Commit Graph

70838 Commits

Author SHA1 Message Date
Brian Foster
84d8949e77 xfs: hold buffer across unpin and potential shutdown processing
The special processing used to simulate a buffer I/O failure on fs
shutdown has a difficult to reproduce race that can result in a use
after free of the associated buffer. Consider a buffer that has been
committed to the on-disk log and thus is AIL resident. The buffer
lands on the writeback delwri queue, but is subsequently locked,
committed and pinned by another transaction before submitted for
I/O. At this point, the buffer is stuck on the delwri queue as it
cannot be submitted for I/O until it is unpinned. A log checkpoint
I/O failure occurs sometime later, which aborts the bli. The unpin
handler is called with the aborted log item, drops the bli reference
count, the pin count, and falls into the I/O failure simulation
path.

The potential problem here is that once the pin count falls to zero
in ->iop_unpin(), xfsaild is free to retry delwri submission of the
buffer at any time, before the unpin handler even completes. If
delwri queue submission wins the race to the buffer lock, it
observes the shutdown state and simulates the I/O failure itself.
This releases both the bli and delwri queue holds and frees the
buffer while xfs_buf_item_unpin() sits on xfs_buf_lock() waiting to
run through the same failure sequence. This problem is rare and
requires many iterations of fstest generic/019 (which simulates disk
I/O failures) to reproduce.

To avoid this problem, grab a hold on the buffer before the log item
is unpinned if the associated item has been aborted and will require
a simulated I/O failure. The hold is already required for the
simulated I/O failure, so the ordering simply guarantees the unpin
handler access to the buffer before it is unpinned and thus
processed by the AIL. This particular ordering is required so long
as the AIL does not acquire a reference on the bli, which is the
long term solution to this problem.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:14:24 -07:00
Darrick J. Wong
4e6b8270c8 xfs: force the log offline when log intent item recovery fails
If any part of log intent item recovery fails, we should shut down the
log immediately to stop the log from writing a clean unmount record to
disk, because the metadata is not consistent.  The inability to cancel a
dirty transaction catches most of these cases, but there are a few
things that have slipped through the cracks, such as ENOSPC from a
transaction allocation, or runtime errors that result in cancellation of
a non-dirty transaction.

This solves some weird behaviors reported by customers where a system
goes down, the first mount fails, the second succeeds, but then the fs
goes down later because of inconsistent metadata.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2021-06-21 10:14:24 -07:00
Darrick J. Wong
81ed94751b xfs: fix log intent recovery ENOSPC shutdowns when inactivating inodes
During regular operation, the xfs_inactive operations create
transactions with zero block reservation because in general we're
freeing space, not asking for more.  The per-AG space reservations
created at mount time enable us to handle expansions of the refcount
btree without needing to reserve blocks to the transaction.

Unfortunately, log recovery doesn't create the per-AG space reservations
when intent items are being recovered.  This isn't an issue for intent
item recovery itself because they explicitly request blocks, but any
inode inactivation that can happen during log recovery uses the same
xfs_inactive paths as regular runtime.  If a refcount btree expansion
happens, the transaction will fail due to blk_res_used > blk_res, and we
shut down the filesystem unnecessarily.

Fix this problem by making per-AG reservations temporarily so that we
can handle the inactivations, and releasing them at the end.  This
brings the recovery environment closer to the runtime environment.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2021-06-21 10:14:24 -07:00
Darrick J. Wong
c06ad17cfa xfs: shorten the shutdown messages to a single line
Consolidate the shutdown messages to a single line containing the
reason, the passed-in flags, the source of the shutdown, and the end
result.  This means we now only have one line to look for when
debugging, which is useful when the fs goes down while something else is
flooding dmesg.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
2021-06-21 10:14:13 -07:00
Darrick J. Wong
3a1c3abe89 xfs: print name of function causing fs shutdown instead of hex pointer
In xfs_do_force_shutdown, print the symbolic name of the function that
called us to shut down the filesystem instead of a raw hex pointer.
This makes debugging a lot easier:

XFS (sda): xfs_do_force_shutdown(0x2) called from line 2440 of file
	fs/xfs/xfs_log.c. Return address = ffffffffa038bc38

becomes:

XFS (sda): xfs_do_force_shutdown(0x2) called from line 2440 of file
	fs/xfs/xfs_log.c. Return address = xfs_trans_mod_sb+0x25

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
2021-06-21 10:13:57 -07:00
Darrick J. Wong
10be350b8c xfs: fix type mismatches in the inode reclaim functions
It's currently unlikely that we will ever end up with more than 4
billion inodes waiting for reclamation, but the fs object code uses long
int for object counts and we're certainly capable of generating that
many.  Instead of truncating the internal counters, widen them and
report the object counts correctly.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-21 10:12:46 -07:00
Darrick J. Wong
77b4d2861e xfs: separate primary inode selection criteria in xfs_iget_cache_hit
During review of the v6 deferred inode inactivation patchset[1], Dave
commented that _cache_hit should have a clear separation between inode
selection criteria and actions performed on a selected inode.  Move a
hunk to make this true, and compact the shrink cases in the function.

[1] https://lore.kernel.org/linux-xfs/162310469340.3465262.504398465311182657.stgit@locust/T/#mca6d958521cb88bbc1bfe1a30767203328d410b5
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Brian Foster <bfoster@redhat.com>
2021-06-21 10:12:46 -07:00
Darrick J. Wong
ff7bebeb91 xfs: refactor the inode recycling code
Hoist the code in xfs_iget_cache_hit that restores the VFS inode state
to an xfs_inode that was previously vfs-destroyed.  The next patch will
add a new set of state flags, so we need the helper to avoid
duplication.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-21 10:12:46 -07:00
Dave Chinner
956f6daa84 xfs: add iclog state trace events
For the DEBUGS!

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: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:12:38 -07:00
Dave Chinner
5f9b4b0de8 xfs: xfs_log_force_lsn isn't passed a LSN
In doing an investigation into AIL push stalls, I was looking at the
log force code to see if an async CIL push could be done instead.
This lead me to xfs_log_force_lsn() and looking at how it works.

xfs_log_force_lsn() is only called from inode synchronisation
contexts such as fsync(), and it takes the ip->i_itemp->ili_last_lsn
value as the LSN to sync the log to. This gets passed to
xlog_cil_force_lsn() via xfs_log_force_lsn() to flush the CIL to the
journal, and then used by xfs_log_force_lsn() to flush the iclogs to
the journal.

The problem is that ip->i_itemp->ili_last_lsn does not store a
log sequence number. What it stores is passed to it from the
->iop_committing method, which is called by xfs_log_commit_cil().
The value this passes to the iop_committing method is the CIL
context sequence number that the item was committed to.

As it turns out, xlog_cil_force_lsn() converts the sequence to an
actual commit LSN for the related context and returns that to
xfs_log_force_lsn(). xfs_log_force_lsn() overwrites it's "lsn"
variable that contained a sequence with an actual LSN and then uses
that to sync the iclogs.

This caused me some confusion for a while, even though I originally
wrote all this code a decade ago. ->iop_committing is only used by
a couple of log item types, and only inode items use the sequence
number it is passed.

Let's clean up the API, CIL structures and inode log item to call it
a sequence number, and make it clear that the high level code is
using CIL sequence numbers and not on-disk LSNs for integrity
synchronisation purposes.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:12:33 -07:00
Dave Chinner
19f4e7cc81 xfs: Fix CIL throttle hang when CIL space used going backwards
A hang with tasks stuck on the CIL hard throttle was reported and
largely diagnosed by Donald Buczek, who discovered that it was a
result of the CIL context space usage decrementing in committed
transactions once the hard throttle limit had been hit and processes
were already blocked.  This resulted in the CIL push not waking up
those waiters because the CIL context was no longer over the hard
throttle limit.

The surprising aspect of this was the CIL space usage going
backwards regularly enough to trigger this situation. Assumptions
had been made in design that the relogging process would only
increase the size of the objects in the CIL, and so that space would
only increase.

This change and commit message fixes the issue and documents the
result of an audit of the triggers that can cause the CIL space to
go backwards, how large the backwards steps tend to be, the
frequency in which they occur, and what the impact on the CIL
accounting code is.

Even though the CIL ctx->space_used can go backwards, it will only
do so if the log item is already logged to the CIL and contains a
space reservation for it's entire logged state. This is tracked by
the shadow buffer state on the log item. If the item is not
previously logged in the CIL it has no shadow buffer nor log vector,
and hence the entire size of the logged item copied to the log
vector is accounted to the CIL space usage. i.e.  it will always go
up in this case.

If the item has a log vector (i.e. already in the CIL) and the size
decreases, then the existing log vector will be overwritten and the
space usage will go down. This is the only condition where the space
usage reduces, and it can only occur when an item is already tracked
in the CIL. Hence we are safe from CIL space usage underruns as a
result of log items decreasing in size when they are relogged.

Typically this reduction in CIL usage occurs from metadata blocks
being free, such as when a btree block merge occurs or a directory
enter/xattr entry is removed and the da-tree is reduced in size.
This generally results in a reduction in size of around a single
block in the CIL, but also tends to increase the number of log
vectors because the parent and sibling nodes in the tree needs to be
updated when a btree block is removed. If a multi-level merge
occurs, then we see reduction in size of 2+ blocks, but again the
log vector count goes up.

The other vector is inode fork size changes, which only log the
current size of the fork and ignore the previously logged size when
the fork is relogged. Hence if we are removing items from the inode
fork (dir/xattr removal in shortform, extent record removal in
extent form, etc) the relogged size of the inode for can decrease.

No other log items can decrease in size either because they are a
fixed size (e.g. dquots) or they cannot be relogged (e.g. relogging
an intent actually creates a new intent log item and doesn't relog
the old item at all.) Hence the only two vectors for CIL context
size reduction are relogging inode forks and marking buffers active
in the CIL as stale.

Long story short: the majority of the code does the right thing and
handles the reduction in log item size correctly, and only the CIL
hard throttle implementation is problematic and needs fixing. This
patch makes that fix, as well as adds comments in the log item code
that result in items shrinking in size when they are relogged as a
clear reminder that this can and does happen frequently.

The throttle fix is based upon the change Donald proposed, though it
goes further to ensure that once the throttle is activated, it
captures all tasks until the CIL push issues a wakeup, regardless of
whether the CIL space used has gone back under the throttle
threshold.

This ensures that we prevent tasks reducing the CIL slightly under
the throttle threshold and then making more changes that push it
well over the throttle limit. This is acheived by checking if the
throttle wait queue is already active as a condition of throttling.
Hence once we start throttling, we continue to apply the throttle
until the CIL context push wakes everything on the wait queue.

We can use waitqueue_active() for the waitqueue manipulations and
checks as they are all done under the ctx->xc_push_lock. Hence the
waitqueue has external serialisation and we can safely peek inside
the wait queue without holding the internal waitqueue locks.

Many thanks to Donald for his diagnostic and analysis work to
isolate the cause of this hang.

Reported-and-tested-by: Donald Buczek <buczek@molgen.mpg.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:06:14 -07:00
Dave Chinner
eef983ffea xfs: journal IO cache flush reductions
Currently every journal IO is issued as REQ_PREFLUSH | REQ_FUA to
guarantee the ordering requirements the journal has w.r.t. metadata
writeback. THe two ordering constraints are:

1. we cannot overwrite metadata in the journal until we guarantee
that the dirty metadata has been written back in place and is
stable.

2. we cannot write back dirty metadata until it has been written to
the journal and guaranteed to be stable (and hence recoverable) in
the journal.

The ordering guarantees of  are provided by REQ_PREFLUSH. This
causes the journal IO to issue a cache flush and wait for it to
complete before issuing the write IO to the journal. Hence all
completed metadata IO is guaranteed to be stable before the journal
overwrites the old metadata.

The ordering guarantees of  are provided by the REQ_FUA, which
ensures the journal writes do not complete until they are on stable
storage. Hence by the time the last journal IO in a checkpoint
completes, we know that the entire checkpoint is on stable storage
and we can unpin the dirty metadata and allow it to be written back.

This is the mechanism by which ordering was first implemented in XFS
way back in 2002 by commit 95d97c36e5155075ba2eb22b17562cfcc53fcf96
("Add support for drive write cache flushing") in the xfs-archive
tree.

A lot has changed since then, most notably we now use delayed
logging to checkpoint the filesystem to the journal rather than
write each individual transaction to the journal. Cache flushes on
journal IO are necessary when individual transactions are wholly
contained within a single iclog. However, CIL checkpoints are single
transactions that typically span hundreds to thousands of individual
journal writes, and so the requirements for device cache flushing
have changed.

That is, the ordering rules I state above apply to ordering of
atomic transactions recorded in the journal, not to the journal IO
itself. Hence we need to ensure metadata is stable before we start
writing a new transaction to the journal (guarantee ), and we need
to ensure the entire transaction is stable in the journal before we
start metadata writeback (guarantee ).

Hence we only need a REQ_PREFLUSH on the journal IO that starts a
new journal transaction to provide , and it is not on any other
journal IO done within the context of that journal transaction.

The CIL checkpoint already issues a cache flush before it starts
writing to the log, so we no longer need the iclog IO to issue a
REQ_REFLUSH for us. Hence if XLOG_START_TRANS is passed
to xlog_write(), we no longer need to mark the first iclog in
the log write with REQ_PREFLUSH for this case. As an added bonus,
this ordering mechanism works for both internal and external logs,
meaning we can remove the explicit data device cache flushes from
the iclog write code when using external logs.

Given the new ordering semantics of commit records for the CIL, we
need iclogs containing commit records to issue a REQ_PREFLUSH. We
also require unmount records to do this. Hence for both
XLOG_COMMIT_TRANS and XLOG_UNMOUNT_TRANS xlog_write() calls we need
to mark the first iclog being written with REQ_PREFLUSH.

For both commit records and unmount records, we also want them
immediately on stable storage, so we want to also mark the iclogs
that contain these records to be marked REQ_FUA. That means if a
record is split across multiple iclogs, they are all marked REQ_FUA
and not just the last one so that when the transaction is completed
all the parts of the record are on stable storage.

And for external logs, unmount records need a pre-write data device
cache flush similar to the CIL checkpoint cache pre-flush as the
internal iclog write code does not do this implicitly anymore.

As an optimisation, when the commit record lands in the same iclog
as the journal transaction starts, we don't need to wait for
anything and can simply use REQ_FUA to provide guarantee .  This
means that for fsync() heavy workloads, the cache flush behaviour is
completely unchanged and there is no degradation in performance as a
result of optimise the multi-IO transaction case.

The most notable sign that there is less IO latency on my test
machine (nvme SSDs) is that the "noiclogs" rate has dropped
substantially. This metric indicates that the CIL push is blocking
in xlog_get_iclog_space() waiting for iclog IO completion to occur.
With 8 iclogs of 256kB, the rate is appoximately 1 noiclog event to
every 4 iclog writes. IOWs, every 4th call to xlog_get_iclog_space()
is blocking waiting for log IO. With the changes in this patch, this
drops to 1 noiclog event for every 100 iclog writes. Hence it is
clear that log IO is completing much faster than it was previously,
but it is also clear that for large iclog sizes, this isn't the
performance limiting factor on this hardware.

With smaller iclogs (32kB), however, there is a substantial
difference. With the cache flush modifications, the journal is now
running at over 4000 write IOPS, and the journal throughput is
largely identical to the 256kB iclogs and the noiclog event rate
stays low at about 1:50 iclog writes. The existing code tops out at
about 2500 IOPS as the number of cache flushes dominate performance
and latency. The noiclog event rate is about 1:4, and the
performance variance is quite large as the journal throughput can
fall to less than half the peak sustained rate when the cache flush
rate prevents metadata writeback from keeping up and the log runs
out of space and throttles reservations.

As a result:

	logbsize	fsmark create rate	rm -rf
before	32kb		152851+/-5.3e+04	5m28s
patched	32kb		221533+/-1.1e+04	5m24s

before	256kb		220239+/-6.2e+03	4m58s
patched	256kb		228286+/-9.2e+03	5m06s

The rm -rf times are included because I ran them, but the
differences are largely noise. This workload is largely metadata
read IO latency bound and the changes to the journal cache flushing
doesn't really make any noticable difference to behaviour apart from
a reduction in noiclog events from background CIL pushing.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:06:08 -07:00
Dave Chinner
3468bb1ca6 xfs: remove need_start_rec parameter from xlog_write()
The CIL push is the only call to xlog_write that sets this variable
to true. The other callers don't need a start rec, and they tell
xlog_write what to do by passing the type of ophdr they need written
in the flags field. The need_start_rec parameter essentially tells
xlog_write to to write an extra ophdr with a XLOG_START_TRANS type,
so get rid of the variable to do this and pass XLOG_START_TRANS as
the flag value into xlog_write() from the CIL push.

$ size fs/xfs/xfs_log.o*
  text	   data	    bss	    dec	    hex	filename
 27595	    560	      8	  28163	   6e03	fs/xfs/xfs_log.o.orig
 27454	    560	      8	  28022	   6d76	fs/xfs/xfs_log.o.patched

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:06:03 -07:00
Dave Chinner
bad77c375e xfs: CIL checkpoint flushes caches unconditionally
Currently every journal IO is issued as REQ_PREFLUSH | REQ_FUA to
guarantee the ordering requirements the journal has w.r.t. metadata
writeback. THe two ordering constraints are:

1. we cannot overwrite metadata in the journal until we guarantee
that the dirty metadata has been written back in place and is
stable.

2. we cannot write back dirty metadata until it has been written to
the journal and guaranteed to be stable (and hence recoverable) in
the journal.

These rules apply to the atomic transactions recorded in the
journal, not to the journal IO itself. Hence we need to ensure
metadata is stable before we start writing a new transaction to the
journal (guarantee ), and we need to ensure the entire transaction
is stable in the journal before we start metadata writeback
(guarantee ).

The ordering guarantees of  are currently provided by REQ_PREFLUSH
being added to every iclog IO. This causes the journal IO to issue a
cache flush and wait for it to complete before issuing the write IO
to the journal. Hence all completed metadata IO is guaranteed to be
stable before the journal overwrites the old metadata.

However, for long running CIL checkpoints that might do a thousand
journal IOs, we don't need every single one of these iclog IOs to
issue a cache flush - the cache flush done before the first iclog is
submitted is sufficient to cover the entire range in the log that
the checkpoint will overwrite because the CIL space reservation
guarantees the tail of the log (completed metadata) is already
beyond the range of the checkpoint write.

Hence we only need a full cache flush between closing off the CIL
checkpoint context (i.e. when the push switches it out) and issuing
the first journal IO. Rather than plumbing this through to the
journal IO, we can start this cache flush the moment the CIL context
is owned exclusively by the push worker. The cache flush can be in
progress while we process the CIL ready for writing, hence
reducing the latency of the initial iclog write. This is especially
true for large checkpoints, where we might have to process hundreds
of thousands of log vectors before we issue the first iclog write.
In these cases, it is likely the cache flush has already been
completed by the time we have built the CIL log vector chain.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:05:57 -07:00
Dave Chinner
0431d926b3 xfs: async blkdev cache flush
The new checkpoint cache flush mechanism requires us to issue an
unconditional cache flush before we start a new checkpoint. We don't
want to block for this if we can help it, and we have a fair chunk
of CPU work to do between starting the checkpoint and issuing the
first journal IO.

Hence it makes sense to amortise the latency cost of the cache flush
by issuing it asynchronously and then waiting for it only when we
need to issue the first IO in the transaction.

To do this, we need async cache flush primitives to submit the cache
flush bio and to wait on it. The block layer has no such primitives
for filesystems, so roll our own for the moment.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:05:51 -07:00
Dave Chinner
b5071ada51 xfs: remove xfs_blkdev_issue_flush
It's a one line wrapper around blkdev_issue_flush(). Just replace it
with direct calls to blkdev_issue_flush().

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-21 10:05:46 -07:00
Dave Chinner
a79b28c284 xfs: separate CIL commit record IO
To allow for iclog IO device cache flush behaviour to be optimised,
we first need to separate out the commit record iclog IO from the
rest of the checkpoint so we can wait for the checkpoint IO to
complete before we issue the commit record.

This separation is only necessary if the commit record is being
written into a different iclog to the start of the checkpoint as the
upcoming cache flushing changes requires completion ordering against
the other iclogs submitted by the checkpoint.

If the entire checkpoint and commit is in the one iclog, then they
are both covered by the one set of cache flush primitives on the
iclog and hence there is no need to separate them for ordering.

Otherwise, we need to wait for all the previous iclogs to complete
so they are ordered correctly and made stable by the REQ_PREFLUSH
that the commit record iclog IO issues. This guarantees that if a
reader sees the commit record in the journal, they will also see the
entire checkpoint that commit record closes off.

This also provides the guarantee that when the commit record IO
completes, we can safely unpin all the log items in the checkpoint
so they can be written back because the entire checkpoint is stable
in the journal.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-18 08:24:23 -07:00
Geert Uytterhoeven
18842e0a4f xfs: Fix 64-bit division on 32-bit in xlog_state_switch_iclogs()
On 32-bit (e.g. m68k):

    ERROR: modpost: "__udivdi3" [fs/xfs/xfs.ko] undefined!

Fix this by using a uint32_t intermediate, like before.

Reported-by: noreply@ellerman.id.au
Fixes: 7660a5b48fbef958 ("xfs: log stripe roundoff is a property of the log")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-18 08:24:19 -07:00
Dave Chinner
a6a65fef5e xfs: log stripe roundoff is a property of the log
We don't need to look at the xfs_mount and superblock every time we
need to do an iclog roundoff calculation. The property is fixed for
the life of the log, so store the roundoff in the log at mount time
and use that everywhere.

On a debug build:

$ size fs/xfs/xfs_log.o.*
   text	   data	    bss	    dec	    hex	filename
  27360	    560	      8	  27928	   6d18	fs/xfs/xfs_log.o.orig
  27219	    560	      8	  27787	   6c8b	fs/xfs/xfs_log.o.patched

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
2021-06-18 08:21:48 -07:00
Shaokun Zhang
9bb38aa080 xfs: remove redundant initialization of variable error
'error' will be initialized, so clean up the redundant initialization.

Cc: "Darrick J. Wong" <djwong@kernel.org>
Signed-off-by: Shaokun Zhang <zhangshaokun@hisilicon.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-18 08:14:31 -07:00
Dave Chinner
90e2c1c20a xfs: perag may be null in xfs_imap()
Dan Carpenter's static checker reported:

The patch 7b13c51551: "xfs: use perag for ialloc btree cursors"
from Jun 2, 2021, leads to the following Smatch complaint:

    fs/xfs/libxfs/xfs_ialloc.c:2403 xfs_imap()
    error: we previously assumed 'pag' could be null (see line 2294)

And it's right. Fix it.

Fixes: 7b13c51551 ("xfs: use perag for ialloc btree cursors")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
2021-06-18 08:14:20 -07:00
Darrick J. Wong
d1015e2ebd Merge tag 'xfs-delay-ready-attrs-v20.1' of https://github.com/allisonhenderson/xfs_work into xfs-5.14-merge4
xfs: Delay Ready Attributes

Hi all,

This set is a subset of a larger series for Dealyed Attributes. Which is a
subset of a yet larger series for parent pointers. Delayed attributes allow
attribute operations (set and remove) to be logged and committed in the same
way that other delayed operations do. This allows more complex operations (like
parent pointers) to be broken up into multiple smaller transactions. To do
this, the existing attr operations must be modified to operate as a delayed
operation.  This means that they cannot roll, commit, or finish transactions.
Instead, they return -EAGAIN to allow the calling function to handle the
transaction.  In this series, we focus on only the delayed attribute portion.
We will introduce parent pointers in a later set.

The set as a whole is a bit much to digest at once, so I usually send out the
smaller sub series to reduce reviewer burn out.  But the entire extended series
is visible through the included github links.

Updates since v19: Added Darricks fix for the remote block accounting as well as
some minor nits about the default assert in xfs_attr_set_iter.  Spent quite
a bit of time testing this cycle to weed out any more unexpected bugs.  No new
test failures were observed with the addition of this set.

xfs: Fix default ASSERT in xfs_attr_set_iter
  Replaced the assert with ASSERT(0);

xfs: Add delay ready attr remove routines
  Added Darricks fix for remote block accounting

This series can be viewed on github here:
https://github.com/allisonhenderson/xfs_work/tree/delay_ready_attrs_v20

As well as the extended delayed attribute and parent pointer series:
https://github.com/allisonhenderson/xfs_work/tree/delay_ready_attrs_v20_extended

And the test cases:
https://github.com/allisonhenderson/xfs_work/tree/pptr_xfstestsv3
In order to run the test cases, you will need have the corresponding xfsprogs

changes as well.  Which can be found here:
https://github.com/allisonhenderson/xfs_work/tree/delay_ready_attrs_xfsprogs_v20
https://github.com/allisonhenderson/xfs_work/tree/delay_ready_attrs_xfsprogs_v20_extended

To run the xfs attributes tests run:
check -g attr

To run as delayed attributes run:
export MOUNT_OPTIONS="-o delattr"
check -g attr

To run parent pointer tests:
check -g parent

I've also made the corresponding updates to the user space side as well, and ported anything
they need to seat correctly.

Questions, comment and feedback appreciated!

Thanks all!
Allison

* tag 'xfs-delay-ready-attrs-v20.1' of https://github.com/allisonhenderson/xfs_work:
  xfs: Make attr name schemes consistent
  xfs: Fix default ASSERT in xfs_attr_set_iter
  xfs: Clean up xfs_attr_node_addname_clear_incomplete
  xfs: Remove xfs_attr_rmtval_set
  xfs: Add delay ready attr set routines
  xfs: Add delay ready attr remove routines
  xfs: Hoist node transaction handling
  xfs: Hoist xfs_attr_leaf_addname
  xfs: Hoist xfs_attr_node_addname
  xfs: Add helper xfs_attr_node_addname_find_attr
  xfs: Separate xfs_attr_node_addname and xfs_attr_node_addname_clear_incomplete
  xfs: Refactor xfs_attr_set_shortform
  xfs: Add xfs_attr_node_remove_name
  xfs: Reverse apply 72b97ea40d
2021-06-18 08:13:22 -07:00
Allison Henderson
816c8e39b7 xfs: Make attr name schemes consistent
This patch renames the following functions to make the nameing scheme more consistent:
xfs_attr_shortform_remove -> xfs_attr_sf_removename
xfs_attr_node_remove_name -> xfs_attr_node_removename
xfs_attr_set_fmt -> xfs_attr_sf_addname

Suggested-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2021-06-09 09:34:05 -07:00
Allison Henderson
4a4957c16d xfs: Fix default ASSERT in xfs_attr_set_iter
This ASSERT checks for the state value of RM_SHRINK in the set path
which should never happen.  Change to ASSERT(0);

Suggested-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2021-06-09 09:33:14 -07:00
Darrick J. Wong
b26b2bf14f xfs: rename struct xfs_eofblocks to xfs_icwalk
The xfs_eofblocks structure is no longer well-named -- nowadays it
provides optional filtering criteria to any walk of the incore inode
cache.  Only one of the cache walk goals has anything to do with
clearing of speculative post-EOF preallocations, so change the name to
be more appropriate.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-08 09:30:20 -07:00
Darrick J. Wong
2d53f66baf xfs: change the prefix of XFS_EOF_FLAGS_* to XFS_ICWALK_FLAG_
In preparation for renaming struct xfs_eofblocks to struct xfs_icwalk,
change the prefix of the existing XFS_EOF_FLAGS_* flags to
XFS_ICWALK_FLAG_ and convert all the existing users.  This adds a degree
of interface separation between the ioctl definitions and the incore
parameters.  Since FLAGS_UNION is only used in xfs_icache.c, move it
there as a private flag.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
2021-06-08 09:30:20 -07:00
Darrick J. Wong
9492750a8b xfs: selectively keep sick inodes in memory
It's important that the filesystem retain its memory of sick inodes for
a little while after problems are found so that reports can be collected
about what was wrong.  Don't let inode reclamation free sick inodes
unless we're unmounting or the fs already went down.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
2021-06-08 09:30:20 -07:00
Darrick J. Wong
7975e465af xfs: drop IDONTCACHE on inodes when we mark them sick
When we decide to mark an inode sick, clear the DONTCACHE flag so that
the incore inode will be kept around until memory pressure forces it out
of memory.  This increases the chances that the sick status will be
caught by someone compiling a health report later on.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
2021-06-08 09:30:20 -07:00
Darrick J. Wong
255794c7ed xfs: only reset incore inode health state flags when reclaiming an inode
While running some fuzz tests on inode metadata, I noticed that the
filesystem health report (as provided by xfs_spaceman) failed to report
the file corruption even when spaceman was run immediately after running
xfs_scrub to detect the corruption.  That isn't the intended behavior;
one ought to be able to run scrub to detect errors in the ondisk
metadata and be able to access to those reports for some time after the
scrub.

After running the same sequence through an instrumented kernel, I
discovered the reason why -- scrub igets the file, scans it, marks it
sick, and ireleases the inode.  When the VFS lets go of the incore
inode, it moves to RECLAIMABLE state.  If spaceman igets the incore
inode before it moves to RECLAIM state, iget reinitializes the VFS
state, clears the sick and checked masks, and hands back the inode.  At
this point, the caller has the exact same incore inode, but with all the
health state erased.

In other words, we're erasing the incore inode's health state flags when
we've decided NOT to sever the link between the incore inode and the
ondisk inode.  This is wrong, so we need to remove the lines that zero
the fields from xfs_iget_cache_hit.

As a precaution, we add the same lines into xfs_reclaim_inode just after
we sever the link between incore and ondisk inode.  Strictly speaking
this isn't necessary because once an inode has gone through reclaim it
must go through xfs_inode_alloc (which also zeroes the state) and
xfs_iget is careful to check for mismatches between the inode it pulls
out of the radix tree and the one it wants.

Fixes: 6772c1f112 ("xfs: track metadata health status")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
2021-06-08 09:30:20 -07:00
Darrick J. Wong
ffc18582ed xfs: clean up incore inode walk functions
This ambitious series aims to cleans up redundant inode walk code in
 xfs_icache.c, hide implementation details of the quotaoff dquot release
 code, and eliminates indirect function calls from incore inode walks.
 
 The first thing it does is to move all the code that quotaoff calls to
 release dquots from all incore inodes into xfs_icache.c.  Next, it
 separates the goal of an inode walk from the actual radix tree tags that
 may or may not be involved and drops the kludgy XFS_ICI_NO_TAG thing.
 Finally, we split the speculative preallocation (blockgc) and quotaoff
 dquot release code paths into separate functions so that we can keep the
 implementations cohesive.
 
 Christoph suggested last cycle that we 'simply' change quotaoff not to
 allow deactivating quota entirely, but as these cleanups are to enable
 one major change in behavior (deferred inode inactivation) I do not want
 to add a second behavior change (quotaoff) as a dependency.
 
 To be blunt: Additional cleanups are not in scope for this series.
 
 Next, I made two observations about incore inode radix tree walks --
 since there's a 1:1 mapping between the walk goal and the per-inode
 processing function passed in, we can use the goal to make a direct call
 to the processing function.  Furthermore, the only caller to supply a
 nonzero iter_flags argument is quotaoff, and there's only one INEW flag.
 
 From that observation, I concluded that it's quite possible to remove
 two parameters from the xfs_inode_walk* function signatures -- the
 iter_flags, and the execute function pointer.  The middle of the series
 moves the INEW functionality into the one piece (quotaoff) that wants
 it, and removes the indirect calls.
 
 The final observation is that the inode reclaim walk loop is now almost
 the same as xfs_inode_walk, so it's silly to maintain two copies.  Merge
 the reclaim loop code into xfs_inode_walk.
 
 Lastly, refactor the per-ag radix tagging functions since there's
 duplicated code that can be consolidated.
 
 This series is a prerequisite for the next two patchsets, since deferred
 inode inactivation will add another inode radix tree tag and iterator
 function to xfs_inode_walk.
 
 v2: walk the vfs inode list when running quotaoff instead of the radix
     tree, then rework the (now completely internal) inode walk function
     to take the tag as the main parameter.
 v3: merge the reclaim loop into xfs_inode_walk, then consolidate the
     radix tree tagging functions
 v4: rebase to 5.13-rc4
 v5: combine with the quotaoff patchset, reorder functions to minimize
     forward declarations, split inode walk goals from radix tree tags
     to reduce conceptual confusion
 v6: start moving the inode cache code towards the xfs_icwalk prefix
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEEUzaAxoMeQq6m2jMV+H93GTRKtOsFAmC5Yv0ACgkQ+H93GTRK
 tOv7Fg//Z7cKph0zSg6qsukMEMZxscuNcEBydCW1bu9gSx1NpszDpiGqAiO5ZB3X
 wP2XkCqjuatbNGGvkNLHS/M4sbLX3ELogvYmMRvUhDoaSFxT/KKgxvsyNffiCSS7
 xRB/rvWRp9MGRpBWPF0ZUxFU6VBzhCrYdMsNhvW95AEup8S/j+NplwoIif0gzaZZ
 Q6Fl4Ca9VEBvJQPV+/zkLih19iFItmARJhPHUs4BO1nZv+CzZBFQHg7Ijw7nW92j
 eSY68W4LH/IQ5cqm+HrD/+Z6ns0P7J2viewzVymkNEGnuX4a0xrQrzQ8ydRsAxTi
 9EDrpIe3MbSI5YjJfmRe8G3LX5p7vBpqc8TeyZdRDMGWkFjT33HPlQNb6WxKLQbA
 mjKdfr8AYZR/UQKW/7oZFrJnOoMpYRAQ4Sn/9BAYZQYm7tiLzuZsrEZ7JBwiUA56
 XHmlsDDeLzJeKvjmUu8M3H4oh4Nwf5/I2vJwHjueTfhl83uJP04igIXC4rnq56bM
 AAAjH9uV11Fo3q0ywAnRtN2HYj8PEJlCMK5CNskILrGeMITsBPGht0SbaA6hDI2h
 GYmltKInHzuPhHC9NfyPVrVr3BrmPR5cBsVFESiz5A4E9rbuKmmna6Yk8MFlMyl8
 FRIA3zVatJ2qQXtsAcdI8AZzMd7ciYhkAgCqFKxv8qK/qxITHh4=
 =Rxdn
 -----END PGP SIGNATURE-----

Merge tag 'inode-walk-cleanups-5.14_2021-06-03' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-5.14-merge2

xfs: clean up incore inode walk functions

This ambitious series aims to cleans up redundant inode walk code in
xfs_icache.c, hide implementation details of the quotaoff dquot release
code, and eliminates indirect function calls from incore inode walks.

The first thing it does is to move all the code that quotaoff calls to
release dquots from all incore inodes into xfs_icache.c.  Next, it
separates the goal of an inode walk from the actual radix tree tags that
may or may not be involved and drops the kludgy XFS_ICI_NO_TAG thing.
Finally, we split the speculative preallocation (blockgc) and quotaoff
dquot release code paths into separate functions so that we can keep the
implementations cohesive.

Christoph suggested last cycle that we 'simply' change quotaoff not to
allow deactivating quota entirely, but as these cleanups are to enable
one major change in behavior (deferred inode inactivation) I do not want
to add a second behavior change (quotaoff) as a dependency.

To be blunt: Additional cleanups are not in scope for this series.

Next, I made two observations about incore inode radix tree walks --
since there's a 1:1 mapping between the walk goal and the per-inode
processing function passed in, we can use the goal to make a direct call
to the processing function.  Furthermore, the only caller to supply a
nonzero iter_flags argument is quotaoff, and there's only one INEW flag.

From that observation, I concluded that it's quite possible to remove
two parameters from the xfs_inode_walk* function signatures -- the
iter_flags, and the execute function pointer.  The middle of the series
moves the INEW functionality into the one piece (quotaoff) that wants
it, and removes the indirect calls.

The final observation is that the inode reclaim walk loop is now almost
the same as xfs_inode_walk, so it's silly to maintain two copies.  Merge
the reclaim loop code into xfs_inode_walk.

Lastly, refactor the per-ag radix tagging functions since there's
duplicated code that can be consolidated.

This series is a prerequisite for the next two patchsets, since deferred
inode inactivation will add another inode radix tree tag and iterator
function to xfs_inode_walk.

v2: walk the vfs inode list when running quotaoff instead of the radix
    tree, then rework the (now completely internal) inode walk function
    to take the tag as the main parameter.
v3: merge the reclaim loop into xfs_inode_walk, then consolidate the
    radix tree tagging functions
v4: rebase to 5.13-rc4
v5: combine with the quotaoff patchset, reorder functions to minimize
    forward declarations, split inode walk goals from radix tree tags
    to reduce conceptual confusion
v6: start moving the inode cache code towards the xfs_icwalk prefix

* tag 'inode-walk-cleanups-5.14_2021-06-03' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
  xfs: refactor per-AG inode tagging functions
  xfs: merge xfs_reclaim_inodes_ag into xfs_inode_walk_ag
  xfs: pass struct xfs_eofblocks to the inode scan callback
  xfs: fix radix tree tag signs
  xfs: make the icwalk processing functions clean up the grab state
  xfs: clean up inode state flag tests in xfs_blockgc_igrab
  xfs: remove indirect calls from xfs_inode_walk{,_ag}
  xfs: remove iter_flags parameter from xfs_inode_walk_*
  xfs: move xfs_inew_wait call into xfs_dqrele_inode
  xfs: separate the dqrele_all inode grab logic from xfs_inode_walk_ag_grab
  xfs: pass the goal of the incore inode walk to xfs_inode_walk()
  xfs: rename xfs_inode_walk functions to xfs_icwalk
  xfs: move the inode walk functions further down
  xfs: detach inode dquots at the end of inactivation
  xfs: move the quotaoff dqrele inode walk into xfs_icache.c

[djwong: added variable names to function declarations while fixing
merge conflicts]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-08 09:26:44 -07:00
Darrick J. Wong
8b943d21d4 xfs: assorted fixes for 5.14, part 1
This branch contains the first round of various small fixes for 5.14.
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEEUzaAxoMeQq6m2jMV+H93GTRKtOsFAmC5YvwACgkQ+H93GTRK
 tOuDTxAAmW+Mc4oPx/Fsa2xqvkDYCGaM1QtkIY9McrGsMdzF0JbB7bR5rrUET2+N
 /FsFTIQ31vZfMEJ69HM6NjQ4xXIKhdEj+PlGsjvg62BqhpPBRuFH7dLBWChiRcf2
 M5SN1P9kcAjIoMR2tKzh8FMDmJTcJlJxxBi9kxb3YK1+UesUHsPS8/jSW3OuIov3
 AJzAZ08SrA8ful5uCMy9Mf5uBgfRuAUjDzA5CM5kA1xqlHREQi+oHl62E81N33mu
 RR8tdHQzvPO5rHyX84GzV5cu2CsmDuOPF2nA5SUxRhIZFfMo5mEezA/nxqqACYti
 rnRGxZVwIG9YYBESVxXFQXIjn5lHoQ2jomk/CraszeVEJCteNRnpbVGzd1xczM3u
 0iuDuHy+aVB/3QhfA6/0vjfttCzkMEld9U9c3WEjIaw5iUCxe531yfrvVEyF9blx
 NBjnQyHGbt+y26BzBjD33NJEdDoZqS3UIQ/rmb4f2mitGN5d9faAcJ754uRJt3o4
 K9HXGjuR+iOH/tCZKDL1hBc4M/pFgNdeBWyFYdQhh8eSj9HSCTCG58zAyQ2WPOSr
 6D/f4BMqivKzzz0HicZPJAoazrtcKGrWTHTxidHIkI4le367NOwv6YJquJ0pFMBs
 8L7OdRqYT3yw6+qErCjEn03WkP9O7V8lHf8hFxt7+dNxDukIj40=
 =6eZB
 -----END PGP SIGNATURE-----

Merge tag 'assorted-fixes-5.14-1_2021-06-03' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-5.14-merge2

xfs: assorted fixes for 5.14, part 1

This branch contains the first round of various small fixes for 5.14.

* tag 'assorted-fixes-5.14-1_2021-06-03' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
  xfs: don't take a spinlock unconditionally in the DIO fastpath
  xfs: mark xfs_bmap_set_attrforkoff static
  xfs: Remove redundant assignment to busy
  xfs: sort variable alphabetically to avoid repeated declaration
2021-06-08 09:22:34 -07:00
Darrick J. Wong
f52edf6c54 xfs: various unit conversions
Crafting the realtime file extent size hint fixes revealed various
 opportunities to clean up unit conversions, so now that gets its own
 series.
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEEUzaAxoMeQq6m2jMV+H93GTRKtOsFAmC5YvYACgkQ+H93GTRK
 tOtRihAAh5wKVV9Fwafkk3l2zTCw+K/SfciWvZqF8BClAjqzWUDaw4DJ2HS0lSmj
 ezOG+/HhJzZB4udC9WDM4TSFA/elR1pJBu5lSBpAt/K78Mpw/i1aFs1SyAfw6Qsp
 1A8um80jaCyzvcSxU1Hx/35CN0MQwpjJ23LziEDUVBvj/qV7Qj8DmI61CfWWYzs9
 jEJFzIvbKoRqs8Wqws7n4a7fw3VOyqC2EQV290xzumkYKOSAmpWQGkbiU2/YtqKY
 7w/GvjS1zRE7o1cGeKBYRA/35zbw5mqoHD5vJmdlUIsWK22fnN8h5vqYo5dpFPiz
 nFyK+MHVrV6WxKBKECHtgm7Jya+jjH0TFu/9js4k31ehe4LJc/nhEqmSrFiH3eGS
 4AXrhhZqtDZ1skjcPym83dCayW1uE2cWbHqYJ+ztQW5PrLafoWQ9Ld/vtrbL+b/U
 VIgh3LQmF371jGcE0twERNQPIb5F96w9mS6F+vq0JuvrGftxoa4tKVjE8jrNWJjo
 750/KT0Nupti0AYKq9WMD47BiSf5BbdnhYTN15X0mc5TyHo/EyXIl/I/hFfEMFp9
 AH/qWaQlt3LjMXvr1xF6R4RNSI+xxai7PHfW69Zi7tBea+7k9wTZjubgc2FFaMOQ
 Jf8n5L07IXVYDy0mGUGaokenQb2KPvfXVK7yYtaMj1qDh7Sjqjw=
 =HGL1
 -----END PGP SIGNATURE-----

Merge tag 'unit-conversion-cleanups-5.14_2021-06-03' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-5.14-merge2

xfs: various unit conversions

Crafting the realtime file extent size hint fixes revealed various
opportunities to clean up unit conversions, so now that gets its own
series.

* tag 'unit-conversion-cleanups-5.14_2021-06-03' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
  xfs: remove unnecessary shifts
  xfs: clean up open-coded fs block unit conversions
2021-06-08 09:21:24 -07:00
Dave Chinner
9ba0889e22 xfs: drop the AGI being passed to xfs_check_agi_freecount
From: Dave Chinner <dchinner@redhat.com>

Stephen Rothwell reported this compiler warning from linux-next:

fs/xfs/libxfs/xfs_ialloc.c: In function 'xfs_difree_finobt':
fs/xfs/libxfs/xfs_ialloc.c:2032:20: warning: unused variable 'agi' [-Wunused-variable]
 2032 |  struct xfs_agi   *agi = agbp->b_addr;

Which is fallout from agno -> perag conversions that were done in
this function. xfs_check_agi_freecount() is the only user of "agi"
in xfs_difree_finobt() now, and it only uses the agi to get the
current free inode count. We hold that in the perag structure, so
there's not need to directly reference the raw AGI to get this
information.

The btree cursor being passed to xfs_check_agi_freecount() has a
reference to the perag being operated on, so use that directly in
xfs_check_agi_freecount() rather than passing an AGI.

Fixes: 7b13c51551 ("xfs: use perag for ialloc btree cursors")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
2021-06-08 09:19:22 -07:00
Darrick J. Wong
c3eabd3650 xfs: initial agnumber -> perag conversions for shrink
If we want to use active references to the perag to be able to gate
 shrink removing AGs and hence perags safely, we've got a fair bit of
 work to do actually use perags in all the places we need to.
 
 There's a lot of code that iterates ag numbers and then
 looks up perags from that, often multiple times for the same perag
 in the one operation. If we want to use reference counted perags for
 access control, then we need to convert all these uses to perag
 iterators, not agno iterators.
 
 [Patches 1-4]
 
 The first step of this is consolidating all the perag management -
 init, free, get, put, etc into a common location. THis is spread all
 over the place right now, so move it all into libxfs/xfs_ag.[ch].
 This does expose kernel only bits of the perag to libxfs and hence
 userspace, so the structures and code is rearranged to minimise the
 number of ifdefs that need to be added to the userspace codebase.
 The perag iterator in xfs_icache.c is promoted to a first class API
 and expanded to the needs of the code as required.
 
 [Patches 5-10]
 
 These are the first basic perag iterator conversions and changes to
 pass the perag down the stack from those iterators where
 appropriate. A lot of this is obvious, simple changes, though in
 some places we stop passing the perag down the stack because the
 code enters into an as yet unconverted subsystem that still uses raw
 AGs.
 
 [Patches 11-16]
 
 These replace the agno passed in the btree cursor for per-ag btree
 operations with a perag that is passed to the cursor init function.
 The cursor takes it's own reference to the perag, and the reference
 is dropped when the cursor is deleted. Hence we get reference
 coverage for the entire time the cursor is active, even if the code
 that initialised the cursor drops it's reference before the cursor
 or any of it's children (duplicates) have been deleted.
 
 The first patch adds the perag infrastructure for the cursor, the
 next four patches convert a btree cursor at a time, and the last
 removes the agno from the cursor once it is unused.
 
 [Patches 17-21]
 
 These patches are a demonstration of the simplifications and
 cleanups that come from plumbing the perag through interfaces that
 select and then operate on a specific AG. In this case the inode
 allocation algorithm does up to three walks across all AGs before it
 either allocates an inode or fails. Two of these walks are purely
 just to select the AG, and even then it doesn't guarantee inode
 allocation success so there's a third walk if the selected AG
 allocation fails.
 
 These patches collapse the selection and allocation into a single
 loop, simplifies the error handling because xfs_dir_ialloc() always
 returns ENOSPC if no AG was selected for inode allocation or we fail
 to allocate an inode in any AG, gets rid of xfs_dir_ialloc()
 wrapper, converts inode allocation to run entirely from a single
 perag instance, and then factors xfs_dialloc() into a much, much
 simpler loop which is easy to understand.
 
 Hence we end up with the same inode allocation logic, but it only
 needs two complete iterations at worst, makes AG selection and
 allocation atomic w.r.t. shrink and chops out out over 100 lines of
 code from this hot code path.
 
 [Patch 22]
 
 Converts the unlink path to pass perags through it.
 
 There's more conversion work to be done, but this patchset gets
 through a large chunk of it in one hit. Most of the iterators are
 converted, so once this is solidified we can move on to converting
 these to active references for being able to free perags while the
 fs is still active.
 -----BEGIN PGP SIGNATURE-----
 
 iQJIBAABCgAyFiEEmJOoJ8GffZYWSjj/regpR/R1+h0FAmC3HUgUHGRhdmlkQGZy
 b21vcmJpdC5jb20ACgkQregpR/R1+h2yaw/+P0JzpI+6n06Ei00mjgE/Du/WhMLi
 0JQ93Grlj+miuGGT9DgGCiRpoZnefhEk+BH6JqoEw1DQ3T5ilmAzrHLUUHSQC3+S
 dv85sJduheQ6yHuoO+4MzkaSq6JWKe7E9gZwAsVyBul5aSjdmaJaQdPwYMTXSXo0
 5Uqq8ECFkMcaHVNjcBfasgR/fdyWy2Qe4PFTHTHdQpd+DNZ9UXgFKHW2og+1iry/
 zDIvdIppJULA09TvVcZuFjd/1NzHQ/fLj5PAzz8GwagB4nz2x3s78Zevmo5yW/jK
 3/+50vXa8ldhiHDYGTS3QXvS0xJRyqUyD47eyWOOiojZw735jEvAlCgjX6+0X1HC
 k3gCkQLv8l96fRkvUpgnLf/fjrUnlCuNBkm9d1Eq2Tied8dvLDtiEzoC6L05Nqob
 yd/nIUb1zwJFa9tsoheHhn0bblTGX1+zP0lbRJBje0LotpNO9DjGX5JoIK4GR7F8
 y1VojcdgRI14HlxUnbF3p8wmQByN+M2tnp6GSdv9BA65bjqi05Rj/steFdZHBV6x
 wiRs8Yh6BTvMwKgufHhRQHfRahjNHQ/T/vOE+zNbWqemS9wtEUDop+KvPhC36R/k
 o/cmr23cF8ESX2eChk7XM4On3VEYpcvp2zSFgrFqZYl6RWOwEis3Htvce3KuSTPp
 8Xq70te0gr2DVUU=
 =YNzW
 -----END PGP SIGNATURE-----

Merge tag 'xfs-perag-conv-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs into xfs-5.14-merge2

xfs: initial agnumber -> perag conversions for shrink

If we want to use active references to the perag to be able to gate
shrink removing AGs and hence perags safely, we've got a fair bit of
work to do actually use perags in all the places we need to.

There's a lot of code that iterates ag numbers and then
looks up perags from that, often multiple times for the same perag
in the one operation. If we want to use reference counted perags for
access control, then we need to convert all these uses to perag
iterators, not agno iterators.

[Patches 1-4]

The first step of this is consolidating all the perag management -
init, free, get, put, etc into a common location. THis is spread all
over the place right now, so move it all into libxfs/xfs_ag.[ch].
This does expose kernel only bits of the perag to libxfs and hence
userspace, so the structures and code is rearranged to minimise the
number of ifdefs that need to be added to the userspace codebase.
The perag iterator in xfs_icache.c is promoted to a first class API
and expanded to the needs of the code as required.

[Patches 5-10]

These are the first basic perag iterator conversions and changes to
pass the perag down the stack from those iterators where
appropriate. A lot of this is obvious, simple changes, though in
some places we stop passing the perag down the stack because the
code enters into an as yet unconverted subsystem that still uses raw
AGs.

[Patches 11-16]

These replace the agno passed in the btree cursor for per-ag btree
operations with a perag that is passed to the cursor init function.
The cursor takes it's own reference to the perag, and the reference
is dropped when the cursor is deleted. Hence we get reference
coverage for the entire time the cursor is active, even if the code
that initialised the cursor drops it's reference before the cursor
or any of it's children (duplicates) have been deleted.

The first patch adds the perag infrastructure for the cursor, the
next four patches convert a btree cursor at a time, and the last
removes the agno from the cursor once it is unused.

[Patches 17-21]

These patches are a demonstration of the simplifications and
cleanups that come from plumbing the perag through interfaces that
select and then operate on a specific AG. In this case the inode
allocation algorithm does up to three walks across all AGs before it
either allocates an inode or fails. Two of these walks are purely
just to select the AG, and even then it doesn't guarantee inode
allocation success so there's a third walk if the selected AG
allocation fails.

These patches collapse the selection and allocation into a single
loop, simplifies the error handling because xfs_dir_ialloc() always
returns ENOSPC if no AG was selected for inode allocation or we fail
to allocate an inode in any AG, gets rid of xfs_dir_ialloc()
wrapper, converts inode allocation to run entirely from a single
perag instance, and then factors xfs_dialloc() into a much, much
simpler loop which is easy to understand.

Hence we end up with the same inode allocation logic, but it only
needs two complete iterations at worst, makes AG selection and
allocation atomic w.r.t. shrink and chops out out over 100 lines of
code from this hot code path.

[Patch 22]

Converts the unlink path to pass perags through it.

There's more conversion work to be done, but this patchset gets
through a large chunk of it in one hit. Most of the iterators are
converted, so once this is solidified we can move on to converting
these to active references for being able to free perags while the
fs is still active.

* tag 'xfs-perag-conv-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs: (23 commits)
  xfs: remove xfs_perag_t
  xfs: use perag through unlink processing
  xfs: clean up and simplify xfs_dialloc()
  xfs: inode allocation can use a single perag instance
  xfs: get rid of xfs_dir_ialloc()
  xfs: collapse AG selection for inode allocation
  xfs: simplify xfs_dialloc_select_ag() return values
  xfs: remove agno from btree cursor
  xfs: use perag for ialloc btree cursors
  xfs: convert allocbt cursors to use perags
  xfs: convert refcount btree cursor to use perags
  xfs: convert rmap btree cursor to using a perag
  xfs: add a perag to the btree cursor
  xfs: pass perags around in fsmap data dev functions
  xfs: push perags through the ag reservation callouts
  xfs: pass perags through to the busy extent code
  xfs: convert secondary superblock walk to use perags
  xfs: convert xfs_iwalk to use perag references
  xfs: convert raw ag walks to use for_each_perag
  xfs: make for_each_perag... a first class citizen
  ...
2021-06-08 09:13:13 -07:00
Darrick J. Wong
ebf2e33723 xfs: buffer cache bulk page allocation
This patchset makes use of the new bulk page allocation interface to
 reduce the overhead of allocating large numbers of pages in a
 loop.
 
 The first two patches are refactoring buffer memory allocation and
 converting the uncached buffer path to use the same page allocation
 path, followed by converting the page allocation path to use bulk
 allocation.
 
 The rest of the patches are then consolidation of the page
 allocation and freeing code to simplify the code and remove a chunk
 of unnecessary abstraction. This is largely based on a series of
 changes made by Christoph Hellwig.
 -----BEGIN PGP SIGNATURE-----
 
 iQJIBAABCgAyFiEEmJOoJ8GffZYWSjj/regpR/R1+h0FAmC+6OwUHGRhdmlkQGZy
 b21vcmJpdC5jb20ACgkQregpR/R1+h21QQ/8C0f7wq1OKwNI2oRubf6J8jtttiRS
 SD2TA03AP2OIOKx4y2G0h0dJeX9tgnbLerIlpfT80nHDoBgKHbZCYSEHQT0DscYo
 fuwQTVR8RCklKspjUlCGR+Gbm6vI8HakK1lAppw168e4c6t8wX1KiSibwaVTQdaZ
 NaXUqTUzGiNq+iiLS6fW3mJ3PKWFJYyrDOSR2jIPbUGIJdejRCGe0xVnu+hIsz+y
 c2gSGCB+j3cYaazhlJTDYPGja3Wq3eR+Ya9i1GcA1tJiJLsu0ZjaVQ69Bl4dud2F
 c3OyhFK0El1VMSEVb3hY8gTpAO02jNWSnB2Zlidt0h4ZJVAxKus0xe2w3eS4uST2
 hcMI3lwjdzRQuoBwOgXQ+CpYVv2wI8HPNLTSR+NYcC2IZaCNieFRWdTYwXrAJBB3
 H09m04GT/7TkkrYHFD1zRtIedP4DZ6MZn/33bufNxEt1NRCFw5AFAEUFfjDA317A
 4nByCmU6XjmmpI/XLixwu0BYCfKVB4UsrgOyzXBy7ZU0+pIser+ynP1V4d9Bb43Y
 xVQ8S0QirT7gqXjx75mD4B4qkXZ5nrz5Z7fSn6YU4TwqsYtZYlsBauLlWmmHp9MT
 CP4PA4j+CQORhfZzWXw2ViXYGoIssc1cw5i4JB6a4u/OaDi19dYkE6SO8P3b9GSm
 khHqWgcTC4VGpmc=
 =JsrV
 -----END PGP SIGNATURE-----

Merge tag 'xfs-buf-bulk-alloc-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs into xfs-5.14-merge2

xfs: buffer cache bulk page allocation

This patchset makes use of the new bulk page allocation interface to
reduce the overhead of allocating large numbers of pages in a
loop.

The first two patches are refactoring buffer memory allocation and
converting the uncached buffer path to use the same page allocation
path, followed by converting the page allocation path to use bulk
allocation.

The rest of the patches are then consolidation of the page
allocation and freeing code to simplify the code and remove a chunk
of unnecessary abstraction. This is largely based on a series of
changes made by Christoph Hellwig.

* tag 'xfs-buf-bulk-alloc-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs:
  xfs: merge xfs_buf_allocate_memory
  xfs: cleanup error handling in xfs_buf_get_map
  xfs: get rid of xb_to_gfp()
  xfs: simplify the b_page_count calculation
  xfs: remove ->b_offset handling for page backed buffers
  xfs: move page freeing into _xfs_buf_free_pages()
  xfs: merge _xfs_buf_get_pages()
  xfs: use alloc_pages_bulk_array() for buffers
  xfs: use xfs_buf_alloc_pages for uncached buffers
  xfs: split up xfs_buf_allocate_memory
2021-06-08 09:10:01 -07:00
Dave Chinner
8bcac7448a xfs: merge xfs_buf_allocate_memory
It only has one caller and is now a simple function, so merge it
into the caller.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2021-06-07 11:50:48 +10:00
Christoph Hellwig
170041f715 xfs: cleanup error handling in xfs_buf_get_map
Use a single goto label for freeing the buffer and returning an
error.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
2021-06-07 11:50:47 +10:00
Dave Chinner
289ae7b48c xfs: get rid of xb_to_gfp()
Only used in one place, so just open code the logic in the macro.
Based on a patch from Christoph Hellwig.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2021-06-07 11:50:17 +10:00
Christoph Hellwig
934d1076bb xfs: simplify the b_page_count calculation
Ever since we stopped using the Linux page cache to back XFS buffers
there is no need to take the start sector into account for
calculating the number of pages in a buffer, as the data always
start from the beginning of the buffer.

Signed-off-by: Christoph Hellwig <hch@lst.de>
[dgc: modified to suit this series]
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2021-06-07 11:50:00 +10:00
Christoph Hellwig
54cd3aa6f8 xfs: remove ->b_offset handling for page backed buffers
->b_offset can only be non-zero for _XBF_KMEM backed buffers, so
remove all code dealing with it for page backed buffers.

Signed-off-by: Christoph Hellwig <hch@lst.de>
[dgc: modified to fit this patchset]
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2021-06-07 11:49:50 +10:00
Darrick J. Wong
c076ae7a93 xfs: refactor per-AG inode tagging functions
In preparation for adding another incore inode tree tag, refactor the
code that sets and clears tags from the per-AG inode tree and the tree
of per-AG structures, and remove the open-coded versions used by the
blockgc code.

Note: For reclaim, we now rely on the radix tree tags instead of the
reclaimable inode count more heavily than we used to.  The conversion
should be fine, but the logic isn't 100% identical.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:04 -07:00
Darrick J. Wong
f1bc5c5630 xfs: merge xfs_reclaim_inodes_ag into xfs_inode_walk_ag
Merge these two inode walk loops together, since they're pretty similar
now.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:04 -07:00
Darrick J. Wong
9d5ee83759 xfs: pass struct xfs_eofblocks to the inode scan callback
Pass a pointer to the actual eofb structure around the inode scanner
functions instead of a void pointer, now that none of the functions is
used as a callback.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:04 -07:00
Darrick J. Wong
919a4ddb68 xfs: fix radix tree tag signs
Radix tree tags are supposed to be unsigned ints, so fix the callers.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:04 -07:00
Darrick J. Wong
594ab00b76 xfs: make the icwalk processing functions clean up the grab state
Soon we're going to be adding two new callers to the incore inode walk
code: reclaim of incore inodes, and (later) inactivation of inodes.
Both states operate on inodes that no longer have any VFS state, so we
need to move the xfs_irele calls into the processing functions.

In other words, icwalk processing functions are responsible for cleaning
up whatever state changes are made by the corresponding icwalk igrab
function that picked the inode for processing.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:03 -07:00
Darrick J. Wong
d20d5edcf9 xfs: clean up inode state flag tests in xfs_blockgc_igrab
Clean up the definition of which inode states are not eligible for
speculative preallocation garbage collecting by creating a private
#define.  The deferred inactivation patchset will add two new entries to
the set of flags-to-ignore, so we want the definition not to end up a
cluttered mess.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:03 -07:00
Darrick J. Wong
f427cf5c62 xfs: remove indirect calls from xfs_inode_walk{,_ag}
It turns out that there is a 1:1 mapping between the execute and goal
parameters that are passed to xfs_inode_walk_ag:

	xfs_blockgc_scan_inode <=> XFS_ICWALK_BLOCKGC
	xfs_dqrele_inode <=> XFS_ICWALK_DQRELE

Because of this exact correspondence, we don't need the execute function
pointer and can replace it with a direct call.

For the price of a forward static declaration, we can eliminate the
indirect function call.  This likely has a negligible impact on
performance (since the execute function runs transactions), but it also
simplifies the function signature.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:03 -07:00
Darrick J. Wong
7fdff52623 xfs: remove iter_flags parameter from xfs_inode_walk_*
The sole iter_flags is XFS_INODE_WALK_INEW_WAIT, and there are no users.
Remove the flag, and the parameter, and all the code that used it.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:03 -07:00
Darrick J. Wong
9d2793ceec xfs: move xfs_inew_wait call into xfs_dqrele_inode
Move the INEW wait into xfs_dqrele_inode so that we can drop the
iter_flags parameter in the next patch.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:03 -07:00
Darrick J. Wong
b9baaef42f xfs: separate the dqrele_all inode grab logic from xfs_inode_walk_ag_grab
Disentangle the dqrele_all inode grab code from the "generic" inode walk
grabbing code, and and use the opportunity to document why the dqrele
grab function does what it does.  Since xfs_inode_walk_ag_grab is now
only used for blockgc, rename it to reflect that.

Ultimately, there will be four reasons to perform a walk of incore
inodes: quotaoff dquote releasing (dqrele), garbage collection of
speculative preallocations (blockgc), reclamation of incore inodes
(reclaim), and deferred inactivation (inodegc).  Each of these four have
their own slightly different criteria for deciding if they want to
handle an inode, so it makes more sense to have four cohesive igrab
functions than one confusing parameteric grab function like we do now.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
2021-06-03 15:56:03 -07:00