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>
In the current nilfs, page cache for btree nodes and meta data files
do not set a valid back pointer to the host inode in mapping->host.
This will change it so that every address space in nilfs uses
mapping->host to hold its host inode.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
After resizing the filesystem, the secondary super block must be moved
to a new location. This adds a helper function for this.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This directly uses sb->s_fs_info to keep a nilfs filesystem object and
fully removes the intermediate nilfs_sb_info structure. With this
change, the hierarchy of on-memory structures of nilfs will be
simplified as follows:
Before:
super_block
-> nilfs_sb_info
-> the_nilfs
-> cptree --+-> nilfs_root (current file system)
+-> nilfs_root (snapshot A)
+-> nilfs_root (snapshot B)
:
-> nilfs_sc_info (log writer structure)
After:
super_block
-> the_nilfs
-> cptree --+-> nilfs_root (current file system)
+-> nilfs_root (snapshot A)
+-> nilfs_root (snapshot B)
:
-> nilfs_sc_info (log writer structure)
The reason why we didn't design so from the beginning is because the
initial shape also differed from the above. The early hierachy was
composed of "per-mount-point" super_block -> nilfs_sb_info pairs and a
shared nilfs object. On the kernel 2.6.37, it was changed to the
current shape in order to unify super block instances into one per
device, and this cleanup became applicable as the result.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Moves s_next_generation counter and a spinlock protecting it to nilfs
object from nilfs_sb_info structure.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Moves s_inode_lock spinlock and s_dirty_files list to nilfs object
from nilfs_sb_info structure.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This moves four parameter variables on nilfs_sb_info s_resuid,
s_resgid, s_interval and s_watermark to the nilfs object.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Michael Leun reported that running parallel opens on a fuse filesystem
can trigger a "kernel BUG at mm/truncate.c:475"
Gurudas Pai reported the same bug on NFS.
The reason is, unmap_mapping_range() is not prepared for more than
one concurrent invocation per inode. For example:
thread1: going through a big range, stops in the middle of a vma and
stores the restart address in vm_truncate_count.
thread2: comes in with a small (e.g. single page) unmap request on
the same vma, somewhere before restart_address, finds that the
vma was already unmapped up to the restart address and happily
returns without doing anything.
Another scenario would be two big unmap requests, both having to
restart the unmapping and each one setting vm_truncate_count to its
own value. This could go on forever without any of them being able to
finish.
Truncate and hole punching already serialize with i_mutex. Other
callers of unmap_mapping_range() do not, and it's difficult to get
i_mutex protection for all callers. In particular ->d_revalidate(),
which calls invalidate_inode_pages2_range() in fuse, may be called
with or without i_mutex.
This patch adds a new mutex to 'struct address_space' to prevent
running multiple concurrent unmap_mapping_range() on the same mapping.
[ We'll hopefully get rid of all this with the upcoming mm
preemptibility series by Peter Zijlstra, the "mm: Remove i_mmap_mutex
lockbreak" patch in particular. But that is for 2.6.39 ]
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Reported-by: Michael Leun <lkml20101129@newton.leun.net>
Reported-by: Gurudas Pai <gurudas.pai@oracle.com>
Tested-by: Gurudas Pai <gurudas.pai@oracle.com>
Acked-by: Hugh Dickins <hughd@google.com>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for-2.6.38/core' of git://git.kernel.dk/linux-2.6-block: (43 commits)
block: ensure that completion error gets properly traced
blktrace: add missing probe argument to block_bio_complete
block cfq: don't use atomic_t for cfq_group
block cfq: don't use atomic_t for cfq_queue
block: trace event block fix unassigned field
block: add internal hd part table references
block: fix accounting bug on cross partition merges
kref: add kref_test_and_get
bio-integrity: mark kintegrityd_wq highpri and CPU intensive
block: make kblockd_workqueue smarter
Revert "sd: implement sd_check_events()"
block: Clean up exit_io_context() source code.
Fix compile warnings due to missing removal of a 'ret' variable
fs/block: type signature of major_to_index(int) to major_to_index(unsigned)
block: convert !IS_ERR(p) && p to !IS_ERR_NOR_NULL(p)
cfq-iosched: don't check cfqg in choose_service_tree()
fs/splice: Pull buf->ops->confirm() from splice_from_pipe actors
cdrom: export cdrom_check_events()
sd: implement sd_check_events()
sr: implement sr_check_events()
...
Remove kobject.h from files which don't need it, notably,
sched.h and fs.h.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Only mount_opt member is used in the nilfs_mount_options structure,
and we can simplify it.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Using %pV reduces the number of printk calls and
eliminates any possible message interleaving from
other printk calls.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
RCU free the struct inode. This will allow:
- Subsequent store-free path walking patch. The inode must be consulted for
permissions when walking, so an RCU inode reference is a must.
- sb_inode_list_lock to be moved inside i_lock because sb list walkers who want
to take i_lock no longer need to take sb_inode_list_lock to walk the list in
the first place. This will simplify and optimize locking.
- Could remove some nested trylock loops in dcache code
- Could potentially simplify things a bit in VM land. Do not need to take the
page lock to follow page->mapping.
The downsides of this is the performance cost of using RCU. In a simple
creat/unlink microbenchmark, performance drops by about 10% due to inability to
reuse cache-hot slab objects. As iterations increase and RCU freeing starts
kicking over, this increases to about 20%.
In cases where inode lifetimes are longer (ie. many inodes may be allocated
during the average life span of a single inode), a lot of this cache reuse is
not applicable, so the regression caused by this patch is smaller.
The cache-hot regression could largely be avoided by using SLAB_DESTROY_BY_RCU,
however this adds some complexity to list walking and store-free path walking,
so I prefer to implement this at a later date, if it is shown to be a win in
real situations. I haven't found a regression in any non-micro benchmark so I
doubt it will be a problem.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
Make d_count non-atomic and protect it with d_lock. This allows us to ensure a
0 refcount dentry remains 0 without dcache_lock. It is also fairly natural when
we start protecting many other dentry members with d_lock.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
After recent blkdev_get() modifications, open_by_devnum() and
open_bdev_exclusive() are simple wrappers around blkdev_get().
Replace them with blkdev_get_by_dev() and blkdev_get_by_path().
blkdev_get_by_dev() is identical to open_by_devnum().
blkdev_get_by_path() is slightly different in that it doesn't
automatically add %FMODE_EXCL to @mode.
All users are converted. Most conversions are mechanical and don't
introduce any behavior difference. There are several exceptions.
* btrfs now sets FMODE_EXCL in btrfs_device->mode, so there's no
reason to OR it explicitly on blkdev_put().
* gfs2, nilfs2 and the generic mount_bdev() now set FMODE_EXCL in
sb->s_mode.
* With the above changes, sb->s_mode now always should contain
FMODE_EXCL. WARN_ON_ONCE() added to kill_block_super() to detect
errors.
The new blkdev_get_*() functions are with proper docbook comments.
While at it, add function description to blkdev_get() too.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Philipp Reisner <philipp.reisner@linbit.com>
Cc: Neil Brown <neilb@suse.de>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Joern Engel <joern@lazybastard.org>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Jan Kara <jack@suse.cz>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp>
Cc: reiserfs-devel@vger.kernel.org
Cc: xfs-masters@oss.sgi.com
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Over time, block layer has accumulated a set of APIs dealing with bdev
open, close, claim and release.
* blkdev_get/put() are the primary open and close functions.
* bd_claim/release() deal with exclusive open.
* open/close_bdev_exclusive() are combination of open and claim and
the other way around, respectively.
* bd_link/unlink_disk_holder() to create and remove holder/slave
symlinks.
* open_by_devnum() wraps bdget() + blkdev_get().
The interface is a bit confusing and the decoupling of open and claim
makes it impossible to properly guarantee exclusive access as
in-kernel open + claim sequence can disturb the existing exclusive
open even before the block layer knows the current open if for another
exclusive access. Reorganize the interface such that,
* blkdev_get() is extended to include exclusive access management.
@holder argument is added and, if is @FMODE_EXCL specified, it will
gain exclusive access atomically w.r.t. other exclusive accesses.
* blkdev_put() is similarly extended. It now takes @mode argument and
if @FMODE_EXCL is set, it releases an exclusive access. Also, when
the last exclusive claim is released, the holder/slave symlinks are
removed automatically.
* bd_claim/release() and close_bdev_exclusive() are no longer
necessary and either made static or removed.
* bd_link_disk_holder() remains the same but bd_unlink_disk_holder()
is no longer necessary and removed.
* open_bdev_exclusive() becomes a simple wrapper around lookup_bdev()
and blkdev_get(). It also has an unexpected extra bdev_read_only()
test which probably should be moved into blkdev_get().
* open_by_devnum() is modified to take @holder argument and pass it to
blkdev_get().
Most of bdev open/close operations are unified into blkdev_get/put()
and most exclusive accesses are tested atomically at the open time (as
it should). This cleans up code and removes some, both valid and
invalid, but unnecessary all the same, corner cases.
open_bdev_exclusive() and open_by_devnum() can use further cleanup -
rename to blkdev_get_by_path() and blkdev_get_by_devt() and drop
special features. Well, let's leave them for another day.
Most conversions are straight-forward. drbd conversion is a bit more
involved as there was some reordering, but the logic should stay the
same.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Neil Brown <neilb@suse.de>
Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Acked-by: Mike Snitzer <snitzer@redhat.com>
Acked-by: Philipp Reisner <philipp.reisner@linbit.com>
Cc: Peter Osterlund <petero2@telia.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <joel.becker@oracle.com>
Cc: Alex Elder <aelder@sgi.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: dm-devel@redhat.com
Cc: drbd-dev@lists.linbit.com
Cc: Leo Chen <leochen@broadcom.com>
Cc: Scott Branden <sbranden@broadcom.com>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
Cc: Joern Engel <joern@logfs.org>
Cc: reiserfs-devel@vger.kernel.org
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2: (36 commits)
nilfs2: eliminate sparse warning - "context imbalance"
nilfs2: eliminate sparse warnings - "symbol not declared"
nilfs2: get rid of bdi from nilfs object
nilfs2: change license of exported header file
nilfs2: add bdev freeze/thaw support
nilfs2: accept 64-bit checkpoint numbers in cp mount option
nilfs2: remove own inode allocator and destructor for metadata files
nilfs2: get rid of back pointer to writable sb instance
nilfs2: get rid of mi_nilfs back pointer to nilfs object
nilfs2: see state of root dentry for mount check of snapshots
nilfs2: use iget for all metadata files
nilfs2: get rid of GCDAT inode
nilfs2: add routines to redirect access to buffers of DAT file
nilfs2: add routines to roll back state of DAT file
nilfs2: add routines to save and restore bmap state
nilfs2: do not allocate nilfs_mdt_info structure to gc-inodes
nilfs2: allow nilfs_clear_inode to clear metadata file inodes
nilfs2: get rid of snapshot mount flag
nilfs2: simplify life cycle management of nilfs object
nilfs2: do not allocate multiple super block instances for a device
...
change nilfs_dat_commit_free and nilfs_inode_cachep static
to fix following warnings
fs/nilfs2/super.c:72:19: warning: symbol 'nilfs_inode_cachep' was not declared. Should it be static?
fs/nilfs2/dat.c:106:6: warning: symbol 'nilfs_dat_commit_free' was not declared. Should it be static?
Signed-off-by: Jiro SEKIBA <jir@unicus.jp>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Nilfs now can use sb->s_bdi to get backing_dev_info, so we use it
instead of ns_bdi on the nilfs object and remove ns_bdi.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Nilfs hasn't supported the freeze/thaw feature because it didn't work
due to the peculiar design that multiple super block instances could
be allocated for a device. This limitation was removed by the patch
"nilfs2: do not allocate multiple super block instances for a device".
So now this adds the freeze/thaw support to nilfs.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
The current implementation doesn't mount snapshots with checkpoint
numbers larger than INT_MAX since it uses match_int() for parsing
"cp=" mount option.
This uses simple_strtoull() for the conversion to resolve the issue.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This finally removes own inode allocator and destructor functions for
metadata files. Several routines, nilfs_mdt_new(),
nilfs_mdt_new_common(), nilfs_mdt_clear(), nilfs_mdt_destroy(), and
nilfs_alloc_inode_common() will be gone.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
After applied the patch that unified sb instances, root dentry of
snapshots can be left in dcache even after their trees are unmounted.
The orphan root dentry/inode keeps a root object, and this causes
false positive of nilfs_checkpoint_is_mounted function.
This resolves the issue by having nilfs_checkpoint_is_mounted test
whether the root dentry is busy or not.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This makes use of iget5_locked to allocate or get inode for metadata
files to stop using own inode allocator.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This stops pre-allocating nilfs object in nilfs_get_sb routine, and
stops managing its life cycle by reference counting.
nilfs_find_or_create_nilfs() function, nilfs->ns_mount_mutex,
nilfs_objects list, and the reference counter will be removed through
the simplification.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This stops allocating multiple super block instances for a device.
All snapshots and a current mode mount (i.e. latest tree) will be
controlled with nilfs_root objects that are kept within an sb
instance.
nilfs_get_sb() is rewritten so that it always has a root object for
the latest tree and snapshots make additional root objects.
The root dentry of the latest tree is binded to sb->s_root even if it
isn't attached on a directory. Root dentries of snapshots or the
latest tree are binded to mnt->mnt_root on which they are mounted.
With this patch, nilfs_find_sbinfo() function, nilfs->ns_supers list,
and nilfs->ns_current back pointer, are deleted. In addition,
init_nilfs() and load_nilfs() are simplified since they will be called
once for a device, not repeatedly called for mount points.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This splits the code to allocate root dentry into a separate routine
for convenience in successive changes.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This rewrites functions using ifile so that they get ifile from
nilfs_root object, and will remove sbi->s_ifile. Some functions that
don't know the root object are extended to receive it from caller.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
The previous export operations cannot handle multiple versions of
a filesystem if they belong to the same sb instance.
This adds a new type of file handle and extends export operations so
that they can get the inode specified by a checkpoint number as well
as an inode number and a generation number.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This puts a pointer to nilfs_root object in the private part of
on-memory inode, and makes nilfs_iget function pick up the inode with
the same root object.
Non-root inodes inherit its nilfs_root object from parent inode. That
of the root inode is allocated through nilfs_attach_checkpoint()
function.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
This uses iget5_locked instead of iget_locked so that gc cache can
look up inodes with an inode number and an optional checkpoint number.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
The current nilfs_destroy_inode() doesn't handle metadata file inodes
including gc inodes (dummy inodes used for garbage collection).
This allows nilfs_destroy_inode() to destroy inodes of metadata files.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
* 'for-2.6.37/barrier' of git://git.kernel.dk/linux-2.6-block: (46 commits)
xen-blkfront: disable barrier/flush write support
Added blk-lib.c and blk-barrier.c was renamed to blk-flush.c
block: remove BLKDEV_IFL_WAIT
aic7xxx_old: removed unused 'req' variable
block: remove the BH_Eopnotsupp flag
block: remove the BLKDEV_IFL_BARRIER flag
block: remove the WRITE_BARRIER flag
swap: do not send discards as barriers
fat: do not send discards as barriers
ext4: do not send discards as barriers
jbd2: replace barriers with explicit flush / FUA usage
jbd2: Modify ASYNC_COMMIT code to not rely on queue draining on barrier
jbd: replace barriers with explicit flush / FUA usage
nilfs2: replace barriers with explicit flush / FUA usage
reiserfs: replace barriers with explicit flush / FUA usage
gfs2: replace barriers with explicit flush / FUA usage
btrfs: replace barriers with explicit flush / FUA usage
xfs: replace barriers with explicit flush / FUA usage
block: pass gfp_mask and flags to sb_issue_discard
dm: convey that all flushes are processed as empty
...
The BKL is only used in put_super, fill_super and remount_fs that are all
three protected by the superblocks s_umount rw_semaphore. Therefore it is
safe to remove the BKL entirely.
Signed-off-by: Jan Blunck <jblunck@infradead.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
This patch is a preparation necessary to remove the BKL from do_new_mount().
It explicitly adds calls to lock_kernel()/unlock_kernel() around
get_sb/fill_super operations for filesystems that still uses the BKL.
I've read through all the code formerly covered by the BKL inside
do_kern_mount() and have satisfied myself that it doesn't need the BKL
any more.
do_kern_mount() is already called without the BKL when mounting the rootfs
and in nfsctl. do_kern_mount() calls vfs_kern_mount(), which is called
from various places without BKL: simple_pin_fs(), nfs_do_clone_mount()
through nfs_follow_mountpoint(), afs_mntpt_do_automount() through
afs_mntpt_follow_link(). Both later functions are actually the filesystems
follow_link inode operation. vfs_kern_mount() is calling the specified
get_sb function and lets the filesystem do its job by calling the given
fill_super function.
Therefore I think it is safe to push down the BKL from the VFS to the
low-level filesystems get_sb/fill_super operation.
[arnd: do not add the BKL to those file systems that already
don't use it elsewhere]
Signed-off-by: Jan Blunck <jblunck@infradead.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: Christoph Hellwig <hch@infradead.org>
Switch to the WRITE_FLUSH_FUA flag for log writes, remove the EOPNOTSUPP
detection for barriers and stop setting the barrier flag for discards.
tj: nilfs is now fixed to wait for discard completion. Updated this
patch accordingly and dropped warning about it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
fs: brlock vfsmount_lock
fs: scale files_lock
lglock: introduce special lglock and brlock spin locks
tty: fix fu_list abuse
fs: cleanup files_lock locking
fs: remove extra lookup in __lookup_hash
fs: fs_struct rwlock to spinlock
apparmor: use task path helpers
fs: dentry allocation consolidation
fs: fix do_lookup false negative
mbcache: Limit the maximum number of cache entries
hostfs ->follow_link() braino
hostfs: dumb (and usually harmless) tpyo - strncpy instead of strlcpy
remove SWRITE* I/O types
kill BH_Ordered flag
vfs: update ctime when changing the file's permission by setfacl
cramfs: only unlock new inodes
fix reiserfs_evict_inode end_writeback second call
Instead of abusing a buffer_head flag just add a variant of
sync_dirty_buffer which allows passing the exact type of write
flag required.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (96 commits)
no need for list_for_each_entry_safe()/resetting with superblock list
Fix sget() race with failing mount
vfs: don't hold s_umount over close_bdev_exclusive() call
sysv: do not mark superblock dirty on remount
sysv: do not mark superblock dirty on mount
btrfs: remove junk sb_dirt change
BFS: clean up the superblock usage
AFFS: wait for sb synchronization when needed
AFFS: clean up dirty flag usage
cifs: truncate fallout
mbcache: fix shrinker function return value
mbcache: Remove unused features
add f_flags to struct statfs(64)
pass a struct path to vfs_statfs
update VFS documentation for method changes.
All filesystems that need invalidate_inode_buffers() are doing that explicitly
convert remaining ->clear_inode() to ->evict_inode()
Make ->drop_inode() just return whether inode needs to be dropped
fs/inode.c:clear_inode() is gone
fs/inode.c:evict() doesn't care about delete vs. non-delete paths now
...
Fix up trivial conflicts in fs/nilfs2/super.c
This forces nilfs to check compatibility of feature flags so as to
reject a filesystem with unknown features when it mounts or remounts
the filesystem.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Al Viro