Create an incore object that will contain information about a realtime
allocation group. This will eventually enable us to shard the realtime
section in a similar manner to how we shard the data section, but for
now just a single object for the entire RT subvolume is created.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Move the allocation of the computed values into generate_rtinfo, and thus
make the variables holding them private in rt.c, and clean up a few
formatting nits.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>
[djwong: move functions to fix build errors] Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:21 +0000 (16:24 -0800)]
xfs_repair: do not count metadata directory files when doing quotacheck
Previously, we stated that files in the metadata directory tree are not
counted in the dquot information. Fix the offline quotacheck code in
xfs_repair and xfs_check to reflect this.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:21 +0000 (16:24 -0800)]
xfs_repair: truncate and unmark orphaned metadata inodes
If an inode claims to be a metadata inode but wasn't linked in either
directory tree, remove the attr fork and reset the data fork if the
contents weren't regular extent mappings before moving the inode to the
lost+found.
We don't ifree the inode, because it's possible that the inode was not
actually a metadata inode but simply got corrupted due to bitflips or
something, and we'd rather let the sysadmin examine what's left of the
file instead of photorec'ing it.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:21 +0000 (16:24 -0800)]
xfs_repair: drop all the metadata directory files during pass 4
Drop the entire metadata directory tree during pass 4 so that we can
reinitialize the entire tree in phase 6. The existing metadata files
(rtbitmap, rtsummary, quotas) will be reattached to the newly rebuilt
directory tree.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:21 +0000 (16:24 -0800)]
xfs_repair: metadata dirs are never plausible root dirs
Metadata directories are never candidates to be the root of the
user-accessible directory tree. Update has_plausible_rootdir to ignore
them all, as well as detecting the case where the superblock incorrectly
thinks both trees have the same root.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:20 +0000 (16:24 -0800)]
xfs_repair: adjust keep_fsinos to handle metadata directories
In keep_fsinos, mark the root of the metadata directory tree as inuse.
The realtime bitmap and summary files still come after the root
directories, so this is a fairly simple change to the loop test.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:20 +0000 (16:24 -0800)]
xfs_repair: mark space used by metadata files
Track space used by metadata files as a separate incore extent type.
This ensures that we can warn about cross-linked metadata files, even
though we are going to rebuild the entire metadata directory tree in the
end.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:20 +0000 (16:24 -0800)]
xfs_repair: don't let metadata and regular files mix
Track whether or not inodes thought they were metadata inodes. We
cannot allow metadata inodes to appear in the regular directory tree,
and we cannot allow regular inodes to appear in the metadata directory
tree.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:19 +0000 (16:24 -0800)]
xfs_repair: rebuild the metadata directory
Check the dirents in metadata directories for problems and repair them
if necessary. Also make sure that the sb-rooted inodes (root, metadir
root, rt bitmap, rt summary) are always allocated in that order.
Note that xfs_repair will always rebuild the metadata directory tree
itself, so we only need to report problems, not fix them.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Create a helper function to grab a realtime metadata inode. When
metadir arrives, the bitmap and summary inodes can float, so we'll
turn this function into a "load or allocate" function.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:18 +0000 (16:24 -0800)]
xfs_repair: dont check metadata directory dirent inumbers
Phase 6 always rebuilds the entire metadata directory tree, and repair
quietly ignores all the DIFLAG2_METADATA directory inodes that it finds.
As a result, none of the metadata directories are marked inuse in the
incore data. Therefore, the is_inode_free checks are not valid for
anything we find in a metadata directory.
Therefore, avoid checking is_inode_free when scanning metadata directory
dirents.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:17 +0000 (16:24 -0800)]
xfs_scrub: tread zero-length read verify as an IO error
While doing some chaos testing on the xfs_scrub read verify code, I
noticed that if the device under a live filesystem gets resized while
scrub is running a media scan, reads will start returning 0. This
causes read_verify() to run around in an infinite loop instead of
erroring out like it should.
Cc: <linux-xfs@vger.kernel.org> # v5.3.0 Fixes: 27464242956fac ("xfs_scrub: fix read verify disk error handling strategy") Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 11 Dec 2024 22:48:33 +0000 (14:48 -0800)]
xfs_db: drop the metadata checking code from blockget
Drop the check subcommand and all the metadata checking code from
xfs_db. We haven't shipped xfs_check in xfsprogs in a decade and the
last known user (fstests) stopped calling it back in July 2024.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Andrey Albershteyn <aalbersh@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:15 +0000 (16:24 -0800)]
xfs_db: disable xfs_check when metadir is enabled
As of July 2024, xfs_repair can detect more types of corruptions than
xfs_check does. I don't think it makes sense to maintain the xfs_check
code anymore, so let's just turn it off for any filesystem that has
metadata directory trees.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 21 Nov 2024 00:24:12 +0000 (16:24 -0800)]
libxfs: enforce metadata inode flag
Add checks for the metadata inode flag so that we don't ever leak
metadata inodes out to userspace, and we don't ever try to read a
regular inode as metadata.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Create a new scrubber type that checks that well known metadata
directory paths are connected to the metadata inode that the incore
structures think is in use. For example, check that "/quota/user" in
the metadata directory tree actually points to
mp->m_quotainfo->qi_uquotaip->i_ino.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Online repair might use the xfs_bmap_add_attrfork to repair a file in
the metadata directory tree if (say) the metadata file lacks the correct
parent pointers. In that case, it is not correct to check that the file
is dqattached -- metadata files must be not have /any/ dquot attached at
all. Adjust the assertions appropriately.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Allow the V5 bulkstat ioctl to return information about metadata
directory files so that xfs_scrub can find and scrub them, since they
are otherwise ordinary directories.
(Metadata files of course require per-file scrub code and hence do not
need exposure.)
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Ideally, we'd put all the metadata inodes in one place if we could, so
that the metadata all stay reasonably close together instead of
spreading out over the disk. Furthermore, if the log is internal we'd
probably prefer to keep the metadata near the log. Therefore, disable
AGI rotoring for metadata inode allocations.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Plumb in the bits we need to load metadata inodes from a named entry in
a metadir directory, create (or hardlink) inodes into a metadir
directory, create metadir directories, and flag inodes as being metadata
files.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Add checks for the metadata inode flag so that we don't ever leak
metadata inodes out to userspace, and we don't ever try to read a
regular inode as metadata.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Create a xfs_trans_metafile_iget function for metadata inodes to ensure
that when we try to iget a metadata file, the inode is allocated and its
file mode matches the metadata file type the caller expects.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Define the on-disk layout and feature flags for the metadata inode
directory feature. Add a xfs_sb_version_hasmetadir for benefit of
xfs_repair, which needs to know where the new end of the superblock
lies.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
The predicate xfs_internal_inum tells us if an inumber refers to one of
the inodes rooted in the superblock. Soon we're going to have internal
inodes in a metadata directory tree, so this helper should be renamed
to capture its limited scope.
Ondisk inodes will soon have a flag to indicate that they're metadata
inodes. Head off some confusion by renaming the xfs_is_metadata_inode
predicate to xfs_is_internal_inode.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Replace the pag pointers in the extent free, bmap, rmap and refcount
intent structures with a pointer to the generic group to prepare
for adding intents for realtime groups.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Add/move the blocks, blklog and blkmask fields to the generic groups
structure so that code can work with AGs and RTGs by just using the
right index into the array.
Then, add convenience helpers to convert block numbers based on the
generic group. This will allow writing code that doesn't care if it is
used on AGs or the upcoming realtime groups.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Replace the pag pointers in the type specific union with a generic
xfs_group pointer. This prepares for adding realtime group support.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Split busy extent tracking from struct xfs_perag into its own private
structure, which can be pointed to by the generic group structure.
Note that this structure is now dynamically allocated instead of embedded
as the upcoming zone XFS code doesn't need it and will also have an
unusually high number of groups due to hardware constraints. Dynamically
allocating the structure this is a big memory saver for this case.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Prepare for the upcoming realtime groups feature by moving the online
repair rmap hooks to based to the generic xfs_group structure.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Prepare supporting the upcoming realtime groups feature by moving the
deferred operation draining to the generic xfs_group structure.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Prepare for also tracking the health status of the upcoming realtime
groups by moving the health tracking code to the generic xfs_group
structure.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
The current for_each_perag* macros are a bit annoying in that they
require the caller to both provide an object and an index iterator, and
also somewhat obsfucate the underlying control flow mechanism.
Switch to open coded while loops using new xfs_perag_next{,_from,_range}
helpers that return the next pag structure to iterate on based on the
previous one or NULL for the loop start.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Add a helper to iterate over iterate over all groups, which can be used
as a simple while loop:
struct xfs_group *xg = NULL;
while ((xg = xfs_group_next_range(mp, xg, 0, MAX_GROUP))) {
...
}
This will be wrapped by the realtime group code first, and eventually
replace the for_each_rtgroup_from and for_each_rtgroup_range helpers.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Split the lookup and refcount handling of struct xfs_perag into an
embedded xfs_group structure that can be reused for the upcoming
realtime groups.
It will be extended with more features later.
Note that he xg_type field will only need a single bit even with
realtime group support. For now it fills a hole, but it might be
worth to fold it into another field if we can use this space better.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Cleaning up is much easier if a structure can't be looked up yet, so only
insert the pag once it is fully set up.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Factor out a xfs_perag_alloc helper that allocates a single perag
structure.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Convert all tracepoints that take [mp,agno] tuples to take a pag argument
instead so that decoding only happens when tracepoints are enabled and to
clean up the callers.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Pass the perag structure and the irec to these tracepoints so that the
decoding is only done when tracing is actually enabled and the call sites
look a lot neater.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
And remove the single instance class indirection for it.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Replace the [mp,agno] tuple with the perag structure, which will become
more useful later.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Add a helpers to convert an agino to an ino based on a pag structure.
This provides a simpler conversion and better type safety compared to the
existing code that passes the mount structure and the agno separately.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Add helpers to convert an agbno to a daddr or fsbno based on a pag
structure.
This provides a simpler conversion and better type safety compared to the
existing code that passes the mount structure and the agno separately.
Signed-off-by: Christoph Hellwig <hch@lst.de> 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_free_ag_extent already has a pointer to the pag structure through
the agf buffer. Use that instead of passing the redundant argument,
and do the same for the tracepoint.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
We'll want to use more than just the agno field in a bit.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
pag_active_wq is only woken, but never waited for.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
There are two invocations of xfs_alloc_log_agf in xfs_alloc_put_freelist.
The AGF does not change between the two calls. Although this does not pose
any practical problems, it seems like a small mistake. Therefore, fix it
by removing the first xfs_alloc_log_agf invocation.
Signed-off-by: Long Li <leo.lilong@huawei.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Carlos Maiolino <cem@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Jan Palus [Thu, 5 Dec 2024 20:43:54 +0000 (12:43 -0800)]
man: fix ioctl_xfs_commit_range man page install
INSTALL_MAN uses first symbol in .SH NAME section for both source and
destination filename hence it needs to match current filename. since
ioctl_xfs_commit_range.2 documents both ioctl_xfs_start_commit as well
as ioctl_xfs_commit_range ensure they are listed in order INSTALL_MAN
expects.
Signed-off-by: Jan Palus <jpalus@fastmail.com> Reviewed-by: "Darrick J. Wong" <djwong@kernel.org> Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
This is due to an incorrect check in xfs_repair, which tries to validate
that a block mapping cannot exceed what it thinks is the maximum file
offset. Unfortunately, the check is wrong, because only br_startoff is
subject to the 2^52-1 limit -- not br_startoff + br_blockcount.
Nowadays libxfs provides a symbol XFS_MAX_FILEOFF for the maximum
allowable file block offset that can be mapped into a file. Use this
instead of the open-coded logic in versions.c and correct all the other
checks. Note that this problem only surfaced when rtgroups were enabled
because hch changed xfs_repair to use the same tree-based block state
data structure that we use for AGs when rtgroups are enabled.
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Catherine Hoang [Wed, 20 Nov 2024 02:35:44 +0000 (18:35 -0800)]
xfs_io: add support for atomic write statx fields
Add support for the new atomic_write_unit_min, atomic_write_unit_max, and
atomic_write_segments_max fields in statx for xfs_io. In order to support builds
against old kernel headers, define our own internal statx structs. If the
system's struct statx does not have the required atomic write fields, override
the struct definitions with the internal definitions in statx.h.
Signed-off-by: Catherine Hoang <catherine.hoang@oracle.com> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: John Garry <john.g.garry@oracle.com>
Darrick J. Wong [Thu, 14 Nov 2024 23:53:24 +0000 (15:53 -0800)]
xfs_repair: synthesize incore inode tree records when required
On a filesystem with 64k fsblock size, xfs/093 fails with the following:
Phase 3 - for each AG...
- scan and clear agi unlinked lists...
found inodes not in the inode allocation tree
found inodes not in the inode allocation tree
- process known inodes and perform inode discovery...
- agno = 0
xfs_repair: dino_chunks.c:1166: process_aginodes: Assertion `num_inos == igeo->ialloc_inos' failed.
./common/xfs: line 392: 361225 Aborted (core dumped) $XFS_REPAIR_PROG $SCRATCH_OPTIONS $* $SCRATCH_DEV
In this situation, the inode size is 512b, which means that two inobt
records map to a single fs block. However, the inobt walk didn't find
the second record, so it didn't create a second incore ino_tree_node_t
object. The assertion trips, and we fail to repair the filesystem.
To fix this, synthesize incore inode records when we know that they must
exist. Mark the inodes as in use so that they will not be purged from
parent directories or moved to lost+found if the directory tree is also
compromised.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Thu, 14 Nov 2024 23:53:24 +0000 (15:53 -0800)]
xfs_repair: fix crasher in pf_queuing_worker
Don't walk off the end of the inode records when we're skipping inodes
for prefetching. The skip loop doesn't make sense to me -- why we
ignore the first N inodes but don't care what number they are makes
little sense to me. But let's fix xfs/155 to crash less, eh?
Cc: <linux-xfs@vger.kernel.org> # v2.10.0 Fixes: 2556c98bd9e6b2 ("Perform true sequential bulk read prefetching in xfs_repair Merge of master-melb:xfs-cmds:29147a by kenmcd.") Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Recently, we found that the CPU spent a lot of time in
xfs_alloc_ag_vextent_size when the filesystem has millions of fragmented
spaces.
The reason is that we conducted much extra searching for extents that
could not yield a better result, and these searches would cost a lot of
time when there were millions of extents to search through. Even if we
get the same result length, we don't switch our choice to the new one,
so we can definitely terminate the search early.
Since the result length cannot exceed the found length, when the found
length equals the best result length we already have, we can conclude
the search.
We did a test in that filesystem:
[root@localhost ~]# xfs_db -c freesp /dev/vdb
from to extents blocks pct
1 1 215 215 0.01
2 3 994476 1988952 99.99
Before this patch:
0) | xfs_alloc_ag_vextent_size [xfs]() {
0) * 15597.94 us | }
After this patch:
0) | xfs_alloc_ag_vextent_size [xfs]() {
0) 19.176 us | }
Signed-off-by: Chi Zhiling <chizhiling@kylinos.cn> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Carlos Maiolino <cem@kernel.org>
Darrick J. Wong [Tue, 29 Oct 2024 00:03:34 +0000 (17:03 -0700)]
xfs_scrub_all: wait for services to start activating
It seems that the function call to start a systemd unit completes
asynchronously from any change in that unit's active state. On a
lightly loaded system, a Start() call followed by an ActiveState()
call actually sees the change in state from inactive to activating.
Unfortunately, on a heavily loaded system, the state change may take a
few seconds. If this is the case, the wait() call can see that the unit
state is "inactive", decide that the service already finished, and exit
early, when in reality it hasn't even gotten to 'activating'.
Fix this by adding a second method that watches either for the inactive
-> activating state transition or for the last exit from inactivation
timestamp to change before waiting for the unit to reach inactive state.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Cc: <linux-xfs@vger.kernel.org> # v6.10.0 Fixes: 6d831e770359ff ("xfs_scrub_all: convert systemctl calls to dbus") Reviewed-by: Christoph Hellwig <hch@lst.de>
xfs_repair: stop preallocating blocks in mk_rbmino and mk_rsumino
Now that repair is using libxfs_rtfile_initialize_blocks to write to the
rtbitmap and rtsummary inodes, space allocation is already taken care of
that helper and there is no need to preallocate it. Remove the code to
do so.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Use libxfs_rtfile_initialize_blocks to write the re-computed rtbitmap
and rtsummary contents. This removes duplicate code and prepares for
even more sharing once the rtgroup features adds a metadata header to
the rtbitmap and rtsummary blocks.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Use the new libxfs helper for initializing the rtbitmap/summary files
for rtgroup-enabled file systems. Also skip the zeroing of the blocks
for rtgroup file systems as we'll overwrite every block instantly.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Darrick J. Wong [Tue, 29 Oct 2024 00:03:33 +0000 (17:03 -0700)]
xfs_repair: checking rt free space metadata must happen during phase 4
Back in the really old days, xfs_repair would generate the new free
space information for the realtime section during phase 5, and write the
contents to the rtbitmap and summary files during phase 6. This was ok
because the incore information isn't used until phase 6.
Then I changed the behavior to check the generated information against
what was on disk and complain about the discrepancies. Unfortunately,
there was a subtle flaw here -- for a non -n run, we'll have regenerated
the AG metadata before we actually check the rt free space information.
If the AG btree regeneration should clobber one of the old rtbitmap or
summary blocks, this will be reported as a corruption even though
nothing's wrong.
Move check_rtmetadata to the end of phase 4 so that this doesn't happen.
Cc: <linux-xfs@vger.kernel.org> # v5.19.0 Fixes: f2e388616d7491 ("xfs_repair: check free rt extent count") Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
projects / thirdparty / xfsprogs-dev.git / log
