vfs pile, part 2
Al Viro
maintainer Mostly assorted fixes + fsync API change + infrastructure and
simple conversions of "new truncate scheme" series. BTW, long-standing abuse of file refcounting by aio.c is gone; now we have a new helper (fput_atomic(file)) suitable for calling in atomic contexts. It drops the reference if it's not the last one (i.e. if nothing potentially blocking needs to be done) and tells the caller whether it had succeeded or not (in the latter case you still need to do normal fput() outside of atomic context, of course).
Please, pull fromgit://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6.git/ for-linus
Shortlog:
Al Viro (5):
Revert "anon_inode: set S_IFREG on the anon_inode" get rid of the magic around f_count in aio
Fix racy use of anon_inode_getfd() in perf_event.c fix fs/sysv s_dirt handling
remove detritus left by "mm: make read_cache_page synchronous"
Christoph Hellwig (2):
drop unused dentry argument to ->fsync
rename the generic fsync implementations
Erik van der Kouwe (1):
fs/minix: bugfix, number of indirect block ptrs per block depends on block size
Julia Lawall (1):
fs: Add missing mutex_unlock
Neil Brown (1):
VFS: fix recent breakage of FS_REVAL_DOT
Nick Piggin (1):
fs: convert simple fs to new truncate
Randy Dunlap (1):
fs/super: fix kernel-doc warning
[email protected] (5):
fs: introduce new truncate sequence
kill spurious reference to vmtruncate
tmpfs: convert to use the new truncate convention
ext2: convert to use the new truncate convention.
fat: convert to use the new truncate convention.
Diffstat:
Documentation/filesystems/Locking | 2 +- Documentation/filesystems/vfs.txt | 9 ++- arch/powerpc/platforms/cell/spufs/file.c | 3 +- arch/powerpc/platforms/cell/spufs/inode.c | 2 +- drivers/char/ps3flash.c | 3 +- drivers/mtd/ubi/cdev.c | 3 +- drivers/staging/pohmelfs/inode.c | 2 +- drivers/usb/gadget/printer.c | 2 +- drivers/video/fb_defio.c | 2 +- fs/9p/vfs_file.c | 6 +- fs/adfs/dir.c | 2 +- fs/adfs/file.c | 2 +- fs/adfs/inode.c | 3 +- fs/affs/affs.h | 2 +- fs/affs/file.c | 4 +- fs/afs/internal.h | 2 +- fs/afs/write.c | 3 +- fs/aio.c | 6 +- fs/anon_inodes.c | 2 +- fs/attr.c | 50 ++++++++-- fs/bad_inode.c | 3 +- fs/bfs/dir.c | 2 +- fs/block_dev.c | 16 +-- fs/btrfs/ctree.h | 2 +- fs/btrfs/file.c | 3 +- fs/buffer.c | 123 ++++++++++++++++++----- fs/ceph/caps.c | 4 +- fs/ceph/dir.c | 5 +- fs/ceph/super.h | 2 +- fs/cifs/cifsfs.h | 2 +- fs/cifs/file.c | 4 +- fs/coda/coda_int.h | 3 +- fs/coda/file.c | 4 +- fs/configfs/inode.c | 9 +-- fs/direct-io.c | 61 ++++++++---- fs/ecryptfs/file.c | 2 +- fs/ecryptfs/inode.c | 4 +- fs/exofs/file.c | 7 +- fs/ext2/ext2.h | 3 +- fs/ext2/file.c | 7 +- fs/ext2/inode.c | 153 ++++++++++++++++++++++------- fs/ext3/fsync.c | 4 +- fs/ext4/ext4.h | 2 +- fs/ext4/fsync.c | 8 +- fs/fat/fat.h | 6 +- fs/fat/file.c | 40 ++++++-- fs/fat/inode.c | 35 ++++++- fs/file_table.c | 21 ++-- fs/fuse/dir.c | 5 +- fs/fuse/file.c | 9 +- fs/fuse/fuse_i.h | 3 +- fs/gfs2/aops.c | 8 ++- fs/gfs2/file.c | 4 +- fs/gfs2/ops_inode.c | 5 +- fs/hostfs/hostfs_kern.c | 4 +- fs/hpfs/file.c | 4 +- fs/hpfs/hpfs_fn.h | 2 +- fs/hppfs/hppfs.c | 2 +- fs/hugetlbfs/inode.c | 2 +- fs/jffs2/file.c | 4 +- fs/jffs2/fs.c | 4 +- fs/jffs2/os-linux.h | 2 +- fs/jfs/file.c | 4 +- fs/jfs/jfs_inode.h | 2 +- fs/libfs.c | 108 ++++++++++++++++++-- fs/logfs/file.c | 4 +- fs/logfs/logfs.h | 2 +- fs/minix/dir.c | 7 +- fs/minix/file.c | 2 +- fs/minix/itree_v2.c | 27 +++--- fs/namei.c | 2 +- fs/ncpfs/file.c | 2 +- fs/nfs/dir.c | 6 +- fs/nfs/file.c | 5 +- fs/nilfs2/file.c | 4 +- fs/nilfs2/nilfs.h | 2 +- fs/ntfs/dir.c | 5 +- fs/ntfs/file.c | 9 +-- fs/ocfs2/file.c | 15 ++- fs/omfs/file.c | 2 +- fs/pipe.c | 13 ++- fs/qnx4/dir.c | 2 +- fs/ramfs/file-mmu.c | 3 +- fs/ramfs/file-nommu.c | 9 +- fs/reiserfs/dir.c | 8 +- fs/reiserfs/file.c | 5 +- fs/smbfs/file.c | 3 +- fs/smbfs/inode.c | 2 +- fs/super.c | 4 +- fs/sync.c | 8 +- fs/sysfs/inode.c | 8 +- fs/sysv/dir.c | 2 +- fs/sysv/file.c | 2 +- fs/sysv/inode.c | 1 + fs/ubifs/file.c | 17 ++-- fs/ubifs/ubifs.h | 4 +- fs/udf/dir.c | 2 +- fs/udf/file.c | 2 +- fs/ufs/dir.c | 2 +- fs/ufs/file.c | 2 +- fs/ufs/truncate.c | 10 +- fs/xfs/linux-2.6/xfs_file.c | 10 +- include/linux/buffer_head.h | 11 ++- include/linux/ext3_fs.h | 2 +- include/linux/fb.h | 5 +- include/linux/file.h | 1 - include/linux/fs.h | 36 ++++++- include/trace/events/ext4.h | 6 +- ipc/shm.c | 11 +-- kernel/perf_event.c | 40 ++++---- mm/shmem.c | 45 +++++---- mm/truncate.c | 10 +- 112 files changed, 782 insertions(+), 415 deletions(-)
Hi Al,
There are two patches
1f24668c vfs: make real_lookup do dentry revalidation with i_mutex held 982ddae4 vfs: clean up do_lookup()
that were in your for-next a few weeks back but aren't anymore (they're still in your autofs branch). I just want to make sure those don't get lost if they're not going to make it upstream this time around? I think they've been in vfs-2.6.git limbo for two release cycles now.
Thanks!
sage
Quoted 166 lines · Al Viro
Mostly assorted fixes + fsync API change + infrastructure and
simple conversions of "new truncate scheme" series. BTW, long-standing abuse of file refcounting byaio.cis gone; now we have a new helper (fput_atomic(file)) suitable for calling in atomic contexts. It drops the reference if it's not the last one (i.e. if nothing potentially blocking needs to be done) and tells the caller whether it had succeeded or not (in the latter case you still need to do normalfput()outside of atomic context, of course).Please, pull from
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6.git/for-linusShortlog:
Al Viro (5):
Revert "anon_inode: setS_IFREGon theanon_inode" get rid of the magic aroundf_countin aio
Fix racy use ofanon_inode_getfd()inperf_event.cfix fs/sysvs_dirthandling
remove detritus left by "mm: makeread_cache_pagesynchronous"Christoph Hellwig (2):
drop unused dentry argument to ->fsync
rename the generic fsync implementationsErik van der Kouwe (1):
fs/minix: bugfix, number of indirect block ptrs per block depends on block sizeJulia Lawall (1):
fs: Add missingmutex_unlockNeil Brown (1):
VFS: fix recent breakage ofFS_REVAL_DOTNick Piggin (1):
fs: convert simple fs to new truncateRandy Dunlap (1):
fs/super: fix kernel-doc warning[email protected] (5):
fs: introduce new truncate sequence
kill spurious reference to vmtruncate
tmpfs: convert to use the new truncate convention
ext2: convert to use the new truncate convention.
fat: convert to use the new truncate convention.Diffstat:
Documentation/filesystems/Locking | 2 +- Documentation/filesystems/vfs.txt | 9 ++- arch/powerpc/platforms/cell/spufs/file.c | 3 +- arch/powerpc/platforms/cell/spufs/inode.c | 2 +- drivers/char/ps3flash.c | 3 +- drivers/mtd/ubi/cdev.c | 3 +- drivers/staging/pohmelfs/inode.c | 2 +- drivers/usb/gadget/printer.c | 2 +- drivers/video/fb_defio.c | 2 +- fs/9p/vfs_file.c | 6 +- fs/adfs/dir.c | 2 +- fs/adfs/file.c | 2 +- fs/adfs/inode.c | 3 +- fs/affs/affs.h | 2 +- fs/affs/file.c | 4 +- fs/afs/internal.h | 2 +- fs/afs/write.c | 3 +- fs/aio.c | 6 +- fs/anon_inodes.c | 2 +- fs/attr.c | 50 ++++++++-- fs/bad_inode.c | 3 +- fs/bfs/dir.c | 2 +- fs/block_dev.c | 16 +-- fs/btrfs/ctree.h | 2 +- fs/btrfs/file.c | 3 +- fs/buffer.c | 123 ++++++++++++++++++----- fs/ceph/caps.c | 4 +- fs/ceph/dir.c | 5 +- fs/ceph/super.h | 2 +- fs/cifs/cifsfs.h | 2 +- fs/cifs/file.c | 4 +- fs/coda/coda_int.h | 3 +- fs/coda/file.c | 4 +- fs/configfs/inode.c | 9 +-- fs/direct-io.c | 61 ++++++++---- fs/ecryptfs/file.c | 2 +- fs/ecryptfs/inode.c | 4 +- fs/exofs/file.c | 7 +- fs/ext2/ext2.h | 3 +- fs/ext2/file.c | 7 +- fs/ext2/inode.c | 153 ++++++++++++++++++++++------- fs/ext3/fsync.c | 4 +- fs/ext4/ext4.h | 2 +- fs/ext4/fsync.c | 8 +- fs/fat/fat.h | 6 +- fs/fat/file.c | 40 ++++++-- fs/fat/inode.c | 35 ++++++- fs/file_table.c | 21 ++-- fs/fuse/dir.c | 5 +- fs/fuse/file.c | 9 +- fs/fuse/fuse_i.h | 3 +- fs/gfs2/aops.c | 8 ++- fs/gfs2/file.c | 4 +- fs/gfs2/ops_inode.c | 5 +- fs/hostfs/hostfs_kern.c | 4 +- fs/hpfs/file.c | 4 +- fs/hpfs/hpfs_fn.h | 2 +- fs/hppfs/hppfs.c | 2 +- fs/hugetlbfs/inode.c | 2 +- fs/jffs2/file.c | 4 +- fs/jffs2/fs.c | 4 +- fs/jffs2/os-linux.h | 2 +- fs/jfs/file.c | 4 +- fs/jfs/jfs_inode.h | 2 +- fs/libfs.c | 108 ++++++++++++++++++-- fs/logfs/file.c | 4 +- fs/logfs/logfs.h | 2 +- fs/minix/dir.c | 7 +- fs/minix/file.c | 2 +- fs/minix/itree_v2.c | 27 +++--- fs/namei.c | 2 +- fs/ncpfs/file.c | 2 +- fs/nfs/dir.c | 6 +- fs/nfs/file.c | 5 +- fs/nilfs2/file.c | 4 +- fs/nilfs2/nilfs.h | 2 +- fs/ntfs/dir.c | 5 +- fs/ntfs/file.c | 9 +-- fs/ocfs2/file.c | 15 ++- fs/omfs/file.c | 2 +- fs/pipe.c | 13 ++- fs/qnx4/dir.c | 2 +- fs/ramfs/file-mmu.c | 3 +- fs/ramfs/file-nommu.c | 9 +- fs/reiserfs/dir.c | 8 +- fs/reiserfs/file.c | 5 +- fs/smbfs/file.c | 3 +- fs/smbfs/inode.c | 2 +- fs/super.c | 4 +- fs/sync.c | 8 +- fs/sysfs/inode.c | 8 +- fs/sysv/dir.c | 2 +- fs/sysv/file.c | 2 +- fs/sysv/inode.c | 1 + fs/ubifs/file.c | 17 ++-- fs/ubifs/ubifs.h | 4 +- fs/udf/dir.c | 2 +- fs/udf/file.c | 2 +- fs/ufs/dir.c | 2 +- fs/ufs/file.c | 2 +- fs/ufs/truncate.c | 10 +- fs/xfs/linux-2.6/xfs_file.c | 10 +- include/linux/buffer_head.h | 11 ++- include/linux/ext3_fs.h | 2 +- include/linux/fb.h | 5 +- include/linux/file.h | 1 - include/linux/fs.h | 36 ++++++- include/trace/events/ext4.h | 6 +- ipc/shm.c | 11 +-- kernel/perf_event.c | 40 ++++---- mm/shmem.c | 45 +++++---- mm/truncate.c | 10 +- 112 files changed, 782 insertions(+), 415 deletions(-)--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Al Viro
maintainerQuoted 11 lines · Sage Weil
Hi Al,
There are two patches
1f24668cvfs: makereal_lookupdo dentry revalidation withi_mutexheld982ddae4vfs: clean updo_lookup()that were in your for-next a few weeks back but aren't anymore (they're still in your autofs branch). I just want to make sure those don't get lost if they're not going to make it upstream this time around? I think they've been in vfs-2.6.git limbo for two release cycles now.
My apologies, I thought you'd been Cc'd on the discussions with Ian concerning that stuff, but...
The bottom line: "autofs4: always use lookup for lookup" is needed for them and it had to be reverted since it'd been FUBAR.
Details: we really, *really* can not do things in ->d_revalidate() and count on ->lookup() to come and finish them. Think for a minute and you'll see why - between these two calls we allocate a dentry to pass to ->lookup(). What happens if that allocation fails?
And without that patch your stuff breaks autofs4. So it's waiting for the ->d_revalidate() interface redesign. Hopefully in the coming cycle...
Quoted 6 lines · Al Viro
My apologies, I thought you'd been Cc'd on the discussions with Ian concerning that stuff, but...
The bottom line: "autofs4: always use lookup for lookup" is needed for them and it had to be reverted since it'd been FUBAR.
Quoted 7 lines
Details: we really, *really* can not do things in ->d_revalidate() and count on ->lookup() to come and finish them. Think for a minute and you'll see why - between these two calls we allocate a dentry to pass to ->lookup(). What happens if that allocation fails?
And without that patch your stuff breaks autofs4. So it's waiting for the ->d_revalidate() interface redesign. Hopefully in the coming cycle...
In addition to that autofs also needs either major surgery or to be dropped. If we want to do the latter we'd be better of to strongly deprecate it now.
Quoted 2 lines · Al Viro
Al Viro (5):
remove detritus left by "mm: makeread_cache_pagesynchronous"
This introduced:
| fs/minix/dir.c: In function ‘dir_get_page’:
| fs/minix/dir.c:79: warning: label ‘fail’ defined but not used
As the whole fail block is now dead code.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds