platform/kernel/linux-starfive.git
2 years agoxfs: Fix double unlock in defer capture code
Allison Henderson [Wed, 4 May 2022 02:39:02 +0000 (12:39 +1000)]
xfs: Fix double unlock in defer capture code

The new deferred attr patch set uncovered a double unlock in the
recent port of the defer ops capture and continue code.  During log
recovery, we're allowed to hold buffers to a transaction that's being
used to replay an intent item.  When we capture the resources as part
of scheduling a continuation of an intent chain, we call xfs_buf_hold
to retain our reference to the buffer beyond the transaction commit,
but we do /not/ call xfs_trans_bhold to maintain the buffer lock.
This means that xfs_defer_ops_continue needs to relock the buffers
before xfs_defer_restore_resources joins then tothe new transaction.

Additionally, the buffers should not be passed back via the dres
structure since they need to remain locked unlike the inodes.  So
simply set dr_bufs to zero after populating the dres structure.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoMerge branch 'guilt/xfs-5.19-fuzz-fixes' into xfs-5.19-for-next
Dave Chinner [Wed, 4 May 2022 02:38:02 +0000 (12:38 +1000)]
Merge branch 'guilt/xfs-5.19-fuzz-fixes' into xfs-5.19-for-next

2 years agoMerge tag 'reflink-speedups-5.19_2022-04-28' of git://git.kernel.org/pub/scm/linux...
Dave Chinner [Wed, 4 May 2022 02:37:40 +0000 (12:37 +1000)]
Merge tag 'reflink-speedups-5.19_2022-04-28' of git://git./linux/kernel/git/djwong/xfs-linux into xfs-5.19-for-next

xfs: fix reflink inefficiencies

As Dave Chinner has complained about on IRC, there are a couple of
things about reflink that are very inefficient.  First of all, we
limited the size of all bunmapi operations to avoid flooding the log
with defer ops in the worst case, but recent changes to the defer
ops code have solved that problem, so get rid of the bunmapi length
clamp.

Second, the log reservations for reflink operations are far far
larger than they need to be.  Shrink them to exactly what we need to
handle each deferred RUI and CUI log item, and no more.  Also reduce
logcount because we don't need 8 rolls per operation.  Introduce a
transaction reservation compatibility layer to avoid changing the
minimum log size calculations.

Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoMerge tag 'rmap-speedups-5.19_2022-04-28' of git://git.kernel.org/pub/scm/linux/kerne...
Dave Chinner [Wed, 4 May 2022 02:37:18 +0000 (12:37 +1000)]
Merge tag 'rmap-speedups-5.19_2022-04-28' of git://git./linux/kernel/git/djwong/xfs-linux into xfs-5.19-for-next

xfs: fix rmap inefficiencies

Reduce the performance impact of the reverse mapping btree when
reflink is enabled by using the much faster non-overlapped btree
lookup functions when we're searching the rmap index with a fully
specified key.  If we find the exact record we're looking for,
great!  We don't have to perform the full overlapped scan.  For
filesystems with high sharing factors this reduces the xfs_scrub
runtime by a good 15%%.

This has been shown to reduce the fstests runtime for realtime rmap
configurations by 30%%, since the lack of AGs severely limits
scalability.

Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoMerge branch 'guilt/xlog-intent-whiteouts' into xfs-5.19-for-next
Dave Chinner [Wed, 4 May 2022 02:37:02 +0000 (12:37 +1000)]
Merge branch 'guilt/xlog-intent-whiteouts' into xfs-5.19-for-next

2 years agoMerge branch 'guilt/xfs-5.19-misc-2' into xfs-5.19-for-next
Dave Chinner [Wed, 4 May 2022 02:36:54 +0000 (12:36 +1000)]
Merge branch 'guilt/xfs-5.19-misc-2' into xfs-5.19-for-next

2 years agoxfs: validate v5 feature fields
Dave Chinner [Wed, 4 May 2022 02:17:18 +0000 (12:17 +1000)]
xfs: validate v5 feature fields

We don't check that the v4 feature flags taht v5 requires to be set
are actually set anywhere. Do this check when we see that the
filesystem is a v5 filesystem.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: set XFS_FEAT_NLINK correctly
Dave Chinner [Wed, 4 May 2022 02:14:13 +0000 (12:14 +1000)]
xfs: set XFS_FEAT_NLINK correctly

While xfs_has_nlink() is not used in kernel, it is used in userspace
(e.g. by xfs_db) so we need to set the XFS_FEAT_NLINK flag correctly
in xfs_sb_version_to_features().

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: validate inode fork size against fork format
Dave Chinner [Wed, 4 May 2022 02:13:53 +0000 (12:13 +1000)]
xfs: validate inode fork size against fork format

xfs_repair catches fork size/format mismatches, but the in-kernel
verifier doesn't, leading to null pointer failures when attempting
to perform operations on the fork. This can occur in the
xfs_dir_is_empty() where the in-memory fork format does not match
the size and so the fork data pointer is accessed incorrectly.

Note: this causes new failures in xfs/348 which is testing mode vs
ftype mismatches. We now detect a regular file that has been changed
to a directory or symlink mode as being corrupt because the data
fork is for a symlink or directory should be in local form when
there are only 3 bytes of data in the data fork. Hence the inode
verify for the regular file now fires w/ -EFSCORRUPTED because
the inode fork format does not match the format the corrupted mode
says it should be in.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: detect self referencing btree sibling pointers
Dave Chinner [Wed, 4 May 2022 02:13:35 +0000 (12:13 +1000)]
xfs: detect self referencing btree sibling pointers

To catch the obvious graph cycle problem and hence potential endless
looping.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: intent item whiteouts
Dave Chinner [Wed, 4 May 2022 01:50:29 +0000 (11:50 +1000)]
xfs: intent item whiteouts

When we log modifications based on intents, we add both intent
and intent done items to the modification being made. These get
written to the log to ensure that the operation is re-run if the
intent done is not found in the log.

However, for operations that complete wholly within a single
checkpoint, the change in the checkpoint is atomic and will never
need replay. In this case, we don't need to actually write the
intent and intent done items to the journal because log recovery
will never need to manually restart this modification.

Log recovery currently handles intent/intent done matching by
inserting the intent into the AIL, then removing it when a matching
intent done item is found. Hence for all the intent-based operations
that complete within a checkpoint, we spend all that time parsing
the intent/intent done items just to cancel them and do nothing with
them.

Hence it follows that the only time we actually need intents in the
log is when the modification crosses checkpoint boundaries in the
log and so may only be partially complete in the journal. Hence if
we commit and intent done item to the CIL and the intent item is in
the same checkpoint, we don't actually have to write them to the
journal because log recovery will always cancel the intents.

We've never really worried about the overhead of logging intents
unnecessarily like this because the intents we log are generally
very much smaller than the change being made. e.g. freeing an extent
involves modifying at lease two freespace btree blocks and the AGF,
so the EFI/EFD overhead is only a small increase in space and
processing time compared to the overall cost of freeing an extent.

However, delayed attributes change this cost equation dramatically,
especially for inline attributes. In the case of adding an inline
attribute, we only log the inode core and attribute fork at present.
With delayed attributes, we now log the attr intent which includes
the name and value, the inode core adn attr fork, and finally the
attr intent done item. We increase the number of items we log from 1
to 3, and the number of log vectors (regions) goes up from 3 to 7.
Hence we tripple the number of objects that the CIL has to process,
and more than double the number of log vectors that need to be
written to the journal.

At scale, this means delayed attributes cause a non-pipelined CIL to
become CPU bound processing all the extra items, resulting in a > 40%
performance degradation on 16-way file+xattr create worklaods.
Pipelining the CIL (as per 5.15) reduces the performance degradation
to 20%, but now the limitation is the rate at which the log items
can be written to the iclogs and iclogs be dispatched for IO and
completed.

Even log IO completion is slowed down by these intents, because it
now has to process 3x the number of items in the checkpoint.
Processing completed intents is especially inefficient here, because
we first insert the intent into the AIL, then remove it from the AIL
when the intent done is processed. IOWs, we are also doing expensive
operations in log IO completion we could completely avoid if we
didn't log completed intent/intent done pairs.

Enter log item whiteouts.

When an intent done is committed, we can check to see if the
associated intent is in the same checkpoint as we are currently
committing the intent done to. If so, we can mark the intent log
item with a whiteout and immediately free the intent done item
rather than committing it to the CIL. We can basically skip the
entire formatting and CIL insertion steps for the intent done item.

However, we cannot remove the intent item from the CIL at this point
because the unlocked per-cpu CIL item lists do not permit removal
without holding the CIL context lock exclusively. Transaction commit
only holds the context lock shared, hence the best we can do is mark
the intent item with a whiteout so that the CIL push can release it
rather than writing it to the log.

This means we never write the intent to the log if the intent done
has also been committed to the same checkpoint, but we'll always
write the intent if the intent done has not been committed or has
been committed to a different checkpoint. This will result in
correct log recovery behaviour in all cases, without the overhead of
logging unnecessary intents.

This intent whiteout concept is generic - we can apply it to all
intent/intent done pairs that have a direct 1:1 relationship. The
way deferred ops iterate and relog intents mean that all intents
currently have a 1:1 relationship with their done intent, and hence
we can apply this cancellation to all existing intent/intent done
implementations.

For delayed attributes with a 16-way 64kB xattr create workload,
whiteouts reduce the amount of journalled metadata from ~2.5GB/s
down to ~600MB/s and improve the creation rate from 9000/s to
14000/s.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: whiteouts release intents that are not in the AIL
Dave Chinner [Wed, 4 May 2022 01:46:47 +0000 (11:46 +1000)]
xfs: whiteouts release intents that are not in the AIL

When we release an intent that a whiteout applies to, it will not
have been committed to the journal and so won't be in the AIL. Hence
when we drop the last reference to the intent, we do not want to try
to remove it from the AIL as that will trigger a filesystem
shutdown. Hence make the removal of intents from the AIL conditional
on them actually being in the AIL so we do the correct thing.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: add log item method to return related intents
Dave Chinner [Wed, 4 May 2022 01:46:39 +0000 (11:46 +1000)]
xfs: add log item method to return related intents

To apply a whiteout to an intent item when an intent done item is
committed, we need to be able to retrieve the intent item from the
the intent done item. Add a log item op method for doing this, and
wire all the intent done items up to it.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: factor and move some code in xfs_log_cil.c
Dave Chinner [Wed, 4 May 2022 01:46:30 +0000 (11:46 +1000)]
xfs: factor and move some code in xfs_log_cil.c

In preparation for adding support for intent item whiteouts.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: tag transactions that contain intent done items
Dave Chinner [Wed, 4 May 2022 01:46:21 +0000 (11:46 +1000)]
xfs: tag transactions that contain intent done items

Intent whiteouts will require extra work to be done during
transaction commit if the transaction contains an intent done item.

To determine if a transaction contains an intent done item, we want
to avoid having to walk all the items in the transaction to check if
they are intent done items. Hence when we add an intent done item to
a transaction, tag the transaction to indicate that it contains such
an item.

We don't tag the transaction when the defer ops is relogging an
intent to move it forward in the log. Whiteouts will never apply to
these cases, so we don't need to bother looking for them.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: add log item flags to indicate intents
Dave Chinner [Wed, 4 May 2022 01:46:09 +0000 (11:46 +1000)]
xfs: add log item flags to indicate intents

We currently have a couple of helper functions that try to infer
whether the log item is an intent or intent done item from the
combinations of operations it supports.  This is incredibly fragile
and not very efficient as it requires checking specific combinations
of ops.

We need to be able to identify intent and intent done items quickly
and easily in upcoming patches, so simply add intent and intent done
type flags to the log item ops flags. These are static flags to
begin with, so intent items should have been typed like this from
the start.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: don't commit the first deferred transaction without intents
Dave Chinner [Wed, 4 May 2022 01:46:00 +0000 (11:46 +1000)]
xfs: don't commit the first deferred transaction without intents

If the first operation in a string of defer ops has no intents,
then there is no reason to commit it before running the first call
to xfs_defer_finish_one(). This allows the defer ops to be used
effectively for non-intent based operations without requiring an
unnecessary extra transaction commit when first called.

This fixes a regression in per-attribute modification transaction
count when delayed attributes are not being used.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: hide log iovec alignment constraints
Dave Chinner [Wed, 4 May 2022 01:45:50 +0000 (11:45 +1000)]
xfs: hide log iovec alignment constraints

Callers currently have to round out the size of buffers to match the
aligment constraints of log iovecs and xlog_write(). They should not
need to know this detail, so introduce a new function to calculate
the iovec length (for use in ->iop_size implementations). Also
modify xlog_finish_iovec() to round up the length to the correct
alignment so the callers don't need to do this, either.

Convert the only user - inode forks - of this alignment rounding to
use the new interface.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: fix potential log item leak
Dave Chinner [Wed, 4 May 2022 01:45:11 +0000 (11:45 +1000)]
xfs: fix potential log item leak

Ever since we added shadown format buffers to the log items, log
items need to handle the item being released with shadow buffers
attached. Due to the fact this requirement was added at the same
time we added new rmap/reflink intents, we missed the cleanup of
those items.

In theory, this means shadow buffers can be leaked in a very small
window when a shutdown is initiated. Testing with KASAN shows this
leak does not happen in practice - we haven't identified a single
leak in several years of shutdown testing since ~v4.8 kernels.

However, the intent whiteout cleanup mechanism results in every
cancelled intent in exactly the same state as this tiny race window
creates and so if intents down clean up shadow buffers on final
release we will leak the shadow buffer for just about every intent
we create.

Hence we start with this patch to close this condition off and
ensure that when whiteouts start to be used we don't leak lots of
memory.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: zero inode fork buffer at allocation
Dave Chinner [Wed, 4 May 2022 01:44:55 +0000 (11:44 +1000)]
xfs: zero inode fork buffer at allocation

When we first allocate or resize an inline inode fork, we round up
the allocation to 4 byte alingment to make journal alignment
constraints. We don't clear the unused bytes, so we can copy up to
three uninitialised bytes into the journal. Zero those bytes so we
only ever copy zeros into the journal.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: rename xfs_*alloc*_log_count to _block_count
Darrick J. Wong [Tue, 26 Apr 2022 01:38:24 +0000 (18:38 -0700)]
xfs: rename xfs_*alloc*_log_count to _block_count

These functions return the maximum number of blocks that could be logged
in a particular transaction.  "log count" is confusing since there's a
separate concept of a log (operation) count in the reservation code, so
let's change it to "block count" to be less confusing.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: rewrite xfs_reflink_end_cow to use intents
Darrick J. Wong [Tue, 26 Apr 2022 01:38:15 +0000 (18:38 -0700)]
xfs: rewrite xfs_reflink_end_cow to use intents

Currently, the code that performs CoW remapping after a write has this
odd behavior where it walks /backwards/ through the data fork to remap
extents in reverse order.  Earlier, we rewrote the reflink remap
function to use deferred bmap log items instead of trying to cram as
much into the first transaction that we could.  Now do the same for the
CoW remap code.  There doesn't seem to be any performance impact; we're
just making better use of code that we added for the benefit of reflink.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: reduce transaction reservations with reflink
Darrick J. Wong [Tue, 26 Apr 2022 01:38:14 +0000 (18:38 -0700)]
xfs: reduce transaction reservations with reflink

Before to the introduction of deferred refcount operations, reflink
would try to cram refcount btree updates into the same transaction as an
allocation or a free event.  Mainline XFS has never actually done that,
but we never refactored the transaction reservations to reflect that we
now do all refcount updates in separate transactions.  Fix this to
reduce the transaction reservation size even farther, so that between
this patch and the previous one, we reduce the tr_write and tr_itruncate
sizes by 66%.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: reduce the absurdly large log operation count
Darrick J. Wong [Tue, 26 Apr 2022 01:38:14 +0000 (18:38 -0700)]
xfs: reduce the absurdly large log operation count

Back in the early days of reflink and rmap development I set the
transaction reservation sizes to be overly generous for rmap+reflink
filesystems, and a little under-generous for rmap-only filesystems.

Since we don't need *eight* transaction rolls to handle three new log
intent items, decrease the logcounts to what we actually need, and amend
the shadow reservation computation function to reflect what we used to
do so that the minimum log size doesn't change.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: report "max_resp" used for min log size computation
Darrick J. Wong [Tue, 26 Apr 2022 01:38:13 +0000 (18:38 -0700)]
xfs: report "max_resp" used for min log size computation

Move the tracepoint that computes the size of the transaction used to
compute the minimum log size into xfs_log_get_max_trans_res so that we
only have to compute this stuff once.

Leave xfs_log_get_max_trans_res as a non-static function so that xfs_db
can call it to report the results of the userspace computation of the
same value to diagnose mkfs/kernel misinteractions.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: create shadow transaction reservations for computing minimum log size
Darrick J. Wong [Tue, 26 Apr 2022 01:38:13 +0000 (18:38 -0700)]
xfs: create shadow transaction reservations for computing minimum log size

Every time someone changes the transaction reservation sizes, they
introduce potential compatibility problems if the changes affect the
minimum log size that we validate at mount time.  If the minimum log
size gets larger (which should be avoided because doing so presents a
serious risk of log livelock), filesystems created with old mkfs will
not mount on a newer kernel; if the minimum size shrinks, filesystems
created with newer mkfs will not mount on older kernels.

Therefore, enable the creation of a shadow log reservation structure
where we can "undo" the effects of tweaks when computing minimum log
sizes.  These shadow reservations should never be used in practice, but
they insulate us from perturbations in minimum log size.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: remove a __xfs_bunmapi call from reflink
Darrick J. Wong [Tue, 26 Apr 2022 01:38:12 +0000 (18:38 -0700)]
xfs: remove a __xfs_bunmapi call from reflink

This raw call isn't necessary since we can always remove a full delalloc
extent.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: stop artificially limiting the length of bunmap calls
Darrick J. Wong [Tue, 26 Apr 2022 01:37:15 +0000 (18:37 -0700)]
xfs: stop artificially limiting the length of bunmap calls

In commit e1a4e37cc7b6, we clamped the length of bunmapi calls on the
data forks of shared files to avoid two failure scenarios: one where the
extent being unmapped is so sparsely shared that we exceed the
transaction reservation with the sheer number of refcount btree updates
and EFI intent items; and the other where we attach so many deferred
updates to the transaction that we pin the log tail and later the log
head meets the tail, causing the log to livelock.

We avoid triggering the first problem by tracking the number of ops in
the refcount btree cursor and forcing a requeue of the refcount intent
item any time we think that we might be close to overflowing.  This has
been baked into XFS since before the original e1a4 patch.

A recent patchset fixed the second problem by changing the deferred ops
code to finish all the work items created by each round of trying to
complete a refcount intent item, which eliminates the long chains of
deferred items (27dad); and causing long-running transactions to relog
their intent log items when space in the log gets low (74f4d).

Because this clamp affects /any/ unmapping request regardless of the
sharing factors of the component blocks, it degrades the performance of
all large unmapping requests -- whereas with an unshared file we can
unmap millions of blocks in one go, shared files are limited to
unmapping a few thousand blocks at a time, which causes the upper level
code to spin in a bunmapi loop even if it wasn't needed.

This also eliminates one more place where log recovery behavior can
differ from online behavior, because bunmapi operations no longer need
to requeue.  The fstest generic/447 was created to test the old fix, and
it still passes with this applied.

Partial-revert-of: e1a4e37cc7b6 ("xfs: try to avoid blowing out the transaction reservation when bunmaping a shared extent")
Depends: 27dada070d59 ("xfs: change the order in which child and parent defer ops ar finished")
Depends: 74f4d6a1e065 ("xfs: only relog deferred intent items if free space in the log gets low")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: count EFIs when deciding to ask for a continuation of a refcount update
Darrick J. Wong [Tue, 26 Apr 2022 22:29:54 +0000 (15:29 -0700)]
xfs: count EFIs when deciding to ask for a continuation of a refcount update

A long time ago, I added to XFS the ability to use deferred reference
count operations as part of a transaction chain.  This enabled us to
avoid blowing out the transaction reservation when the blocks in a
physical extent all had different reference counts because we could ask
the deferred operation manager for a continuation, which would get us a
clean transaction.

The refcount code asks for a continuation when the number of refcount
record updates reaches the point where we think that the transaction has
logged enough full btree blocks due to refcount (and free space) btree
shape changes and refcount record updates that we're in danger of
overflowing the transaction.

We did not previously count the EFIs logged to the refcount update
transaction because the clamps on the length of a bunmap operation were
sufficient to avoid overflowing the transaction reservation even in the
worst case situation where every other block of the unmapped extent is
shared.

Unfortunately, the restrictions on bunmap length avoid failure in the
worst case by imposing a maximum unmap length of ~3000 blocks, even for
non-pathological cases.  This seriously limits performance when freeing
large extents.

Therefore, track EFIs with the same counter as refcount record updates,
and use that information as input into when we should ask for a
continuation.  This enables the next patch to drop the clumsy bunmap
limitation.

Depends: 27dada070d59 ("xfs: change the order in which child and parent defer ops ar finished")
Depends: 74f4d6a1e065 ("xfs: only relog deferred intent items if free space in the log gets low")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: speed up write operations by using non-overlapped lookups when possible
Darrick J. Wong [Tue, 26 Apr 2022 01:37:06 +0000 (18:37 -0700)]
xfs: speed up write operations by using non-overlapped lookups when possible

Reverse mapping on a reflink-capable filesystem has some pretty high
overhead when performing file operations.  This is because the rmap
records for logically and physically adjacent extents might not be
adjacent in the rmap index due to data block sharing.  As a result, we
use expensive overlapped-interval btree search, which walks every record
that overlaps with the supplied key in the hopes of finding the record.

However, profiling data shows that when the index contains a record that
is an exact match for a query key, the non-overlapped btree search
function can find the record much faster than the overlapped version.
Try the non-overlapped lookup first when we're trying to find the left
neighbor rmap record for a given file mapping, which makes unwritten
extent conversion and remap operations run faster if data block sharing
is minimal in this part of the filesystem.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: speed up rmap lookups by using non-overlapped lookups when possible
Darrick J. Wong [Tue, 26 Apr 2022 01:37:06 +0000 (18:37 -0700)]
xfs: speed up rmap lookups by using non-overlapped lookups when possible

Reverse mapping on a reflink-capable filesystem has some pretty high
overhead when performing file operations.  This is because the rmap
records for logically and physically adjacent extents might not be
adjacent in the rmap index due to data block sharing.  As a result, we
use expensive overlapped-interval btree search, which walks every record
that overlaps with the supplied key in the hopes of finding the record.

However, profiling data shows that when the index contains a record that
is an exact match for a query key, the non-overlapped btree search
function can find the record much faster than the overlapped version.
Try the non-overlapped lookup first, which will make scrub run much
faster.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: simplify xfs_rmap_lookup_le call sites
Darrick J. Wong [Tue, 26 Apr 2022 01:37:06 +0000 (18:37 -0700)]
xfs: simplify xfs_rmap_lookup_le call sites

Most callers of xfs_rmap_lookup_le will retrieve the btree record
immediately if the lookup succeeds.  The overlapped version of this
function (xfs_rmap_lookup_le_range) will return the record if the lookup
succeeds, so make the regular version do it too.  Get rid of the useless
len argument, since it's not part of the lookup key.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoxfs: capture buffer ops in the xfs_buf tracepoints
Darrick J. Wong [Tue, 26 Apr 2022 01:37:05 +0000 (18:37 -0700)]
xfs: capture buffer ops in the xfs_buf tracepoints

Record the buffer ops in the xfs_buf tracepoints so that we can monitor
the alleged type of the buffer.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2 years agoxfs: revert "xfs: actually bump warning counts when we send warnings"
Eric Sandeen [Tue, 26 Apr 2022 03:35:26 +0000 (13:35 +1000)]
xfs: revert "xfs: actually bump warning counts when we send warnings"

This reverts commit 4b8628d57b725b32616965e66975fcdebe008fe7.

XFS quota has had the concept of a "quota warning limit" since
the earliest Irix implementation, but a mechanism for incrementing
the warning counter was never implemented, as documented in the
xfs_quota(8) man page. We do know from the historical archive that
it was never incremented at runtime during quota reservation
operations.

With this commit, the warning counter quickly increments for every
allocation attempt after the user has crossed a quote soft
limit threshold, and this in turn transitions the user to hard
quota failures, rendering soft quota thresholds and timers useless.
This was reported as a regression by users.

Because the intended behavior of this warning counter has never been
understood or documented, and the result of this change is a regression
in soft quota functionality, revert this commit to make soft quota
limits and timers operable again.

Fixes: 4b8628d57b72 ("xfs: actually bump warning counts when we send warnings)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: fix soft lockup via spinning in filestream ag selection loop
Brian Foster [Tue, 26 Apr 2022 03:34:54 +0000 (13:34 +1000)]
xfs: fix soft lockup via spinning in filestream ag selection loop

The filestream AG selection loop uses pagf data to aid in AG
selection, which depends on pagf initialization. If the in-core
structure is not initialized, the caller invokes the AGF read path
to do so and carries on. If another task enters the loop and finds
a pagf init already in progress, the AGF read returns -EAGAIN and
the task continues the loop. This does not increment the current ag
index, however, which means the task spins on the current AGF buffer
until unlocked.

If the AGF read I/O submitted by the initial task happens to be
delayed for whatever reason, this results in soft lockup warnings
via the spinning task. This is reproduced by xfs/170. To avoid this
problem, fix the AGF trylock failure path to properly iterate to the
next AG. If a task iterates all AGs without making progress, the
trylock behavior is dropped in favor of blocking locks and thus a
soft lockup is no longer possible.

Fixes: f48e2df8a877ca1c ("xfs: make xfs_*read_agf return EAGAIN to ALLOC_FLAG_TRYLOCK callers")
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: improve __xfs_set_acl
Yang Xu [Tue, 26 Apr 2022 03:34:42 +0000 (13:34 +1000)]
xfs: improve __xfs_set_acl

Provide a proper stub for the !CONFIG_XFS_POSIX_ACL case.

Also use a easy way for xfs_get_acl stub.

Suggested-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoMerge tag 'large-extent-counters-v9' of https://github.com/chandanr/linux into xfs...
Dave Chinner [Thu, 21 Apr 2022 06:46:17 +0000 (16:46 +1000)]
Merge tag 'large-extent-counters-v9' of https://github.com/chandanr/linux into xfs-5.19-for-next

xfs: Large extent counters

The commit xfs: fix inode fork extent count overflow
(3f8a4f1d876d3e3e49e50b0396eaffcc4ba71b08) mentions that 10 billion
data fork extents should be possible to create. However the
corresponding on-disk field has a signed 32-bit type. Hence this
patchset extends the per-inode data fork extent counter to 64 bits
(out of which 48 bits are used to store the extent count).

Also, XFS has an attribute fork extent counter which is 16 bits
wide. A workload that,
1. Creates 1 million 255-byte sized xattrs,
2. Deletes 50% of these xattrs in an alternating manner,
3. Tries to insert 400,000 new 255-byte sized xattrs
   causes the xattr extent counter to overflow.

Dave tells me that there are instances where a single file has more
than 100 million hardlinks. With parent pointers being stored in
xattrs, we will overflow the signed 16-bits wide attribute extent
counter when large number of hardlinks are created. Hence this
patchset extends the on-disk field to 32-bits.

The following changes are made to accomplish this,
1. A 64-bit inode field is carved out of existing di_pad and
   di_flushiter fields to hold the 64-bit data fork extent counter.
2. The existing 32-bit inode data fork extent counter will be used to
   hold the attribute fork extent counter.
3. A new incompat superblock flag to prevent older kernels from mounting
   the filesystem.

Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoMerge branch 'guilt/xlog-write-rework' into xfs-5.19-for-next
Dave Chinner [Thu, 21 Apr 2022 06:45:52 +0000 (16:45 +1000)]
Merge branch 'guilt/xlog-write-rework' into xfs-5.19-for-next

2 years agoMerge branch 'guilt/xfs-unsigned-flags-5.18' into xfs-5.19-for-next
Dave Chinner [Thu, 21 Apr 2022 06:45:03 +0000 (16:45 +1000)]
Merge branch 'guilt/xfs-unsigned-flags-5.18' into xfs-5.19-for-next

2 years agoMerge branch 'guilt/5.19-miscellaneous' into xfs-5.19-for-next
Dave Chinner [Thu, 21 Apr 2022 01:40:17 +0000 (11:40 +1000)]
Merge branch 'guilt/5.19-miscellaneous' into xfs-5.19-for-next

2 years agoxfs: convert log ticket and iclog flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:48:01 +0000 (10:48 +1000)]
xfs: convert log ticket and iclog flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert shutdown reasons to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:47:38 +0000 (10:47 +1000)]
xfs: convert shutdown reasons to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert quota options flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:47:32 +0000 (10:47 +1000)]
xfs: convert quota options flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert ptag flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:47:25 +0000 (10:47 +1000)]
xfs: convert ptag flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert inode lock flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:47:16 +0000 (10:47 +1000)]
xfs: convert inode lock flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert log item tracepoint flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:47:07 +0000 (10:47 +1000)]
xfs: convert log item tracepoint flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert dquot flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:55 +0000 (10:46 +1000)]
xfs: convert dquot flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert da btree operations flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:47 +0000 (10:46 +1000)]
xfs: convert da btree operations flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert buffer log item flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:40 +0000 (10:46 +1000)]
xfs: convert buffer log item flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert btree buffer log flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:33 +0000 (10:46 +1000)]
xfs: convert btree buffer log flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

We also pass the fields to log to xfs_btree_offsets() as a uint32_t
all cases now. I have no idea why we made that parameter a int64_t
in the first place, but while we are fixing this up change it to
a uint32_t field, too.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert AGI log flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:24 +0000 (10:46 +1000)]
xfs: convert AGI log flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert AGF log flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:16 +0000 (10:46 +1000)]
xfs: convert AGF log flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert bmapi flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:09 +0000 (10:46 +1000)]
xfs: convert bmapi flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert bmap extent type flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:46:01 +0000 (10:46 +1000)]
xfs: convert bmap extent type flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert scrub type flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:45:52 +0000 (10:45 +1000)]
xfs: convert scrub type flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

This touches xfs_fs.h so affects the user API, but the user API
fields are also unsigned so the flags should really be unsigned,
too.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert attr type flags to unsigned.
Dave Chinner [Thu, 21 Apr 2022 00:45:41 +0000 (10:45 +1000)]
xfs: convert attr type flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: CIL context doesn't need to count iovecs
Dave Chinner [Thu, 21 Apr 2022 00:36:56 +0000 (10:36 +1000)]
xfs: CIL context doesn't need to count iovecs

Now that we account for log opheaders in the log item formatting
code, we don't actually use the aggregated count of log iovecs in
the CIL for anything. Remove it and the tracking code that
calculates it.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: xlog_write() doesn't need optype anymore
Dave Chinner [Thu, 21 Apr 2022 00:36:48 +0000 (10:36 +1000)]
xfs: xlog_write() doesn't need optype anymore

So remove it from the interface and callers.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: xlog_write() no longer needs contwr state
Dave Chinner [Thu, 21 Apr 2022 00:36:37 +0000 (10:36 +1000)]
xfs: xlog_write() no longer needs contwr state

The rework of xlog_write() no longer requires xlog_get_iclog_state()
to tell it about internal iclog space reservation state to direct it
on what to do. Remove this parameter.

$ size fs/xfs/xfs_log.o.*
   text    data     bss     dec     hex filename
  26520     560       8   27088    69d0 fs/xfs/xfs_log.o.orig
  26384     560       8   26952    6948 fs/xfs/xfs_log.o.patched

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: remove xlog_verify_dest_ptr
Christoph Hellwig [Thu, 21 Apr 2022 00:36:27 +0000 (10:36 +1000)]
xfs: remove xlog_verify_dest_ptr

Just check that the offset in xlog_write_vec is smaller than the iclog
size and remove the expensive cycling through all iclogs.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: introduce xlog_write_partial()
Dave Chinner [Thu, 21 Apr 2022 00:36:15 +0000 (10:36 +1000)]
xfs: introduce xlog_write_partial()

Re-implement writing of a log vector that does not fit into the
current iclog. The iclog will already be in XLOG_STATE_WANT_SYNC
because xlog_get_iclog_space() will have reserved all the remaining
iclog space for us, hence we can simply iterate over the iovecs in
the log vector getting more iclog space until the entire log vector
is written.

Handling this partial write case separately means we do need to pass
unnecessary state around for the common, fast path case when the log
vector fits entirely within the current iclog. It isolates the
complexity and allows us to modify and improve the partial write
case without impacting the simple fast path.

This change includes several improvements incorporated from patches
written by Christoph Hellwig.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: introduce xlog_write_full()
Dave Chinner [Thu, 21 Apr 2022 00:36:05 +0000 (10:36 +1000)]
xfs: introduce xlog_write_full()

Introduce an optimised version of xlog_write() that is used when the
entire write will fit in a single iclog. This greatly simplifies the
implementation of writing a log vector chain into an iclog, and sets
the ground work for a much more understandable xlog_write()
implementation.

This incorporates some factoring and simplifications proposed by
Christoph Hellwig.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: change the type of ic_datap
Christoph Hellwig [Thu, 21 Apr 2022 00:35:53 +0000 (10:35 +1000)]
xfs: change the type of ic_datap

Turn ic_datap from a char into a void pointer given that it points
to arbitrary data.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[dgc: also remove (char *) cast in xlog_alloc_log()]
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: pass lv chain length into xlog_write()
Dave Chinner [Thu, 21 Apr 2022 00:35:19 +0000 (10:35 +1000)]
xfs: pass lv chain length into xlog_write()

The caller of xlog_write() usually has a close accounting of the
aggregated vector length contained in the log vector chain passed to
xlog_write(). There is no need to iterate the chain to calculate he
length of the data in xlog_write_calculate_len() if the caller is
already iterating that chain to build it.

Passing in the vector length avoids doing an extra chain iteration,
which can be a significant amount of work given that large CIL
commits can have hundreds of thousands of vectors attached to the
chain.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: log ticket region debug is largely useless
Dave Chinner [Thu, 21 Apr 2022 00:35:09 +0000 (10:35 +1000)]
xfs: log ticket region debug is largely useless

xlog_tic_add_region() is used to trace the regions being added to a
log ticket to provide information in the situation where a ticket
reservation overrun occurs. The information gathered is stored int
the ticket, and dumped if xlog_print_tic_res() is called.

For a front end struct xfs_trans overrun, the ticket only contains
reservation tracking information - the ticket is never handed to the
log so has no regions attached to it. The overrun debug information in this
case comes from xlog_print_trans(), which walks the items attached
to the transaction and dumps their attached formatted log vectors
directly. It also dumps the ticket state, but that only contains
reservation accounting and nothing else. Hence xlog_print_tic_res()
never dumps region or overrun information from this path.

xlog_tic_add_region() is actually called from xlog_write(), which
means it is being used to track the regions seen in a
CIL checkpoint log vector chain. In looking at CIL behaviour
recently, I've seen 32MB checkpoints regularly exceed 250,000
regions in the LV chain. The log ticket debug code can track *15*
regions. IOWs, if there is a ticket overrun in the CIL code, the
ticket region tracking code is going to be completely useless for
determining what went wrong. The only thing it can tell us is how
much of an overrun occurred, and we really don't need extra debug
information in the log ticket to tell us that.

Indeed, the main place we call xlog_tic_add_region() is also adding
up the number of regions and the space used so that xlog_write()
knows how much will be written to the log. This is exactly the same
information that log ticket is storing once we take away the useless
region tracking array. Hence xlog_tic_add_region() is not useful,
but can be called 250,000 times a CIL push...

Just strip all that debug "information" out of the of the log ticket
and only have it report reservation space information when an
overrun occurs. This also reduces the size of a log ticket down by
about 150 bytes...

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: reserve space and initialise xlog_op_header in item formatting
Dave Chinner [Thu, 21 Apr 2022 00:34:59 +0000 (10:34 +1000)]
xfs: reserve space and initialise xlog_op_header in item formatting

Current xlog_write() adds op headers to the log manually for every
log item region that is in the vector passed to it. While
xlog_write() needs to stamp the transaction ID into the ophdr, we
already know it's length, flags, clientid, etc at CIL commit time.

This means the only time that xlog write really needs to format and
reserve space for a new ophdr is when a region is split across two
iclogs. Adding the opheader and accounting for it as part of the
normal formatted item region means we simplify the accounting
of space used by a transaction and we don't have to special case
reserving of space in for the ophdrs in xlog_write(). It also means
we can largely initialise the ophdr in transaction commit instead
of xlog_write, making the xlog_write formatting inner loop much
tighter.

xlog_prepare_iovec() is now too large to stay as an inline function,
so we move it out of line and into xfs_log.c.

Object sizes:
text    data     bss     dec     hex filename
1125934  305951     484 1432369  15db31 fs/xfs/built-in.a.before
1123360  305951     484 1429795  15d123 fs/xfs/built-in.a.after

So the code is a roughly 2.5kB smaller with xlog_prepare_iovec() now
out of line, even though it grew in size itself.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: move log iovec alignment to preparation function
Dave Chinner [Thu, 21 Apr 2022 00:34:49 +0000 (10:34 +1000)]
xfs: move log iovec alignment to preparation function

To include log op headers directly into the log iovec regions that
the ophdrs wrap, we need to move the buffer alignment code from
xlog_finish_iovec() to xlog_prepare_iovec(). This is because the
xlog_op_header is only 12 bytes long, and we need the buffer that
the caller formats their data into to be 8 byte aligned.

Hence once we start prepending the ophdr in xlog_prepare_iovec(), we
are going to need to manage the padding directly to ensure that the
buffer pointer returned is correctly aligned.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: log tickets don't need log client id
Dave Chinner [Thu, 21 Apr 2022 00:34:33 +0000 (10:34 +1000)]
xfs: log tickets don't need log client id

We currently set the log ticket client ID when we reserve a
transaction. This client ID is only ever written to the log by
a CIL checkpoint or unmount records, and so anything using a high
level transaction allocated through xfs_trans_alloc() does not need
a log ticket client ID to be set.

For the CIL checkpoint, the client ID written to the journal is
always XFS_TRANSACTION, and for the unmount record it is always
XFS_LOG, and nothing else writes to the log. All of these operations
tell xlog_write() exactly what they need to write to the log (the
optype) and build their own opheaders for start, commit and unmount
records. Hence we no longer need to set the client id in either the
log ticket or the xfs_trans.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: embed the xlog_op_header in the commit record
Dave Chinner [Thu, 21 Apr 2022 00:34:15 +0000 (10:34 +1000)]
xfs: embed the xlog_op_header in the commit record

Remove the final case where xlog_write() has to prepend an opheader
to a log transaction. Similar to the start record, the commit record
is just an empty opheader with a XLOG_COMMIT_TRANS type, so we can
just make this the payload for the region being passed to
xlog_write() and remove the special handling in xlog_write() for
the commit record.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: embed the xlog_op_header in the unmount record
Dave Chinner [Thu, 21 Apr 2022 00:34:04 +0000 (10:34 +1000)]
xfs: embed the xlog_op_header in the unmount record

Remove another case where xlog_write() has to prepend an opheader to
a log transaction. The unmount record + ophdr is smaller than the
minimum amount of space guaranteed to be free in an iclog (2 *
sizeof(ophdr)) and so we don't have to care about an unmount record
being split across 2 iclogs.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: only CIL pushes require a start record
Dave Chinner [Thu, 21 Apr 2022 00:33:48 +0000 (10:33 +1000)]
xfs: only CIL pushes require a start record

So move the one-off start record writing in xlog_write() out into
the static header that the CIL push builds to write into the log
initially. This simplifes the xlog_write() logic a lot.

pahole on x86-64 confirms that the xlog_cil_trans_hdr is correctly
32 bit aligned and packed for copying the log op and transaction
headers directly into the log as a single log region copy.

struct xlog_cil_trans_hdr {
        struct xlog_op_header      oph[2];               /*     0    24 */
        struct xfs_trans_header    thdr;                 /*    24    16 */
        struct xfs_log_iovec       lhdr[2];              /*    40    32 */

        /* size: 72, cachelines: 2, members: 3 */
        /* last cacheline: 8 bytes */
};

A wart is needed to handle the fact that length of the region the
opheader points to doesn't include the opheader length. hence if
we embed the opheader, we have to substract the opheader length from
the length written into the opheader by the generic copying code.
This will eventually go away when everything is converted to
embedded opheaders.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: factor out the CIL transaction header building
Dave Chinner [Thu, 21 Apr 2022 00:33:23 +0000 (10:33 +1000)]
xfs: factor out the CIL transaction header building

It is static code deep in the middle of the CIL push logic. Factor
it out into a helper so that it is clear and easy to modify
separately.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: simplify local variable assignment in file write code
Kaixu Xia [Wed, 20 Apr 2022 22:47:54 +0000 (08:47 +1000)]
xfs: simplify local variable assignment in file write code

Get the struct inode pointer from iocb->ki_filp->f_mapping->host
directly and the other variables are unnecessary, so simplify the
local variables assignment.

Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: reorder iunlink remove operation in xfs_ifree
Dave Chinner [Wed, 20 Apr 2022 22:45:16 +0000 (08:45 +1000)]
xfs: reorder iunlink remove operation in xfs_ifree

The O_TMPFILE creation implementation creates a specific order of
operations for inode allocation/freeing and unlinked list
modification. Currently both are serialised by the AGI, so the order
doesn't strictly matter as long as the are both in the same
transaction.

However, if we want to move the unlinked list insertions largely out
from under the AGI lock, then we have to be concerned about the
order in which we do unlinked list modification operations.
O_TMPFILE creation tells us this order is inode allocation/free,
then unlinked list modification.

Change xfs_ifree() to use this same ordering on unlinked list
removal. This way we always guarantee that when we enter the
iunlinked list removal code from this path, we already have the AGI
locked and we don't have to worry about lock nesting AGI reads
inside unlink list locks because it's already locked and attached to
the transaction.

We can do this safely as the inode freeing and unlinked list removal
are done in the same transaction and hence are atomic operations
with respect to log recovery.

Reported-by: Frank Hofmann <fhofmann@cloudflare.com>
Fixes: 298f7bec503f ("xfs: pin inode backing buffer to the inode log item")
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoMAINTAINERS: update IOMAP FILESYSTEM LIBRARY and XFS FILESYSTEM
Tiezhu Yang [Wed, 20 Apr 2022 22:45:14 +0000 (08:45 +1000)]
MAINTAINERS: update IOMAP FILESYSTEM LIBRARY and XFS FILESYSTEM

In IOMAP FILESYSTEM LIBRARY and XFS FILESYSTEM, the M(ail): entry is
redundant with the L(ist): entry, remove the redundant M(ail): entry.

Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: convert buffer flags to unsigned.
Dave Chinner [Wed, 20 Apr 2022 22:44:59 +0000 (08:44 +1000)]
xfs: convert buffer flags to unsigned.

5.18 w/ std=gnu11 compiled with gcc-5 wants flags stored in unsigned
fields to be unsigned. This manifests as a compiler error such as:

/kisskb/src/fs/xfs/./xfs_trace.h:432:2: note: in expansion of macro 'TP_printk'
  TP_printk("dev %d:%d daddr 0x%llx bbcount 0x%x hold %d pincount %d "
  ^
/kisskb/src/fs/xfs/./xfs_trace.h:440:5: note: in expansion of macro '__print_flags'
     __print_flags(__entry->flags, "|", XFS_BUF_FLAGS),
     ^
/kisskb/src/fs/xfs/xfs_buf.h:67:4: note: in expansion of macro 'XBF_UNMAPPED'
  { XBF_UNMAPPED,  "UNMAPPED" }
    ^
/kisskb/src/fs/xfs/./xfs_trace.h:440:40: note: in expansion of macro 'XFS_BUF_FLAGS'
     __print_flags(__entry->flags, "|", XFS_BUF_FLAGS),
                                        ^
/kisskb/src/fs/xfs/./xfs_trace.h: In function 'trace_raw_output_xfs_buf_flags_class':
/kisskb/src/fs/xfs/xfs_buf.h:46:23: error: initializer element is not constant
 #define XBF_UNMAPPED  (1 << 31)/* do not map the buffer */

as __print_flags assigns XFS_BUF_FLAGS to a structure that uses an
unsigned long for the flag. Since this results in the value of
XBF_UNMAPPED causing a signed integer overflow, the result is
technically undefined behavior, which gcc-5 does not accept as an
integer constant.

This is based on a patch from Arnd Bergman <arnd@arndb.de>.

Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandan.babu@oracle.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: Add XFS_SB_FEAT_INCOMPAT_NREXT64 to the list of supported flags
Chandan Babu R [Wed, 11 Aug 2021 05:03:20 +0000 (10:33 +0530)]
xfs: Add XFS_SB_FEAT_INCOMPAT_NREXT64 to the list of supported flags

This commit enables XFS module to work with fs instances having 64-bit
per-inode extent counters by adding XFS_SB_FEAT_INCOMPAT_NREXT64 flag to the
list of supported incompat feature flags.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Enable bulkstat ioctl to support 64-bit per-inode extent counters
Chandan Babu R [Wed, 9 Mar 2022 12:58:37 +0000 (12:58 +0000)]
xfs: Enable bulkstat ioctl to support 64-bit per-inode extent counters

The following changes are made to enable userspace to obtain 64-bit extent
counters,
1. Carve out a new 64-bit field xfs_bulkstat->bs_extents64 from
   xfs_bulkstat->bs_pad[] to hold 64-bit extent counter.
2. Define the new flag XFS_BULK_IREQ_BULKSTAT for userspace to indicate that
   it is capable of receiving 64-bit extent counters.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Suggested-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Decouple XFS_IBULK flags from XFS_IWALK flags
Chandan Babu R [Wed, 9 Mar 2022 12:34:04 +0000 (12:34 +0000)]
xfs: Decouple XFS_IBULK flags from XFS_IWALK flags

A future commit will add a new XFS_IBULK flag which will not have a
corresponding XFS_IWALK flag. In preparation for the change, this commit
separates XFS_IBULK_* flags from XFS_IWALK_* flags.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Conditionally upgrade existing inodes to use large extent counters
Chandan Babu R [Wed, 9 Mar 2022 07:49:36 +0000 (07:49 +0000)]
xfs: Conditionally upgrade existing inodes to use large extent counters

This commit enables upgrading existing inodes to use large extent counters
provided that underlying filesystem's superblock has large extent counter
feature enabled.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Directory's data fork extent counter can never overflow
Chandan Babu R [Tue, 29 Mar 2022 06:14:00 +0000 (06:14 +0000)]
xfs: Directory's data fork extent counter can never overflow

The maximum file size that can be represented by the data fork extent counter
in the worst case occurs when all extents are 1 block in length and each block
is 1KB in size.

With XFS_MAX_EXTCNT_DATA_FORK_SMALL representing maximum extent count and with
1KB sized blocks, a file can reach upto,
(2^31) * 1KB = 2TB

This is much larger than the theoretical maximum size of a directory
i.e. XFS_DIR2_SPACE_SIZE * 3 = ~96GB.

Since a directory's inode can never overflow its data fork extent counter,
this commit removes all the overflow checks associated with
it. xfs_dinode_verify() now performs a rough check to verify if a diretory's
data fork is larger than 96GB.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: use a separate frextents counter for rt extent reservations
Darrick J. Wong [Mon, 11 Apr 2022 20:49:42 +0000 (06:49 +1000)]
xfs: use a separate frextents counter for rt extent reservations

As mentioned in the previous commit, the kernel misuses sb_frextents in
the incore mount to reflect both incore reservations made by running
transactions as well as the actual count of free rt extents on disk.
This results in the superblock being written to the log with an
underestimate of the number of rt extents that are marked free in the
rtbitmap.

Teaching XFS to recompute frextents after log recovery avoids
operational problems in the current mount, but it doesn't solve the
problem of us writing undercounted frextents which are then recovered by
an older kernel that doesn't have that fix.

Create an incore percpu counter to mirror the ondisk frextents.  This
new counter will track transaction reservations and the only time we
will touch the incore super counter (i.e the one that gets logged) is
when those transactions commit updates to the rt bitmap.  This is in
contrast to the lazysbcount counters (e.g. fdblocks), where we know that
log recovery will always fix any incorrect counter that we log.
As a bonus, we only take m_sb_lock at transaction commit time.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: recalculate free rt extents after log recovery
Darrick J. Wong [Mon, 11 Apr 2022 20:49:42 +0000 (06:49 +1000)]
xfs: recalculate free rt extents after log recovery

I've been observing periodic corruption reports from xfs_scrub involving
the free rt extent counter (frextents) while running xfs/141.  That test
uses an error injection knob to induce a torn write to the log, and an
arbitrary number of recovery mounts, frextents will count fewer free rt
extents than can be found the rtbitmap.

The root cause of the problem is a combination of the misuse of
sb_frextents in the incore mount to reflect both incore reservations
made by running transactions as well as the actual count of free rt
extents on disk.  The following sequence can reproduce the undercount:

Thread 1 Thread 2
xfs_trans_alloc(rtextents=3)
xfs_mod_frextents(-3)
<blocks>
xfs_attr_set()
xfs_bmap_attr_addfork()
xfs_add_attr2()
xfs_log_sb()
xfs_sb_to_disk()
xfs_trans_commit()
<log flushed to disk>
<log goes down>

Note that thread 1 subtracts 3 from sb_frextents even though it never
commits to using that space.  Thread 2 writes the undercounted value to
the ondisk superblock and logs it to the xattr transaction, which is
then flushed to disk.  At next mount, log recovery will find the logged
superblock and write that back into the filesystem.  At the end of log
recovery, we reread the superblock and install the recovered
undercounted frextents value into the incore superblock.  From that
point on, we've effectively leaked thread 1's transaction reservation.

The correct fix for this is to separate the incore reservation from the
ondisk usage, but that's a matter for the next patch.  Because the
kernel has been logging superblocks with undercounted frextents for a
very long time and we don't demand that sysadmins run xfs_repair after a
crash, fix the undercount by recomputing frextents after log recovery.

Gating this on log recovery is a reasonable balance (I think) between
correcting the problem and slowing down every mount attempt.  Note that
xfs_repair will fix undercounted frextents.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: pass explicit mount pointer to rtalloc query functions
Darrick J. Wong [Mon, 11 Apr 2022 20:49:41 +0000 (06:49 +1000)]
xfs: pass explicit mount pointer to rtalloc query functions

Pass an explicit xfs_mount pointer to the rtalloc query functions so
that they can support transactionless queries.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: Use generic_file_open()
Matthew Wilcox (Oracle) [Mon, 11 Apr 2022 20:49:40 +0000 (06:49 +1000)]
xfs: Use generic_file_open()

Remove the open-coded check of O_LARGEFILE.  This changes the errno
to be the same as other filesystems; it was changed generically in
2.6.24 but that fix skipped XFS.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2 years agoxfs: Introduce per-inode 64-bit extent counters
Chandan Babu R [Tue, 8 Mar 2022 09:34:28 +0000 (09:34 +0000)]
xfs: Introduce per-inode 64-bit extent counters

This commit introduces new fields in the on-disk inode format to support
64-bit data fork extent counters and 32-bit attribute fork extent
counters. The new fields will be used only when an inode has
XFS_DIFLAG2_NREXT64 flag set. Otherwise we continue to use the regular 32-bit
data fork extent counters and 16-bit attribute fork extent counters.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
Suggested-by: Dave Chinner <dchinner@redhat.com>
2 years agoxfs: Replace numbered inode recovery error messages with descriptive ones
Chandan Babu R [Tue, 8 Mar 2022 09:19:46 +0000 (09:19 +0000)]
xfs: Replace numbered inode recovery error messages with descriptive ones

This commit also prints inode fields with invalid values instead of printing
addresses of inode and buffer instances.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
Suggested-by: Dave Chinner <dchinner@redhat.com>
2 years agoxfs: Introduce macros to represent new maximum extent counts for data/attr forks
Chandan Babu R [Tue, 16 Nov 2021 09:54:37 +0000 (09:54 +0000)]
xfs: Introduce macros to represent new maximum extent counts for data/attr forks

This commit defines new macros to represent maximum extent counts allowed by
filesystems which have support for large per-inode extent counters.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Use uint64_t to count maximum blocks that can be used by BMBT
Chandan Babu R [Tue, 16 Nov 2021 09:20:01 +0000 (09:20 +0000)]
xfs: Use uint64_t to count maximum blocks that can be used by BMBT

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Introduce XFS_DIFLAG2_NREXT64 and associated helpers
Chandan Babu R [Tue, 16 Nov 2021 09:04:43 +0000 (09:04 +0000)]
xfs: Introduce XFS_DIFLAG2_NREXT64 and associated helpers

This commit adds the new per-inode flag XFS_DIFLAG2_NREXT64 to indicate that
an inode supports 64-bit extent counters. This flag is also enabled by default
on newly created inodes when the corresponding filesystem has large extent
counter feature bit (i.e. XFS_FEAT_NREXT64) set.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Introduce XFS_FSOP_GEOM_FLAGS_NREXT64
Chandan Babu R [Tue, 16 Nov 2021 07:56:54 +0000 (07:56 +0000)]
xfs: Introduce XFS_FSOP_GEOM_FLAGS_NREXT64

XFS_FSOP_GEOM_FLAGS_NREXT64 indicates that the current filesystem instance
supports 64-bit per-inode extent counters.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Introduce XFS_SB_FEAT_INCOMPAT_NREXT64 and associated per-fs feature bit
Chandan Babu R [Tue, 16 Nov 2021 08:39:32 +0000 (08:39 +0000)]
xfs: Introduce XFS_SB_FEAT_INCOMPAT_NREXT64 and associated per-fs feature bit

XFS_SB_FEAT_INCOMPAT_NREXT64 incompat feature bit will be set on filesystems
which support large per-inode extent counters. This commit defines the new
incompat feature bit and the corresponding per-fs feature bit (along with
inline functions to work on it).

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Promote xfs_extnum_t and xfs_aextnum_t to 64 and 32-bits respectively
Chandan Babu R [Tue, 16 Nov 2021 07:28:40 +0000 (07:28 +0000)]
xfs: Promote xfs_extnum_t and xfs_aextnum_t to 64 and 32-bits respectively

A future commit will introduce a 64-bit on-disk data extent counter and a
32-bit on-disk attr extent counter. This commit promotes xfs_extnum_t and
xfs_aextnum_t to 64 and 32-bits in order to correctly handle in-core versions
of these quantities.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Use basic types to define xfs_log_dinode's di_nextents and di_anextents
Chandan Babu R [Tue, 16 Nov 2021 07:20:39 +0000 (07:20 +0000)]
xfs: Use basic types to define xfs_log_dinode's di_nextents and di_anextents

A future commit will increase the width of xfs_extnum_t in order to facilitate
larger per-inode extent counters. Hence this patch now uses basic types to
define xfs_log_dinode->[di_nextents|dianextents].

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Introduce xfs_dfork_nextents() helper
Chandan Babu R [Thu, 27 Aug 2020 10:04:34 +0000 (15:34 +0530)]
xfs: Introduce xfs_dfork_nextents() helper

This commit replaces the macro XFS_DFORK_NEXTENTS() with the helper function
xfs_dfork_nextents(). As of this commit, xfs_dfork_nextents() returns the same
value as XFS_DFORK_NEXTENTS(). A future commit which extends inode's extent
counter fields will add more logic to this helper.

This commit also replaces direct accesses to xfs_dinode->di_[a]nextents
with calls to xfs_dfork_nextents().

No functional changes have been made.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Use xfs_extnum_t instead of basic data types
Chandan Babu R [Fri, 26 Feb 2021 05:54:31 +0000 (11:24 +0530)]
xfs: Use xfs_extnum_t instead of basic data types

xfs_extnum_t is the type to use to declare variables which have values
obtained from xfs_dinode->di_[a]nextents. This commit replaces basic
types (e.g. uint32_t) with xfs_extnum_t for such variables.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Introduce xfs_iext_max_nextents() helper
Chandan Babu R [Thu, 27 Aug 2020 09:39:10 +0000 (15:09 +0530)]
xfs: Introduce xfs_iext_max_nextents() helper

xfs_iext_max_nextents() returns the maximum number of extents possible for one
of data, cow or attribute fork. This helper will be extended further in a
future commit when maximum extent counts associated with data/attribute forks
are increased.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Define max extent length based on on-disk format definition
Chandan Babu R [Mon, 9 Aug 2021 06:35:22 +0000 (12:05 +0530)]
xfs: Define max extent length based on on-disk format definition

The maximum extent length depends on maximum block count that can be stored in
a BMBT record. Hence this commit defines MAXEXTLEN based on
BMBT_BLOCKCOUNT_BITLEN.

While at it, the commit also renames MAXEXTLEN to XFS_MAX_BMBT_EXTLEN.

Suggested-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Move extent count limits to xfs_format.h
Chandan Babu R [Wed, 7 Oct 2020 09:30:03 +0000 (15:00 +0530)]
xfs: Move extent count limits to xfs_format.h

Maximum values associated with extent counters i.e. Maximum extent length,
Maximum data extents and Maximum xattr extents are dictated by the on-disk
format. Hence move these definitions over to xfs_format.h.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Chandan Babu R <chandan.babu@oracle.com>
2 years agoxfs: Add XFS messages to printk index
Jonathan Lassoff [Mon, 11 Apr 2022 03:06:39 +0000 (13:06 +1000)]
xfs: Add XFS messages to printk index

In order for end users to quickly react to new issues that come up in
production, it is proving useful to leverage the printk indexing system.

This printk index enables kernel developers to use calls to printk()
with changeable format strings (as they always have; no change of
expectations), while enabling end users to examine format strings to
detect changes.
Since end users are using regular expressions to match messages printed
through printk(), being able to detect changes in chosen format strings
from release to release provides a useful signal to review
printk()-matching regular expressions for any necessary updates.

So that detailed XFS messages are captures by this printk index, this
patch wraps the xfs_<level> and xfs_alert_tag functions.

Signed-off-by: Jonathan Lassoff <jof@thejof.com>
Reviewed-by: Chris Down <chris@chrisdown.name>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>