| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Use a more generic form for __section that requires quotes to avoid
complications with clang and gcc differences.
Remove the quote operator # from compiler_attributes.h __section macro.
Convert all unquoted __section(foo) uses to quoted __section("foo").
Also convert __attribute__((section("foo"))) uses to __section("foo")
even if the __attribute__ has multiple list entry forms.
Conversion done using the script at:
https://lore.kernel.org/lkml/75393e5ddc272dc7403de74d645e6c6e0f4e70eb.camel@perches.com/2-convert_section.pl
Signed-off-by: Joe Perches <joe@perches.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@gooogle.com>
Reviewed-by: Miguel Ojeda <ojeda@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull misc vfs updates from Al Viro:
"Assorted stuff all over the place (the largest group here is
Christoph's stat cleanups)"
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fs: remove KSTAT_QUERY_FLAGS
fs: remove vfs_stat_set_lookup_flags
fs: move vfs_fstatat out of line
fs: implement vfs_stat and vfs_lstat in terms of vfs_fstatat
fs: remove vfs_statx_fd
fs: omfs: use kmemdup() rather than kmalloc+memcpy
[PATCH] reduce boilerplate in fsid handling
fs: Remove duplicated flag O_NDELAY occurring twice in VALID_OPEN_FLAGS
selftests: mount: add nosymfollow tests
Add a "nosymfollow" mount option.
|
|
Pull xfs fixes from Darrick Wong:
"Two bug fixes that trickled in during the merge window:
- Make fallocate check the alignment of its arguments against the
fundamental allocation unit of the volume the file lives on, so
that we don't trigger the fs' alignment checks.
- Cancel unprocessed log intents immediately when log recovery fails,
to avoid a log deadlock"
* tag 'xfs-5.10-merge-7' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
xfs: cancel intents immediately if process_intents fails
xfs: fix fallocate functions when rtextsize is larger than 1
|
|
If processing recovered log intent items fails, we need to cancel all
the unprocessed recovered items immediately so that a subsequent AIL
push in the bail out path won't get wedged on the pinned intent items
that didn't get processed.
This can happen if the log contains (1) an intent that gets and releases
an inode, (2) an intent that cannot be recovered successfully, and (3)
some third intent item. When recovery of (2) fails, we leave (3) pinned
in memory. Inode reclamation is called in the error-out path of
xfs_mountfs before xfs_log_cancel_mount. Reclamation calls
xfs_ail_push_all_sync, which gets stuck waiting for (3).
Therefore, call xlog_recover_cancel_intents if _process_intents fails.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
In commit fe341eb151ec, I forgot that xfs_free_file_space isn't strictly
a "remove mapped blocks" function. It is actually a function to zero
file space by punching out the middle and writing zeroes to the
unaligned ends of the specified range. Therefore, putting a rtextsize
alignment check in that function is wrong because that breaks unaligned
ZERO_RANGE on the realtime volume.
Furthermore, xfs_file_fallocate already has alignment checks for the
functions require the file range to be aligned to the size of a
fundamental allocation unit (which is 1 FSB on the data volume and 1 rt
extent on the realtime volume). Create a new helper to check fallocate
arguments against the realtiem allocation unit size, fix the fallocate
frontend to use it, fix free_file_space to delete the correct range, and
remove a now redundant check from insert_file_space.
NOTE: The realtime extent size is not required to be a power of two!
Fixes: fe341eb151ec ("xfs: ensure that fpunch, fcollapse, and finsert operations are aligned to rt extent size")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
Pull more xfs updates from Darrick Wong:
"The second large pile of new stuff for 5.10, with changes even more
monumental than last week!
We are formally announcing the deprecation of the V4 filesystem format
in 2030. All users must upgrade to the V5 format, which contains
design improvements that greatly strengthen metadata validation,
supports reflink and online fsck, and is the intended vehicle for
handling timestamps past 2038. We're also deprecating the old Irix
behavioral tweaks in September 2025.
Coming along for the ride are two design changes to the deferred
metadata ops subsystem. One of the improvements is to retain correct
logical ordering of tasks and subtasks, which is a more logical design
for upper layers of XFS and will become necessary when we add atomic
file range swaps and commits. The second improvement to deferred ops
improves the scalability of the log by helping the log tail to move
forward during long-running operations. This reduces log contention
when there are a large number of threads trying to run transactions.
In addition to that, this fixes numerous small bugs in log recovery;
refactors logical intent log item recovery to remove the last
remaining place in XFS where we could have nested transactions; fixes
a couple of ways that intent log item recovery could fail in ways that
wouldn't have happened in the regular commit paths; fixes a deadlock
vector in the GETFSMAP implementation (which improves its performance
by 20%); and fixes serious bugs in the realtime growfs, fallocate, and
bitmap handling code.
Summary:
- Deprecate the V4 filesystem format, some disused mount options, and
some legacy sysctl knobs now that we can support dates into the
25th century. Note that removal of V4 support will not happen until
the early 2030s.
- Fix some probles with inode realtime flag propagation.
- Fix some buffer handling issues when growing a rt filesystem.
- Fix a problem where a BMAP_REMAP unmap call would free rt extents
even though the purpose of BMAP_REMAP is to avoid freeing the
blocks.
- Strengthen the dabtree online scrubber to check hash values on
child dabtree blocks.
- Actually log new intent items created as part of recovering log
intent items.
- Fix a bug where quotas weren't attached to an inode undergoing bmap
intent item recovery.
- Fix a buffer overrun problem with specially crafted log buffer
headers.
- Various cleanups to type usage and slightly inaccurate comments.
- More cleanups to the xattr, log, and quota code.
- Don't run the (slower) shared-rmap operations on attr fork
mappings.
- Fix a bug where we failed to check the LSN of finobt blocks during
replay and could therefore overwrite newer data with older data.
- Clean up the ugly nested transaction mess that log recovery uses to
stage intent item recovery in the correct order by creating a
proper data structure to capture recovered chains.
- Use the capture structure to resume intent item chains with the
same log space and block reservations as when they were captured.
- Fix a UAF bug in bmap intent item recovery where we failed to
maintain our reference to the incore inode if the bmap operation
needed to relog itself to continue.
- Rearrange the defer ops mechanism to finish newly created subtasks
of a parent task before moving on to the next parent task.
- Automatically relog intent items in deferred ops chains if doing so
would help us avoid pinning the log tail. This will help fix some
log scaling problems now and will facilitate atomic file updates
later.
- Fix a deadlock in the GETFSMAP implementation by using an internal
memory buffer to reduce indirect calls and copies to userspace,
thereby improving its performance by ~20%.
- Fix various problems when calling growfs on a realtime volume would
not fully update the filesystem metadata.
- Fix broken Kconfig asking about deprecated XFS when XFS is
disabled"
* tag 'xfs-5.10-merge-5' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (48 commits)
xfs: fix Kconfig asking about XFS_SUPPORT_V4 when XFS_FS=n
xfs: fix high key handling in the rt allocator's query_range function
xfs: annotate grabbing the realtime bitmap/summary locks in growfs
xfs: make xfs_growfs_rt update secondary superblocks
xfs: fix realtime bitmap/summary file truncation when growing rt volume
xfs: fix the indent in xfs_trans_mod_dquot
xfs: do the ASSERT for the arguments O_{u,g,p}dqpp
xfs: fix deadlock and streamline xfs_getfsmap performance
xfs: limit entries returned when counting fsmap records
xfs: only relog deferred intent items if free space in the log gets low
xfs: expose the log push threshold
xfs: periodically relog deferred intent items
xfs: change the order in which child and parent defer ops are finished
xfs: fix an incore inode UAF in xfs_bui_recover
xfs: clean up xfs_bui_item_recover iget/trans_alloc/ilock ordering
xfs: clean up bmap intent item recovery checking
xfs: xfs_defer_capture should absorb remaining transaction reservation
xfs: xfs_defer_capture should absorb remaining block reservations
xfs: proper replay of deferred ops queued during log recovery
xfs: remove XFS_LI_RECOVERED
...
|
|
Pavel Machek complained that the question about supporting deprecated
XFS v4 comes up even when XFS is disabled. This clearly makes no sense,
so fix Kconfig.
Reported-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
|
|
Fix some off-by-one errors in xfs_rtalloc_query_range. The highest key
in the realtime bitmap is always one less than the number of rt extents,
which means that the key clamp at the start of the function is wrong.
The 4th argument to xfs_rtfind_forw is the highest rt extent that we
want to probe, which means that passing 1 less than the high key is
wrong. Finally, drop the rem variable that controls the loop because we
can compare the iteration point (rtstart) against the high key directly.
The sordid history of this function is that the original commit (fb3c3)
incorrectly passed (high_rec->ar_startblock - 1) as the 'limit' parameter
to xfs_rtfind_forw. This was wrong because the "high key" is supposed
to be the largest key for which the caller wants result rows, not the
key for the first row that could possibly be outside the range that the
caller wants to see.
A subsequent attempt (8ad56) to strengthen the parameter checking added
incorrect clamping of the parameters to the number of rt blocks in the
system (despite the bitmap functions all taking units of rt extents) to
avoid querying ranges past the end of rt bitmap file but failed to fix
the incorrect _rtfind_forw parameter. The original _rtfind_forw
parameter error then survived the conversion of the startblock and
blockcount fields to rt extents (a0e5c), and the most recent off-by-one
fix (a3a37) thought it was patching a problem when the end of the rt
volume is not in use, but none of these fixes actually solved the
original problem that the author was confused about the "limit" argument
to xfs_rtfind_forw.
Sadly, all four of these patches were written by this author and even
his own usage of this function and rt testing were inadequate to get
this fixed quickly.
Original-problem: fb3c3de2f65c ("xfs: add a couple of queries to iterate free extents in the rtbitmap")
Not-fixed-by: 8ad560d2565e ("xfs: strengthen rtalloc query range checks")
Not-fixed-by: a0e5c435babd ("xfs: fix xfs_rtalloc_rec units")
Fixes: a3a374bf1889 ("xfs: fix off-by-one error in xfs_rtalloc_query_range")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
Pull xfs updates from Darrick Wong:
"The biggest changes are two new features for the ondisk metadata: one
to record the sizes of the inode btrees in the AG to increase
redundancy checks and to improve mount times; and a second new feature
to support timestamps until the year 2486.
We also fixed a problem where reflinking into a file that requires
synchronous writes wouldn't actually flush the updates to disk; clean
up a fair amount of cruft; and started fixing some bugs in the
realtime volume code.
Summary:
- Clean up the buffer ioend calling path so that the retry strategy
isn't quite so scattered everywhere.
- Clean up m_sb_bp handling.
- New feature: storing inode btree counts in the AGI to speed up
certain mount time per-AG block reservation operatoins and add a
little more metadata redundancy.
- New feature: Widen inode timestamps and quota grace expiration
timestamps to support dates through the year 2486.
- Get rid of more of our custom buffer allocation API wrappers.
- Use a proper VLA for shortform xattr structure namevals.
- Force the log after reflinking or deduping into a file that is
opened with O_SYNC or O_DSYNC.
- Fix some math errors in the realtime allocator"
* tag 'xfs-5.10-merge-2' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (42 commits)
xfs: ensure that fpunch, fcollapse, and finsert operations are aligned to rt extent size
xfs: make sure the rt allocator doesn't run off the end
xfs: Remove unneeded semicolon
xfs: force the log after remapping a synchronous-writes file
xfs: Convert xfs_attr_sf macros to inline functions
xfs: Use variable-size array for nameval in xfs_attr_sf_entry
xfs: Remove typedef xfs_attr_shortform_t
xfs: remove typedef xfs_attr_sf_entry_t
xfs: Remove kmem_zalloc_large()
xfs: enable big timestamps
xfs: trace timestamp limits
xfs: widen ondisk quota expiration timestamps to handle y2038+
xfs: widen ondisk inode timestamps to deal with y2038+
xfs: redefine xfs_ictimestamp_t
xfs: redefine xfs_timestamp_t
xfs: move xfs_log_dinode_to_disk to the log recovery code
xfs: refactor quota timestamp coding
xfs: refactor default quota grace period setting code
xfs: refactor quota expiration timer modification
xfs: explicitly define inode timestamp range
...
|
|
Use XFS_ILOCK_RT{BITMAP,SUM} to annotate grabbing the rt bitmap and
summary locks when we grow the realtime volume, just like we do most
everywhere else. This shuts up lockdep warnings about grabbing the
ILOCK class of locks recursively:
============================================
WARNING: possible recursive locking detected
5.9.0-rc4-djw #rc4 Tainted: G O
--------------------------------------------
xfs_growfs/4841 is trying to acquire lock:
ffff888035acc230 (&xfs_nondir_ilock_class){++++}-{3:3}, at: xfs_ilock+0xac/0x1a0 [xfs]
but task is already holding lock:
ffff888035acedb0 (&xfs_nondir_ilock_class){++++}-{3:3}, at: xfs_ilock+0xac/0x1a0 [xfs]
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&xfs_nondir_ilock_class);
lock(&xfs_nondir_ilock_class);
*** DEADLOCK ***
May be due to missing lock nesting notation
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
When we call growfs on the data device, we update the secondary
superblocks to reflect the updated filesystem geometry. We need to do
this for growfs on the realtime volume too, because a future xfs_repair
run could try to fix the filesystem using a backup superblock.
This was observed by the online superblock scrubbers while running
xfs/233. One can also trigger this by growing an rt volume, cycling the
mount, and creating new rt files.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
The realtime bitmap and summary files are regular files that are hidden
away from the directory tree. Since they're regular files, inode
inactivation will try to purge what it thinks are speculative
preallocations beyond the incore size of the file. Unfortunately,
xfs_growfs_rt forgets to update the incore size when it resizes the
inodes, with the result that inactivating the rt inodes at unmount time
will cause their contents to be truncated.
Fix this by updating the incore size when we change the ondisk size as
part of updating the superblock. Note that we don't do this when we're
allocating blocks to the rt inodes because we actually want those blocks
to get purged if the growfs fails.
This fixes corruption complaints from the online rtsummary checker when
running xfs/233. Since that test requires rmap, one can also trigger
this by growing an rt volume, cycling the mount, and creating rt files.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
The formatting is strange in xfs_trans_mod_dquot, so do a reindent.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
If we pass in XFS_QMOPT_{U,G,P}QUOTA flags and different uid/gid/prid
than them currently associated with the inode, the arguments
O_{u,g,p}dqpp shouldn't be NULL, so add the ASSERT for them.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
Refactor xfs_getfsmap to improve its performance: instead of indirectly
calling a function that copies one record to userspace at a time, create
a shadow buffer in the kernel and copy the whole array once at the end.
On the author's computer, this reduces the runtime on his /home by ~20%.
This also eliminates a deadlock when running GETFSMAP against the
realtime device. The current code locks the rtbitmap to create
fsmappings and copies them into userspace, having not released the
rtbitmap lock. If the userspace buffer is an mmap of a sparse file that
itself resides on the realtime device, the write page fault will recurse
into the fs for allocation, which will deadlock on the rtbitmap lock.
Fixes: 4c934c7dd60c ("xfs: report realtime space information via the rtbitmap")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
If userspace asked fsmap to count the number of entries, we cannot
return more than UINT_MAX entries because fmh_entries is u32.
Therefore, stop counting if we hit this limit or else we will waste time
to return truncated results.
Fixes: e89c041338ed ("xfs: implement the GETFSMAP ioctl")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
|
|
Now that we have the ability to ask the log how far the tail needs to be
pushed to maintain its free space targets, augment the decision to relog
an intent item so that we only do it if the log has hit the 75% full
threshold. There's no point in relogging an intent into the same
checkpoint, and there's no need to relog if there's plenty of free space
in the log.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
Separate the computation of the log push threshold and the push logic in
xlog_grant_push_ail. This enables higher level code to determine (for
example) that it is holding on to a logged intent item and the log is so
busy that it is more than 75% full. In that case, it would be desirable
to move the log item towards the head to release the tail, which we will
cover in the next patch.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
There's a subtle design flaw in the deferred log item code that can lead
to pinning the log tail. Taking up the defer ops chain examples from
the previous commit, we can get trapped in sequences like this:
Caller hands us a transaction t0 with D0-D3 attached. The defer ops
chain will look like the following if the transaction rolls succeed:
t1: D0(t0), D1(t0), D2(t0), D3(t0)
t2: d4(t1), d5(t1), D1(t0), D2(t0), D3(t0)
t3: d5(t1), D1(t0), D2(t0), D3(t0)
...
t9: d9(t7), D3(t0)
t10: D3(t0)
t11: d10(t10), d11(t10)
t12: d11(t10)
In transaction 9, we finish d9 and try to roll to t10 while holding onto
an intent item for D3 that we logged in t0.
The previous commit changed the order in which we place new defer ops in
the defer ops processing chain to reduce the maximum chain length. Now
make xfs_defer_finish_noroll capable of relogging the entire chain
periodically so that we can always move the log tail forward. Most
chains will never get relogged, except for operations that generate very
long chains (large extents containing many blocks with different sharing
levels) or are on filesystems with small logs and a lot of ongoing
metadata updates.
Callers are now required to ensure that the transaction reservation is
large enough to handle logging done items and new intent items for the
maximum possible chain length. Most callers are careful to keep the
chain lengths low, so the overhead should be minimal.
The decision to relog an intent item is made based on whether the intent
was logged in a previous checkpoint, since there's no point in relogging
an intent into the same checkpoint.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
The defer ops code has been finishing items in the wrong order -- if a
top level defer op creates items A and B, and finishing item A creates
more defer ops A1 and A2, we'll put the new items on the end of the
chain and process them in the order A B A1 A2. This is kind of weird,
since it's convenient for programmers to be able to think of A and B as
an ordered sequence where all the sub-tasks for A must finish before we
move on to B, e.g. A A1 A2 D.
Right now, our log intent items are not so complex that this matters,
but this will become important for the atomic extent swapping patchset.
In order to maintain correct reference counting of extents, we have to
unmap and remap extents in that order, and we want to complete that work
before moving on to the next range that the user wants to swap. This
patch fixes defer ops to satsify that requirement.
The primary symptom of the incorrect order was noticed in an early
performance analysis of the atomic extent swap code. An astonishingly
large number of deferred work items accumulated when userspace requested
an atomic update of two very fragmented files. The cause of this was
traced to the same ordering bug in the inner loop of
xfs_defer_finish_noroll.
If the ->finish_item method of a deferred operation queues new deferred
operations, those new deferred ops are appended to the tail of the
pending work list. To illustrate, say that a caller creates a
transaction t0 with four deferred operations D0-D3. The first thing
defer ops does is roll the transaction to t1, leaving us with:
t1: D0(t0), D1(t0), D2(t0), D3(t0)
Let's say that finishing each of D0-D3 will create two new deferred ops.
After finish D0 and roll, we'll have the following chain:
t2: D1(t0), D2(t0), D3(t0), d4(t1), d5(t1)
d4 and d5 were logged to t1. Notice that while we're about to start
work on D1, we haven't actually completed all the work implied by D0
being finished. So far we've been careful (or lucky) to structure the
dfops callers such that D1 doesn't depend on d4 or d5 being finished,
but this is a potential logic bomb.
There's a second problem lurking. Let's see what happens as we finish
D1-D3:
t3: D2(t0), D3(t0), d4(t1), d5(t1), d6(t2), d7(t2)
t4: D3(t0), d4(t1), d5(t1), d6(t2), d7(t2), d8(t3), d9(t3)
t5: d4(t1), d5(t1), d6(t2), d7(t2), d8(t3), d9(t3), d10(t4), d11(t4)
Let's say that d4-d11 are simple work items that don't queue any other
operations, which means that we can complete each d4 and roll to t6:
t6: d5(t1), d6(t2), d7(t2), d8(t3), d9(t3), d10(t4), d11(t4)
t7: d6(t2), d7(t2), d8(t3), d9(t3), d10(t4), d11(t4)
...
t11: d10(t4), d11(t4)
t12: d11(t4)
<done>
When we try to roll to transaction #12, we're holding defer op d11,
which we logged way back in t4. This means that the tail of the log is
pinned at t4. If the log is very small or there are a lot of other
threads updating metadata, this means that we might have wrapped the log
and cannot get roll to t11 because there isn't enough space left before
we'd run into t4.
Let's shift back to the original failure. I mentioned before that I
discovered this flaw while developing the atomic file update code. In
that scenario, we have a defer op (D0) that finds a range of file blocks
to remap, creates a handful of new defer ops to do that, and then asks
to be continued with however much work remains.
So, D0 is the original swapext deferred op. The first thing defer ops
does is rolls to t1:
t1: D0(t0)
We try to finish D0, logging d1 and d2 in the process, but can't get all
the work done. We log a done item and a new intent item for the work
that D0 still has to do, and roll to t2:
t2: D0'(t1), d1(t1), d2(t1)
We roll and try to finish D0', but still can't get all the work done, so
we log a done item and a new intent item for it, requeue D0 a second
time, and roll to t3:
t3: D0''(t2), d1(t1), d2(t1), d3(t2), d4(t2)
If it takes 48 more rolls to complete D0, then we'll finally dispense
with D0 in t50:
t50: D<fifty primes>(t49), d1(t1), ..., d102(t50)
We then try to roll again to get a chain like this:
t51: d1(t1), d2(t1), ..., d101(t50), d102(t50)
...
t152: d102(t50)
<done>
Notice that in rolling to transaction #51, we're holding on to a log
intent item for d1 that was logged in transaction #1. This means that
the tail of the log is pinned at t1. If the log is very small or there
are a lot of other threads updating metadata, this means that we might
have wrapped the log and cannot roll to t51 because there isn't enough
space left before we'd run into t1. This is of course problem #2 again.
But notice the third problem with this scenario: we have 102 defer ops
tied to this transaction! Each of these items are backed by pinned
kernel memory, which means that we risk OOM if the chains get too long.
Yikes. Problem #1 is a subtle logic bomb that could hit someone in the
future; problem #2 applies (rarely) to the current upstream, and problem
#3 applies to work under development.
This is not how incremental deferred operations were supposed to work.
The dfops design of logging in the same transaction an intent-done item
and a new intent item for the work remaining was to make it so that we
only have to juggle enough deferred work items to finish that one small
piece of work. Deferred log item recovery will find that first
unfinished work item and restart it, no matter how many other intent
items might follow it in the log. Therefore, it's ok to put the new
intents at the start of the dfops chain.
For the first example, the chains look like this:
t2: d4(t1), d5(t1), D1(t0), D2(t0), D3(t0)
t3: d5(t1), D1(t0), D2(t0), D3(t0)
...
t9: d9(t7), D3(t0)
t10: D3(t0)
t11: d10(t10), d11(t10)
t12: d11(t10)
For the second example, the chains look like this:
t1: D0(t0)
t2: d1(t1), d2(t1), D0'(t1)
t3: d2(t1), D0'(t1)
t4: D0'(t1)
t5: d1(t4), d2(t4), D0''(t4)
...
t148: D0<50 primes>(t147)
t149: d101(t148), d102(t148)
t150: d102(t148)
<done>
This actually sucks more for pinning the log tail (we try to roll to t10
while holding an intent item that was logged in t1) but we've solved
problem #1. We've also reduced the maximum chain length from:
sum(all the new items) + nr_original_items
to:
max(new items that each original item creates) + nr_original_items
This solves problem #3 by sharply reducing the number of defer ops that
can be attached to a transaction at any given time. The change makes
the problem of log tail pinning worse, but is improvement we need to
solve problem #2. Actually solving #2, however, is left to the next
patch.
Note that a subsequent analysis of some hard-to-trigger reflink and COW
livelocks on extremely fragmented filesystems (or systems running a lot
of IO threads) showed the same symptoms -- uncomfortably large numbers
of incore deferred work items and occasional stalls in the transaction
grant code while waiting for log reservations. I think this patch and
the next one will also solve these problems.
As originally written, the code used list_splice_tail_init instead of
list_splice_init, so change that, and leave a short comment explaining
our actions.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
In xfs_bui_item_recover, there exists a use-after-free bug with regards
to the inode that is involved in the bmap replay operation. If the
mapping operation does not complete, we call xfs_bmap_unmap_extent to
create a deferred op to finish the unmapping work, and we retain a
pointer to the incore inode.
Unfortunately, the very next thing we do is commit the transaction and
drop the inode. If reclaim tears down the inode before we try to finish
the defer ops, we dereference garbage and blow up. Therefore, create a
way to join inodes to the defer ops freezer so that we can maintain the
xfs_inode reference until we're done with the inode.
Note: This imposes the requirement that there be enough memory to keep
every incore inode in memory throughout recovery.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
In most places in XFS, we have a specific order in which we gather
resources: grab the inode, allocate a transaction, then lock the inode.
xfs_bui_item_recover doesn't do it in that order, so fix it to be more
consistent. This also makes the error bailout code a bit less weird.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
The bmap intent item checking code in xfs_bui_item_recover is spread all
over the function. We should check the recovered log item at the top
before we allocate any resources or do anything else, so do that.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
When xfs_defer_capture extracts the deferred ops and transaction state
from a transaction, it should record the transaction reservation type
from the old transaction so that when we continue the dfops chain, we
still use the same reservation parameters.
Doing this means that the log item recovery functions get to determine
the transaction reservation instead of abusing tr_itruncate in yet
another part of xfs.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
When xfs_defer_capture extracts the deferred ops and transaction state
from a transaction, it should record the remaining block reservations so
that when we continue the dfops chain, we can reserve the same number of
blocks to use. We capture the reservations for both data and realtime
volumes.
This adds the requirement that every log intent item recovery function
must be careful to reserve enough blocks to handle both itself and all
defer ops that it can queue. On the other hand, this enables us to do
away with the handwaving block estimation nonsense that was going on in
xlog_finish_defer_ops.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
When we replay unfinished intent items that have been recovered from the
log, it's possible that the replay will cause the creation of more
deferred work items. As outlined in commit 509955823cc9c ("xfs: log
recovery should replay deferred ops in order"), later work items have an
implicit ordering dependency on earlier work items. Therefore, recovery
must replay the items (both recovered and created) in the same order
that they would have been during normal operation.
For log recovery, we enforce this ordering by using an empty transaction
to collect deferred ops that get created in the process of recovering a
log intent item to prevent them from being committed before the rest of
the recovered intent items. After we finish committing all the
recovered log items, we allocate a transaction with an enormous block
reservation, splice our huge list of created deferred ops into that
transaction, and commit it, thereby finishing all those ops.
This is /really/ hokey -- it's the one place in XFS where we allow
nested transactions; the splicing of the defer ops list is is inelegant
and has to be done twice per recovery function; and the broken way we
handle inode pointers and block reservations cause subtle use-after-free
and allocator problems that will be fixed by this patch and the two
patches after it.
Therefore, replace the hokey empty transaction with a structure designed
to capture each chain of deferred ops that are created as part of
recovering a single unfinished log intent. Finally, refactor the loop
that replays those chains to do so using one transaction per chain.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
The ->iop_recover method of a log intent item removes the recovered
intent item from the AIL by logging an intent done item and committing
the transaction, so it's superfluous to have this flag check. Nothing
else uses it, so get rid of the flag entirely.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
Remove this one-line helper since the assert is trivially true in one
call site and the rest obscures a bitmask operation.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
Nathan popped up on #xfs and pointed out that we fail to handle
finobt btree blocks in xlog_recover_get_buf_lsn(). This means they
always fall through the entire magic number matching code to "recover
immediately". Whilst most of the time this is the correct behaviour,
occasionally it will be incorrect and could potentially overwrite
more recent metadata because we don't check the LSN in the on disk
metadata at all.
This bug has been present since the finobt was first introduced, and
is a potential cause of the occasional xfs_iget_check_free_state()
failures we see that indicate that the inode btree state does not
match the on disk inode state.
Fixes: aafc3c246529 ("xfs: support the XFS_BTNUM_FINOBT free inode btree type")
Reported-by: Nathan Scott <nathans@redhat.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
These optionr were for Irix compatibility, probably for clustered XFS
clients in a heterogenous cluster which contained both Irix & Linux
machines, so that behavior would be consistent. That doesn't exist anymore
and it's no longer needed.
Signed-off-by: Pavel Reichl <preichl@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
[darrick: actually state when the sysctls go away]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
ikeep/noikeep was a workaround for old DMAPI code which is no longer
relevant.
attr2/noattr2 - is for controlling upgrade behaviour from fixed attribute
fork sizes in the inode (attr1) and dynamic attribute fork sizes (attr2).
mkfs has defaulted to setting attr2 since 2007, hence just about every
XFS filesystem out there in production right now uses attr2.
Signed-off-by: Pavel Reichl <preichl@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
[darrick: fix minor typos]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
The current create and mkdir handlers both call the xfs_vn_mknod()
which is a wrapper routine around xfs_generic_create() function.
Actually the create and mkdir handlers can directly call
xfs_generic_create() function and reduce the call chain.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
During code review, I noticed that the rmap code uses the (slower)
shared mappings rmap functions for any extent of a reflinked file, even
if those extents are for the attr fork, which doesn't support sharing.
We can speed up rmap a tiny bit by optimizing out this case.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
Since commit 1c1c6ebcf52 ("xfs: Replace per-ag array with a radix
tree"), there is no m_peraglock anymore, so it's hard to understand
the described situation since per-ag is no longer an array and no
need to reallocate, call xfs_filestream_flush() in growfs.
In addition, the race condition for shrink feature is quite confusing
to me currently as well. Get rid of it instead.
Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
Cleanup the typedef usage, the unnecessary parentheses, the unnecessary
backslash and use the open-coded round_up call in
xfs_attr_leaf_entsize_{remote,local}.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
We should do the assert for all the log intent-done items if they appear
here. This patch detect intent-done items by the fact that their item ops
don't have iop_unpin and iop_push methods and also move the helper
xlog_item_is_intent to xfs_trans.h.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
Since we never use the second parameter id, so remove it from
xfs_qm_dqattach_one() function.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
We already check whether the crc feature is enabled before calling
xfs_attr3_rmt_verify(), so remove the redundant feature check in that
function.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
Fix the comments to help people understand the code.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
[darrick: fix the indenting problems too]
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
Since the type prid_t and xfs_dqid_t both are uint32_t, seems the
type cast is unnecessary, so remove it.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
We have already defined the project ID type prid_t, so maybe should
use it here.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
There are no callers of the SYNCHRONIZE() macro, so remove it.
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
Let's use DIV_ROUND_UP() to calculate log record header
blocks as what did in xlog_get_iclog_buffer_size() and
wrap up a common helper for log recovery.
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
|
Currently, crafted h_len has been blocked for the log
header of the tail block in commit a70f9fe52daa ("xfs:
detect and handle invalid iclog size set by mkfs").
However, each log record could still have crafted h_len
and cause log record buffer overrun. So let's check
h_len vs buffer size for each log record as well.
Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
Nowadays, log recovery will call ->release on the recovered intent items
if recovery fails. Therefore, it's redundant to release them from
inside the ->recover functions when they're about to return an error.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
In the bmap intent item recovery code, we must be careful to attach the
inode to its dquots (if quotas are enabled) so that a change in the
shape of the bmap btree doesn't cause the quota counters to be
incorrect.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
During a code inspection, I found a serious bug in the log intent item
recovery code when an intent item cannot complete all the work and
decides to requeue itself to get that done. When this happens, the
item recovery creates a new incore deferred op representing the
remaining work and attaches it to the transaction that it allocated. At
the end of _item_recover, it moves the entire chain of deferred ops to
the dummy parent_tp that xlog_recover_process_intents passed to it, but
fail to log a new intent item for the remaining work before committing
the transaction for the single unit of work.
xlog_finish_defer_ops logs those new intent items once recovery has
finished dealing with the intent items that it recovered, but this isn't
sufficient. If the log is forced to disk after a recovered log item
decides to requeue itself and the system goes down before we call
xlog_finish_defer_ops, the second log recovery will never see the new
intent item and therefore has no idea that there was more work to do.
It will finish recovery leaving the filesystem in a corrupted state.
The same logic applies to /any/ deferred ops added during intent item
recovery, not just the one handling the remaining work.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
When xchk_da_btree_block is loading a non-root dabtree block, we know
that the parent block had to have a (hashval, address) pointer to the
block that we just loaded. Check that the hashval in the parent matches
the block we just loaded.
This was found by fuzzing nbtree[3].hashval = ones in xfs/394.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
When callers pass XFS_BMAPI_REMAP into xfs_bunmapi, they want the extent
to be unmapped from the given file fork without the extent being freed.
We do this for non-rt files, but we forgot to do this for realtime
files. So far this isn't a big deal since nobody makes a bunmapi call
to a rt file with the REMAP flag set, but don't leave a logic bomb.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
In xfs_growfs_rt(), we enlarge bitmap and summary files by allocating
new blocks for both files. For each of the new blocks allocated, we
allocate an xfs_buf, zero the payload, log the contents and commit the
transaction. Hence these buffers will eventually find themselves
appended to list at xfs_ail->ail_buf_list.
Later, xfs_growfs_rt() loops across all of the new blocks belonging to
the bitmap inode to set the bitmap values to 1. In doing so, it
allocates a new transaction and invokes the following sequence of
functions,
- xfs_rtfree_range()
- xfs_rtmodify_range()
- xfs_rtbuf_get()
We pass '&xfs_rtbuf_ops' as the ops pointer to xfs_trans_read_buf().
- xfs_trans_read_buf()
We find the xfs_buf of interest in per-ag hash table, invoke
xfs_buf_reverify() which ends up assigning '&xfs_rtbuf_ops' to
xfs_buf->b_ops.
On the other hand, if xfs_growfs_rt_alloc() had allocated a few blocks
for the bitmap inode and returned with an error, all the xfs_bufs
corresponding to the new bitmap blocks that have been allocated would
continue to be on xfs_ail->ail_buf_list list without ever having a
non-NULL value assigned to their b_ops members. An AIL flush operation
would then trigger the following warning message to be printed on the
console,
XFS (loop0): _xfs_buf_ioapply: no buf ops on daddr 0x58 len 8
00000000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000040: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000050: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000060: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000070: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
CPU: 3 PID: 449 Comm: xfsaild/loop0 Not tainted 5.8.0-rc4-chandan-00038-g4d8c2b9de9ab-dirty #37
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
Call Trace:
dump_stack+0x57/0x70
_xfs_buf_ioapply+0x37c/0x3b0
? xfs_rw_bdev+0x1e0/0x1e0
? xfs_buf_delwri_submit_buffers+0xd4/0x210
__xfs_buf_submit+0x6d/0x1f0
xfs_buf_delwri_submit_buffers+0xd4/0x210
xfsaild+0x2c8/0x9e0
? __switch_to_asm+0x42/0x70
? xfs_trans_ail_cursor_first+0x80/0x80
kthread+0xfe/0x140
? kthread_park+0x90/0x90
ret_from_fork+0x22/0x30
This message indicates that the xfs_buf had its b_ops member set to
NULL.
This commit fixes the issue by assigning "&xfs_rtbuf_ops" to b_ops
member of each of the xfs_bufs logged by xfs_growfs_rt_alloc().
Signed-off-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
|
