27 May, 2011
1 commit
-
* 'for-linus' of git://oss.sgi.com/xfs/xfs:
xfs: correctly decrement the extent buffer index in xfs_bmap_del_extent
xfs: check for valid indices in xfs_iext_get_ext and xfs_iext_idx_to_irec
xfs: fix up asserts in xfs_iflush_fork
xfs: do not do pointer arithmetic on extent records
xfs: do not use unchecked extent indices in xfs_bunmapi
xfs: do not use unchecked extent indices in xfs_bmapi
xfs: do not use unchecked extent indices in xfs_bmap_add_extent_*
xfs: remove if_lastex
xfs: remove the unused XFS_BMAPI_RSVBLOCKS flag
xfs: do not discard alloc btree blocks
xfs: add online discard support
25 May, 2011
12 commits
-
The code in xfs_bmap_del_extent does not correctly decrement the
extent buffer index when deleting a whole extent. Most of the time
this gets caught by checks in xfs_bmapi that work around it and
decrement it manually and thus wasn't noticed so far.Based on an earlier patch from Lachlan McIlroy.
Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
Based on an earlier patch from Lachlan McIlroy.
Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
Remove asserts in xfs_iflush_fork that would call xfs_iext_get_ext
with a potentially invalid extent buffer index.Based on an earlier patch from Lachlan McIlroy.
Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
We need to call xfs_iext_get_ext for the previous extent to get a
valid pointer, and can't just do pointer arithmetics as they might
be in different pages.Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
Make sure to only call xfs_iext_get_ext after we've validate the
extent index when moving on to the next index in xfs_bunmapi. Also
remove the old workaround for too large indices that has been
superceeded by the proper fix in xfs_bmap_del_extent.Based on an earlier patch from Lachlan McIlroy.
Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
Make sure to only call xfs_iext_get_ext after we've validate the
extent index when moving on to the next index in xfs_bmapi.Based on an earlier patch from Lachlan McIlroy.
Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
Make sure to only call xfs_iext_get_ext after we've validate the
extent index in the various xfs_bmap_add_extent_* helpers.Based on an earlier patch from Lachlan McIlroy.
Signed-off-by: Christoph Hellwig
Reviewed-by: Lachlan McIlroy
Signed-off-by: Alex Elder -
The if_lastex field in struct xfs_ifork is only used as a temporary
index during xfs_bmapi and xfs_bunmapi. Instead of using the inode
fork to store it keep it local in the callchain. Fortunately this
is very easy as we already pass a stack copy of it down the whole
chain which can simplify be changed to be passed by reference.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
The XFS_BMAPI_RSVBLOCKS is unused, and as far as I can see has
always been. Remove it to simplify the bmapi implementation and
conserve stack space.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
Change each shrinker's API by consolidating the existing parameters into
shrink_control struct. This will simplify any further features added w/o
touching each file of shrinker.[akpm@linux-foundation.org: fix build]
[akpm@linux-foundation.org: fix warning]
[kosaki.motohiro@jp.fujitsu.com: fix up new shrinker API]
[akpm@linux-foundation.org: fix xfs warning]
[akpm@linux-foundation.org: update gfs2]
Signed-off-by: Ying Han
Cc: KOSAKI Motohiro
Cc: Minchan Kim
Acked-by: Pavel Emelyanov
Cc: KAMEZAWA Hiroyuki
Cc: Mel Gorman
Acked-by: Rik van Riel
Cc: Johannes Weiner
Cc: Hugh Dickins
Cc: Dave Hansen
Cc: Steven Whitehouse
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Blocks for the allocation btree are allocated from and released to
the AGFL, and thus frequently reused. Even worse we do not have an
easy way to avoid using an AGFL block when it is discarded due to
the simple FILO list of free blocks, and thus can frequently stall
on blocks that are currently undergoing a discard.Add a flag to the busy extent tracking structure to skip the discard
for allocation btree blocks. In normal operation these blocks are
reused frequently enough that there is no need to discard them
anyway, but if they spill over to the allocation btree as part of a
balance we "leak" blocks that we would otherwise discard. We could
fix this by adding another flag and keeping these block in the
rbtree even after they aren't busy any more so that we could discard
them when they migrate out of the AGFL. Given that this would cause
significant overhead I don't think it's worthwile for now.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
Now that we have reliably tracking of deleted extents in a
transaction we can easily implement "online" discard support
which calls blkdev_issue_discard once a transaction commits.The actual discard is a two stage operation as we first have
to mark the busy extent as not available for reuse before we
can start the actual discard. Note that we don't bother
supporting discard for the non-delaylog mode.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder
24 May, 2011
2 commits
-
* 'for-linus' of git://oss.sgi.com/xfs/xfs:
xfs: obey minleft values during extent allocation correctly
xfs: reset buffer pointers before freeing them
xfs: avoid getting stuck during async inode flushes
xfs: fix xfs_itruncate_start tracing
xfs: fix duplicate workqueue initialisation
xfs: kill off xfs_printk()
xfs: fix race condition in AIL push trigger
xfs: make AIL target updates and compares 32bit safe.
xfs: always push the AIL to the target
xfs: exit AIL push work correctly when AIL is empty
xfs: ensure reclaim cursor is reset correctly at end of AG
xfs: add an x86 compat handler for XFS_IOC_ZERO_RANGE
xfs: fix compiler warning in xfs_trace.h
xfs: cleanup duplicate initializations
xfs: reduce the number of pagb_lock roundtrips in xfs_alloc_clear_busy
xfs: exact busy extent tracking
xfs: do not immediately reuse busy extent ranges
xfs: optimize AGFL refills -
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (39 commits)
b43: fix comment typo reqest -> request
Haavard Skinnemoen has left Atmel
cris: typo in mach-fs Makefile
Kconfig: fix copy/paste-ism for dell-wmi-aio driver
doc: timers-howto: fix a typo ("unsgined")
perf: Only include annotate.h once in tools/perf/util/ui/browsers/annotate.c
md, raid5: Fix spelling error in comment ('Ofcourse' --> 'Of course').
treewide: fix a few typos in comments
regulator: change debug statement be consistent with the style of the rest
Revert "arm: mach-u300/gpio: Fix mem_region resource size miscalculations"
audit: acquire creds selectively to reduce atomic op overhead
rtlwifi: don't touch with treewide double semicolon removal
treewide: cleanup continuations and remove logging message whitespace
ath9k_hw: don't touch with treewide double semicolon removal
include/linux/leds-regulator.h: fix syntax in example code
tty: fix typo in descripton of tty_termios_encode_baud_rate
xtensa: remove obsolete BKL kernel option from defconfig
m68k: fix comment typo 'occcured'
arch:Kconfig.locks Remove unused config option.
treewide: remove extra semicolons
...
20 May, 2011
6 commits
-
When allocating an extent that is long enough to consume the
remaining free space in an AG, we need to ensure that the allocation
leaves enough space in the AG for any subsequent bmap btree blocks
that are needed to track the new extent. These have to be allocated
in the same AG as we only reserve enough blocks in an allocation
transaction for modification of the freespace trees in a single AG.xfs_alloc_fix_minleft() has been considering blocks on the AGFL as
free blocks available for extent and bmbt block allocation, which is
not correct - blocks on the AGFL are there exclusively for the use
of the free space btrees. As a result, when minleft is less than the
number of blocks on the AGFL, xfs_alloc_fix_minleft() does not trim
the given extent to leave minleft blocks available for bmbt
allocation, and hence we can fail allocation during bmbt record
insertion.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Signed-off-by: Alex Elder -
When we free a vmapped buffer, we need to ensure the vmap address
and length we free is the same as when it was allocated. In various
places in the log code we change the memory the buffer is pointing
to before issuing IO, but we never reset the buffer to point back to
it's original memory (or no memory, if that is the case for the
buffer).As a result, when we free the buffer it points to memory that is
owned by something else and attempts to unmap and free it. Because
the range does not match any known mapped range, it can trigger
BUG_ON() traps in the vmap code, and potentially corrupt the vmap
area tracking.Fix this by always resetting these buffers to their original state
before freeing them.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Signed-off-by: Alex Elder -
When the underlying inode buffer is locked and xfs_sync_inode_attr()
is doing a non-blocking flush, xfs_iflush() can return EAGAIN. When
this happens, clear the error rather than returning it to
xfs_inode_ag_walk(), as returning EAGAIN will result in the AG walk
delaying for a short while and trying again. This can result in
background walks getting stuck on the one AG until inode buffer is
unlocked by some other means.This behaviour was noticed when analysing event traces followed by
code inspection and verification of the fix via further traces.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Signed-off-by: Alex Elder -
Variables are ordered incorrectly in trace call.
Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Signed-off-by: Alex Elder -
The workqueue initialisation function is called twice when
initialising the XFS subsystem. Remove the second initialisation
call.Signed-off-by: Dave Chinner
Signed-off-by: Alex Elder -
xfs_alert_tag() can be defined using xfs_alert(), and thereby avoid
using xfs_printk() altogether. This is the only remaining use of
xfs_printk(), so changing it this way means xfs_printk() can simply
be eliminated.can simply be eliminated.can simply be eliminated.can
simply be eliminated.can simply be eliminated.can simply be
eliminated.can simply be eliminated.can simply be eliminated.can
simply be eliminated.Also add format checking to the non-debug inline function xfs_debug.
Miscellaneous function prototype argument alignment.(Updated to delete the definition of xfs_printk(), which is
no longer used or needed.)Signed-off-by: Joe Perches
Signed-off-by: Alex Elder
10 May, 2011
11 commits
-
- kenrel -> kernel
- whetehr -> whether
- ttt -> tt
- sss -> ssSigned-off-by: Justin P. Mattock
Signed-off-by: Jiri Kosina -
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. One is caused by a
race condition in determining whether there is a psh in progress or
not.The XFS_AIL_PUSHING_BIT is used to determine whether a push is
currently in progress. When the AIL push work completes, it checked
whether the target changed and cleared the PUSHING bit to allow a
new push to be requeued. The race condition is as follows:Thread 1 push work
smp_wmb()
smp_rmb()
check ailp->xa_target unchanged
update ailp->xa_target
test/set PUSHING bit
does not queue
clear PUSHING bit
does not requeueNow that the push target is updated, new attempts to push the AIL
will not trigger as the push target will be the same, and hence
despite trying to push the AIL we won't ever wake it again.The fix is to ensure that the AIL push work clears the PUSHING bit
before it checks if the target is unchanged.As a result, both push triggers operate on the same test/set bit
criteria, so even if we race in the push work and miss the target
update, the thread requesting the push will still set the PUSHING
bit and queue the push work to occur. For safety sake, the same
queue check is done if the push work detects the target change,
though only one of the two will will queue new work due to the use
of test_and_set_bit() checks.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder(cherry picked from commit e4d3c4a43b595d5124ae824d300626e6489ae857)
-
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. One of the problems
noticed was that updates of the push target are not 32 bit safe as
the target is a 64 bit value.We cannot copy a 64 bit LSN without the possibility of corrupting
the result when racing with another updating thread. We have
function to do this update safely without needing to care about
32/64 bit issues - xfs_trans_ail_copy_lsn() - so use that when
updating the AIL push target.Also move the reading of the target in the push work inside the AIL
lock, and use XFS_LSN_CMP() for the unlocked comparison during work
termination to close read holes as well.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder(cherry picked from commit fd5670f22fce247754243cf2ed41941e5762d990)
-
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. One of the problems
discovered is a target mismatch between the item pushing loop and
the target itself.The push trigger checks for the target increasing (i.e. new target >
current) while the push loop only pushes items that have a LSN <
current. As a result, we can get the situation where the push target
is X, the items at the tail of the AIL have LSN X and they don't get
pushed. The push work then completes thinking it is done, and cannot
be restarted until the push target increases to >= X + 1. If the
push target then never increases (because the tail is not moving),
then we never run the push work again and we stall.Fix it by making sure log items with a LSN that matches the target
exactly are pushed during the loop.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder(cherry picked from commit cb64026b6e8af50db598ec7c3f59d504259b00bb)
-
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. The main cause is a
regression where a work exit path fails to clear the PUSHING state
and recheck the target correctly.Make both exit paths do the same PUSHING bit clearing and target
checking when the "no more work to be done" condition is hit.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder(cherry picked from commit ea35a20021f8497390d05b93271b4d675516c654)
-
On a 32 bit highmem PowerPC machine, the XFS inode cache was growing
without bound and exhausting low memory causing the OOM killer to be
triggered. After some effort, the problem was reproduced on a 32 bit
x86 highmem machine.The problem is that the per-ag inode reclaim index cursor was not
getting reset to the start of the AG if the radix tree tag lookup
found no more reclaimable inodes. Hence every further reclaim
attempt started at the same index beyond where any reclaimable
inodes lay, and no further background reclaim ever occurred from the
AG.Without background inode reclaim the VM driven cache shrinker
simply cannot keep up with cache growth, and OOM is the result.While the change that exposed the problem was the conversion of the
inode reclaim to use work queues for background reclaim, it was not
the cause of the bug. The bug was introduced when the cursor code
was added, just waiting for some weird configuration to strike....Signed-off-by: Dave Chinner
Tested-By: Christian Kujau
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder(cherry picked from commit b223221956675ce8a7b436d198ced974bb388571)
-
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. One is caused by a
race condition in determining whether there is a psh in progress or
not.The XFS_AIL_PUSHING_BIT is used to determine whether a push is
currently in progress. When the AIL push work completes, it checked
whether the target changed and cleared the PUSHING bit to allow a
new push to be requeued. The race condition is as follows:Thread 1 push work
smp_wmb()
smp_rmb()
check ailp->xa_target unchanged
update ailp->xa_target
test/set PUSHING bit
does not queue
clear PUSHING bit
does not requeueNow that the push target is updated, new attempts to push the AIL
will not trigger as the push target will be the same, and hence
despite trying to push the AIL we won't ever wake it again.The fix is to ensure that the AIL push work clears the PUSHING bit
before it checks if the target is unchanged.As a result, both push triggers operate on the same test/set bit
criteria, so even if we race in the push work and miss the target
update, the thread requesting the push will still set the PUSHING
bit and queue the push work to occur. For safety sake, the same
queue check is done if the push work detects the target change,
though only one of the two will will queue new work due to the use
of test_and_set_bit() checks.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder -
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. One of the problems
noticed was that updates of the push target are not 32 bit safe as
the target is a 64 bit value.We cannot copy a 64 bit LSN without the possibility of corrupting
the result when racing with another updating thread. We have
function to do this update safely without needing to care about
32/64 bit issues - xfs_trans_ail_copy_lsn() - so use that when
updating the AIL push target.Also move the reading of the target in the push work inside the AIL
lock, and use XFS_LSN_CMP() for the unlocked comparison during work
termination to close read holes as well.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder -
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. One of the problems
discovered is a target mismatch between the item pushing loop and
the target itself.The push trigger checks for the target increasing (i.e. new target >
current) while the push loop only pushes items that have a LSN <
current. As a result, we can get the situation where the push target
is X, the items at the tail of the AIL have LSN X and they don't get
pushed. The push work then completes thinking it is done, and cannot
be restarted until the push target increases to >= X + 1. If the
push target then never increases (because the tail is not moving),
then we never run the push work again and we stall.Fix it by making sure log items with a LSN that matches the target
exactly are pushed during the loop.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder -
The recent conversion of the xfsaild functionality to a work queue
introduced a hard-to-hit log space grant hang. The main cause is a
regression where a work exit path fails to clear the PUSHING state
and recheck the target correctly.Make both exit paths do the same PUSHING bit clearing and target
checking when the "no more work to be done" condition is hit.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder -
On a 32 bit highmem PowerPC machine, the XFS inode cache was growing
without bound and exhausting low memory causing the OOM killer to be
triggered. After some effort, the problem was reproduced on a 32 bit
x86 highmem machine.The problem is that the per-ag inode reclaim index cursor was not
getting reset to the start of the AG if the radix tree tag lookup
found no more reclaimable inodes. Hence every further reclaim
attempt started at the same index beyond where any reclaimable
inodes lay, and no further background reclaim ever occurred from the
AG.Without background inode reclaim the VM driven cache shrinker
simply cannot keep up with cache growth, and OOM is the result.While the change that exposed the problem was the conversion of the
inode reclaim to use work queues for background reclaim, it was not
the cause of the bug. The bug was introduced when the cursor code
was added, just waiting for some weird configuration to strike....Signed-off-by: Dave Chinner
Tested-By: Christian Kujau
Reviewed-by: Christoph Hellwig
Reviewed-by: Alex Elder
29 Apr, 2011
7 commits
-
XFS_IOC_ZERO_RANGE uses struct xfs_flock64, and thus requires argument
translation for 32-bit binaries on x86. Add the required
XFS_IOC_ZERO_RANGE_32 defined and add it to the list of commands that
require xfs_flock64 translation.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
xfs_fsblock_t may be a 32-bit type on if XFS_BIG_BLKNOS is not set,
make sure to cast a value of this type to an unsigned long long
before using the ll printk qualifier.Reported-by: Randy Dunlap
Signed-off-by: Christoph Hellwig
Reviewed-by: Dave Chinner
Signed-off-by: Alex Elder -
follow these guidelines:
- leave initialization in the declaration block if it fits the line
- move to the code where it's more suitable ('for' init block)The last chunk was modified from David's original to be a correct
fix for what appeared to be a duplicate initialization.Signed-off-by: David Sterba
Signed-off-by: Alex Elder
Reviewed-by: Dave Chinner -
Instead of finding the per-ag and then taking and releasing the pagb_lock
for every single busy extent completed sort the list of busy extents and
only switch betweens AGs where nessecary. This becomes especially important
with the online discard support which will hit this lock more often.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
Update the extent tree in case we have to reuse a busy extent, so that it
always is kept uptodate. This is done by replacing the busy list searches
with a new xfs_alloc_busy_reuse helper, which updates the busy extent tree
in case of a reuse. This allows us to allow reusing metadata extents
unconditionally, and thus avoid log forces especially for allocation btree
blocks.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
Every time we reallocate a busy extent, we cause a synchronous log force
to occur to ensure the freeing transaction is on disk before we continue
and use the newly allocated extent. This is extremely sub-optimal as we
have to mark every transaction with blocks that get reused as synchronous.Instead of searching the busy extent list after deciding on the extent to
allocate, check each candidate extent during the allocation decisions as
to whether they are in the busy list. If they are in the busy list, we
trim the busy range out of the extent we have found and determine if that
trimmed range is still OK for allocation. In many cases, this check can
be incorporated into the allocation extent alignment code which already
does trimming of the found extent before determining if it is a valid
candidate for allocation.Based on earlier patches from Dave Chinner.
Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder -
While we need to make sure we do not reuse busy extents, there is no need
to force out busy extents when moving them between the AGFL and the
freespace btree as we still take care of that when doing the real allocation.To avoid the log force when just moving extents from the different free
space tracking structures, move the busy search out of
xfs_alloc_get_freelist into the callers that need it, and move the busy
list insert from xfs_free_ag_extent which is used both by AGFL refills
and real allocation to xfs_free_extent, which is only used by the latter.Signed-off-by: Christoph Hellwig
Signed-off-by: Alex Elder
21 Apr, 2011
1 commit
-
Commit 957935dc ("xfs: fix xfs_debug warnings" broke the logic in
__xfs_printk(). Instead of only printing one of two possible output
strings based on whether the fs has a name or not, it outputs both.
Fix it to only output one message again.Signed-off-by: Dave Chinner
Reviewed-by: Christoph Hellwig
Signed-off-by: Alex Elder