Age | Commit message (Collapse) | Author |
|
Add a new error injection knob so that we can arbitrarily slow down
pagecahe writes to test for race conditions and aberrant reclaim
behavior if the writeback mechanisms are slow to issue writeback. This
will enable functional testing for the ifork sequence counters
introduced in commit XXXXXXXXXXXX that fixes write racing with reclaim
writeback.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Add a new error injection knob so that we can arbitrarily slow down
writeback to test for race conditions and aberrant reclaim behavior if
the writeback mechanisms are slow to issue writeback. This will enable
functional testing for the ifork sequence counters introduced in commit
745b3f76d1c8 ("xfs: maintain a sequence count for inode fork
manipulations").
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
In the previous patch, we transformed the iomap revalidation code to use
an explicit context object where data and cow fork sequence counters are
tracked explicitly. The existing validation function only validated the
data fork sequence counter, so now let's make it validate both.
I /think/ this isn't actually necessary here because we're writing to
the page cache, and the page state does not track or care about cow
status. However, this question came up when Dave and I were chatting
about this patchset on IRC, so here it is for formal consideration.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
To avoid racing with writeback, XFS needs to revalidate (io)mappings
after grabbing each folio lock. Right now, XFS stashes the sequence
counter values in the (void *pointer) field of the iomap, which is not
the greatest solution. First, we don't record which fork was sampled,
which means that the current code which samples either fork and compares
it to the data fork seqcount is wrong. Second, if another thread
touches the cow fork, we (conservatively) want to revalidate because
*something* has changed.
The previous three patches reorganized the iomap callbacks to pass the
iomap_iter to the ->iomap_{begin,end} functions and provided a way to
set the iter->private field for a buffered write. Now we can update the
buffered write paths in XFS to allocate the necessary memory to sample
both forks' sequence counters and revalidate the data fork during a
write operation.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Allow filesystems to pass a filesystem-private pointer into iomap for
writes into the pagecache. This will now be accessible from
->iomap_begin implementations, which is key to being able to revalidate
mappings after taking folio locks.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Clean up the ->iomap_end call sites by passing a pointer to the iter
structure into the iomap_end functions. This isn't strictly needed,
but it cleans up the callsites neatly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Clean up the ->iomap_begin call sites by passing a pointer to the iter
structure into the iomap_begin functions. This will be needed to clean
up the xfs race condition fixes in the next patch, and will hopefully
reduce register pressure as well.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
With the changes to scan the page cache for dirty data to avoid data
corruptions from partial write cleanup racing with other page cache
operations, the drop writes error injection no longer works the same
way it used to and causes xfs/196 to fail. This is because xfs/196
writes to the file and populates the page cache before it turns on
the error injection and starts failing -overwrites-.
The result is that the original drop-writes code failed writes only
-after- overwriting the data in the cache, followed by invalidates
the cached data, then punching out the delalloc extent from under
that data.
On the surface, this looks fine. The problem is that page cache
invalidation *doesn't guarantee that it removes anything from the
page cache* and it doesn't change the dirty state of the folio. When
block size == page size and we do page aligned IO (as xfs/196 does)
everything happens to align perfectly and page cache invalidation
removes the single page folios that span the written data. Hence the
followup delalloc punch pass does not find cached data over that
range and it can punch the extent out.
IOWs, xfs/196 "works" for block size == page size with the new
code. I say "works", because it actually only works for the case
where IO is page aligned, and no data was read from disk before
writes occur. Because the moment we actually read data first, the
readahead code allocates multipage folios and suddenly the
invalidate code goes back to zeroing subfolio ranges without
changing dirty state.
Hence, with multipage folios in play, block size == page size is
functionally identical to block size < page size behaviour, and
drop-writes is manifestly broken w.r.t to this case. Invalidation of
a subfolio range doesn't result in the folio being removed from the
cache, just the range gets zeroed. Hence after we've sequentially
walked over a folio that we've dirtied (via write data) and then
invalidated, we end up with a dirty folio full of zeroed data.
And because the new code skips punching ranges that have dirty
folios covering them, we end up leaving the delalloc range intact
after failing all the writes. Hence failed writes now end up
writing zeroes to disk in the cases where invalidation zeroes folios
rather than removing them from cache.
This is a fundamental change of behaviour that is needed to avoid
the data corruption vectors that exist in the old write fail path,
and it renders the drop-writes injection non-functional and
unworkable as it stands.
As it is, I think the error injection is also now unnecessary, as
partial writes that need delalloc extent are going to be a lot more
common with stale iomap detection in place. Hence this patch removes
the drop-writes error injection completely. xfs/196 can remain for
testing kernels that don't have this data corruption fix, but those
that do will report:
xfs/196 3s ... [not run] XFS error injection drop_writes unknown on this kernel.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Now that iomap supports a mechanism to validate cached iomaps for
buffered write operations, hook it up to the XFS buffered write ops
so that we can avoid data corruptions that result from stale cached
iomaps. See:
https://lore.kernel.org/linux-xfs/20220817093627.GZ3600936@dread.disaster.area/
or the ->iomap_valid() introduction commit for exact details of the
corruption vector.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
A recent multithreaded write data corruption has been uncovered in
the iomap write code. The core of the problem is partial folio
writes can be flushed to disk while a new racing write can map it
and fill the rest of the page:
writeback new write
allocate blocks
blocks are unwritten
submit IO
.....
map blocks
iomap indicates UNWRITTEN range
loop {
lock folio
copyin data
.....
IO completes
runs unwritten extent conv
blocks are marked written
<iomap now stale>
get next folio
}
Now add memory pressure such that memory reclaim evicts the
partially written folio that has already been written to disk.
When the new write finally gets to the last partial page of the new
write, it does not find it in cache, so it instantiates a new page,
sees the iomap is unwritten, and zeros the part of the page that
it does not have data from. This overwrites the data on disk that
was originally written.
The full description of the corruption mechanism can be found here:
https://lore.kernel.org/linux-xfs/20220817093627.GZ3600936@dread.disaster.area/
To solve this problem, we need to check whether the iomap is still
valid after we lock each folio during the write. We have to do it
after we lock the page so that we don't end up with state changes
occurring while we wait for the folio to be locked.
Hence we need a mechanism to be able to check that the cached iomap
is still valid (similar to what we already do in buffered
writeback), and we need a way for ->begin_write to back out and
tell the high level iomap iterator that we need to remap the
remaining write range.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
xfs_buffered_write_iomap_end() currently invalidates the page cache
over the unused range of the delalloc extent it allocated. While the
write allocated the delalloc extent, it does not own it exclusively
as the write does not hold any locks that prevent either writeback
or mmap page faults from changing the state of either the page cache
or the extent state backing this range.
Whilst xfs_bmap_punch_delalloc_range() already handles races in
extent conversion - it will only punch out delalloc extents and it
ignores any other type of extent - the page cache truncate does not
discriminate between data written by this write or some other task.
As a result, truncating the page cache can result in data corruption
if the write races with mmap modifications to the file over the same
range.
generic/346 exercises this workload, and if we randomly fail writes
(as will happen when iomap gets stale iomap detection later in the
patchset), it will randomly corrupt the file data because it removes
data written by mmap() in the same page as the write() that failed.
Hence we do not want to punch out the page cache over the range of
the extent we failed to write to - what we actually need to do is
detect the ranges that have dirty data in cache over them and *not
punch them out*.
TO do this, we have to walk the page cache over the range of the
delalloc extent we want to remove. This is made complex by the fact
we have to handle partially up-to-date folios correctly and this can
happen even when the FSB size == PAGE_SIZE because we now support
multi-page folios in the page cache.
Because we are only interested in discovering the edges of data
ranges in the page cache (i.e. hole-data boundaries) we can make use
of mapping_seek_hole_data() to find those transitions in the page
cache. As we hold the invalidate_lock, we know that the boundaries
are not going to change while we walk the range. This interface is
also byte-based and is sub-page block aware, so we can find the data
ranges in the cache based on byte offsets rather than page, folio or
fs block sized chunks. This greatly simplifies the logic of finding
dirty cached ranges in the page cache.
Once we've identified a range that contains cached data, we can then
iterate the range folio by folio. This allows us to determine if the
data is dirty and hence perform the correct delalloc extent punching
operations. The seek interface we use to iterate data ranges will
give us sub-folio start/end granularity, so we may end up looking up
the same folio multiple times as the seek interface iterates across
each discontiguous data region in the folio.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
xfs_buffered_write_iomap_end() currently converts the byte ranges
passed to it to filesystem blocks to pass them to the bmap code to
punch out delalloc blocks, but then has to convert filesytem
blocks back to byte ranges for page cache truncate.
We're about to make the page cache truncate go away and replace it
with a page cache walk, so having to convert everything to/from/to
filesystem blocks is messy and error-prone. It is much easier to
pass around byte ranges and convert to page indexes and/or
filesystem blocks only where those units are needed.
In preparation for the page cache walk being added, add a helper
that converts byte ranges to filesystem blocks and calls
xfs_bmap_punch_delalloc_range() and convert
xfs_buffered_write_iomap_end() to calculate limits in byte ranges.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
xfs_buffered_write_iomap_end() has a comment about the safety of
punching delalloc extents based holding the IOLOCK_EXCL. This
comment is wrong, and punching delalloc extents is not race free.
When we punch out a delalloc extent after a write failure in
xfs_buffered_write_iomap_end(), we punch out the page cache with
truncate_pagecache_range() before we punch out the delalloc extents.
At this point, we only hold the IOLOCK_EXCL, so there is nothing
stopping mmap() write faults racing with this cleanup operation,
reinstantiating a folio over the range we are about to punch and
hence requiring the delalloc extent to be kept.
If this race condition is hit, we can end up with a dirty page in
the page cache that has no delalloc extent or space reservation
backing it. This leads to bad things happening at writeback time.
To avoid this race condition, we need the page cache truncation to
be atomic w.r.t. the extent manipulation. We can do this by holding
the mapping->invalidate_lock exclusively across this operation -
this will prevent new pages from being inserted into the page cache
whilst we are removing the pages and the backing extent and space
reservation.
Taking the mapping->invalidate_lock exclusively in the buffered
write IO path is safe - it naturally nests inside the IOLOCK (see
truncate and fallocate paths). iomap_zero_range() can be called from
under the mapping->invalidate_lock (from the truncate path via
either xfs_zero_eof() or xfs_truncate_page(), but iomap_zero_iter()
will not instantiate new delalloc pages (because it skips holes) and
hence will not ever need to punch out delalloc extents on failure.
Fix the locking issue, and clean up the code logic a little to avoid
unnecessary work if we didn't allocate the delalloc extent or wrote
the entire region we allocated.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
When we reserve a delalloc region in xfs_buffered_write_iomap_begin,
we mark the iomap as IOMAP_F_NEW so that the the write context
understands that it allocated the delalloc region.
If we then fail that buffered write, xfs_buffered_write_iomap_end()
checks for the IOMAP_F_NEW flag and if it is set, it punches out
the unused delalloc region that was allocated for the write.
The assumption this code makes is that all buffered write operations
that can allocate space are run under an exclusive lock (i_rwsem).
This is an invalid assumption: page faults in mmap()d regions call
through this same function pair to map the file range being faulted
and this runs only holding the inode->i_mapping->invalidate_lock in
shared mode.
IOWs, we can have races between page faults and write() calls that
fail the nested page cache write operation that result in data loss.
That is, the failing iomap_end call will punch out the data that
the other racing iomap iteration brought into the page cache. This
can be reproduced with generic/34[46] if we arbitrarily fail page
cache copy-in operations from write() syscalls.
Code analysis tells us that the iomap_page_mkwrite() function holds
the already instantiated and uptodate folio locked across the iomap
mapping iterations. Hence the folio cannot be removed from memory
whilst we are mapping the range it covers, and as such we do not
care if the mapping changes state underneath the iomap iteration
loop:
1. if the folio is not already dirty, there is no writeback races
possible.
2. if we allocated the mapping (delalloc or unwritten), the folio
cannot already be dirty. See #1.
3. If the folio is already dirty, it must be up to date. As we hold
it locked, it cannot be reclaimed from memory. Hence we always
have valid data in the page cache while iterating the mapping.
4. Valid data in the page cache can exist when the underlying
mapping is DELALLOC, UNWRITTEN or WRITTEN. Having the mapping
change from DELALLOC->UNWRITTEN or UNWRITTEN->WRITTEN does not
change the data in the page - it only affects actions if we are
initialising a new page. Hence #3 applies and we don't care
about these extent map transitions racing with
iomap_page_mkwrite().
5. iomap_page_mkwrite() checks for page invalidation races
(truncate, hole punch, etc) after it locks the folio. We also
hold the mapping->invalidation_lock here, and hence the mapping
cannot change due to extent removal operations while we are
iterating the folio.
As such, filesystems that don't use bufferheads will never fail
the iomap_folio_mkwrite_iter() operation on the current mapping,
regardless of whether the iomap should be considered stale.
Further, the range we are asked to iterate is limited to the range
inside EOF that the folio spans. Hence, for XFS, we will only map
the exact range we are asked for, and we will only do speculative
preallocation with delalloc if we are mapping a hole at the EOF
page. The iterator will consume the entire range of the folio that
is within EOF, and anything beyond the EOF block cannot be accessed.
We never need to truncate this post-EOF speculative prealloc away in
the context of the iomap_page_mkwrite() iterator because if it
remains unused we'll remove it when the last reference to the inode
goes away.
Hence we don't actually need an .iomap_end() cleanup/error handling
path at all for iomap_page_mkwrite() for XFS. This means we can
separate the page fault processing from the complexity of the
.iomap_end() processing in the buffered write path. This also means
that the buffered write path will also be able to take the
mapping->invalidate_lock as necessary.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
If we tried to repair something but the repair failed with -EDEADLOCK,
that means that the repair function couldn't grab some resource it
needed and wants us to try again. If we try again (with TRY_HARDER) but
still can't get all the resources we need, the repair fails and errors
remain on the filesystem.
Right now, repair returns the -EDEADLOCK to the caller as -EFSCORRUPTED,
which results in XFS_SCRUB_OFLAG_CORRUPT being passed out to userspace.
This is not correct because repair has not determined that anything is
corrupt. If the repair had been invoked on an object that could be
optimized but wasn't corrupt (OFLAG_PREEN), the inability to grab
resources will be reported to userspace as corrupt metadata, and users
will be unnecessarily alarmed that their suboptimal metadata turned into
a corruption.
Fix this by returning zero so that the results of the actual scrub will
be copied back out to userspace.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Don't transform the logical block size to a bit shift only to shift it
back to the original block size. Just use the size.
Cc: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
|
|
As of now only device names are printed out over __xfs_printk().
The device names are not persistent across reboots which in case
of searching for origin of corruption brings another task to properly
identify the devices. This patch add XFS UUID upon every mount/umount
event which will make the identification much easier.
Signed-off-by: Lukas Herbolt <lukas@herbolt.com>
[sandeen: rebase onto current upstream kernel]
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Clean up resources if resetting the dotdot entry doesn't succeed.
Observed through code inspection.
Fixes: 5838d0356bb3 ("xfs: reset child dir '..' entry when unlinking child")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Andrey Albershteyn <aalbersh@redhat.com>
|
|
CoW forks only exist in memory, which means that they can only ever have
an incore extent tree. Hence they must always be FMT_EXTENTS, so check
this when we're scrubbing them.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Ensure that extents in an inode's CoW fork are not marked as shared in
the refcount btree.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Teach scrub to flag quota files containing unwritten extents.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Enhance the block map scrubber to check delayed allocation reservations.
Though there are no physical space allocations to check, we do need to
make sure that the range of file offsets being mapped are correct, and
to bump the lastoff cursor so that key order checking works correctly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
When scrub is checking file fork mappings against rmap records and
the rmap record starts before or ends after the bmap record, check the
adjacent bmap records to make sure that they're adjacent to the one
we're checking. This helps us to detect cases where the rmaps cover
territory that the bmaps do not.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
sparse complains that we can return an uninitialized error from this
function and that pag could be uninitialized. We know that there are no
zero-AG filesystems and hence we had to call xchk_bmap_check_ag_rmaps at
least once, so this is not actually possible, but I'm too worn out from
automated complaints from unsophisticated AIs so let's just fix this and
move on to more interesting problems, eh?
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Teach the summary count checker to count the number of free realtime
extents and compare that to the superblock copy.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
If any part of the per-AG summary counter scan loop aborts without
collecting all of the data we need, the scrubber's observation data will
be invalid. Set the incomplete flag so that we abort the scrub without
reporting false corruptions. Document the data dependency here too.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
xfs_rtalloc_query_range scans the realtime bitmap file in order of
increasing file offset, so this caller can take ILOCK_SHARED on the rt
bitmap inode instead of ILOCK_EXCL. This isn't going to yield any
practical benefits at mount time, but we'd like to make the locking
usage consistent around xfs_rtalloc_query_all calls. Make all the
places we do this use the same xfs_ilock lockflags for consistency.
Fixes: 4c934c7dd60c ("xfs: report realtime space information via the rtbitmap")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
It turns out that GETFSMAP and online fsck have had a bug for years due
to their use of ILOCK_SHARED to coordinate their linear scans of the
realtime bitmap. If the bitmap file's data fork happens to be in BTREE
format and the scan occurs immediately after mounting, the incore bmbt
will not be populated, leading to ASSERTs tripping over the incorrect
inode state. Because the bitmap scans always lock bitmap buffers in
increasing order of file offset, it is appropriate for these two callers
to take a shared ILOCK to improve scalability.
To fix this problem, load both data and attr fork state into memory when
mounting the realtime inodes. Realtime metadata files aren't supposed
to have an attr fork so the second step is likely a nop.
On most filesystems this is unlikely since the rtbitmap data fork is
usually in extents format, but it's possible to craft a filesystem that
will by fragmenting the free space in the data section and growfsing the
rt section.
Fixes: 4c934c7dd60c ("xfs: report realtime space information via the rtbitmap")
Also-Fixes: 46d9bfb5e706 ("xfs: cross-reference the realtime bitmap")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Repair functions will not return EAGAIN -- if they were not able to
obtain resources, they should return EDEADLOCK (like the rest of online
fsck) to signal that we need to grab all the resources and try again.
Hence we don't need to deal with this case except as a debugging
assertion.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
If the scrub process is sent a fatal signal while we're checking dquots,
the predicate for this will set the error code to -EINTR. Don't then
squash that into -ECANCELED, because the wrong errno turns up in the
trace output.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
If the program calling online fsck is terminated with a fatal signal,
bail out to userspace by returning EINTR, not EAGAIN. EAGAIN is used by
scrubbers to indicate that we should try again with more resources
locked, and not to indicate that the operation was cancelled. The
miswiring is mostly harmless, but it shows up in the trace data.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Convert all the online scrub code to use the Linux slab allocator
functions directly instead of going through the kmem wrappers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Initialize the check_owner list head so that we don't corrupt the list.
Reduce the scope of the object pointer.
Fixes: 858333dcf021 ("xfs: check btree block ownership with bnobt/rmapbt when scrubbing btree")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Memory allocation usage is the same throughout online fsck -- we want
kernel memory, we have to be able to back out if we can't allocate
memory, and we don't want to spray dmesg with memory allocation failure
reports. Standardize the GFP flag usage and document these requirements.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Teach the AGFL repair function to check each block of the proposed AGFL
against the rmap btree. If the rmapbt finds any mappings that are not
OWN_AG, strike that block from the list.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Currently, the only way to lock an allocation group is to hold the AGI
and AGF buffers. If a repair needs to roll the transaction while
repairing some AG metadata, it maintains that lock by holding the two
buffers across the transaction roll and joins them afterwards.
However, repair is not like other parts of XFS that employ the bhold -
roll - bjoin sequence because it's possible that the AGI or AGF buffers
are not actually dirty before the roll. This presents two problems --
First, we need to redirty those buffers to keep them moving along in the
log to avoid pinning the log tail. Second, a clean buffer log item can
detach from the buffer. If this happens, the buffer type state is
discarded along with the bli and must be reattached before the next time
the buffer is logged. If it is not, the logging code will complain and
log recovery will not work properly.
An earlier version of this patch tried to fix the second problem by
re-setting the buffer type in the bli after joining the buffer to the
new transaction, but that looked weird and didn't solve the first
problem. Instead, solve both problems by logging the buffer before
rolling the transaction.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
While scrubbing an allocation group, we don't need to hold the AGFL
buffer as part of the scrub context. All that is necessary to lock an
AG is to hold the AGI and AGF buffers, so fix all the existing users of
the AGFL buffer to grab them only when necessary.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
While running the online fsck test suite, I noticed the following
assertion in the kernel log (edited for brevity):
XFS: Assertion failed: 0, file: fs/xfs/xfs_health.c, line: 571
------------[ cut here ]------------
WARNING: CPU: 3 PID: 11667 at fs/xfs/xfs_message.c:104 assfail+0x46/0x4a [xfs]
CPU: 3 PID: 11667 Comm: xfs_scrub Tainted: G W 5.19.0-rc7-xfsx #rc7 6e6475eb29fd9dda3181f81b7ca7ff961d277a40
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.15.0-1 04/01/2014
RIP: 0010:assfail+0x46/0x4a [xfs]
Call Trace:
<TASK>
xfs_dir2_isblock+0xcc/0xe0
xchk_directory_blocks+0xc7/0x420
xchk_directory+0x53/0xb0
xfs_scrub_metadata+0x2b6/0x6b0
xfs_scrubv_metadata+0x35e/0x4d0
xfs_ioc_scrubv_metadata+0x111/0x160
xfs_file_ioctl+0x4ec/0xef0
__x64_sys_ioctl+0x82/0xa0
do_syscall_64+0x2b/0x80
entry_SYSCALL_64_after_hwframe+0x46/0xb0
This assertion triggers in xfs_dirattr_mark_sick when the caller passes
in a whichfork value that is neither of XFS_{DATA,ATTR}_FORK. The cause
of this is that xchk_directory_blocks only partially initializes the
xfs_da_args structure that is passed to xfs_dir2_isblock. If the data
fork is not correct, the XFS_IS_CORRUPT clause will trigger. My
development branch reports this failure to the health monitoring
subsystem, which accesses the uninitialized args->whichfork field,
leading the the assertion tripping. We really shouldn't be passing
random stack contents around, so the solution here is to force the
compiler to zero-initialize the struct.
Found by fuzzing u3.bmx[0].blockcount = middlebit on xfs/1554.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
When lazysbcount is enabled, fsstress and loop mount/unmount test report
the following problems:
XFS (loop0): SB summary counter sanity check failed
XFS (loop0): Metadata corruption detected at xfs_sb_write_verify+0x13b/0x460,
xfs_sb block 0x0
XFS (loop0): Unmount and run xfs_repair
XFS (loop0): First 128 bytes of corrupted metadata buffer:
00000000: 58 46 53 42 00 00 10 00 00 00 00 00 00 28 00 00 XFSB.........(..
00000010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000020: 69 fb 7c cd 5f dc 44 af 85 74 e0 cc d4 e3 34 5a i.|._.D..t....4Z
00000030: 00 00 00 00 00 20 00 06 00 00 00 00 00 00 00 80 ..... ..........
00000040: 00 00 00 00 00 00 00 81 00 00 00 00 00 00 00 82 ................
00000050: 00 00 00 01 00 0a 00 00 00 00 00 04 00 00 00 00 ................
00000060: 00 00 0a 00 b4 b5 02 00 02 00 00 08 00 00 00 00 ................
00000070: 00 00 00 00 00 00 00 00 0c 09 09 03 14 00 00 19 ................
XFS (loop0): Corruption of in-memory data (0x8) detected at _xfs_buf_ioapply
+0xe1e/0x10e0 (fs/xfs/xfs_buf.c:1580). Shutting down filesystem.
XFS (loop0): Please unmount the filesystem and rectify the problem(s)
XFS (loop0): log mount/recovery failed: error -117
XFS (loop0): log mount failed
This corruption will shutdown the file system and the file system will
no longer be mountable. The following script can reproduce the problem,
but it may take a long time.
#!/bin/bash
device=/dev/sda
testdir=/mnt/test
round=0
function fail()
{
echo "$*"
exit 1
}
mkdir -p $testdir
while [ $round -lt 10000 ]
do
echo "******* round $round ********"
mkfs.xfs -f $device
mount $device $testdir || fail "mount failed!"
fsstress -d $testdir -l 0 -n 10000 -p 4 >/dev/null &
sleep 4
killall -w fsstress
umount $testdir
xfs_repair -e $device > /dev/null
if [ $? -eq 2 ];then
echo "ERR CODE 2: Dirty log exception during repair."
exit 1
fi
round=$(($round+1))
done
With lazysbcount is enabled, There is no additional lock protection for
reading m_ifree and m_icount in xfs_log_sb(), if other cpu modifies the
m_ifree, this will make the m_ifree greater than m_icount. For example,
consider the following sequence and ifreedelta is postive:
CPU0 CPU1
xfs_log_sb xfs_trans_unreserve_and_mod_sb
---------- ------------------------------
percpu_counter_sum(&mp->m_icount)
percpu_counter_add_batch(&mp->m_icount,
idelta, XFS_ICOUNT_BATCH)
percpu_counter_add(&mp->m_ifree, ifreedelta);
percpu_counter_sum(&mp->m_ifree)
After this, incorrect inode count (sb_ifree > sb_icount) will be writen to
the log. In the subsequent writing of sb, incorrect inode count (sb_ifree >
sb_icount) will fail to pass the boundary check in xfs_validate_sb_write()
that cause the file system shutdown.
When lazysbcount is enabled, we don't need to guarantee that Lazy sb
counters are completely correct, but we do need to guarantee that sb_ifree
<= sb_icount. On the other hand, the constraint that m_ifree <= m_icount
must be satisfied any time that there /cannot/ be other threads allocating
or freeing inode chunks. If the constraint is violated under these
circumstances, sb_i{count,free} (the ondisk superblock inode counters)
maybe incorrect and need to be marked sick at unmount, the count will
be rebuilt on the next mount.
Fixes: 8756a5af1819 ("libxfs: add more bounds checking to sb sanity checks")
Signed-off-by: Long Li <leo.lilong@huawei.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl
Pull cxl fixes from Dan Williams:
"Several fixes for CXL region creation crashes, leaks and failures.
This is mainly fallout from the original implementation of dynamic CXL
region creation (instantiate new physical memory pools) that arrived
in v6.0-rc1.
Given the theme of "failures in the presence of pass-through decoders"
this also includes new regression test infrastructure for that case.
Summary:
- Fix region creation crash with pass-through decoders
- Fix region creation crash when no decoder allocation fails
- Fix region creation crash when scanning regions to enforce the
increasing physical address order constraint that CXL mandates
- Fix a memory leak for cxl_pmem_region objects, track 1:N instead of
1:1 memory-device-to-region associations.
- Fix a memory leak for cxl_region objects when regions with active
targets are deleted
- Fix assignment of NUMA nodes to CXL regions by CFMWS (CXL Window)
emulated proximity domains.
- Fix region creation failure for switch attached devices downstream
of a single-port host-bridge
- Fix false positive memory leak of cxl_region objects by recycling
recently used region ids rather than freeing them
- Add regression test infrastructure for a pass-through decoder
configuration
- Fix some mailbox payload handling corner cases"
* tag 'cxl-fixes-for-6.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl:
cxl/region: Recycle region ids
cxl/region: Fix 'distance' calculation with passthrough ports
tools/testing/cxl: Add a single-port host-bridge regression config
tools/testing/cxl: Fix some error exits
cxl/pmem: Fix cxl_pmem_region and cxl_memdev leak
cxl/region: Fix cxl_region leak, cleanup targets at region delete
cxl/region: Fix region HPA ordering validation
cxl/pmem: Use size_add() against integer overflow
cxl/region: Fix decoder allocation crash
ACPI: NUMA: Add CXL CFMWS 'nodes' to the possible nodes set
cxl/pmem: Fix failure to account for 8 byte header for writes to the device LSA.
cxl/region: Fix null pointer dereference due to pass through decoder commit
cxl/mbox: Add a check on input payload size
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging
Pull hwmon fixes from Guenter Roeck:
"Fix two regressions:
- Commit 54cc3dbfc10d ("hwmon: (pmbus) Add regulator supply into
macro") resulted in regulator undercount when disabling regulators.
Revert it.
- The thermal subsystem rework caused the scmi driver to no longer
register with the thermal subsystem because index values no longer
match. To fix the problem, the scmi driver now directly registers
with the thermal subsystem, no longer through the hwmon core"
* tag 'hwmon-for-v6.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
Revert "hwmon: (pmbus) Add regulator supply into macro"
hwmon: (scmi) Register explicitly with Thermal Framework
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf fixes from Borislav Petkov:
- Add Cooper Lake's stepping to the PEBS guest/host events isolation
fixed microcode revisions checking quirk
- Update Icelake and Sapphire Rapids events constraints
- Use the standard energy unit for Sapphire Rapids in RAPL
- Fix the hw_breakpoint test to fail more graciously on !SMP configs
* tag 'perf_urgent_for_v6.1_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel: Add Cooper Lake stepping to isolation_ucodes[]
perf/x86/intel: Fix pebs event constraints for SPR
perf/x86/intel: Fix pebs event constraints for ICL
perf/x86/rapl: Use standard Energy Unit for SPR Dram RAPL domain
perf/hw_breakpoint: test: Skip the test if dependencies unmet
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Borislav Petkov:
- Add new Intel CPU models
- Enforce that TDX guests are successfully loaded only on TDX hardware
where virtualization exception (#VE) delivery on kernel memory is
disabled because handling those in all possible cases is "essentially
impossible"
- Add the proper include to the syscall wrappers so that BTF can see
the real pt_regs definition and not only the forward declaration
* tag 'x86_urgent_for_v6.1_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/cpu: Add several Intel server CPU model numbers
x86/tdx: Panic on bad configs that #VE on "private" memory access
x86/tdx: Prepare for using "INFO" call for a second purpose
x86/syscall: Include asm/ptrace.h in syscall_wrapper header
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- Use POSIX-compatible grep options
- Document git-related tips for reproducible builds
- Fix a typo in the modpost rule
- Suppress SIGPIPE error message from gcc-ar and llvm-ar
- Fix segmentation fault in the menuconfig search
* tag 'kbuild-fixes-v6.1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: fix segmentation fault in menuconfig search
kbuild: fix SIGPIPE error message for AR=gcc-ar and AR=llvm-ar
kbuild: fix typo in modpost
Documentation: kbuild: Add description of git for reproducible builds
kbuild: use POSIX-compatible grep option
|
|
Pull kvm fixes from Paolo Bonzini:
"ARM:
- Fix the pKVM stage-1 walker erronously using the stage-2 accessor
- Correctly convert vcpu->kvm to a hyp pointer when generating an
exception in a nVHE+MTE configuration
- Check that KVM_CAP_DIRTY_LOG_* are valid before enabling them
- Fix SMPRI_EL1/TPIDR2_EL0 trapping on VHE
- Document the boot requirements for FGT when entering the kernel at
EL1
x86:
- Use SRCU to protect zap in __kvm_set_or_clear_apicv_inhibit()
- Make argument order consistent for kvcalloc()
- Userspace API fixes for DEBUGCTL and LBRs"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: x86: Fix a typo about the usage of kvcalloc()
KVM: x86: Use SRCU to protect zap in __kvm_set_or_clear_apicv_inhibit()
KVM: VMX: Ignore guest CPUID for host userspace writes to DEBUGCTL
KVM: VMX: Fold vmx_supported_debugctl() into vcpu_supported_debugctl()
KVM: VMX: Advertise PMU LBRs if and only if perf supports LBRs
arm64: booting: Document our requirements for fine grained traps with SME
KVM: arm64: Fix SMPRI_EL1/TPIDR2_EL0 trapping on VHE
KVM: Check KVM_CAP_DIRTY_LOG_{RING, RING_ACQ_REL} prior to enabling them
KVM: arm64: Fix bad dereference on MTE-enabled systems
KVM: arm64: Use correct accessor to parse stage-1 PTEs
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen fixes from Juergen Gross:
"One fix for silencing a smatch warning, and a small cleanup patch"
* tag 'for-linus-6.1-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
x86/xen: simplify sysenter and syscall setup
x86/xen: silence smatch warning in pmu_msr_chk_emulated()
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
Pull ext4 fixes from Ted Ts'o:
"Fix a number of bugs, including some regressions, the most serious of
which was one which would cause online resizes to fail with file
systems with metadata checksums enabled.
Also fix a warning caused by the newly added fortify string checker,
plus some bugs that were found using fuzzed file systems"
* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
ext4: fix fortify warning in fs/ext4/fast_commit.c:1551
ext4: fix wrong return err in ext4_load_and_init_journal()
ext4: fix warning in 'ext4_da_release_space'
ext4: fix BUG_ON() when directory entry has invalid rec_len
ext4: update the backup superblock's at the end of the online resize
|
|
Pull cifs fixes from Steve French:
"One symlink handling fix and two fixes foir multichannel issues with
iterating channels, including for oplock breaks when leases are
disabled"
* tag '6.1-rc4-smb3-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: fix use-after-free on the link name
cifs: avoid unnecessary iteration of tcp sessions
cifs: always iterate smb sessions using primary channel
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull `lTracing fixes for 6.1-rc3:
- Fixed NULL pointer dereference in the ring buffer wait-waiters code
for machines that have less CPUs than what nr_cpu_ids returns.
The buffer array is of size nr_cpu_ids, but only the online CPUs get
initialized.
- Fixed use after free call in ftrace_shutdown.
- Fix accounting of if a kprobe is enabled
- Fix NULL pointer dereference on error path of fprobe rethook_alloc().
- Fix unregistering of fprobe_kprobe_handler
- Fix memory leak in kprobe test module
* tag 'trace-v6.1-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
tracing: kprobe: Fix memory leak in test_gen_kprobe/kretprobe_cmd()
tracing/fprobe: Fix to check whether fprobe is registered correctly
fprobe: Check rethook_alloc() return in rethook initialization
kprobe: reverse kp->flags when arm_kprobe failed
ftrace: Fix use-after-free for dynamic ftrace_ops
ring-buffer: Check for NULL cpu_buffer in ring_buffer_wake_waiters()
|