf_sys.c | 8 +-
arch/mips/include/asm/statfs.h | 12 +-
arch/parisc/hpux/sys_hpux.c | 10 +-
arch/powerpc/platforms/cell/spufs/inode.c | 12 +-
arch/s390/hypfs/inode.c | 6 +-
arch/s390/include/asm/statfs.h | 9 +-
arch/um/include/shared/os.h | 3 +
arch/um/kernel/ksyms.c | 3 +
arch/um/os-Linux/file.c | 15 +
arch/um/os-Linux/user_syms.c | 4 +
drivers/staging/pohmelfs/inode.c | 18 +-
fs/9p/v9fs_vfs.h | 2 +-
fs/9p/vfs_inode.c | 19 +-
fs/9p/vfs_super.c | 4 +-
fs/adfs/inode.c | 16 +-
fs/affs/affs.h | 3 +-
fs/affs/file.c | 11 +-
fs/affs/inode.c | 38 ++-
fs/affs/super.c | 32 +--
fs/afs/inode.c | 5 +-
fs/afs/internal.h | 2 +-
fs/afs/super.c | 2 +-
fs/attr.c | 88 +++---
fs/bfs/bfs.h | 1 -
fs/bfs/file.c | 17 +-
fs/bfs/inode.c | 116 +++-----
fs/binfmt_misc.c | 5 +-
fs/block_dev.c | 17 +-
fs/btrfs/ctree.h | 4 +-
fs/btrfs/inode.c | 34 ++-
fs/btrfs/super.c | 2 +-
fs/buffer.c | 180 ++--------
fs/cachefiles/bind.c | 2 +-
fs/cachefiles/daemon.c | 6 +-
fs/cifs/cifsfs.c | 15 +-
fs/cifs/inode.c | 86 +++---
fs/coda/inode.c | 8 +-
fs/compat.c | 10 +-
fs/cramfs/inode.c | 88 +++---
fs/dcache.c | 39 ++-
fs/direct-io.c | 74 ++---
fs/drop_caches.c | 2 +-
fs/ecryptfs/inode.c | 18 +-
fs/ecryptfs/super.c | 14 +-
fs/exofs/exofs.h | 3 +-
fs/exofs/file.c | 1 -
fs/exofs/inode.c | 127 +++----
fs/exofs/super.c | 2 +-
fs/ext2/balloc.c | 11 +-
fs/ext2/dir.c | 23 +-
fs/ext2/ext2.h | 5 +-
fs/ext2/ialloc.c | 13 +-
fs/ext2/inode.c | 87 +++---
fs/ext2/super.c | 14 +-
fs/ext2/xattr.c | 25 +-
fs/ext3/ialloc.c | 12 -
fs/ext3/inode.c | 63 +++-
fs/ext3/super.c | 14 +-
fs/ext3/xattr.c | 12 +-
fs/ext4/ext4.h | 3 +-
fs/ext4/ialloc.c | 2 +-
fs/ext4/inode.c | 53 ++-
fs/ext4/super.c | 10 +-
fs/ext4/xattr.c | 12 +-
fs/fat/fat.h | 1 -
fs/fat/file.c | 49 +--
fs/fat/inode.c | 26 +-
fs/freevxfs/vxfs_extern.h | 2 +-
fs/freevxfs/vxfs_inode.c | 8 +-
fs/freevxfs/vxfs_super.c | 2 +-
fs/fs-writeback.c | 8 +-
fs/fuse/dir.c | 17 +-
fs/fuse/inode.c | 6 +-
fs/gfs2/aops.c | 10 +-
fs/gfs2/inode.c | 27 +-
fs/gfs2/ops_inode.c | 18 +-
fs/gfs2/super.c | 43 +--
fs/gfs2/xattr.c | 24 +-
fs/hfs/hfs_fs.h | 2 +-
fs/hfs/inode.c | 70 ++++-
fs/hfs/super.c | 2 +-
fs/hfsplus/hfsplus_fs.h | 1 +
fs/hfsplus/inode.c | 77 ++++-
fs/hfsplus/super.c | 10 +-
fs/hostfs/hostfs.h | 22 +-
fs/hostfs/hostfs_kern.c | 517 +++++++++++++----------------
fs/hostfs/hostfs_user.c | 112 ++-----
fs/hpfs/file.c | 11 +-
fs/hpfs/hpfs_fn.h | 2 +-
fs/hpfs/inode.c | 24 +-
fs/hpfs/super.c | 2 +-
fs/hppfs/hppfs.c | 8 +-
fs/hugetlbfs/inode.c | 41 +--
fs/inode.c | 177 ++++-------
fs/jffs2/dir.c | 16 +-
fs/jffs2/fs.c | 10 +-
fs/jffs2/os-linux.h | 2 +-
fs/jffs2/super.c | 2 +-
fs/jffs2/xattr.c | 2 +-
fs/jfs/file.c | 14 +-
fs/jfs/inode.c | 63 +++-
fs/jfs/jfs_inode.h | 2 +-
fs/jfs/super.c | 8 +-
fs/libfs.c | 70 +---
fs/logfs/dir.c | 5 +-
fs/logfs/file.c | 18 +-
fs/logfs/inode.c | 51 +--
fs/logfs/journal.c | 2 -
fs/logfs/logfs.h | 4 +-
fs/logfs/readwrite.c | 62 ++--
fs/logfs/segment.c | 1 -
fs/logfs/super.c | 23 +-
fs/mbcache.c | 168 +++-------
fs/minix/bitmap.c | 6 +-
fs/minix/dir.c | 21 +-
fs/minix/file.c | 22 ++
fs/minix/inode.c | 35 ++-
fs/minix/minix.h | 4 +-
fs/namespace.c | 2 +-
fs/ncpfs/inode.c | 36 +-
fs/nfs/inode.c | 13 +-
fs/nfs/internal.h | 4 +-
fs/nfs/super.c | 4 +-
fs/nfsd/nfs4xdr.c | 6 +-
fs/nfsd/vfs.c | 10 +-
fs/nilfs2/dir.c | 25 +-
fs/nilfs2/gcdat.c | 2 +-
fs/nilfs2/inode.c | 78 ++++-
fs/nilfs2/nilfs.h | 2 +-
fs/nilfs2/recovery.c | 11 +-
fs/nilfs2/super.c | 20 +-
fs/notify/inode_mark.c | 6 +-
fs/notify/inotify/inotify.c | 7 +-
fs/ntfs/inode.c | 10 +-
fs/ntfs/inode.h | 2 +-
fs/ntfs/super.c | 2 +-
fs/ocfs2/aops.c | 9 +-
fs/ocfs2/dlmfs/dlmfs.c | 15 +-
fs/ocfs2/file.c | 22 +-
fs/ocfs2/inode.c | 29 +-
fs/ocfs2/inode.h | 5 +-
fs/ocfs2/super.c | 3 +-
fs/omfs/file.c | 36 ++-
fs/omfs/inode.c | 9 +-
fs/proc/base.c | 16 +-
fs/proc/generic.c | 18 +-
fs/proc/inode.c | 6 +-
fs/proc/proc_sysctl.c | 15 +-
fs/qnx4/inode.c | 11 +-
fs/quota/dquot.c | 2 +-
fs/ramfs/file-nommu.c | 7 +-
fs/reiserfs/file.c | 50 ++--
fs/reiserfs/inode.c | 134 +++++---
fs/reiserfs/super.c | 10 +-
fs/smbfs/inode.c | 12 +-
fs/statfs.c | 95 ++++--
fs/super.c | 51 ++-
fs/sync.c | 25 --
fs/sysfs/inode.c | 8 +-
fs/sysfs/mount.c | 2 +-
fs/sysfs/sysfs.h | 2 +-
fs/sysv/dir.c | 21 +-
fs/sysv/file.c | 22 ++
fs/sysv/ialloc.c | 1 -
fs/sysv/inode.c | 19 +-
fs/sysv/itree.c | 19 +-
fs/sysv/super.c | 1 -
fs/sysv/sysv.h | 4 +-
fs/ubifs/file.c | 23 +-
fs/ubifs/super.c | 12 +-
fs/ubifs/ubifs.h | 2 +-
fs/udf/file.c | 22 ++
fs/udf/ialloc.c | 2 -
fs/udf/inode.c | 61 ++--
fs/udf/super.c | 3 +-
fs/udf/udfdecl.h | 3 +-
fs/ufs/dir.c | 13 +-
fs/ufs/ialloc.c | 2 -
fs/ufs/inode.c | 63 +++--
fs/ufs/super.c | 2 +-
fs/ufs/truncate.c | 16 +-
fs/ufs/ufs.h | 2 +-
fs/ufs/util.h | 4 +-
fs/xfs/linux-2.6/xfs_aops.c | 62 +++-
fs/xfs/linux-2.6/xfs_iops.c | 20 +-
fs/xfs/linux-2.6/xfs_linux.h | 2 -
fs/xfs/linux-2.6/xfs_super.c | 8 +-
fs/xfs/linux-2.6/xfs_trace.h | 2 +-
fs/xfs/xfs_vnodeops.c | 38 ++-
include/asm-generic/statfs.h | 9 +-
include/linux/buffer_head.h | 20 +-
include/linux/dcache.h | 1 +
include/linux/ext3_fs.h | 2 +-
include/linux/fs.h | 66 +---
include/linux/mbcache.h | 20 +-
include/linux/mm.h | 1 +
include/linux/reiserfs_fs.h | 2 +-
include/linux/reiserfs_fs_i.h | 4 +-
include/linux/statfs.h | 25 ++-
ipc/mqueue.c | 16 +-
kernel/acct.c | 2 +-
mm/shmem.c | 23 +-
mm/truncate.c | 38 ++-
205 files changed, 2584 insertions(+), 2475 deletions(-)
Hi Al,
On Tue, 10 Aug 2010 04:35:36 +0100 Al Viro <[email protected]> wrote:
>
> Assorted work from the last cycle; among other things, there's
> a filesystem API change (->evict_inode() series), a lot of truncate
> rework stuff (mostly by Chris) and a bunch of fixes.
Did you remember to apply the three fix patches I sent you? Two were for
evict_inode fixups (cifs and xfs) and one was inode_setattr removal fixup
(for v9fs).
--
Cheers,
Stephen Rothwell [email protected]
http://www.canb.auug.org.au/~sfr/
Hi Al,
On Tue, 10 Aug 2010 14:24:34 +1000 Stephen Rothwell <[email protected]> wrote:
>
> On Tue, 10 Aug 2010 04:35:36 +0100 Al Viro <[email protected]> wrote:
> >
> > Assorted work from the last cycle; among other things, there's
> > a filesystem API change (->evict_inode() series), a lot of truncate
> > rework stuff (mostly by Chris) and a bunch of fixes.
>
> Did you remember to apply the three fix patches I sent you? Two were for
> evict_inode fixups (cifs and xfs) and one was inode_setattr removal fixup
> (for v9fs).
Having just fetched your for-next branch, they appear to have been rolled
in. Thanks.
--
Cheers,
Stephen Rothwell [email protected]
http://www.canb.auug.org.au/~sfr/
On Mon, Aug 9, 2010 at 11:35 PM, Al Viro <[email protected]> wrote:
> ? ? ? ?Assorted work from the last cycle; among other things, there's
> a filesystem API change (->evict_inode() series), a lot of truncate
> rework stuff (mostly by Chris) and a bunch of fixes.
>
> ? ? ? ?This is _not_ all; in particular, there's pending freeze/thaw
> locking rework and I would really like to see inode_lock splitup
> series separated and posted. ?Along with vfsmount_lock and files_lock
> work. ?Nick, could you do that ASAP, please?
My laptop is dead so it will have to wait until the end of the week
when I get home.
I'll do that as soon as possible.
On Tue, Aug 10, 2010 at 6:35 AM, Al Viro <[email protected]> wrote:
> Assorted work from the last cycle; among other things, there's
> a filesystem API change (->evict_inode() series), a lot of truncate
> rework stuff (mostly by Chris) and a bunch of fixes.
Linux 2.6.36-rc1 seemed to break my UML hostfs setup or well, make it
behave in very odd ways.. No matter what directory I'm trying to read
from hostfs in UML, the root directory seems to be returned. I have
hostfs mounted in UML into /mnt/host and "ls /mnt/host" and "ls
/mnt/host/etc" show the same directory. Actually, any unknown
directory like "/mnt/host/does-not-exist" shows the same root
directory.. Obviously something is going quite wrong in mapping the
path name to directory entry.
Based on a quick git bisect, the main part of the issue seems to be
caused or at least triggered by a change in this set:
> hostfs: fix races in dentry_name() and inode_name()
e9193059b1b3733695d5b80e667778311695aa73 is the first bad commit
commit e9193059b1b3733695d5b80e667778311695aa73
Author: Al Viro <[email protected]>
Date: Sun Jun 6 23:16:34 2010 -0400
hostfs: fix races in dentry_name() and inode_name()
calculating size, then doing allocation, then filling the
path is a Bad Idea(tm), since the ancestors can be renamed,
leading to buffer overrun.
Signed-off-by: Al Viro <[email protected]>
In addition, reverting just that commit allows hostfs to be used with
2.6.36-rc1. Though, there seems to be some additional issues with
symlinks, so there could be something else related to this change that
is actually causing the problems. The other issues result in memory
allocation failing when reading symlinks:
# ls /mnt/host/
ls: cannot read symbolic link /mnt/host/vmlinuz: Cannot allocate memory
ls: cannot read symbolic link /mnt/host/vmlinuz.old: Cannot allocate memory
(this is not only limited to ls; "ln -s p.patch a.patch; cat a.patch"
results in same from cat)
Other than that, hostfs seems to be working, though.
Any idea what caused this regression and how to fix it properly? I can
work around this by reverting the patch pointed out by git bisect, but
surely that is not the proper fix for the issues.
- Jouni