Darrick J. Wong [Thu, 15 Aug 2024 18:48:07 +0000 (11:48 -0700)]
xfs: define the format of rt groups
Define the ondisk format of realtime group metadata, and a superblock
for realtime volumes. rt supers are conditionally enabled by a
predicate function so that they can be disabled if we ever implement
zoned storage support for the realtime volume.
For rt group enabled file systems there is a separate bitmap and summary
file for each group and thus the number of bitmap and summary blocks
needs to be calculated differently.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Christoph Hellwig [Wed, 7 Aug 2024 23:03:59 +0000 (16:03 -0700)]
iomap: add a merge boundary flag
File systems might have boundaries over which merges aren't possible.
In fact these are very common, although most of the time some kind of
header at the beginning of this region (e.g. XFS alloation groups, ext4
block groups) automatically create a merge barrier. But if that is
not present, say for a device purely used for data we need to manually
communicate that to iomap.
Add a IOMAP_F_BOUNDARY flag to never merge I/O into a previous mapping.
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>
Christoph Hellwig [Sun, 22 Sep 2024 03:46:17 +0000 (05:46 +0200)]
xfs: make RT extent numbers relative to the rtgroup
To prepare for adding per-rtgroup bitmap files, make the xfs_rtxnum_t
type encode the RT extent number relative to the rtgroup. The biggest
part of this to clearly distinguish between the relative extent number
that gets masked when converting from a global block number and length
values that just have a factor applied to them when converting from
file system 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>
Christoph Hellwig [Thu, 15 Aug 2024 18:48:06 +0000 (11:48 -0700)]
xfs: refactor xfs_rtsummary_blockcount
Make xfs_rtsummary_blockcount take all the required information from
the mount structure and return the number of summary levels from it
as well. This cleans up many of the callers and prepares for making the
rtsummary files per-rtgroup where they need to look at different value.
This means we recalculate some values in some callers, but as all these
calculations are outside the fast path and cheap that seems like a price
worth paying.
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>
Christoph Hellwig [Thu, 15 Aug 2024 18:48:05 +0000 (11:48 -0700)]
xfs: refactor xfs_rtbitmap_blockcount
Rename the existing xfs_rtbitmap_blockcount to
xfs_rtbitmap_blockcount_len and add a new xfs_rtbitmap_blockcount wrapper
around it that takes the number of extents from the mount structure.
This will simplify the move to per-rtgroup bitmaps as those will need to
pass in the number of extents per rtgroup instead.
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>
Christoph Hellwig [Thu, 15 Aug 2024 18:48:04 +0000 (11:48 -0700)]
xfs: factor out a xfs_growfs_check_rtgeom helper
Split the check that the rtsummary fits into the log into a separate
helper, and use xfs_growfs_rt_alloc_fake_mount to calculate the new RT
geometry.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: avoid division for the 0-rtx growfs check] Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Darrick J. Wong [Fri, 20 Sep 2024 17:06:00 +0000 (19:06 +0200)]
xfs: factor out a xfs_growfs_rt_alloc_fake_mount helper
Split the code to set up a fake mount point to calculate new RT
geometry out of xfs_growfs_rt_bmblock so that it can be reused.
Note that this changes the rmblocks calculation method to be based
on the passed in rblocks and extsize and not the explicitly passed
one, but both methods will always lead to the same result. The new
version just does a little bit more math while being more general.
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>
Christoph Hellwig [Thu, 15 Aug 2024 18:48:01 +0000 (11:48 -0700)]
xfs: calculate RT bitmap and summary blocks based on sb_rextents
Use the on-disk rextents to calculate the bitmap and summary blocks
instead of the calculated one so that we can refactor the helpers for
calculating them.
As the RT bitmap and summary scrubbers already check that sb_rextents
match the block count this does not change coverage of the scrubber.
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>
Christoph Hellwig [Sun, 22 Sep 2024 08:00:16 +0000 (10:00 +0200)]
xfs: move RT bitmap and summary information to the rtgroup
Move the pointers to the RT bitmap and summary inodes as well as the
summary cache to the rtgroups structure to prepare for having a
separate bitmap and summary inodes for each rtgroup.
Code using the inodes now needs to operate on a rtgroup. Where easily
possible such code is converted to iterate over all rtgroups, else
rtgroup 0 (the only one that can currently exist) is hardcoded.
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>
Create a pair of helpers to deal with setting up the necessary incore
context to check metadata records against the realtime metadata. Right
now this is limited to locking the realtime bitmap and summary inodes,
but as we add rmap and reflink to the realtime device this will grow to
include btree cursors.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 23:04:02 +0000 (16:04 -0700)]
xfs: add a lockdep class key for rtgroup inodes
Add a dynamic lockdep class key for rtgroup inodes. This will enable
lockdep to deduce inconsistencies in the rtgroup metadata ILOCK locking
order. Each class can have 8 subclasses, and for now we will only have
2 inodes per group. This enables rtgroup order and inode order checks
when nesting ILOCKs.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 23:04:01 +0000 (16:04 -0700)]
xfs: define locking primitives for realtime groups
Define helper functions to lock all metadata inodes related to a
realtime group. There's not much to look at now, but this will become
important when we add per-rtgroup metadata files and online fsck code
for them.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Fri, 20 Sep 2024 16:31:36 +0000 (18:31 +0200)]
xfs: create incore realtime group structures
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>
Christoph Hellwig [Fri, 20 Sep 2024 16:29:01 +0000 (18:29 +0200)]
xfs: clean up xfs_getfsmap_helper arguments
The calling conventions for xfs_getfsmap_helper are confusing -- callers
pass in an rmap record, but they must also supply startblock and
blockcount in daddr units. This was bolted onto the original fsmap
implementation so that we could report *something* for realtime
volumes, which do not support rmap and hence can draw only from the rt
free space bitmap. Free space on the rt volume can be more than 2^32
fsblocks long, which means that we can't use the rmap startblock or
blockcount fields.
This is confusing for callers, because they must supplying redundant
data, but not all of it is used. Streamline this by creating a separate
fsmap irec structure that contains exactly the data we need, once.
Note that we actually do need rm_startblock for rmap key comparisons
when we're actually querying an rmap btree, so leave that field but
document why it's there.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:26 +0000 (15:54 -0700)]
xfs: confirm dotdot target before replacing it during a repair
xfs_dir_replace trips an assertion if you tell it to change a dirent to
point to an inumber that it already points at. Look up the dotdot entry
directly to confirm that we need to make a change.
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. IOWs, 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>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:25 +0000 (15:54 -0700)]
xfs: move repair temporary files to the metadata directory tree
Due to resource acquisition rules, we have to create the ondisk
temporary files used to stage a filesystem repair before we can acquire
a reference to the inode that we actually want to repair. Therefore,
we do not know at tempfile creation time whether the tempfile will
belong to the regular directory tree or the metadata directory tree.
This distinction becomes important when the swapext code tries to figure
out the quota accounting of the two files whose mappings are being
swapped. The swapext code assumes that accounting updates are required
for a file if dqattach attaches dquots. Metadir files are never
accounted in quota, which means that swapext must not update the quota
accounting when swapping in a repaired directory/xattr/rtbitmap structure.
Prior to the swapext call, therefore, both files must be marked as
METADIR for dqattach so that dqattach will ignore them. Add support for
a repair tempfile to be switched to the metadir tree and switched back
before being released so that ifree will just free the file.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:24 +0000 (15:54 -0700)]
xfs: check the metadata directory inumber in superblocks
When metadata directories are enabled, make sure that the secondary
superblocks point to the metadata directory. This isn't strictly
required because the secondaries are only used to recover damaged
filesystems.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:22 +0000 (15:54 -0700)]
xfs: adjust parent pointer scrubber for sb-rooted metadata files
Starting with the metadata directory feature, we're allowed to call the
directory and parent pointer scrubbers for every metadata file,
including the ones that are children of the superblock.
For these children, checking the link count against the number of parent
pointers is a bit funny -- there's no such thing as a parent pointer for
a child of the superblock since there's no corresponding dirent. For
purposes of validating nlink, we pretend that there is a parent pointer.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:21 +0000 (15:54 -0700)]
xfs: metadata files can have xattrs if metadir is enabled
If parent pointers are enabled, then metadata files will store parent
pointers in xattrs, just like files in the user visible directory tree.
Therefore, scrub and repair need to handle attr forks for metadata files
on metadir filesystems.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:21 +0000 (15:54 -0700)]
xfs: don't fail repairs on metadata files with no attr fork
Fix a minor bug where we fail repairs on metadata files that do not have
attr forks because xrep_metadata_inode_subtype doesn't filter ENOENT.
Fixes: 5a8e07e799721 ("xfs: repair the inode core and forks of a metadata inode") Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:20 +0000 (15:54 -0700)]
xfs: do not count metadata directory files when doing online quotacheck
Previously, we stated that files in the metadata directory tree are not
counted in the dquot information. Fix the online quotacheck code to
reflect this.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:19 +0000 (15:54 -0700)]
xfs: refactor directory tree root predicates
Metadata directory trees make reasoning about the parent of a file more
difficult. Traditionally, user files are children of sb_rootino, and
metadata files are "children" of the superblock. Now, we add a third
possibility -- some metadata files can be children of sb_metadirino, but
the classic ones (rt free space data and quotas) are left alone.
Let's add some helper functions (instead of open-coding the logic
everywhere) to make scrub logic easier to understand.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:17 +0000 (15:54 -0700)]
xfs: adjust xfs_bmap_add_attrfork for metadir
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.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:16 +0000 (15:54 -0700)]
xfs: don't count metadata directory files to quota
Files in the metadata directory tree are internal to the filesystem.
Don't count the inodes or the blocks they use in the root dquot because
users do not need to know about their resource usage. This will also
quiet down complaints about dquot usage not matching du output.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:15 +0000 (15:54 -0700)]
xfs: allow bulkstat to return metadata directories
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>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:13 +0000 (15:54 -0700)]
xfs: hide metadata inodes from everyone because they are special
Metadata inodes are private files and therefore cannot be exposed to
userspace. This means no bulkstat, no open-by-handle, no linking them
into the directory tree, and no feeding them to LSMs. As such, we mark
them S_PRIVATE, which stops all that.
While we're at it, put them in a separate lockdep class so that it won't
get confused by "recursive" i_rwsem locking such as what happens when we
write to a rt file and need to allocate from the rt bitmap file. The
static function that we use to do this will be exported in the rtgroups
patchset.
Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:13 +0000 (15:54 -0700)]
xfs: disable the agi rotor for metadata inodes
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>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:12 +0000 (15:54 -0700)]
xfs: read and write metadata inode directory tree
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>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:11 +0000 (15:54 -0700)]
xfs: 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>
Darrick J. Wong [Wed, 14 Aug 2024 17:49:00 +0000 (10:49 -0700)]
xfs: iget for metadata inodes
Create a xfs_metafile_iget function for metadata inodes to ensure that
when we try to iget a metadata file, the inobt thinks a metadata inode
is in use and that the metadata type matches what we are expecting.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>Reviewed-by: Christoph Hellwig <hch@lst.de>
Darrick J. Wong [Wed, 7 Aug 2024 22:54:09 +0000 (15:54 -0700)]
xfs: define the on-disk format for the metadir feature
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>
Christoph Hellwig [Sun, 22 Sep 2024 06:04:55 +0000 (08:04 +0200)]
xfs: store a generic group structure in the intents
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.
Christoph Hellwig [Fri, 20 Sep 2024 15:35:42 +0000 (17:35 +0200)]
xfs: add group based bno conversion helpers
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.
Christoph Hellwig [Fri, 20 Sep 2024 13:40:57 +0000 (15:40 +0200)]
xfs: convert busy extent tracking to the generic group structure
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.
Christoph Hellwig [Fri, 20 Sep 2024 13:37:43 +0000 (15:37 +0200)]
xfs: switch perag iteration from the for_each macros to a while based iterator
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.
Christoph Hellwig [Sun, 22 Sep 2024 03:03:22 +0000 (05:03 +0200)]
xfs: factor out a generic xfs_group structure
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.
Christoph Hellwig [Sun, 22 Sep 2024 03:02:42 +0000 (05:02 +0200)]
xfs: factor out a xfs_iwalk_args helper
Add a helper to share more code between xfs_iwalk and xfs_inobt_walk,
and at the same time do away with the extra flags indirect so that
everyone use the same names for the same flags when using the common
iwalk code.
Christoph Hellwig [Sat, 31 Aug 2024 08:10:49 +0000 (11:10 +0300)]
xfs: convert remaining trace points to pass pag structures
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.
Christoph Hellwig [Sun, 1 Sep 2024 05:39:19 +0000 (08:39 +0300)]
xfs: pass the pag to the xrep_newbt_extent_class tracepoints
This requires moving a few of the callsites a little bit to ensure that
we already have the reference, but allows for the decoding to only happen
when tracing is actually enabled, and cleans up the callsites a bit.
Christoph Hellwig [Sun, 1 Sep 2024 05:34:33 +0000 (08:34 +0300)]
xfs: pass the pag to the trace_xrep_calc_ag_resblks{,_btsize} trace points
This requires holding the pag refcount a little longer, but allows for the
decoding to only happen when tracing is actually enabled, and cleans up the
callsites a bit.
Christoph Hellwig [Sun, 1 Sep 2024 05:26:13 +0000 (08:26 +0300)]
xfs: pass objects to the xrep_ibt_walk_rmap tracepoint
Pass the perag structure and the irec so that the decoding is only done
when tracing is actually enabled and the call sites look a lot neater,
and remove the pointless class indirection.
Christoph Hellwig [Sat, 31 Aug 2024 07:39:59 +0000 (10:39 +0300)]
xfs: pass objects to the xfs_irec_merge_{pre,post} trace points
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.
Christoph Hellwig [Thu, 19 Sep 2024 07:02:34 +0000 (09:02 +0200)]
xfs: constify pag arguments to trace points
Trace points never modify their arguments. Mark all the pag objects
passed to trace points. The exception is the xfs_ag_resv_class, which
uses the xfs_perag_resv helper that can't be marked const due to
other users modifying the returned structure.
Christoph Hellwig [Sat, 31 Aug 2024 08:18:02 +0000 (11:18 +0300)]
xfs: remove the agno argument to xfs_free_ag_extent
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.
Christoph Hellwig [Thu, 19 Sep 2024 13:15:10 +0000 (15:15 +0200)]
xfs: don't use __GFP_RETRY_MAYFAIL in xfs_initialize_perag
__GFP_RETRY_MAYFAIL increases the likelyhood of allocations to fail,
which isn't really helpful during log recovery. Remove the flag and
stick to the default GFP_KERNEL policies.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Christoph Hellwig [Sun, 1 Sep 2024 08:09:32 +0000 (11:09 +0300)]
xfs: merge the perag freeing helpers
There is no good reason to have two different routines for freeing perag
structures for the unmount and error cases. Add two arguments to specify
the range of AGs to free to xfs_free_perag, and use that to replace
xfs_free_unused_perag_range.
The addition RCU grace period for the error case is harmless, and the
extra check for the AG to actually exist is not required now that the
callers pass the exact known allocated range.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Christoph Hellwig [Sun, 8 Sep 2024 07:53:41 +0000 (10:53 +0300)]
xfs: pass the exact range to initialize to xfs_initialize_perag
Currently only the new agcount is passed to xfs_initialize_perag, which
requires lookups of existing AGs to skip them and complicates error
handling. Also pass the previous agcount so that the range that
xfs_initialize_perag operates on is exactly defined. That way the
extra lookups can be avoided, and error handling can clean up the
exact range from the old count to the last added perag structure.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Christoph Hellwig [Tue, 27 Aug 2024 05:03:21 +0000 (07:03 +0200)]
xfs: ensure st_blocks never goes to zero during COW writes
COW writes remove the amount overwritten either directly for delalloc
reservations, or in earlier deferred transactions than adding the new
amount back in the bmap map transaction. This means st_blocks on an
inode where all data is overwritten using the COW path can temporarily
show a 0 st_blocks. This can easily be reproduced with the pending
zoned device support where all writes use this path and trips the
check in generic/615, but could also happen on a reflink file without
that.
Fix this by temporarily add the pending blocks to be mapped to
i_delayed_blks while the item is queued.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Thu, 29 Aug 2024 04:08:41 +0000 (07:08 +0300)]
xfs: use xas_for_each_marked in xfs_reclaim_inodes_count
xfs_reclaim_inodes_count iterates over all AGs to sum up the reclaimable
inodes counts. There is no point in grabbing a reference to the them or
unlock the RCU critical section for each iteration, so switch to the
more efficient xas_for_each_marked iterator.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Thu, 29 Aug 2024 04:08:39 +0000 (07:08 +0300)]
xfs: simplify tagged perag iteration
Pass the old perag structure to the tagged loop helpers so that they can
grab the old agno before releasing the reference. This removes the need
to separately track the agno and the iterator macro, and thus also
obsoletes the for_each_perag_tag syntactic sugar.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Thu, 29 Aug 2024 04:08:38 +0000 (07:08 +0300)]
xfs: move the tagged perag lookup helpers to xfs_icache.c
The tagged perag helpers are only used in xfs_icache.c in the kernel code
and not at all in xfsprogs. Move them to xfs_icache.c in preparation for
switching to an xarray, for which I have no plan to implement the tagged
lookup functions for userspace.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Hongbo Li [Wed, 21 Aug 2024 06:43:55 +0000 (14:43 +0800)]
xfs: use LIST_HEAD() to simplify code
list_head can be initialized automatically with LIST_HEAD()
instead of calling INIT_LIST_HEAD().
Signed-off-by: Hongbo Li <lihongbo22@huawei.com> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Dan Carpenter [Fri, 12 Jul 2024 14:07:36 +0000 (09:07 -0500)]
xfs: remove unnecessary check
We checked that "pip" is non-NULL at the start of the if else statement
so there is no need to check again here. Delete the check.
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
John Garry [Wed, 10 Jul 2024 10:31:19 +0000 (10:31 +0000)]
xfs: Use xfs set and clear mp state helpers
Use the set and clear mp state helpers instead of open-coding.
It is noted that in some instances calls to atomic operation set_bit() and
clear_bit() are being replaced with test_and_set_bit() and
test_and_clear_bit(), respectively, as there is no specific helpers for
set_bit() and clear_bit() only. However should be ok, as we are just
ignoring the returned value from those "test" variants.
Signed-off-by: John Garry <john.g.garry@oracle.com> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Tue, 13 Aug 2024 07:39:42 +0000 (09:39 +0200)]
xfs: reclaim speculative preallocations for append only files
The XFS XFS_DIFLAG_APPEND maps to the VFS S_APPEND flag, which forbids
writes that don't append at the current EOF.
But the commit originally adding XFS_DIFLAG_APPEND support (commit a23321e766d in xfs xfs-import repository) also checked it to skip
releasing speculative preallocations, which doesn't make any sense.
Another commit (dd9f438e3290 in the xfs-import repository) later extended
that flag to also report these speculation preallocations which should
not exist in getbmap.
Remove these checks as nothing XFS_DIFLAG_APPEND implies that
preallocations beyond EOF should exist, but explicitly check for
XFS_DIFLAG_APPEND in xfs_file_release to bypass the algorithm that
discard preallocations on the first close as append only files aren't
expected to be written to only once.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Tue, 13 Aug 2024 07:39:41 +0000 (09:39 +0200)]
xfs: simplify extent lookup in xfs_can_free_eofblocks
xfs_can_free_eofblocks just cares if there is an extent beyond EOF.
Replace the call to xfs_bmapi_read with a xfs_iext_lookup_extent
as we've already checked that extents are read in earlier.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Tue, 13 Aug 2024 07:39:40 +0000 (09:39 +0200)]
xfs: check XFS_EOFBLOCKS_RELEASED earlier in xfs_release_eofblocks
If the XFS_EOFBLOCKS_RELEASED flag is set, we are not going to free the
eofblocks, so don't bother locking the inode or performing the checks in
xfs_can_free_eofblocks. Also switch to a test_and_set operation once
the iolock has been acquire so that only the caller that sets it actually
frees the post-EOF blocks.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Darrick J. Wong [Tue, 13 Aug 2024 07:39:39 +0000 (09:39 +0200)]
xfs: only free posteof blocks on first close
Certain workloads fragment files on XFS very badly, such as a software
package that creates a number of threads, each of which repeatedly run
the sequence: open a file, perform a synchronous write, and close the
file, which defeats the speculative preallocation mechanism. We work
around this problem by only deleting posteof blocks the /first/ time a
file is closed to preserve the behavior that unpacking a tarball lays
out files one after the other with no gaps.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
[hch: rebased, updated comment, renamed the flag] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Dave Chinner [Tue, 13 Aug 2024 07:39:38 +0000 (09:39 +0200)]
xfs: don't free post-EOF blocks on read close
When we have a workload that does open/read/close in parallel with other
allocation, the file becomes rapidly fragmented. This is due to close()
calling xfs_file_release() and removing the speculative preallocation
beyond EOF.
Add a check for a writable context to xfs_file_release to skip the
post-EOF block freeing (an the similarly pointless flushing on truncate
down).
Signed-off-by: Dave Chinner <dchinner@redhat.com>
[darrick: wordsmithing, fix commit message] Signed-off-by: Darrick J. Wong <djwong@kernel.org>
[hch: ported to the new ->release code structure] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Tue, 13 Aug 2024 07:39:37 +0000 (09:39 +0200)]
xfs: skip all of xfs_file_release when shut down
There is no point in trying to free post-EOF blocks when the file system
is shutdown, as it will just error out ASAP. Instead return instantly
when xfs_file_release is called on a shut down file system.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
Christoph Hellwig [Tue, 13 Aug 2024 07:39:36 +0000 (09:39 +0200)]
xfs: don't bother returning errors from xfs_file_release
While ->release returns int, the only caller ignores the return value.
As we're only doing cleanup work there isn't much of a point in
return a value to start with, so just document the situation instead.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <djwong@kernel.org> Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>