This makes the performance impact of possible mmap_miss wrap around to be
temporary and tolerable: i.e.  MMAP_LOTSAMISS=100 extra readarounds.

Otherwise if ever mmap_miss wraps around to negative, it takes INT_MAX
cache misses to bring it back to normal state.  During the time mmap
readaround will be _enabled_ for whatever wild random workload.  That's
almost permanent performance impact.
Signed-off-by: NWu Fengguang <fengguang.wu@intel.com>
Cc: Ying Han <yinghan@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

writes associated buffers, then does sync_inode() to write
the inode itself (and to make it clean).  Depends on
->write_inode() honouring the second argument.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>
Cc: linux-fsdevel@vger.kernel.org
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The only user of the i_cindex element in the inode structure is used
is by the firewire drivers.  As part of an attempt to slim down the
inode structure to save memory --- since a typical Linux system will
have hundreds of thousands if not millions of inodes cached, a
reduction in the size inode has high leverage.

The firewire driver does not need i_cindex in any fast path, so it's
simple enough to calculate when it is needed, instead of wasting space
in the inode structure.
Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>
Cc: krh@redhat.com
Cc: stefanr@s5r6.in-berlin.de
Cc: linux-fsdevel@vger.kernel.org
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

do_remount_sb() is fs/internal.h fodder, fsync_no_super() is long gone.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Rename the function so that it better describe what it really does. Also
remove the unnecessary include of buffer_head.h.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Move sync_filesystems(), __fsync_super(), fsync_super() from
super.c to sync.c where it fits better.

[build fixes folded]
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

It is unnecessarily fragile to have two places (fsync_super() and do_sync())
doing data integrity sync of the filesystem. Alter __fsync_super() to
accommodate needs of both callers and use it. So after this patch
__fsync_super() is the only place where we gather all the calls needed to
properly send all data on a filesystem to disk.

Nice bonus is that we get a complete livelock avoidance and write_supers()
is now only used for periodic writeback of superblocks.

sync_blockdevs() introduced a couple of patches ago is gone now.

[build fixes folded]
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

__fsync_super() does the same thing as fsync_super(). So change the only
caller to use fsync_super() and make __fsync_super() static. This removes
unnecessarily duplicated call to sync_blockdev() and prepares ground
for the changes to __fsync_super() in the following patches.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove the unused s_async_list in the superblock, a leftover of the
broken async inode deletion code that leaked into mainline.  Having this
in the middle of the sync/unmount path is not helpful for the following
cleanups.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Reimplement dnotify using fsnotify.
Signed-off-by: NEric Paris <eparis@redhat.com>
Acked-by: NAl Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>

This patch creates a way for fsnotify groups to attach marks to inodes.
These marks have little meaning to the generic fsnotify infrastructure
and thus their meaning should be interpreted by the group that attached
them to the inode's list.

dnotify and inotify  will make use of these markings to indicate which
inodes are of interest to their respective groups.  But this implementation
has the useful property that in the future other listeners could actually
use the marks for the exact opposite reason, aka to indicate which inodes
it had NO interest in.
Signed-off-by: NEric Paris <eparis@redhat.com>
Acked-by: NAl Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>

If f_op->splice_read() is not implemented, fall back to a plain read.
Use vfs_readv() to read into previously allocated pages.

This will allow splice and functions using splice, such as the loop
device, to work on all filesystems.  This includes "direct_io" files
in fuse which bypass the page cache.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

Put generic_show_options read access to s_options under rcu_read_lock,
split save_mount_options() into "we are setting it the first time"
(uses in foo_fill_super()) and "we are relacing and freeing the old one",
synchronize_rcu() before kfree() in the latter.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

... and make path_lookup_open() static
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Does equivalent of up_write(&s->s_umount); deactivate_super(s);
However, it does not does not unlock it until it's all over.
As the result, it's safe to use to dispose of new superblock on ->get_sb()
failure exits - nobody will see the sucker until it's all over.
Equivalent using up_write/deactivate_super is safe for that purpose
if superblock is either	safe to use or has NULL ->s_root when we unlock.
Normally filesystems take the required precautions, but
	a) we do have bugs in that area in some of them.
	b) up_write/deactivate_super sequence is extremely common,
so the helper makes sense anyway.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

"int get_filesystem_list(char * buf)" is called by only
"static void __init get_fs_names(char *page)".
We can mark get_filesystem_list() as "__init".
Signed-off-by: NTetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

There's really no reason to keep vfs_stat_fd and vfs_lstat_fd with
Oleg's vfs_fstatat.  Use vfs_fstatat for the few cases having the
directory fd, and switch all others to vfs_stat / vfs_lstat.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This is a version incorporating Christoph's suggestion.

Separate out common *fstatat functionality into a single function
instead of duplicating it all over the code.
Signed-off-by: NOleg Drokin <green@linuxhacker.ru>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

There are lots of sequences like this, especially in splice code:

	if (pipe->inode)
		mutex_lock(&pipe->inode->i_mutex);
	/* do something */
	if (pipe->inode)
		mutex_unlock(&pipe->inode->i_mutex);

so introduce helpers which do the conditional locking and unlocking.
Also replace the inode_double_lock() call with a pipe_double_lock()
helper to avoid spreading the use of this functionality beyond the
pipe code.

This patch is just a cleanup, and should cause no behavioral changes.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

Remove the now unused generic_file_splice_write_nolock() function.
It's conceptually broken anyway, because splice may need to wait for
pipe events so holding locks across the whole operation is wrong.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

It's a somewhat twisty maze of hints and behavioural modifiers, try
and clear it up a bit with some documentation.
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

The mqueuefs filesystem will use this helper as well.  Proc's main get_sb
could also be made to use it, but that will require a bit more rework.
Signed-off-by: NSerge E. Hallyn <serue@us.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

By default, CFQ will anticipate more IO from a given io context if the
previously completed IO was sync. This used to be fine, since the only
sync IO was reads and O_DIRECT writes. But with more "normal" sync writes
being used now, we don't want to anticipate for those.

Add a bio/request flag that informs the IO scheduler that this is a sync
request that we should not idle for. Introduce WRITE_ODIRECT specifically
for O_DIRECT writes, and make sure that the other sync writes set this
flag.
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

(S)WRITE_SYNC always unplugs the device right after IO submission.
Sometimes we want to build up a queue before doing so, so add
variants that explicitly DON'T unplug the queue. The caller must
then do that after submitting all the IO.
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Impact: cleanup

We want to remove percpu.h from rcupdate.h (for upcoming kmemtrace
changes), but this is not possible currently without breaking the
build because fs.h has an implicit include file depedency: it
uses PAGE_SIZE but does not include asm/page.h which defines it.

This problem gets masked in practice because most fs.h using sites
use rcupreempt.h (and other headers) which includes percpu.h which
brings in asm/page.h indirectly.

We cannot add asm/page.h to asm/fs.h because page.h is not an
exported header.

Move simple_transaction_set() to the other simple-transaction
file helpers in fs/libfs.c.

This removes the include file hell and also reduces
kernel size a bit.
Acked-by: NAl Viro <viro@zeniv.linux.org.uk>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
Cc: paulmck@linux.vnet.ibm.com
LKML-Reference: <1237898630.25315.83.camel@penberg-laptop>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Impact: cleanup

We want to remove percpu.h from rcupdate.h (for upcoming kmemtrace
changes), but this is not possible currently without breaking the
build because fs.h has implicit include file depedencies: it uses
GFP_* types in inlines but does not include gfp.h.

In practice most fs.h using .c files get gfp.h included implicitly,
via an indirect route: via rcupdate.h inclusion - so this underlying
problem gets masked in practice.

So we want to solve fs.h's dependency on gfp.h.

gfp.h can not be included here directly because it is not exported and it
would break the build the following way:

  /home/mingo/tip/usr/include/linux/bsg.h:11: found __[us]{8,16,32,64} type without #include <linux/types.h>
  /home/mingo/tip/usr/include/linux/fs.h:11: included file 'linux/gfp.h' is not exported
  make[3]: *** [/home/mingo/tip/usr/include/linux/.check] Error 1
  make[2]: *** [linux] Error 2

As suggested by Alexey Dobriyan, move alloc_secdata() and free_secdata()
to linux/security.h - they belong there. This also cleans fs.h of GFP_*
usage.
Suggested-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NPekka Enberg <penberg@cs.helsinki.fi>
Cc: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
LKML-Reference: <1237906803.25315.96.camel@penberg-laptop>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Now that the filesystem freeze operation has been elevated to the VFS, and
is just an ioctl away, some sort of safety net for unintentionally frozen
root filesystems may be in order.

The timeout thaw originally proposed did not get merged, but perhaps
something like this would be useful in emergencies.

For example, freeze /path/to/mountpoint may freeze your root filesystem if
you forgot that you had that unmounted.

I chose 'j' as the last remaining character other than 'h' which is sort
of reserved for help (because help is generated on any unknown character).

I've tested this on a non-root fs with multiple (nested) freezers, as well
as on a system rendered unresponsive due to a frozen root fs.

[randy.dunlap@oracle.com: emergency thaw only if CONFIG_BLOCK enabled]
Signed-off-by: NEric Sandeen <sandeen@redhat.com>
Cc: Takashi Sato <t-sato@yk.jp.nec.com>
Signed-off-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

fsync_bdev() export and a bunch of stubs for !CONFIG_BLOCK case had
been left behind
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

current->fs->umask is what most of fs_struct users are doing.
Put that into a helper function.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Lockdep gripes if file->f_lock is taken in a no-IRQ situation, since that
is not always the case.  We don't really want to disable IRQs for every
acquisition of f_lock; instead, just move it outside of fasync_lock.
Reported-by: NBartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Reported-by: NLarry Finger <Larry.Finger@lwfinger.net>
Reported-by: NWu Fengguang <fengguang.wu@intel.com>
Signed-off-by: NJonathan Corbet <corbet@lwn.net>

simple_set_mnt() is defined as returning 'int' but always returns 0.
Callers assume simple_set_mnt() never fails and don't properly cleanup if
it were to _ever_ fail.  For instance, get_sb_single() and get_sb_nodev()
should:

        up_write(sb->s_unmount);
        deactivate_super(sb);

if simple_set_mnt() fails.

Since simple_set_mnt() never fails, would be cleaner if it did not
return anything.

[akpm@linux-foundation.org: fix build]
Signed-off-by: NSukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Acked-by: NSerge Hallyn <serue@us.ibm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Move some block device related code out from buffer.c and put it in
block_dev.c. I'm trying to move non-buffer_head code out of buffer.c
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

There is a second set of macros for when CONFIG_FILE_LOCKING is
not set. This patch updates those to become inline functions
as well.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This avoids various issues which might give rise to compiler warnings
about missing functions and/or unused variable with the previous
macros. This also fixes a bug where one of the macros was returning
0, but it should have been void.
Reported-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>
Tested-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The last user of do_pipe is in arch/alpha/, after replacing it with
do_pipe_flags, the do_pipe can be totally dropped.
Signed-off-by: NCheng Renquan <crquan@gmail.com>
Acked-by: NRichard Henderson <rth@twiddle.net>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Add support for explicitly requesting full atime updates. This makes it
possible for kernels to default to relatime but still allow userspace to
override it.
Signed-off-by: NMatthew Garrett <mjg@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Traditionally, changes to struct file->f_flags have been done under BKL
protection, or with no protection at all.  This patch causes all f_flags
changes after file open/creation time to be done under protection of
f_lock.  This allows the removal of some BKL usage and fixes a number of
longstanding (if microscopic) races.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: NJonathan Corbet <corbet@lwn.net>

This lock moves out of the CONFIG_EPOLL ifdef and becomes f_lock.  For now,
epoll remains the only user, but a future patch will use it to protect
f_flags as well.

Cc: Davide Libenzi <davidel@xmailserver.org>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJonathan Corbet <corbet@lwn.net>