Skip to content
  1. Dec 02, 2010
  3. Dec 01, 2010
    • Dave Chinner's avatar
      xfs: only run xfs_error_test if error injection is active · c76febef
      Dave Chinner authored 
      

Recent tests writing lots of small files showed the flusher thread
being CPU bound and taking a long time to do allocations on a debug
kernel. perf showed this as the prime reason:

             samples  pcnt function                    DSO
             _______ _____ ___________________________ _________________

           224648.00 36.8% xfs_error_test              [kernel.kallsyms]
            86045.00 14.1% xfs_btree_check_sblock      [kernel.kallsyms]
            39778.00  6.5% prandom32                   [kernel.kallsyms]
            37436.00  6.1% xfs_btree_increment         [kernel.kallsyms]
            29278.00  4.8% xfs_btree_get_rec           [kernel.kallsyms]
            27717.00  4.5% random32                    [kernel.kallsyms]

Walking btree blocks during allocation checking them requires each
block (a cache hit, so no I/O) call xfs_error_test(), which then
does a random32() call as the first operation.  IOWs, ~50% of the
CPU is being consumed just testing whether we need to inject an
error, even though error injection is not active.

Kill this overhead when error injection is not active by adding a
global counter of active error traps and only calling into
xfs_error_test when fault injection is active.

Signed-off-by: default avatarDave Chinner <dchinner@redhat.com>
Reviewed-by: default avatarChristoph Hellwig <hch@lst.de>
      c76febef
    • Dave Chinner's avatar
      xfs: avoid moving stale inodes in the AIL · de25c181
      Dave Chinner authored 
      

When an inode has been marked stale because the cluster is being
freed, we don't want to (re-)insert this inode into the AIL. There
is a race condition where the cluster buffer may be unpinned before
the inode is inserted into the AIL during transaction committed
processing. If the buffer is unpinned before the inode item has been
committed and inserted, then it is possible for the buffer to be
released and hence processthe stale inode callbacks before the inode
is inserted into the AIL.

In this case, we then insert a clean, stale inode into the AIL which
will never get removed by an IO completion. It will, however, get
reclaimed and that triggers an assert in xfs_inode_free()
complaining about freeing an inode still in the AIL.

This race can be avoided by not moving stale inodes forward in the AIL
during transaction commit completion processing. This closes the
race condition by ensuring we never insert clean stale inodes into
the AIL. It is safe to do this because a dirty stale inode, by
definition, must already be in the AIL.

Signed-off-by: default avatarDave Chinner <dchinner@redhat.com>
Reviewed-by: default avatarChristoph Hellwig <hch@lst.de>
      de25c181
    • Dave Chinner's avatar
      xfs: delayed alloc blocks beyond EOF are valid after writeback · 309c8480
      Dave Chinner authored 
      

There is an assumption in the parts of XFS that flushing a dirty
file will make all the delayed allocation blocks disappear from an
inode. That is, that after calling xfs_flush_pages() then
ip->i_delayed_blks will be zero.

This is an invalid assumption as we may have specualtive
preallocation beyond EOF and they are recorded in
ip->i_delayed_blks. A flush of the dirty pages of an inode will not
change the state of these blocks beyond EOF, so a non-zero
deeelalloc block count after a flush is valid.

The bmap code has an invalid ASSERT() that needs to be removed, and
the swapext code has a bug in that while it swaps the data forks
around, it fails to swap the i_delayed_blks counter associated with
the fork and hence can get the block accounting wrong.

Signed-off-by: default avatarDave Chinner <dchinner@redhat.com>
Reviewed-by: default avatarChristoph Hellwig <hch@lst.de>
      309c8480
    • Dave Chinner's avatar
      xfs: push stale, pinned buffers on trylock failures · 90810b9e
      Dave Chinner authored 
      

As reported by Nick Piggin, XFS is suffering from long pauses under
highly concurrent workloads when hosted on ramdisks. The problem is
that an inode buffer is stuck in the pinned state in memory and as a
result either the inode buffer or one of the inodes within the
buffer is stopping the tail of the log from being moved forward.

The system remains in this state until a periodic log force issued
by xfssyncd causes the buffer to be unpinned. The main problem is
that these are stale buffers, and are hence held locked until the
transaction/checkpoint that marked them state has been committed to
disk. When the filesystem gets into this state, only the xfssyncd
can cause the async transactions to be committed to disk and hence
unpin the inode buffer.

This problem was encountered when scaling the busy extent list, but
only the blocking lock interface was fixed to solve the problem.
Extend the same fix to the buffer trylock operations - if we fail to
lock a pinned, stale buffer, then force the log immediately so that
when the next attempt to lock it comes around, it will have been
unpinned.

Signed-off-by: default avatarDave Chinner <dchinner@redhat.com>
Reviewed-by: default avatarChristoph Hellwig <hch@lst.de>
      90810b9e
    • Dave Chinner's avatar
      xfs: fix failed write truncation handling. · c726de44
      Dave Chinner authored 
      

Since the move to the new truncate sequence we call xfs_setattr to
truncate down excessively instanciated blocks.  As shown by the testcase
in kernel.org BZ #22452 that doesn't work too well.  Due to the confusion
of the internal inode size, and the VFS inode i_size it zeroes data that
it shouldn't.

But full blown truncate seems like overkill here.  We only instanciate
delayed allocations in the write path, and given that we never released
the iolock we can't have converted them to real allocations yet either.

The only nasty case is pre-existing preallocation which we need to skip.
We already do this for page discard during writeback, so make the delayed
allocation block punching a generic function and call it from the failed
write path as well as xfs_aops_discard_page. The callers are
responsible for ensuring that partial blocks are not truncated away,
and that they hold the ilock.

Based on a fix originally from Christoph Hellwig. This version used
filesystem blocks as the range unit.

Signed-off-by: default avatarDave Chinner <dchinner@redhat.com>
Reviewed-by: default avatarChristoph Hellwig <hch@lst.de>
      c726de44
    • Oleg Nesterov's avatar
      exec: copy-and-paste the fixes into compat_do_execve() paths · 114279be
      Oleg Nesterov authored 
      

Note: this patch targets 2.6.37 and tries to be as simple as possible.
That is why it adds more copy-and-paste horror into fs/compat.c and
uglifies fs/exec.c, this will be cleanuped later.

compat_copy_strings() plays with bprm->vma/mm directly and thus has
two problems: it lacks the RLIMIT_STACK check and argv/envp memory
is not visible to oom killer.

Export acct_arg_size() and get_arg_page(), change compat_copy_strings()
to use get_arg_page(), change compat_do_execve() to do acct_arg_size(0)
as do_execve() does.

Add the fatal_signal_pending/cond_resched checks into compat_count() and
compat_copy_strings(), this matches the code in fs/exec.c and certainly
makes sense.

Signed-off-by: default avatarOleg Nesterov <oleg@redhat.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: stable@kernel.org
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      114279be
    • Oleg Nesterov's avatar
      exec: make argv/envp memory visible to oom-killer · 3c77f845
      Oleg Nesterov authored 
      Brad Spengler published a local memory-allocation DoS that
evades the OOM-killer (though not the virtual memory RLIMIT):
http://www.grsecurity.net/~spender/64bit_dos.c



execve()->copy_strings() can allocate a lot of memory, but
this is not visible to oom-killer, nobody can see the nascent
bprm->mm and take it into account.

With this patch get_arg_page() increments current's MM_ANONPAGES
counter every time we allocate the new page for argv/envp. When
do_execve() succeds or fails, we change this counter back.

Technically this is not 100% correct, we can't know if the new
page is swapped out and turn MM_ANONPAGES into MM_SWAPENTS, but
I don't think this really matters and everything becomes correct
once exec changes ->mm or fails.

Reported-by: default avatarBrad Spengler <spender@grsecurity.net>
Reviewed-and-discussed-by: default avatarKOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: default avatarOleg Nesterov <oleg@redhat.com>
Cc: stable@kernel.org
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      3c77f845
  5. Nov 30, 2010
  7. Nov 29, 2010
  9. Nov 28, 2010
    • Linus Torvalds's avatar
      Export 'get_pipe_info()' to other users · c66fb347
      Linus Torvalds authored 
      

And in particular, use it in 'pipe_fcntl()'.

The other pipe functions do not need to use the 'careful' version, since
they are only ever called for things that are already known to be pipes.

The normal read/write/ioctl functions are called through the file
operations structures, so if a file isn't a pipe, they'd never get
called.  But pipe_fcntl() is special, and called directly from the
generic fcntl code, and needs to use the same careful function that the
splice code is using.

Cc: Jens Axboe <jaxboe@fusionio.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Dave Jones <davej@redhat.com>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      c66fb347
    • Linus Torvalds's avatar
      Rename 'pipe_info()' to 'get_pipe_info()' · 71993e62
      Linus Torvalds authored 
      

.. and change it to take the 'file' pointer instead of an inode, since
that's what all users want anyway.

The renaming is preparatory to exporting it to other users.  The old
'pipe_info()' name was too generic and is already used elsewhere, so
before making the function public we need to use a more specific name.

Cc: Jens Axboe <jaxboe@fusionio.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Dave Jones <davej@redhat.com>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      71993e62
  11. Nov 27, 2010
    • Josef Bacik's avatar
      Btrfs: setup blank root and fs_info for mount time · 450ba0ea
      Josef Bacik authored 
      

There is a problem with how we use sget, it searches through the list of supers
attached to the fs_type looking for a super with the same fs_devices as what
we're trying to mount.  This depends on sb->s_fs_info being filled, but we don't
fill that in until we get to btrfs_fill_super, so we could hit supers on the
fs_type super list that have a null s_fs_info.  In order to fix that we need to
go ahead and setup a blank root with a blank fs_info to hold fs_devices, that
way our test will work out right and then we can set s_fs_info in
btrfs_set_super, and then open_ctree will simply use our pre-allocated root and
fs_info when setting everything up.  Thanks,

Signed-off-by: default avatarJosef Bacik <josef@redhat.com>
Signed-off-by: default avatarChris Mason <chris.mason@oracle.com>
      450ba0ea
    • Josef Bacik's avatar
      Btrfs: fix fiemap · 975f84fe
      Josef Bacik authored 
      

There are two big problems currently with FIEMAP

1) We return extents for holes.  This isn't supposed to happen, we just don't
return extents for holes and then userspace interprets the lack of an extent as
a hole.

2) We sometimes don't set FIEMAP_EXTENT_LAST properly.  This is because we wait
to see a EXTENT_FLAG_VACANCY flag on the em, but this won't happen if say we ask
fiemap to map up to the last extent in a file, and there is nothing but holes up
to the i_size.  To fix this we need to lookup the last extent in this file and
save the logical offset, so if we happen to try and map that extent we can be
sure to set FIEMAP_EXTENT_LAST.

With this patch we now pass xfstest 225, which we never have before.

Signed-off-by: default avatarJosef Bacik <josef@redhat.com>
Signed-off-by: default avatarChris Mason <chris.mason@oracle.com>
      975f84fe
    • Ian Kent's avatar
      Btrfs - fix race between btrfs_get_sb() and umount · 619c8c76
      Ian Kent authored 
      

When mounting a btrfs file system btrfs_test_super() may attempt to
use sb->s_fs_info, the btrfs root, of a super block that is going away
and that has had the btrfs root set to NULL in its ->put_super(). But
if the super block is going away it cannot be an existing super block
so we can return false in this case.

Signed-off-by: default avatarIan Kent <raven@themaw.net>
Signed-off-by: default avatarChris Mason <chris.mason@oracle.com>
      619c8c76
    • Josef Bacik's avatar
      Btrfs: update inode ctime when using links · bc1cbf1f
      Josef Bacik authored 
      

Currently we fail xfstest 236 because we're not updating the inode ctime on
link.  This is a simple fix, and makes it so we pass 236 now.

Signed-off-by: default avatarJosef Bacik <josef@redhat.com>
Signed-off-by: default avatarChris Mason <chris.mason@oracle.com>
      bc1cbf1f
    • Josef Bacik's avatar
      Btrfs: make sure new inode size is ok in fallocate · 0ed42a63
      Josef Bacik authored 
      

We have been failing xfstest 228 forever, because we don't check to make sure
the new inode size is acceptable as far as RLIMIT is concerned.  Just check to
make sure it's ok to create a inode with this new size and error out if not.
With this patch we now pass 228.

Signed-off-by: default avatarJosef Bacik <josef@redhat.com>
Signed-off-by: default avatarChris Mason <chris.mason@oracle.com>
      0ed42a63
    • Josef Bacik's avatar
      Btrfs: fix typo in fallocate to make it honor actual size · 55a61d1d
      Josef Bacik authored 
      

There is a typo in __btrfs_prealloc_file_range() where we set the i_size to
actual_len/cur_offset, and then just set it to cur_offset again, and do the same
with btrfs_ordered_update_i_size().  This fixes it back to keeping i_size in a
local variable and then updating i_size properly.  Tested this with

xfs_io -F -f -c "falloc 0 1" -c "pwrite 0 1" foo

stat'ing foo gives us a size of 1 instead of 4096 like it was.  Thanks,

Signed-off-by: default avatarJosef Bacik <josef@redhat.com>
Signed-off-by: default avatarChris Mason <chris.mason@oracle.com>
      55a61d1d
  13. Nov 24, 2010
  15. Nov 23, 2010
  17. Nov 22, 2010