task #28557760

commit 4b674b9ac852937af1f8c62f730c325fb6eadcdb upstream.

The filesystem freeze sequence in XFS waits on any background
eofblocks or cowblocks scans to complete before the filesystem is
quiesced. At this point, the freezer has already stopped the
transaction subsystem, however, which means a truncate or cowblock
cancellation in progress is likely blocked in transaction
allocation. This results in a deadlock between freeze and the
associated scanner.

Fix this problem by holding superblock write protection across calls
into the block reapers. Since protection for background scans is
acquired from the workqueue task context, trylock to avoid a similar
deadlock between freeze and blocking on the write lock.

Fixes: d6b636eb ("xfs: halt auto-reclamation activities while rebuilding rmap")
Reported-by: NPaul Furtado <paulfurtado91@gmail.com>
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChandan Rajendra <chandanrlinux@gmail.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NAllison Collins <allison.henderson@oracle.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: NJeffle Xu <jefflexu@linux.alibaba.com>
Acked-by: NJoseph Qi <joseph.qi@linux.alibaba.com>

We have a few places that already check if an inode has actual data in
the COW fork to avoid work on reflink inodes that do not actually have
outstanding COW blocks.  There are a few more places that can avoid
working if doing the same check, so add a documented helper for this
condition and use it in all places where it makes sense.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

We only have a few more callers left, so seize the opportunity and kill
it off.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

Replace the IRELE macro with a proper function so that we can do proper
typechecking and so that we can stop open-coding iput in scrub, which
means that we'll be able to ftrace inode lifetimes going through scrub
correctly.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NCarlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>

Remove the verbose license text from XFS files and replace them
with SPDX tags. This does not change the license of any of the code,
merely refers to the common, up-to-date license files in LICENSES/

This change was mostly scripted. fs/xfs/Makefile and
fs/xfs/libxfs/xfs_fs.h were modified by hand, the rest were detected
and modified by the following command:

for f in `git grep -l "GNU General" fs/xfs/` ; do
	echo $f
	cat $f | awk -f hdr.awk > $f.new
	mv -f $f.new $f
done

And the hdr.awk script that did the modification (including
detecting the difference between GPL-2.0 and GPL-2.0+ licenses)
is as follows:

$ cat hdr.awk
BEGIN {
	hdr = 1.0
	tag = "GPL-2.0"
	str = ""
}

/^ \* This program is free software/ {
	hdr = 2.0;
	next
}

/any later version./ {
	tag = "GPL-2.0+"
	next
}

/^ \*\// {
	if (hdr > 0.0) {
		print "// SPDX-License-Identifier: " tag
		print str
		print $0
		str=""
		hdr = 0.0
		next
	}
	print $0
	next
}

/^ \* / {
	if (hdr > 1.0)
		next
	if (hdr > 0.0) {
		if (str != "")
			str = str "\n"
		str = str $0
		next
	}
	print $0
	next
}

/^ \*/ {
	if (hdr > 0.0)
		next
	print $0
	next
}

// {
	if (hdr > 0.0) {
		if (str != "")
			str = str "\n"
		str = str $0
		next
	}
	print $0
}

END { }
$
Signed-off-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

Rebuilding the reverse-mapping tree requires us to quiesce all inodes in
the filesystem, so we must stop background reclamation of post-EOF and
CoW prealloc blocks.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NDave Chinner <dchinner@redhat.com>

The log item flags contain a field that is protected by the AIL
lock - the XFS_LI_IN_AIL flag. We use non-atomic RMW operations to
set and clear these flags, but most of the updates and checks are
not done with the AIL lock held and so are susceptible to update
races.

Fix this by changing the log item flags to use atomic bitops rather
than be reliant on the AIL lock for update serialisation.
Signed-Off-By: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

A recent fuzzed filesystem image cached random dcache corruption
when the reproducer was run. This often showed up as panics in
lookup_slow() on a null inode->i_ops pointer when doing pathwalks.

BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
....
Call Trace:
 lookup_slow+0x44/0x60
 walk_component+0x3dd/0x9f0
 link_path_walk+0x4a7/0x830
 path_lookupat+0xc1/0x470
 filename_lookup+0x129/0x270
 user_path_at_empty+0x36/0x40
 path_listxattr+0x98/0x110
 SyS_listxattr+0x13/0x20
 do_syscall_64+0xf5/0x280
 entry_SYSCALL_64_after_hwframe+0x42/0xb7

but had many different failure modes including deadlocks trying to
lock the inode that was just allocated or KASAN reports of
use-after-free violations.

The cause of the problem was a corrupt INOBT on a v4 fs where the
root inode was marked as free in the inobt record. Hence when we
allocated an inode, it chose the root inode to allocate, found it in
the cache and re-initialised it.

We recently fixed a similar inode allocation issue caused by inobt
record corruption problem in xfs_iget_cache_miss() in commit
ee457001 ("xfs: catch inode allocation state mismatch
corruption"). This change adds similar checks to the cache-hit path
to catch it, and turns the reproducer into a corruption shutdown
situation.
Reported-by: NWen Xu <wen.xu@gatech.edu>
Signed-Off-By: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NCarlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
[darrick: fix typos in comment]
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

We recently came across a V4 filesystem causing memory corruption
due to a newly allocated inode being setup twice and being added to
the superblock inode list twice. From code inspection, the only way
this could happen is if a newly allocated inode was not marked as
free on disk (i.e. di_mode wasn't zero).

Running the metadump on an upstream debug kernel fails during inode
allocation like so:

XFS: Assertion failed: ip->i_d.di_nblocks == 0, file: fs/xfs/xfs_inod=
e.c, line: 838
 ------------[ cut here ]------------
kernel BUG at fs/xfs/xfs_message.c:114!
invalid opcode: 0000 [#1] PREEMPT SMP
CPU: 11 PID: 3496 Comm: mkdir Not tainted 4.16.0-rc5-dgc #442
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/0=
1/2014
RIP: 0010:assfail+0x28/0x30
RSP: 0018:ffffc9000236fc80 EFLAGS: 00010202
RAX: 00000000ffffffea RBX: 0000000000004000 RCX: 0000000000000000
RDX: 00000000ffffffc0 RSI: 000000000000000a RDI: ffffffff8227211b
RBP: ffffc9000236fce8 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000bec R11: f000000000000000 R12: ffffc9000236fd30
R13: ffff8805c76bab80 R14: ffff8805c77ac800 R15: ffff88083fb12e10
FS:  00007fac8cbff040(0000) GS:ffff88083fd00000(0000) knlGS:0000000000000=
000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fffa6783ff8 CR3: 00000005c6e2b003 CR4: 00000000000606e0
Call Trace:
 xfs_ialloc+0x383/0x570
 xfs_dir_ialloc+0x6a/0x2a0
 xfs_create+0x412/0x670
 xfs_generic_create+0x1f7/0x2c0
 ? capable_wrt_inode_uidgid+0x3f/0x50
 vfs_mkdir+0xfb/0x1b0
 SyS_mkdir+0xcf/0xf0
 do_syscall_64+0x73/0x1a0
 entry_SYSCALL_64_after_hwframe+0x42/0xb7

Extracting the inode number we crashed on from an event trace and
looking at it with xfs_db:

xfs_db> inode 184452204
xfs_db> p
core.magic = 0x494e
core.mode = 0100644
core.version = 2
core.format = 2 (extents)
core.nlinkv2 = 1
core.onlink = 0
.....

Confirms that it is not a free inode on disk. xfs_repair
also trips over this inode:

.....
zero length extent (off = 0, fsbno = 0) in ino 184452204
correcting nextents for inode 184452204
bad attribute fork in inode 184452204, would clear attr fork
bad nblocks 1 for inode 184452204, would reset to 0
bad anextents 1 for inode 184452204, would reset to 0
imap claims in-use inode 184452204 is free, would correct imap
would have cleared inode 184452204
.....
disconnected inode 184452204, would move to lost+found

And so we have a situation where the directory structure and the
inobt thinks the inode is free, but the inode on disk thinks it is
still in use. Where this corruption came from is not possible to
diagnose, but we can detect it and prevent the kernel from oopsing
on lookup. The reproducer now results in:

$ sudo mkdir /mnt/scratch/{0,1,2,3,4,5}{0,1,2,3,4,5}
mkdir: cannot create directory =E2=80=98/mnt/scratch/00=E2=80=99: File ex=
ists
mkdir: cannot create directory =E2=80=98/mnt/scratch/01=E2=80=99: File ex=
ists
mkdir: cannot create directory =E2=80=98/mnt/scratch/03=E2=80=99: Structu=
re needs cleaning
mkdir: cannot create directory =E2=80=98/mnt/scratch/04=E2=80=99: Input/o=
utput error
mkdir: cannot create directory =E2=80=98/mnt/scratch/05=E2=80=99: Input/o=
utput error
....

And this corruption shutdown:

[   54.843517] XFS (loop0): Corruption detected! Free inode 0xafe846c not=
 marked free on disk
[   54.845885] XFS (loop0): Internal error xfs_trans_cancel at line 1023 =
of file fs/xfs/xfs_trans.c.  Caller xfs_create+0x425/0x670
[   54.848994] CPU: 10 PID: 3541 Comm: mkdir Not tainted 4.16.0-rc5-dgc #=
443
[   54.850753] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIO=
S 1.10.2-1 04/01/2014
[   54.852859] Call Trace:
[   54.853531]  dump_stack+0x85/0xc5
[   54.854385]  xfs_trans_cancel+0x197/0x1c0
[   54.855421]  xfs_create+0x425/0x670
[   54.856314]  xfs_generic_create+0x1f7/0x2c0
[   54.857390]  ? capable_wrt_inode_uidgid+0x3f/0x50
[   54.858586]  vfs_mkdir+0xfb/0x1b0
[   54.859458]  SyS_mkdir+0xcf/0xf0
[   54.860254]  do_syscall_64+0x73/0x1a0
[   54.861193]  entry_SYSCALL_64_after_hwframe+0x42/0xb7
[   54.862492] RIP: 0033:0x7fb73bddf547
[   54.863358] RSP: 002b:00007ffdaa553338 EFLAGS: 00000246 ORIG_RAX: 0000=
000000000053
[   54.865133] RAX: ffffffffffffffda RBX: 00007ffdaa55449a RCX: 00007fb73=
bddf547
[   54.866766] RDX: 0000000000000001 RSI: 00000000000001ff RDI: 00007ffda=
a55449a
[   54.868432] RBP: 00007ffdaa55449a R08: 00000000000001ff R09: 00005623a=
8670dd0
[   54.870110] R10: 00007fb73be72d5b R11: 0000000000000246 R12: 000000000=
00001ff
[   54.871752] R13: 00007ffdaa5534b0 R14: 0000000000000000 R15: 00007ffda=
a553500
[   54.873429] XFS (loop0): xfs_do_force_shutdown(0x8) called from line 1=
024 of file fs/xfs/xfs_trans.c.  Return address = ffffffff814cd050
[   54.882790] XFS (loop0): Corruption of in-memory data detected.  Shutt=
ing down filesystem
[   54.884597] XFS (loop0): Please umount the filesystem and rectify the =
problem(s)

Note that this crash is only possible on v4 filesystemsi or v5
filesystems mounted with the ikeep mount option. For all other V5
filesystems, this problem cannot occur because we don't read inodes
we are allocating from disk - we simply overwrite them with the new
inode information.
Signed-Off-By: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NCarlos Maiolino <cmaiolino@redhat.com>
Tested-by: NCarlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

Commit 66f36464 ("xfs: remove if_rdev") moved storing of rdev
value for special inodes to VFS inodes, but forgot to preserve the
value of i_rdev when recycling a reclaimable xfs_inode.

This was detected by xfstest overlay/017 with inodex=on mount option
and xfs base fs. The test does a lookup of overlay chardev and blockdev
right after drop caches.

Overlayfs inodes hold a reference on underlying xfs inodes when mount
option index=on is configured. If drop caches reclaim xfs inodes, before
it relclaims overlayfs inodes, that can sometimes leave a reclaimable xfs
inode and that test hits that case quite often.

When that happens, the xfs inode cache remains broken (zere i_rdev)
until the next cycle mount or drop caches.

Fixes: 66f36464 ("xfs: remove if_rdev")
Signed-off-by: NAmir Goldstein <amir73il@gmail.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

Signed-off-by: NJeff Layton <jlayton@redhat.com>
Acked-by: NDarrick J. Wong <darrick.wong@oracle.com>
Acked-by: NDave Chinner <dchinner@redhat.com>

Eryu Guan reported seeing occasional hangs when running generic/269 with
a new fsstress that supports clonerange/deduperange.  The cause of this
hang is an infinite loop when we convert the CoW fork extents from
unwritten to real just prior to writing the pages out; the infinite
loop happens because there's nothing in the CoW fork to convert, and so
it spins forever.

The fundamental issue here is that when we go to perform these CoW fork
conversions, we're supposed to have an extent waiting for us, but the
low space CoW reaper has snuck in and blown them away!  There are four
conditions that can dissuade the reaper from touching our file -- no
reflink iflag; dirty page cache; writeback in progress; or directio in
progress.  We check the four conditions prior to taking the locks, but
we neglect to recheck them once we have the locks, which is how we end
up whacking the writeback that's in progress.

Therefore, refactor the four checks into a helper function and call it
once again once we have the locks to make sure we really want to reap
the inode.  While we're at it, add an ASSERT for this weird condition so
that we'll fail noisily if we ever screw this up again.
Reported-by: NEryu Guan <eguan@redhat.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Tested-by: NEryu Guan <eguan@redhat.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>

Replace the current haphazard dir2 shortform verifier callsites with a
centralized verifier function that can be called either with the default
verifier functions or with a custom set.  This helps us strengthen
integrity checking while providing us with flexibility for repair tools.

xfs_repair wants this to be able to supply its own verifier functions
when trying to fix possibly corrupt metadata.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NDave Chinner <dchinner@redhat.com>

When we're remounting the filesystem readonly, remove all CoW
preallocations prior to going ro.  If the fs goes down after the ro
remount, we never clean up the staging extents, which means xfs_check
will trip over them on a subsequent run.  Practically speaking, the next
mount will clean them up too, so this is unlikely to be seen.  Since we
shut down the cowblocks cleaner on remount-ro, we also have to make sure
we start it back up if/when we remount-rw.

Found by adding clonerange to fsstress and running xfs/017.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>

The EOFBLOCKS/COWBLOCKS tags are totally separate things, so track them
with separate i_flags.  Right now we're abusing IEOFBLOCKS for both,
which is totally bogus because we won't tag the inode with COWBLOCKS if
IEOFBLOCKS was set by a previous tagging of the inode with EOFBLOCKS.
Found by wiring up clonerange to fsstress in xfs/017.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>

For an XFS_IGET_INCORE iget operation, if the inode isn't in the cache,
return ENODATA so that we don't confuse it with the pre-existing ENOENT
cases (inode is in cache, but freed).
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDave Chinner <dchinner@redhat.com>

After xfs_ifree_cluster() finds an inode in the radix tree and verifies
that the inode number is what it expected, xfs_reclaim_inode() can swoop
in and free it. xfs_ifree_cluster() will then happily continue working
on the freed inode. Most importantly, it will mark the inode stale,
which will probably be overwritten when the inode slab object is
reallocated, but if it has already been reallocated then we can end up
with an inode spuriously marked stale.

In 8a17d7dd ("xfs: mark reclaimed inodes invalid earlier") we added
a second check to xfs_iflush_cluster() to detect this race, but the
similar RCU lookup in xfs_ifree_cluster() needs the same treatment.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

Rename 'struct wait_bit_queue::wait' to ::wq_entry, to more clearly
name it as a wait-queue entry.

Propagate it to a couple of usage sites where the wait-bit-queue internals
are exposed.

Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NIngo Molnar <mingo@kernel.org>

Check the inode cache for a particular inode number.  If it's in the
cache, check that it's not currently being reclaimed.  If it's not being
reclaimed, return zero if the inode is allocated.  This function will be
used by various scrubbers to decide if the cache is more up to date
than the disk in terms of checking if an inode is allocated.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>

The 0-day kernel test robot reports assertion failures on
!CONFIG_SMP kernels due to failed spin_is_locked() checks. As it
turns out, spin_is_locked() is hardcoded to return zero on
!CONFIG_SMP kernels and so this function cannot be relied on to
verify spinlock state in this configuration.

To avoid this problem, replace the associated asserts with lockdep
variants that do the right thing regardless of kernel configuration.
Drop the one assert that checks for an unlocked lock as there is no
suitable lockdep variant for that case. This moves the spinlock
checks from XFS debug code to lockdep, but generally provides the
same level of protection.
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

The AG inode iterator currently skips new inodes as such inodes are
inserted into the inode radix tree before they are fully
constructed. Certain contexts require the ability to wait on the
construction of new inodes, however. The fs-wide dquot release from
the quotaoff sequence is an example of this.

Update the AG inode iterator to support the ability to wait on
inodes flagged with XFS_INEW upon request. Create a new
xfs_inode_ag_iterator_flags() interface and support a set of
iteration flags to modify the iteration behavior. When the
XFS_AGITER_INEW_WAIT flag is set, include XFS_INEW flags in the
radix tree inode lookup and wait on them before the callback is
executed.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

Inodes that are inserted into the perag tree but still under
construction are flagged with the XFS_INEW bit. Most contexts either
skip such inodes when they are encountered or have the ability to
handle them.

The runtime quotaoff sequence introduces a context that must wait
for construction of such inodes to correctly ensure that all dquots
in the fs are released. In anticipation of this, support the ability
to wait on new inodes. Wake the appropriate bit when XFS_INEW is
cleared.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

We only want to reclaim preallocations from our periodic work item.
Currently this is archived by looking for a dirty inode, but that check
is rather fragile.  Instead add a flag to xfs_reflink_cancel_cow_* so
that the caller can ask for just cancelling unwritten extents in the COW
fork.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
[darrick: fix typos in commit message]
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

The xfs_eofblocks.eof_scan_owner field is an internal field to
facilitate invoking eofb scans from the kernel while under the iolock.
This is necessary because the eofb scan acquires the iolock of each
inode. Synchronous scans are invoked on certain buffered write failures
while under iolock. In such cases, the scan owner indicates that the
context for the scan already owns the particular iolock and prevents a
double lock deadlock.

eofblocks scans while under iolock are still livelock prone in the event
of multiple parallel scans, however. If multiple buffered writes to
different inodes fail and invoke eofblocks scans at the same time, each
scan avoids a deadlock with its own inode by virtue of the
eof_scan_owner field, but will never be able to acquire the iolock of
the inode from the parallel scan. Because the low free space scans are
invoked with SYNC_WAIT, the scan will not return until it has processed
every tagged inode and thus both scans will spin indefinitely on the
iolock being held across the opposite scan. This problem can be
reproduced reliably by generic/224 on systems with higher cpu counts
(x16).

To avoid this problem, simplify the semantics of eofblocks scans to
never invoke a scan while under iolock. This means that the buffered
write context must drop the iolock before the scan. It must reacquire
the lock before the write retry and also repeat the initial write
checks, as the original state might no longer be valid once the iolock
was dropped.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

xfs_free_eofblocks() requires the IOLOCK_EXCL lock, but is called from
different contexts where the lock may or may not be held. The
need_iolock parameter exists for this reason, to indicate whether
xfs_free_eofblocks() must acquire the iolock itself before it can
proceed.

This is ugly and confusing. Simplify the semantics of
xfs_free_eofblocks() to require the caller to acquire the iolock
appropriately and kill the need_iolock parameter. While here, the mp
param can be removed as well as the xfs_mount is accessible from the
xfs_inode structure. This patch does not change behavior.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

There is a race window between write_cache_pages calling
clear_page_dirty_for_io and XFS calling set_page_writeback, in which
the mapping for an inode is tagged neither as dirty, nor as writeback.

If the COW shrinker hits in exactly that window we'll remove the delayed
COW extents and writepages trying to write it back, which in release
kernels will manifest as corruption of the bmap btree, and in debug
kernels will trip the ASSERT about now calling xfs_bmapi_write with the
COWFORK flag for holes.  A complex customer load manages to hit this
window fairly reliably, probably by always having COW writeback in flight
while the cow shrinker runs.

This patch adds another check for having the I_DIRTY_PAGES flag set,
which is still set during this race window.  While this fixes the problem
I'm still not overly happy about the way the COW shrinker works as it
still seems a bit fragile.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>

This patch drops the XFS-own i_iolock and uses the VFS i_rwsem which
recently replaced i_mutex instead.  This means we only have to take
one lock instead of two in many fast path operations, and we can
also shrink the xfs_inode structure.  Thanks to the xfs_ilock family
there is very little churn, the only thing of note is that we need
to switch to use the lock_two_directory helper for taking the i_rwsem
on two inodes in a few places to make sure our lock order matches
the one used in the VFS.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Tested-by: NJens Axboe <axboe@fb.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

Filesystem shutdown testing on an older distro kernel has uncovered an
imbalanced locking pattern for the inode flush lock in
xfs_reclaim_inode(). Specifically, there is a double unlock sequence
between the call to xfs_iflush_abort() and xfs_reclaim_inode() at the
"reclaim:" label.

This actually does not cause obvious problems on current kernels due to
the current flush lock implementation. Older kernels use a counting
based flush lock mechanism, however, which effectively breaks the lock
indefinitely when an already unlocked flush lock is repeatedly unlocked.
Though this only currently occurs on filesystem shutdown, it has
reproduced the effect of elevating an fs shutdown to a system-wide crash
or hang.

As it turns out, the flush lock is not actually required for the reclaim
logic in xfs_reclaim_inode() because by that time we have already cycled
the flush lock once while holding ILOCK_EXCL. Therefore, remove the
additional flush lock/unlock cycle around the 'reclaim:' label and
update branches into this label to release the flush lock where
appropriate. Add an assert to xfs_ifunlock() to help prevent future
occurences of the same problem.
Reported-by: NZorro Lang <zlang@redhat.com>
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDave Chinner <dchinner@redhat.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

The cowblocks background scanner currently clears the cowblocks tag
for inodes without any real allocations in the cow fork. This
excludes inodes with only delalloc blocks in the cow fork. While we
might never expect to clear delalloc blocks from the cow fork in the
background scanner, it is not necessarily correct to clear the
cowblocks tag from such inodes.

For example, if the background scanner happens to process an inode
between a buffered write and writeback, the scanner catches the
inode in a state after delalloc blocks have been allocated to the
cow fork but before the delalloc blocks have been converted to real
blocks by writeback. The background scanner then incorrectly clears
the cowblocks tag, even if part of the aforementioned delalloc
reservation will not be remapped to the data fork (i.e., extra
blocks due to the cowextsize hint). This means that any such
additional blocks in the cow fork might never be reclaimed by the
background scanner and could persist until the inode itself is
reclaimed.

To address this problem, only skip and clear inodes without any cow
fork allocations whatsoever from the background scanner. While we
generally do not want to cancel delalloc reservations from the
background scanner, the pagecache dirty check following the
cowblocks check should prevent that situation. If we do end up with
delalloc cow fork blocks without a dirty address space mapping, this
is probably an indication that something has gone wrong and the
blocks should be reclaimed, as they may never be converted to a real
allocation.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

These calls are still using the eofblocks tracepoints. The cowblocks
equivalents are already defined, we just aren't actually calling them.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

Trim CoW reservations made on behalf of a cowextsz hint if they get too
old or we run low on quota, so long as we don't have dirty data awaiting
writeback or directio operations in progress.

Garbage collection of the cowextsize extents are kept separate from
prealloc extent reaping because setting the CoW prealloc lifetime to a
(much) higher value than the regular prealloc extent lifetime has been
useful for combatting CoW fragmentation on VM hosts where the VMs
experience bursty write behaviors and we can keep the utilization ratios
low enough that we don't start to run out of space.  IOWs, it benefits
us to keep the CoW fork reservations around for as long as we can unless
we run out of blocks or hit inode reclaim.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>

Introduce a new in-core fork for storing copy-on-write delalloc
reservations and allocated extents that are in the process of being
written out.
Signed-off-by: NDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>

For long growing file writes we will usually already have the
eofblocks tag set when adding more speculative preallocations.  Add
a flag in the inode to allow us to skip the the fairly expensive
AG-wide spinlocks and multiple radix tree operations in that case.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NDave Chinner <dchinner@redhat.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

The filesystem quiesce sequence performs the operations necessary to
drain all background work, push pending transactions through the log
infrastructure and wait on I/O resulting from the final AIL push. We
have had reports of remount,ro hangs in xfs_log_quiesce() ->
xfs_wait_buftarg(), however, and some instrumentation code to detect
transaction commits at this point in the quiesce sequence has inculpated
the eofblocks background scanner as a cause.

While higher level remount code generally prevents user modifications by
the time the filesystem has made it to xfs_log_quiesce(), the background
scanner may still be alive and can perform pending work at any time. If
this occurs between the xfs_log_force() and xfs_wait_buftarg() calls
within xfs_log_quiesce(), this can lead to an indefinite lockup in
xfs_wait_buftarg().

To prevent this problem, cancel the background eofblocks scan worker
during the remount read-only quiesce sequence. This suspends background
trimming when a filesystem is remounted read-only. This is only done in
the remount path because the freeze codepath has already locked out new
transactions by the time the filesystem attempts to quiesce (and thus
waiting on an active work item could deadlock). Kick the eofblocks
worker to pick up where it left off once an fs is remounted back to
read-write.
Signed-off-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NEric Sandeen <sandeen@redhat.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

Rearrange the inode tagging functions so that they are higher up in
xfs_cache.c and so there is no need for forward prototypes to be
defined. This is purely code movement, no other change.
Signed-off-by: NDave Chinner <dchinner@redhat.com>

Inode radix tree tagging for reclaim passes a lot of unnecessary
variables around. Over time the xfs-perag has grown a xfs_mount
backpointer, and an internal agno so we don't need to pass other
variables into the tagging functions to supply this information.

Rework the functions to pass the minimal variable set required
and simplify the internal logic and flow.
Signed-off-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

The last thing we do before using call_rcu() on an xfs_inode to be
freed is mark it as invalid. This means there is a window between
when we know for certain that the inode is going to be freed and
when we do actually mark it as "freed".

This is important in the context of RCU lookups - we can look up the
inode, find that it is valid, and then use it as such not realising
that it is in the final stages of being freed.

As such, mark the inode as being invalid the moment we know it is
going to be reclaimed. This can be done while we still hold the
XFS_ILOCK_EXCL and the flush lock in xfs_inode_reclaim, meaning that
it occurs well before we remove it from the radix tree, and that
the i_flags_lock, the XFS_ILOCK and the inode flush lock all act as
synchronisation points for detecting that an inode is about to go
away.

For defensive purposes, this allows us to add a further check to
xfs_iflush_cluster to ensure we skip inodes that are being freed
after we grab the XFS_ILOCK_SHARED and the flush lock - we know that
if the inode number if valid while we have these locks held we know
that it has not progressed through reclaim to the point where it is
clean and is about to be freed.

[bfoster: fixed __xfs_inode_clear_reclaim() using ip->i_ino after it
	  had already been zeroed.]
Signed-off-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Signed-off-by: NDave Chinner <david@fromorbit.com>

The xfs_inode freed in xfs_inode_free() has multiple allocated
structures attached to it. We free these in xfs_inode_free() before
we mark the inode as invalid, and before we run call_rcu() to queue
the structure for freeing.

Unfortunately, this freeing can race with other accesses that are in
the RCU current grace period that have found the inode in the radix
tree with a valid state.  This includes xfs_iflush_cluster(), which
calls xfs_inode_clean(), and that accesses the inode log item on the
xfs_inode.

The log item structure is freed in xfs_inode_free(), so there is the
possibility we can be accessing freed memory in xfs_iflush_cluster()
after validating the xfs_inode structure as being valid for this RCU
context. Hence we can get spuriously incorrect clean state returned
from such checks. This can lead to use thinking the inode is dirty
when it is, in fact, clean, and so incorrectly attaching it to the
buffer for IO and completion processing.

This then leads to use-after-free situations on the xfs_inode itself
if the IO completes after the current RCU grace period expires. The
buffer callbacks will access the xfs_inode and try to do all sorts
of things it shouldn't with freed memory.

IOWs, xfs_iflush_cluster() only works correctly when racing with
inode reclaim if the inode log item is present and correctly stating
the inode is clean. If the inode is being freed, then reclaim has
already made sure the inode is clean, and hence xfs_iflush_cluster
can skip it. However, we are accessing the inode inode under RCU
read lock protection and so also must ensure that all dynamically
allocated memory we reference in this context is not freed until the
RCU grace period expires.

To fix this, move all the potential memory freeing into
xfs_inode_free_callback() so that we are guarantee RCU protected
lookup code will always have the memory structures it needs
available during the RCU grace period that lookup races can occur
in.
Discovered-by: NBrain Foster <bfoster@redhat.com>
Signed-off-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDave Chinner <david@fromorbit.com>

Move the di_mode value from the xfs_icdinode to the VFS inode, reducing
the xfs_icdinode byte another 2 bytes and collapsing another 2 byte hole
in the structure.
Signed-off-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDave Chinner <david@fromorbit.com>

We can store the di_changecount in the i_version field of the VFS
inode and remove another 8 bytes from the xfs_icdinode.
Signed-off-by: NDave Chinner <dchinner@redhat.com>
Reviewed-by: NBrian Foster <bfoster@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDave Chinner <david@fromorbit.com>