For 3.13-rc1 we have an eclectic assortment of bugfixes, cleanups, and
refactoring. Bugfixes that stand out are the fix for the AGF/AGI
deadlock, incore extent list fixes, verifier fixes for v4 superblocks
and growfs, and memory leaks. There are some asserts, warnings, and
strings that were cleaned up. There was further rearrangement of code
to make libxfs and the kernel sync up more easily, differences between
v2 and v3 directory code were abstracted using an ops vector,
xfs_inactive was reworked, and the preallocation/hole punching code was
refactored.
- simplify kmem_zone_zalloc
- add traces for AGF/AGI read ops
- add additional AIL traces
- fix xfs_remove AGF vs AGI deadlock
- fix the extent count of new incore extent page in the indirection array
- don't fail bad secondary superblocks verification on v4 filesystems
due to unzeroed bits after v4 fields
- fix possible NULL dereference in xlog_verify_iclog
- remove redundant assert in xfs_dir2_leafn_split
- prevent stack overflows from page cache allocation
- fix some sparse warnings
- fix directory block format verifier to check the leaf entry count
- abstract the differences in dir2/dir3 via an ops vector
- continue process of reorganization to make libxfs/kernel code merges easier
- refactor the preallocation and hole punching code
- fix for growfs and verifiers
- remove unnecessary scary corruption error when probing non-xfs filesystems
- remove extra newlines from strings passed to printk
- prevent deadlock trying to cover an active log
- rework xfs_inactive()
- add the inode directory type support to XFS_IOC_FSGEOM
- cleanup (remove) usage of is_bad_inode
- fix miscalculation in xfs_iext_realloc_direct which results in oversized
direct extent list
- remove unnecessary count arg to xfs_iomap_write_allocate
- fix memory leak in xlog_recover_add_to_trans
- check superblock instead of block magic to determine if dtype field
is present
- fix lockdep annotation due to project quotas
- fix regression in xfs_node_toosmall which can lead to incorrect directory
btree node collapse
- make log recovery verify filesystem uuid of recovering blocks
- fix XFS_IOC_FREE_EOFBLOCKS definition
- remove invalid assert in xfs_inode_free
- fix for AIL lock regression
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)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=XMd/
-----END PGP SIGNATURE-----
Merge tag 'xfs-for-linus-v3.13-rc1' of git://oss.sgi.com/xfs/xfs
Pull xfs update from Ben Myers:
"For 3.13-rc1 we have an eclectic assortment of bugfixes, cleanups, and
refactoring. Bugfixes that stand out are the fix for the AGF/AGI
deadlock, incore extent list fixes, verifier fixes for v4 superblocks
and growfs, and memory leaks. There are some asserts, warnings, and
strings that were cleaned up. There was further rearrangement of code
to make libxfs and the kernel sync up more easily, differences between
v2 and v3 directory code were abstracted using an ops vector,
xfs_inactive was reworked, and the preallocation/hole punching code
was refactored.
- simplify kmem_zone_zalloc
- add traces for AGF/AGI read ops
- add additional AIL traces
- fix xfs_remove AGF vs AGI deadlock
- fix the extent count of new incore extent page in the indirection
array
- don't fail bad secondary superblocks verification on v4 filesystems
due to unzeroed bits after v4 fields
- fix possible NULL dereference in xlog_verify_iclog
- remove redundant assert in xfs_dir2_leafn_split
- prevent stack overflows from page cache allocation
- fix some sparse warnings
- fix directory block format verifier to check the leaf entry count
- abstract the differences in dir2/dir3 via an ops vector
- continue process of reorganization to make libxfs/kernel code
merges easier
- refactor the preallocation and hole punching code
- fix for growfs and verifiers
- remove unnecessary scary corruption error when probing non-xfs
filesystems
- remove extra newlines from strings passed to printk
- prevent deadlock trying to cover an active log
- rework xfs_inactive()
- add the inode directory type support to XFS_IOC_FSGEOM
- cleanup (remove) usage of is_bad_inode
- fix miscalculation in xfs_iext_realloc_direct which results in
oversized direct extent list
- remove unnecessary count arg to xfs_iomap_write_allocate
- fix memory leak in xlog_recover_add_to_trans
- check superblock instead of block magic to determine if dtype field
is present
- fix lockdep annotation due to project quotas
- fix regression in xfs_node_toosmall which can lead to incorrect
directory btree node collapse
- make log recovery verify filesystem uuid of recovering blocks
- fix XFS_IOC_FREE_EOFBLOCKS definition
- remove invalid assert in xfs_inode_free
- fix for AIL lock regression"
* tag 'xfs-for-linus-v3.13-rc1' of git://oss.sgi.com/xfs/xfs: (49 commits)
xfs: simplify kmem_{zone_}zalloc
xfs: add tracepoints to AGF/AGI read operations
xfs: trace AIL manipulations
xfs: xfs_remove deadlocks due to inverted AGF vs AGI lock ordering
xfs: fix the extent count when allocating an new indirection array entry
xfs: be more forgiving of a v4 secondary sb w/ junk in v5 fields
xfs: fix possible NULL dereference in xlog_verify_iclog
xfs:xfs_dir2_node.c: pointer use before check for null
xfs: prevent stack overflows from page cache allocation
xfs: fix static and extern sparse warnings
xfs: validity check the directory block leaf entry count
xfs: make dir2 ftype offset pointers explicit
xfs: convert directory vector functions to constants
xfs: convert directory vector functions to constants
xfs: vectorise encoding/decoding directory headers
xfs: vectorise DA btree operations
xfs: vectorise directory leaf operations
xfs: vectorise directory data operations part 2
xfs: vectorise directory data operations
xfs: vectorise remaining shortform dir2 ops
...
When there are processes heavily creating small files while sync(2) is
running, it can easily happen that quite some new files are created
between WB_SYNC_NONE and WB_SYNC_ALL pass of sync(2). That can happen
especially if there are several busy filesystems (remember that sync
traverses filesystems sequentially and waits in WB_SYNC_ALL phase on one
fs before starting it on another fs). Because WB_SYNC_ALL pass is slow
(e.g. causes a transaction commit and cache flush for each inode in
ext3), resulting sync(2) times are rather large.
The following script reproduces the problem:
function run_writers
{
for (( i = 0; i < 10; i++ )); do
mkdir $1/dir$i
for (( j = 0; j < 40000; j++ )); do
dd if=/dev/zero of=$1/dir$i/$j bs=4k count=4 &>/dev/null
done &
done
}
for dir in "$@"; do
run_writers $dir
done
sleep 40
time sync
Fix the problem by disregarding inodes dirtied after sync(2) was called
in the WB_SYNC_ALL pass. To allow for this, sync_inodes_sb() now takes
a time stamp when sync has started which is used for setting up work for
flusher threads.
To give some numbers, when above script is run on two ext4 filesystems
on simple SATA drive, the average sync time from 10 runs is 267.549
seconds with standard deviation 104.799426. With the patched kernel,
the average sync time from 10 runs is 2.995 seconds with standard
deviation 0.096.
Signed-off-by: Jan Kara <jack@suse.cz>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Introduce flag KM_ZERO which is used to alloc zeroed entry, and convert
kmem_{zone_}zalloc to call kmem_{zone_}alloc() with KM_ZERO directly,
in order to avoid the setting to zero step.
And following Dave's suggestion, make kmem_{zone_}zalloc static inline
into kmem.h as they're now just a simple wrapper.
V2:
Make kmem_{zone_}zalloc static inline into kmem.h as Dave suggested.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
To help track down AGI/AGF lock ordering issues, I added these
tracepoints to tell us when an AGI or AGF is read and locked. With
these we can now determine if the lock ordering goes wrong from
tracing captures.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
I debugging a log tail issue on a RHEL6 kernel, I added these trace
points to trace log items being added, moved and removed in the AIL
and how that affected the log tail LSN that was written to the log.
They were very helpful in that they immediately identified the cause
of the problem being seen. Hence I'd like to always have them
available for use.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Removing an inode from the namespace involves removing the directory
entry and dropping the link count on the inode. Removing the
directory entry can result in locking an AGF (directory blocks were
freed) and removing a link count can result in placing the inode on
an unlinked list which results in locking an AGI.
The big problem here is that we have an ordering constraint on AGF
and AGI locking - inode allocation locks the AGI, then can allocate
a new extent for new inodes, locking the AGF after the AGI.
Similarly, freeing the inode removes the inode from the unlinked
list, requiring that we lock the AGI first, and then freeing the
inode can result in an inode chunk being freed and hence freeing
disk space requiring that we lock an AGF.
Hence the ordering that is imposed by other parts of the code is AGI
before AGF. This means we cannot remove the directory entry before
we drop the inode reference count and put it on the unlinked list as
this results in a lock order of AGF then AGI, and this can deadlock
against inode allocation and freeing. Therefore we must drop the
link counts before we remove the directory entry.
This is still safe from a transactional point of view - it is not
until we get to xfs_bmap_finish() that we have the possibility of
multiple transactions in this operation. Hence as long as we remove
the directory entry and drop the link count in the first transaction
of the remove operation, there are no transactional constraints on
the ordering here.
Change the ordering of the operations in the xfs_remove() function
to align the ordering of AGI and AGF locking to match that of the
rest of the code.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
At xfs_iext_add(), if extent(s) are being appended to the last page in
the indirection array and the new extent(s) don't fit in the page, the
number of extents(erp->er_extcount) in a new allocated entry should be
the minimum value between count and XFS_LINEAR_EXTS, instead of count.
For now, there is no existing test case can demonstrates a problem with
the er_extcount being set incorrectly here, but it obviously like a bug.
Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Today, if xfs_sb_read_verify encounters a v4 superblock
with junk past v4 fields which includes data in sb_crc,
it will be treated as a failing checksum and a significant
corruption.
There are known prior bugs which leave junk at the end
of the V4 superblock; we don't need to actually fail the
verification in this case if other checks pan out ok.
So if this is a secondary superblock, and the primary
superblock doesn't indicate that this is a V5 filesystem,
don't treat this as an actual checksum failure.
We should probably check the garbage condition as
we do in xfs_repair, and possibly warn about it
or self-heal, but that's a different scope of work.
Stable folks: This can go back to v3.10, which is what
introduced the sb CRC checking that is tripped up by old,
stale, incorrect V4 superblocks w/ unzeroed bits.
Cc: stable@vger.kernel.org
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Acked-by: Dave Chinner <david@fromorbit.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
In xlog_verify_iclog a debug check of the incore log buffers prints an
error if icptr is null and then goes on to dereference the pointer
regardless. Convert this to an assert so that the intention is clear.
This was reported by Coverty.
Signed-off-by: Ben Myers <bpm@sgi.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
ASSERT on args takes place after args dereference.
This assertion is redundant since we are going to panic anyway.
Found by Linux Driver Verification project (linuxtesting.org) -
PVS-Studio analyzer.
Signed-off-by: Denis Efremov <yefremov.denis@gmail.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Page cache allocation doesn't always go through ->begin_write and
hence we don't always get the opportunity to set the allocation
context to GFP_NOFS. Failing to do this means we open up the direct
relcaim stack to recurse into the filesystem and consume a
significant amount of stack.
On RHEL6.4 kernels we are seeing ra_submit() and
generic_file_splice_read() from an nfsd context recursing into the
filesystem via the inode cache shrinker and evicting inodes. This is
causing truncation to be run (e.g EOF block freeing) and causing
bmap btree block merges and free space btree block splits to occur.
These btree manipulations are occurring with the call chain already
30 functions deep and hence there is not enough stack space to
complete such operations.
To avoid these specific overruns, we need to prevent the page cache
allocation from recursing via direct reclaim. We can do that because
the allocation functions take the allocation context from that which
is stored in the mapping for the inode. We don't set that right now,
so the default is GFP_HIGHUSER_MOVABLE, which is effectively a
GFP_KERNEL context. We need it to be the equivalent of GFP_NOFS, so
when we initialise an inode, set the mapping gfp mask appropriately.
This makes the use of AOP_FLAG_NOFS redundant from other parts of
the XFS IO path, so get rid of it.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
The kbuild test robot indicated that there were some new sparse
warnings in fs/xfs/xfs_dquot_buf.c. Actually, there were a lot more
that is wasn't warning about, so fix them all up.
Reported-by: kbuild test robot
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
The directory block format verifier fails to check that the leaf
entry count is in a valid range, and so if it is corrupted then it
can lead to derefencing a pointer outside the block buffer. While we
can't exactly validate the count without first walking the directory
block, we can ensure the count lands in the valid area within the
directory block and hence avoid out-of-block references.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Rather than hiding the ftype field size accounting inside the dirent
padding for the ".." and first entry offset functions for v2
directory formats, add explicit functions that calculate it
correctly.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Many of the vectorised function calls now take no parameters and
return a constant value. There is no reason for these to be vectored
functions, so convert them to constants
Binary sizes:
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
789005 96802 1096 886903 d8997 fs/xfs/xfs.o.p4
789061 96802 1096 886959 d88af fs/xfs/xfs.o.p5
789733 96802 1096 887631 d8b4f fs/xfs/xfs.o.p6
791421 96802 1096 889319 d91e7 fs/xfs/xfs.o.p7
791701 96802 1096 889599 d92ff fs/xfs/xfs.o.p8
791205 96802 1096 889103 d91cf fs/xfs/xfs.o.p9
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Next step in the vectorisation process is the directory free block
encode/decode operations. There are relatively few of these, though
there are quite a number of calls to them.
Binary sizes:
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
789005 96802 1096 886903 d8997 fs/xfs/xfs.o.p4
789061 96802 1096 886959 d88af fs/xfs/xfs.o.p5
789733 96802 1096 887631 d8b4f fs/xfs/xfs.o.p6
791421 96802 1096 889319 d91e7 fs/xfs/xfs.o.p7
791701 96802 1096 889599 d92ff fs/xfs/xfs.o.p8
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Conversion from on-disk structures to in-core header structures
currently relies on magic number checks. If the magic number is
wrong, but one of the supported values, we do the wrong thing with
the encode/decode operation. Split these functions so that there are
discrete operations for the specific directory format we are
handling.
In doing this, move all the header encode/decode functions to
xfs_da_format.c as they are directly manipulating the on-disk
format. It should be noted that all the growth in binary size is
from xfs_da_format.c - the rest of the code actaully shrinks.
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
789005 96802 1096 886903 d8997 fs/xfs/xfs.o.p4
789061 96802 1096 886959 d88af fs/xfs/xfs.o.p5
789733 96802 1096 887631 d8b4f fs/xfs/xfs.o.p6
791421 96802 1096 889319 d91e7 fs/xfs/xfs.o.p7
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
The remaining non-vectorised code for the directory structure is the
node format blocks. This is shared with the attribute tree, and so
is slightly more complex to vectorise.
Introduce a "non-directory" directory ops structure that is attached
to all non-directory inodes so that attribute operations can be
vectorised for all inodes.
Once we do this, we can vectorise all the da btree operations.
Because this patch adds more infrastructure than it removes the
binary size does not decrease:
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
789005 96802 1096 886903 d8997 fs/xfs/xfs.o.p4
789061 96802 1096 886959 d88af fs/xfs/xfs.o.p5
789733 96802 1096 887631 d8b4f fs/xfs/xfs.o.p6
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Next step in the vectorisation process is the leaf block
encode/decode operations. Most of the operations on leaves are
handled by the data block vectors, so there are relatively few of
them here.
Because of all the shuffling of code and having to pass more state
to some functions, this patch doesn't directly reduce the size of
the binary. It does open up many more opportunities for factoring
and optimisation, however.
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
789005 96802 1096 886903 d8997 fs/xfs/xfs.o.p4
789061 96802 1096 886959 d88af fs/xfs/xfs.o.p5
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Convert the rest of the directory data block encode/decode
operations to vector format.
This further reduces the size of the built binary:
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
789005 96802 1096 886903 d8997 fs/xfs/xfs.o.p4
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Following from the initial patches to vectorise the shortform
directory encode/decode operations, convert half the data block
operations to use the vector. The rest will be done in a second
patch.
This further reduces the size of the built binary:
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
789293 96802 1096 887191 d8997 fs/xfs/xfs.o.p3
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Following from the initial patch to introduce the directory
operations vector, convert the rest of the shortform directory
operations to use vectored ops rather than superblock feature
checks. This further reduces the size of the built binary:
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
792350 96802 1096 890248 d9588 fs/xfs/xfs.o.p2
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Lots of the dir code now goes through switches to determine what is
the correct on-disk format to parse. It generally involves a
"xfs_sbversion_hasfoo" check, deferencing the superblock version and
feature fields and hence touching several cache lines per operation
in the process. Some operations do multiple checks because they nest
conditional operations and they don't pass the information in a
direct fashion between each other.
Hence, add an ops vector to the xfs_inode structure that is
configured when the inode is initialised to point to all the correct
decode and encoding operations. This will significantly reduce the
branchiness and cacheline footprint of the directory object decoding
and encoding.
This is the first patch in a series of conversion patches. It will
introduce the ops structure, the setup of it and add the first
operation to the vector. Subsequent patches will convert directory
ops one at a time to keep the changes simple and obvious.
Just this patch shows the benefit of such an approach on code size.
Just converting the two shortform dir operations as this patch does
decreases the built binary size by ~1500 bytes:
$ size fs/xfs/xfs.o.orig fs/xfs/xfs.o.p1
text data bss dec hex filename
794490 96802 1096 892388 d9de4 fs/xfs/xfs.o.orig
792986 96802 1096 890884 d9804 fs/xfs/xfs.o.p1
$
That's a significant decrease in the instruction cache footprint of
the directory code for such a simple change, and indicates that this
approach is definitely worth pursuing further.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
xfs_rtalloc.c is partially shared with userspace. Split the file up
into two parts - one that is kernel private and the other which is
wholly shared with userspace.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Currently the xfs_inode.h header has a dependency on the definition
of the BMAP btree records as the inode fork includes an array of
xfs_bmbt_rec_host_t objects in it's definition.
Move all the btree format definitions from xfs_btree.h,
xfs_bmap_btree.h, xfs_alloc_btree.h and xfs_ialloc_btree.h to
xfs_format.h to continue the process of centralising the on-disk
format definitions. With this done, the xfs inode definitions are no
longer dependent on btree header files.
The enables a massive culling of unnecessary includes, with close to
200 #include directives removed from the XFS kernel code base.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
xfs_trans.h has a dependency on xfs_log.h for a couple of
structures. Most code that does transactions doesn't need to know
anything about the log, but this dependency means that they have to
include xfs_log.h. Decouple the xfs_trans.h and xfs_log.h header
files and clean up the includes to be in dependency order.
In doing this, remove the direct include of xfs_trans_reserve.h from
xfs_trans.h so that we remove the dependency between xfs_trans.h and
xfs_mount.h. Hence the xfs_trans.h include can be moved to the
indicate the actual dependencies other header files have on it.
Note that these are kernel only header files, so this does not
translate to any userspace changes at all.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
We don't do callbacks at transaction commit time, no do we have any
infrastructure to set up or run such callbacks, so remove the
variables and typedefs for these operations. If we ever need to add
callbacks, we can reintroduce the variables at that time.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Parts of userspace want to be able to read and modify dquot buffers
(e.g. xfs_db) so we need to split out the reading and writing of
these buffers so it is easy to shared code with libxfs in userspace.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
The on-disk format definitions for the directory and attribute
structures are spread across 3 header files right now, only one of
which is dedicated to defining on-disk structures and their
manipulation (xfs_dir2_format.h). Pull all the format definitions
into a single header file - xfs_da_format.h - and switch all the
code over to point at that.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
All of the buffer operations structures are needed to be exported
for xfs_db, so move them all to a common location rather than
spreading them all over the place. They are verifying the on-disk
format, so while xfs_format.h might be a good place, it is not part
of the on disk format.
Hence we need to create a new header file that we centralise these
related definitions. Start by moving the bffer operations
structures, and then also move all the other definitions that have
crept into xfs_log_format.h and xfs_format.h as there was no other
shared header file to put them in.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
Now that only one caller of xfs_change_file_space is left it can be merged
into said caller.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Call xfs_alloc_file_space or xfs_free_file_space directly from
xfs_file_fallocate instead of going through xfs_change_file_space.
This simplified the code by removing the unessecary marshalling of the
arguments into an xfs_flock64_t structure and allows removing checks that
are already done in the VFS code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Currently fallocate always holds the iolock when calling into
xfs_change_file_space, while the ioctl path lets some of the lower level
functions take it, but leave it out in others.
This patch makes sure the ioctl path also always holds the iolock and
thus introduces consistent locking for the preallocation operations while
simplifying the code and allowing to kill the now unused XFS_ATTR_NOLOCK
flag.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
There is no reason to conditionally take the iolock inside xfs_setattr_size
when we can let the caller handle it unconditionally, which just incrases
the lock hold time for the case where it was previously taken internally
by a few instructions.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
When xfs_growfs_data_private() is updating backup superblocks,
it bails out on the first error encountered, whether reading or
writing:
* If we get an error writing out the alternate superblocks,
* just issue a warning and continue. The real work is
* already done and committed.
This can cause a problem later during repair, because repair
looks at all superblocks, and picks the most prevalent one
as correct. If we bail out early in the backup superblock
loop, we can end up with more "bad" matching superblocks than
good, and a post-growfs repair may revert the filesystem to
the old geometry.
With the combination of superblock verifiers and old bugs,
we're more likely to encounter read errors due to verification.
And perhaps even worse, we don't even properly write any of the
newly-added superblocks in the new AGs.
Even with this change, growfs will still say:
xfs_growfs: XFS_IOC_FSGROWFSDATA xfsctl failed: Structure needs cleaning
data blocks changed from 319815680 to 335216640
which might be confusing to the user, but it at least communicates
that something has gone wrong, and dmesg will probably highlight
the need for an xfs_repair.
And this is still best-effort; if verifiers fail on more than
half the backup supers, they may still "win" - but that's probably
best left to repair to more gracefully handle by doing its own
strict verification as part of the backup super "voting."
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Acked-by: Dave Chinner <david@fromorbit.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
If we get EWRONGFS due to probing of non-xfs filesystems,
there's no need to issue the scary corruption error and backtrace.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
__xfs_printk adds its own "\n". Having it in the original string
leads to unintentional blank lines from these messages.
Most format strings have no newline, but a few do, leading to
i.e.:
[ 7347.119911] XFS (sdb2): Access to block zero in inode 132 start_block: 0 start_off: 0 blkcnt: 0 extent-state: 0 lastx: 1a05
[ 7347.119911]
[ 7347.119919] XFS (sdb2): Access to block zero in inode 132 start_block: 0 start_off: 0 blkcnt: 0 extent-state: 0 lastx: 1a05
[ 7347.119919]
Fix them all.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Recent analysis of a deadlocked XFS filesystem from a kernel
crash dump indicated that the filesystem was stuck waiting for log
space. The short story of the hang on the RHEL6 kernel is this:
- the tail of the log is pinned by an inode
- the inode has been pushed by the xfsaild
- the inode has been flushed to it's backing buffer and is
currently flush locked and hence waiting for backing
buffer IO to complete and remove it from the AIL
- the backing buffer is marked for write - it is on the
delayed write queue
- the inode buffer has been modified directly and logged
recently due to unlinked inode list modification
- the backing buffer is pinned in memory as it is in the
active CIL context.
- the xfsbufd won't start buffer writeback because it is
pinned
- xfssyncd won't force the log because it sees the log as
needing to be covered and hence wants to issue a dummy
transaction to move the log covering state machine along.
Hence there is no trigger to force the CIL to the log and hence
unpin the inode buffer and therefore complete the inode IO, remove
it from the AIL and hence move the tail of the log along, allowing
transactions to start again.
Mainline kernels also have the same deadlock, though the signature
is slightly different - the inode buffer never reaches the delayed
write lists because xfs_buf_item_push() sees that it is pinned and
hence never adds it to the delayed write list that the xfsaild
flushes.
There are two possible solutions here. The first is to simply force
the log before trying to cover the log and so ensure that the CIL is
emptied before we try to reserve space for the dummy transaction in
the xfs_log_worker(). While this might work most of the time, it is
still racy and is no guarantee that we don't get stuck in
xfs_trans_reserve waiting for log space to come free. Hence it's not
the best way to solve the problem.
The second solution is to modify xfs_log_need_covered() to be aware
of the CIL. We only should be attempting to cover the log if there
is no current activity in the log - covering the log is the process
of ensuring that the head and tail in the log on disk are identical
(i.e. the log is clean and at idle). Hence, by definition, if there
are items in the CIL then the log is not at idle and so we don't
need to attempt to cover it.
When we don't need to cover the log because it is active or idle, we
issue a log force from xfs_log_worker() - if the log is idle, then
this does nothing. However, if the log is active due to there being
items in the CIL, it will force the items in the CIL to the log and
unpin them.
In the case of the above deadlock scenario, instead of
xfs_log_worker() getting stuck in xfs_trans_reserve() attempting to
cover the log, it will instead force the log, thereby unpinning the
inode buffer, allowing IO to be issued and complete and hence
removing the inode that was pinning the tail of the log from the
AIL. At that point, everything will start moving along again. i.e.
the xfs_log_worker turns back into a watchdog that can alleviate
deadlocks based around pinned items that prevent the tail of the log
from being moved...
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
The xfs_inactive() return value is meaningless. Turn xfs_inactive()
into a void function and clean up the error handling appropriately.
Kill the VN_INACTIVE_[NO]CACHE directives as they are not relevant
to Linux.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Push the inode free work performed during xfs_inactive() down into
a new xfs_inactive_ifree() helper. This clears xfs_inactive() from
all inode locking and transaction management more directly
associated with freeing the inode xattrs, extents and the inode
itself.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Create the new xfs_inactive_truncate() function to handle the
truncate portion of xfs_inactive(). Push the locking and
transaction management into the new function.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Push down the transaction management for remote symlinks from
xfs_inactive() down to xfs_inactive_symlink_rmt(). The latter is
cleaned up to avoid transaction management intended for the
calling context (i.e., trans duplication, reservation, item
attachment).
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Add the inode type directory type support to XFS_IOC_FSGEOM
so that xfs_repair/xfs_info knows if the superblock v4 filesystem
enabled the feature.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
This fixes a build failure caused by calling the free() function which
does not exist in the Linux kernel.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit aaaae98022)
Free the memory in error path of xlog_recover_add_to_trans().
Normally this memory is freed in recovery pass2, but is leaked
in the error path.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit 519ccb81ac)
The determination of whether a directory entry contains a dtype
field originally was dependent on the filesystem having CRCs
enabled. This meant that the format for dtype beign enabled could be
determined by checking the directory block magic number rather than
doing a feature bit check. This was useful in that it meant that we
didn't need to pass a struct xfs_mount around to functions that
were already supplied with a directory block header.
Unfortunately, the introduction of dtype fields into the v4
structure via a feature bit meant this "use the directory block
magic number" method of discriminating the dirent entry sizes is
broken. Hence we need to convert the places that use magic number
checks to use feature bit checks so that they work correctly and not
by chance.
The current code works on v4 filesystems only because the dirent
size roundup covers the extra byte needed by the dtype field in the
places where this problem occurs.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit 367993e7c6)
Michael Semon reported that xfs/299 generated this lockdep warning:
=============================================
[ INFO: possible recursive locking detected ]
3.12.0-rc2+ #2 Not tainted
---------------------------------------------
touch/21072 is trying to acquire lock:
(&xfs_dquot_other_class){+.+...}, at: [<c12902fb>] xfs_trans_dqlockedjoin+0x57/0x64
but task is already holding lock:
(&xfs_dquot_other_class){+.+...}, at: [<c12902fb>] xfs_trans_dqlockedjoin+0x57/0x64
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&xfs_dquot_other_class);
lock(&xfs_dquot_other_class);
*** DEADLOCK ***
May be due to missing lock nesting notation
7 locks held by touch/21072:
#0: (sb_writers#10){++++.+}, at: [<c11185b6>] mnt_want_write+0x1e/0x3e
#1: (&type->i_mutex_dir_key#4){+.+.+.}, at: [<c11078ee>] do_last+0x245/0xe40
#2: (sb_internal#2){++++.+}, at: [<c122c9e0>] xfs_trans_alloc+0x1f/0x35
#3: (&(&ip->i_lock)->mr_lock/1){+.+...}, at: [<c126cd1b>] xfs_ilock+0x100/0x1f1
#4: (&(&ip->i_lock)->mr_lock){++++-.}, at: [<c126cf52>] xfs_ilock_nowait+0x105/0x22f
#5: (&dqp->q_qlock){+.+...}, at: [<c12902fb>] xfs_trans_dqlockedjoin+0x57/0x64
#6: (&xfs_dquot_other_class){+.+...}, at: [<c12902fb>] xfs_trans_dqlockedjoin+0x57/0x64
The lockdep annotation for dquot lock nesting only understands
locking for user and "other" dquots, not user, group and quota
dquots. Fix the annotations to match the locking heirarchy we now
have.
Reported-by: Michael L. Semon <mlsemon35@gmail.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit f112a04971)
XFS never calls mark_inode_bad or iget_failed, so it will never see a
bad inode. Remove all checks for is_bad_inode because they are
unnecessary.
Signed-off-by: Ben Myers <bpm@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
At xfs_iext_realloc_direct(), the new_size is changed by adding
if_bytes if originally the extent records are stored at the inline
extent buffer, and we have to switch from it to a direct extent
list for those new allocated extents, this is wrong. e.g,
Create a file with three extents which was showing as following,
xfs_io -f -c "truncate 100m" /xfs/testme
for i in $(seq 0 5 10); do
offset=$(($i * $((1 << 20))))
xfs_io -c "pwrite $offset 1m" /xfs/testme
done
Inline
------
irec: if_bytes bytes_diff new_size
1st 0 16 16
2nd 16 16 32
Switching
--------- rnew_size
3rd 32 16 48 + 32 = 80 roundup=128
In this case, the desired value of new_size should be 48, and then
it will be roundup to 64 and be assigned to rnew_size.
However, this issue has been covered by resetting the if_bytes to
the new_size which is calculated at the begnning of xfs_iext_add()
before leaving out this function, and in turn make the rnew_size
correctly again. Hence, this can not be detected via xfstestes.
This patch fix above problem and revise the new_size comments at
xfs_iext_realloc_direct() to make it more readable. Also, fix the
comments while switching from the inline extent buffer to a direct
extent list to reflect this change.
Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Linus Torvalds
Jan Kara
Gu Zheng
Dave Chinner
Jie Liu
Eric Sandeen
Geyslan G. Bem
Denis Efremov
Dave Chinner
Christoph Hellwig
Brian Foster
Mark Tinguely
Thierry Reding
Ben Myers