When we are running out of space for updating the chunk tree, that is,
when we are low on available space in the system space info, if we have
many task concurrently allocating block groups, via fallocate for example,
many of them can end up all allocating new system chunks when only one is
needed. In extreme cases this can lead to exhaustion of the system chunk
array, which has a size limit of 2048 bytes, and results in a transaction
abort with errno EFBIG, producing a trace in dmesg like the following,
which was triggered on a PowerPC machine with a node/leaf size of 64K:
[1359.518899] ------------[ cut here ]------------
[1359.518980] BTRFS: Transaction aborted (error -27)
[1359.519135] WARNING: CPU: 3 PID: 16463 at ../fs/btrfs/block-group.c:1968 btrfs_create_pending_block_groups+0x340/0x3c0 [btrfs]
[1359.519152] Modules linked in: (...)
[1359.519239] Supported: Yes, External
[1359.519252] CPU: 3 PID: 16463 Comm: stress-ng Tainted: G X 5.3.18-47-default #1 SLE15-SP3
[1359.519274] NIP: c008000000e36fe8 LR: c008000000e36fe4 CTR: 00000000006de8e8
[1359.519293] REGS: c00000056890b700 TRAP: 0700 Tainted: G X (5.3.18-47-default)
[1359.519317] MSR: 800000000282b033 <SF,VEC,VSX,EE,FP,ME,IR,DR,RI,LE> CR: 48008222 XER: 00000007
[1359.519356] CFAR: c00000000013e170 IRQMASK: 0
[1359.519356] GPR00: c008000000e36fe4 c00000056890b990 c008000000e83200 0000000000000026
[1359.519356] GPR04: 0000000000000000 0000000000000000 0000d52a3b027651 0000000000000007
[1359.519356] GPR08: 0000000000000003 0000000000000001 0000000000000007 0000000000000000
[1359.519356] GPR12: 0000000000008000 c00000063fe44600 000000001015e028 000000001015dfd0
[1359.519356] GPR16: 000000000000404f 0000000000000001 0000000000010000 0000dd1e287affff
[1359.519356] GPR20: 0000000000000001 c000000637c9a000 ffffffffffffffe5 0000000000000000
[1359.519356] GPR24: 0000000000000004 0000000000000000 0000000000000100 ffffffffffffffc0
[1359.519356] GPR28: c000000637c9a000 c000000630e09230 c000000630e091d8 c000000562188b08
[1359.519561] NIP [c008000000e36fe8] btrfs_create_pending_block_groups+0x340/0x3c0 [btrfs]
[1359.519613] LR [c008000000e36fe4] btrfs_create_pending_block_groups+0x33c/0x3c0 [btrfs]
[1359.519626] Call Trace:
[1359.519671] [c00000056890b990] [c008000000e36fe4] btrfs_create_pending_block_groups+0x33c/0x3c0 [btrfs] (unreliable)
[1359.519729] [c00000056890ba90] [c008000000d68d44] __btrfs_end_transaction+0xbc/0x2f0 [btrfs]
[1359.519782] [c00000056890bae0] [c008000000e309ac] btrfs_alloc_data_chunk_ondemand+0x154/0x610 [btrfs]
[1359.519844] [c00000056890bba0] [c008000000d8a0fc] btrfs_fallocate+0xe4/0x10e0 [btrfs]
[1359.519891] [c00000056890bd00] [c0000000004a23b4] vfs_fallocate+0x174/0x350
[1359.519929] [c00000056890bd50] [c0000000004a3cf8] ksys_fallocate+0x68/0xf0
[1359.519957] [c00000056890bda0] [c0000000004a3da8] sys_fallocate+0x28/0x40
[1359.519988] [c00000056890bdc0] [c000000000038968] system_call_exception+0xe8/0x170
[1359.520021] [c00000056890be20] [c00000000000cb70] system_call_common+0xf0/0x278
[1359.520037] Instruction dump:
[1359.520049] 7d0049ad 40c2fff4 7c0004ac 71490004 40820024 2f83fffb 419e0048 3c620000
[1359.520082] e863bcb8 7ec4b378 48010d91 e8410018 <0fe00000> 3c820000 e884bcc8 7ec6b378
[1359.520122] ---[ end trace d6c186e151022e20 ]---
The following steps explain how we can end up in this situation:
1) Task A is at check_system_chunk(), either because it is allocating a
new data or metadata block group, at btrfs_chunk_alloc(), or because
it is removing a block group or turning a block group RO. It does not
matter why;
2) Task A sees that there is not enough free space in the system
space_info object, that is 'left' is < 'thresh'. And at this point
the system space_info has a value of 0 for its 'bytes_may_use'
counter;
3) As a consequence task A calls btrfs_alloc_chunk() in order to allocate
a new system block group (chunk) and then reserves 'thresh' bytes in
the chunk block reserve with the call to btrfs_block_rsv_add(). This
changes the chunk block reserve's 'reserved' and 'size' counters by an
amount of 'thresh', and changes the 'bytes_may_use' counter of the
system space_info object from 0 to 'thresh'.
Also during its call to btrfs_alloc_chunk(), we end up increasing the
value of the 'total_bytes' counter of the system space_info object by
8MiB (the size of a system chunk stripe). This happens through the
call chain:
btrfs_alloc_chunk()
create_chunk()
btrfs_make_block_group()
btrfs_update_space_info()
4) After it finishes the first phase of the block group allocation, at
btrfs_chunk_alloc(), task A unlocks the chunk mutex;
5) At this point the new system block group was added to the transaction
handle's list of new block groups, but its block group item, device
items and chunk item were not yet inserted in the extent, device and
chunk trees, respectively. That only happens later when we call
btrfs_finish_chunk_alloc() through a call to
btrfs_create_pending_block_groups();
Note that only when we update the chunk tree, through the call to
btrfs_finish_chunk_alloc(), we decrement the 'reserved' counter
of the chunk block reserve as we COW/allocate extent buffers,
through:
btrfs_alloc_tree_block()
btrfs_use_block_rsv()
btrfs_block_rsv_use_bytes()
And the system space_info's 'bytes_may_use' is decremented everytime
we allocate an extent buffer for COW operations on the chunk tree,
through:
btrfs_alloc_tree_block()
btrfs_reserve_extent()
find_free_extent()
btrfs_add_reserved_bytes()
If we end up COWing less chunk btree nodes/leaves than expected, which
is the typical case since the amount of space we reserve is always
pessimistic to account for the worst possible case, we release the
unused space through:
btrfs_create_pending_block_groups()
btrfs_trans_release_chunk_metadata()
btrfs_block_rsv_release()
block_rsv_release_bytes()
btrfs_space_info_free_bytes_may_use()
But before task A gets into btrfs_create_pending_block_groups()...
6) Many other tasks start allocating new block groups through fallocate,
each one does the first phase of block group allocation in a
serialized way, since btrfs_chunk_alloc() takes the chunk mutex
before calling check_system_chunk() and btrfs_alloc_chunk().
However before everyone enters the final phase of the block group
allocation, that is, before calling btrfs_create_pending_block_groups(),
new tasks keep coming to allocate new block groups and while at
check_system_chunk(), the system space_info's 'bytes_may_use' keeps
increasing each time a task reserves space in the chunk block reserve.
This means that eventually some other task can end up not seeing enough
free space in the system space_info and decide to allocate yet another
system chunk.
This may repeat several times if yet more new tasks keep allocating
new block groups before task A, and all the other tasks, finish the
creation of the pending block groups, which is when reserved space
in excess is released. Eventually this can result in exhaustion of
system chunk array in the superblock, with btrfs_add_system_chunk()
returning EFBIG, resulting later in a transaction abort.
Even when we don't reach the extreme case of exhausting the system
array, most, if not all, unnecessarily created system block groups
end up being unused since when finishing creation of the first
pending system block group, the creation of the following ones end
up not needing to COW nodes/leaves of the chunk tree, so we never
allocate and deallocate from them, resulting in them never being
added to the list of unused block groups - as a consequence they
don't get deleted by the cleaner kthread - the only exceptions are
if we unmount and mount the filesystem again, which adds any unused
block groups to the list of unused block groups, if a scrub is
run, which also adds unused block groups to the unused list, and
under some circumstances when using a zoned filesystem or async
discard, which may also add unused block groups to the unused list.
So fix this by:
*) Tracking the number of reserved bytes for the chunk tree per
transaction, which is the sum of reserved chunk bytes by each
transaction handle currently being used;
*) When there is not enough free space in the system space_info,
if there are other transaction handles which reserved chunk space,
wait for some of them to complete in order to have enough excess
reserved space released, and then try again. Otherwise proceed with
the creation of a new system chunk.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Tree manipulating operations like merging nodes often release
once-allocated tree nodes. Such nodes are cleaned so that pages in the
node are not uselessly written out. On zoned volumes, however, such
optimization blocks the following IOs as the cancellation of the write
out of the freed blocks breaks the sequential write sequence expected by
the device.
Introduce a list of clean and unwritten extent buffers that have been
released in a transaction. Redirty the buffers so that
btree_write_cache_pages() can send proper bios to the devices.
Besides it clears the entire content of the extent buffer not to confuse
raw block scanners e.g. 'btrfs check'. By clearing the content,
csum_dirty_buffer() complains about bytenr mismatch, so avoid the
checking and checksum using newly introduced buffer flag
EXTENT_BUFFER_NO_CHECK.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Often an fsync needs to fallback to a transaction commit for several
reasons (to ensure consistency after a power failure, a new block group
was allocated or a temporary error such as ENOMEM or ENOSPC happened).
In that case the log is marked as needing a full commit and any concurrent
tasks attempting to log inodes or commit the log will also fallback to the
transaction commit. When this happens they all wait for the task that first
started the transaction commit to finish the transaction commit - however
they wait until the full transaction commit happens, which is not needed,
as they only need to wait for the superblocks to be persisted and not for
unpinning all the extents pinned during the transaction's lifetime, which
even for short lived transactions can be a few thousand and take some
significant amount of time to complete - for dbench workloads I have
observed up to 4~5 milliseconds of time spent unpinning extents in the
worst cases, and the number of pinned extents was between 2 to 3 thousand.
So allow fsync tasks to skip waiting for the unpinning of extents when
they call btrfs_commit_transaction() and they were not the task that
started the transaction commit (that one has to do it, the alternative
would be to offload the transaction commit to another task so that it
could avoid waiting for the extent unpinning or offload the extent
unpinning to another task).
This patch is part of a patchset comprised of the following patches:
btrfs: remove unnecessary directory inode item update when deleting dir entry
btrfs: stop setting nbytes when filling inode item for logging
btrfs: avoid logging new ancestor inodes when logging new inode
btrfs: skip logging directories already logged when logging all parents
btrfs: skip logging inodes already logged when logging new entries
btrfs: remove unnecessary check_parent_dirs_for_sync()
btrfs: make concurrent fsyncs wait less when waiting for a transaction commit
After applying the entire patchset, dbench shows improvements in respect
to throughput and latency. The script used to measure it is the following:
$ cat dbench-test.sh
#!/bin/bash
DEV=/dev/sdk
MNT=/mnt/sdk
MOUNT_OPTIONS="-o ssd"
MKFS_OPTIONS="-m single -d single"
echo "performance" | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
umount $DEV &> /dev/null
mkfs.btrfs -f $MKFS_OPTIONS $DEV
mount $MOUNT_OPTIONS $DEV $MNT
dbench -D $MNT -t 300 64
umount $MNT
The test was run on a physical machine with 12 cores (Intel corei7), 64G
of ram, using a NVMe device and a non-debug kernel configuration (Debian's
default configuration).
Before applying patchset, 32 clients:
Operation Count AvgLat MaxLat
----------------------------------------
NTCreateX 9627107 0.153 61.938
Close 7072076 0.001 3.175
Rename 407633 1.222 44.439
Unlink 1943895 0.658 44.440
Deltree 256 17.339 110.891
Mkdir 128 0.003 0.009
Qpathinfo 8725406 0.064 17.850
Qfileinfo 1529516 0.001 2.188
Qfsinfo 1599884 0.002 1.457
Sfileinfo 784200 0.005 3.562
Find 3373513 0.411 30.312
WriteX 4802132 0.053 29.054
ReadX 15089959 0.002 5.801
LockX 31344 0.002 0.425
UnlockX 31344 0.001 0.173
Flush 674724 5.952 341.830
Throughput 1008.02 MB/sec 32 clients 32 procs max_latency=341.833 ms
After applying patchset, 32 clients:
After patchset, with 32 clients:
Operation Count AvgLat MaxLat
----------------------------------------
NTCreateX 9931568 0.111 25.597
Close 7295730 0.001 2.171
Rename 420549 0.982 49.714
Unlink 2005366 0.497 39.015
Deltree 256 11.149 89.242
Mkdir 128 0.002 0.014
Qpathinfo 9001863 0.049 20.761
Qfileinfo 1577730 0.001 2.546
Qfsinfo 1650508 0.002 3.531
Sfileinfo 809031 0.005 5.846
Find 3480259 0.309 23.977
WriteX 4952505 0.043 41.283
ReadX 15568127 0.002 5.476
LockX 32338 0.002 0.978
UnlockX 32338 0.001 2.032
Flush 696017 7.485 228.835
Throughput 1049.91 MB/sec 32 clients 32 procs max_latency=228.847 ms
--> +4.1% throughput, -39.6% max latency
Before applying patchset, 64 clients:
Operation Count AvgLat MaxLat
----------------------------------------
NTCreateX 8956748 0.342 108.312
Close 6579660 0.001 3.823
Rename 379209 2.396 81.897
Unlink 1808625 1.108 131.148
Deltree 256 25.632 172.176
Mkdir 128 0.003 0.018
Qpathinfo 8117615 0.131 55.916
Qfileinfo 1423495 0.001 2.635
Qfsinfo 1488496 0.002 5.412
Sfileinfo 729472 0.007 8.643
Find 3138598 0.855 78.321
WriteX 4470783 0.102 79.442
ReadX 14038139 0.002 7.578
LockX 29158 0.002 0.844
UnlockX 29158 0.001 0.567
Flush 627746 14.168 506.151
Throughput 924.738 MB/sec 64 clients 64 procs max_latency=506.154 ms
After applying patchset, 64 clients:
Operation Count AvgLat MaxLat
----------------------------------------
NTCreateX 9069003 0.303 43.193
Close 6662328 0.001 3.888
Rename 383976 2.194 46.418
Unlink 1831080 1.022 43.873
Deltree 256 24.037 155.763
Mkdir 128 0.002 0.005
Qpathinfo 8219173 0.137 30.233
Qfileinfo 1441203 0.001 3.204
Qfsinfo 1507092 0.002 4.055
Sfileinfo 738775 0.006 5.431
Find 3177874 0.936 38.170
WriteX 4526152 0.084 39.518
ReadX 14213562 0.002 24.760
LockX 29522 0.002 1.221
UnlockX 29522 0.001 0.694
Flush 635652 14.358 422.039
Throughput 990.13 MB/sec 64 clients 64 procs max_latency=422.043 ms
--> +6.8% throughput, -18.1% max latency
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We love running delayed refs in commit_cowonly_roots, but it is a bit
excessive. I was seeing cases of running 3 or 4 refs a few times in a
row during this time. Instead simply:
- update all of the roots first
- then run delayed refs
- then handle the empty block groups case
- and then if we have any more dirty roots do the whole thing again
This allows us to be much more efficient with our delayed ref running,
as we can batch a few more operations at once.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This was added in commit 361048f586 ("Btrfs: fix full backref problem
when inserting shared block reference") to address a problem where we
hit the following BUG_ON() in alloc_reserved_tree_block
if (node->type == BTRFS_SHARED_BLOCK_REF_KEY) {
BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF));
However this BUG_ON() is bogus, and was removed by previous commit:
btrfs: remove bogus BUG_ON in alloc_reserved_tree_block
We no longer need to run delayed refs because of this, and can remove
this flushing here.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The commit d672633545 ("btrfs: qgroup: Make snapshot accounting work
with new extent-oriented qgroup.") added a flush of the delayed refs
during snapshot creation in order to get the qgroup accounting properly.
However this code has changed and been moved to it's own helper that is
skipped if qgroups are turned off. Move the flushing to the helper, as
we do not need it when qgroups are turned off.
Also add a comment explaining why it exists, and why it doesn't actually
save us. This will be helpful later when we try to fix qgroup
accounting properly.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We try to pre-flush the delayed refs when committing, because we want to
do as little work as possible in the critical section of the transaction
commit.
However doing this twice can lead to very long transaction commit delays
as other threads are allowed to continue to generate more delayed refs,
which potentially delays the commit by multiple minutes in very extreme
cases.
So simply stick to one pre-flush, and then continue the rest of the
transaction commit.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
I've been running a stress test that runs 20 workers in their own
subvolume, which are running an fsstress instance with 4 threads per
worker, which is 80 total fsstress threads. In addition to this I'm
running balance in the background as well as creating and deleting
snapshots. This test takes around 12 hours to run normally, going
slower and slower as the test goes on.
The reason for this is because fsstress is running fsync sometimes, and
because we're messing with block groups we often fall through to
btrfs_commit_transaction, so will often have 20-30 threads all calling
btrfs_commit_transaction at the same time.
These all get stuck contending on the extent tree while they try to run
delayed refs during the initial part of the commit.
This is suboptimal, really because the extent tree is a single point of
failure we only want one thread acting on that tree at once to reduce
lock contention.
Fix this by making the flushing mechanism a bit operation, to make it
easy to use test_and_set_bit() in order to make sure only one task does
this initial flush.
Once we're into the transaction commit we only have one thread doing
delayed ref running, it's just this initial pre-flush that is
problematic. With this patch my stress test takes around 90 minutes to
run, instead of 12 hours.
The memory barrier is not necessary for the flushing bit as it's
ordered, unlike plain int. The transaction state accessed in
btrfs_should_end_transaction could be affected by that too as it's not
always used under transaction lock. Upon Nikolay's analysis in [1]
it's not necessary:
In should_end_transaction it's read without holding any locks. (U)
It's modified in btrfs_cleanup_transaction without holding the
fs_info->trans_lock (U), but the STATE_ERROR flag is going to be set.
set in cleanup_transaction under fs_info->trans_lock (L)
set in btrfs_commit_trans to COMMIT_START under fs_info->trans_lock.(L)
set in btrfs_commit_trans to COMMIT_DOING under fs_info->trans_lock.(L)
set in btrfs_commit_trans to COMMIT_UNBLOCK under
fs_info->trans_lock.(L)
set in btrfs_commit_trans to COMMIT_COMPLETED without locks but at this
point the transaction is finished and fs_info->running_trans is NULL (U
but irrelevant).
So by the looks of it we can have a concurrent READ race with a WRITE,
due to reads not taking a lock. In this case what we want to ensure is
we either see new or old state. I consulted with Will Deacon and he said
that in such a case we'd want to annotate the accesses to ->state with
(READ|WRITE)_ONCE so as to avoid a theoretical tear, in this case I
don't think this could happen but I imagine at some point KCSAN would
flag such an access as racy (which it is).
[1] https://lore.kernel.org/linux-btrfs/e1fd5cc1-0f28-f670-69f4-e9958b4964e6@suse.com
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[ add comments regarding memory barrier ]
Signed-off-by: David Sterba <dsterba@suse.com>
This better reflects the semantics of the function i.e no search is
performed whatsoever.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
While doing error injection I would sometimes get a corrupt file system.
This is because I was injecting errors at btrfs_search_slot, but would
only do it one time per stack. This uncovered a problem in
commit_fs_roots, where if we get an error we would just break. However
we're in a nested loop, the first loop being a loop to find all the
dirty fs roots, and then subsequent root updates would succeed clearing
the error value.
This isn't likely to happen in real scenarios, however we could
potentially get a random ENOMEM once and then not again, and we'd end up
with a corrupted file system. Fix this by moving the error checking
around a bit to the main loop, as this is the only place where something
will fail, and return the error as soon as it occurs.
With this patch my reproducer no longer corrupts the file system.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The inode number cache has been removed in this dev cycle, there's one
more leftover. We don't need to run the delayed refs again after
commit_fs_roots as stated in the comment, because btrfs_save_ino_cache
is no more since 5297199a8b ("btrfs: remove inode number cache
feature").
Nothing else between commit_fs_roots and btrfs_qgroup_account_extents
could create new delayed refs so the qgroup consistency should be safe.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When mounting, btrfs uses the cache_generation in the super block to
determine if space cache v1 is in use. However, by mounting with
nospace_cache or space_cache=v2, it is possible to disable space cache
v1, which does not result in un-setting cache_generation back to 0.
In order to base some logic, like mount option printing in /proc/mounts,
on the current state of the space cache rather than just the values of
the mount option, keep the value of cache_generation consistent with the
status of space cache v1.
We ensure that cache_generation > 0 iff the file system is using
space_cache v1. This requires committing a transaction on any mount
which changes whether we are using v1. (v1->nospace_cache, v1->v2,
nospace_cache->v1, v2->v1).
Since the mechanism for writing out the cache generation is transaction
commit, but we want some finer grained control over when we un-set it,
we can't just rely on the SPACE_CACHE mount option, and introduce an
fs_info flag that mount can use when it wants to unset the generation.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
It's been deprecated since commit b547a88ea5 ("btrfs: start
deprecation of mount option inode_cache") which enumerates the reasons.
A filesystem that uses the feature (mount -o inode_cache) tracks the
inode numbers in bitmaps, that data stay on the filesystem after this
patch. The size is roughly 5MiB for 1M inodes [1], which is considered
small enough to be left there. Removal of the change can be implemented
in btrfs-progs if needed.
[1] https://lore.kernel.org/linux-btrfs/20201127145836.GZ6430@twin.jikos.cz/
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
While writing an explanation for the need of the commit_root_sem for
btrfs_prepare_extent_commit, I realized we have a slight hole that could
result in leaked space if we have to do the old style caching. Consider
the following scenario
commit root
+----+----+----+----+----+----+----+
|\\\\| |\\\\|\\\\| |\\\\|\\\\|
+----+----+----+----+----+----+----+
0 1 2 3 4 5 6 7
new commit root
+----+----+----+----+----+----+----+
| | | |\\\\| | |\\\\|
+----+----+----+----+----+----+----+
0 1 2 3 4 5 6 7
Prior to this patch, we run btrfs_prepare_extent_commit, which updates
the last_byte_to_unpin, and then we subsequently run
switch_commit_roots. In this example lets assume that
caching_ctl->progress == 1 at btrfs_prepare_extent_commit() time, which
means that cache->last_byte_to_unpin == 1. Then we go and do the
switch_commit_roots(), but in the meantime the caching thread has made
some more progress, because we drop the commit_root_sem and re-acquired
it. Now caching_ctl->progress == 3. We swap out the commit root and
carry on to unpin.
The race can happen like:
1) The caching thread was running using the old commit root when it
found the extent for [2, 3);
2) Then it released the commit_root_sem because it was in the last
item of a leaf and the semaphore was contended, and set ->progress
to 3 (value of 'last'), as the last extent item in the current leaf
was for the extent for range [2, 3);
3) Next time it gets the commit_root_sem, will start using the new
commit root and search for a key with offset 3, so it never finds
the hole for [2, 3).
So the caching thread never saw [2, 3) as free space in any of the
commit roots, and by the time finish_extent_commit() was called for
the range [0, 3), ->last_byte_to_unpin was 1, so it only returned the
subrange [0, 1) to the free space cache, skipping [2, 3).
In the unpin code we have last_byte_to_unpin == 1, so we unpin [0,1),
but do not unpin [2,3). However because caching_ctl->progress == 3 we
do not see the newly freed section of [2,3), and thus do not add it to
our free space cache. This results in us missing a chunk of free space
in memory (on disk too, unless we have a power failure before writing
the free space cache to disk).
Fix this by making sure the ->last_byte_to_unpin is set at the same time
that we swap the commit roots, this ensures that we will always be
consistent.
CC: stable@vger.kernel.org # 5.8+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[ update changelog with Filipe's review comments ]
Signed-off-by: David Sterba <dsterba@suse.com>
Now that we're using a rw_semaphore we no longer need to indicate if a
lock is blocking or not, nor do we need to flip the entire path from
blocking to spinning. Remove these helpers and all the places they are
called.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We do not need anymore to start writeback for delalloc of roots that are
being snapshotted and wait for it to complete. This was done in commit
609e804d77 ("Btrfs: fix file corruption after snapshotting due to mix
of buffered/DIO writes") to fix a type of file corruption where files in a
snapshot end up having their i_size updated in a non-ordered way, leaving
implicit file holes, when buffered IO writes that increase a file's size
are followed by direct IO writes that also increase the file's size.
This is not needed anymore because we now have a more generic mechanism
to prevent a non-ordered i_size update since commit 9ddc959e80
("btrfs: use the file extent tree infrastructure"), which addresses this
scenario involving snapshots as well.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When we COW a block we are holding a lock on the original block, and
then we lock the new COW block. Because our lockdep maps are based on
root + level, this will make lockdep complain. We need a way to
indicate a subclass for locking the COW'ed block, so plumb through our
btrfs_lock_nesting from btrfs_cow_block down to the btrfs_init_buffer,
and then introduce BTRFS_NESTING_COW to be used for cow'ing blocks.
The reason I've added all this extra infrastructure is because there
will be need of different nesting classes in follow up patches.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently regardless of a full or a fast fsync we always wait for ordered
extents to complete, and then start logging the inode after that. However
for fast fsyncs we can just wait for the writeback to complete, we don't
need to wait for the ordered extents to complete since we use the list of
modified extents maps to figure out which extents we must log and we can
get their checksums directly from the ordered extents that are still in
flight, otherwise look them up from the checksums tree.
Until commit b5e6c3e170 ("btrfs: always wait on ordered extents at
fsync time"), for fast fsyncs, we used to start logging without even
waiting for the writeback to complete first, we would wait for it to
complete after logging, while holding a transaction open, which lead to
performance issues when using cgroups and probably for other cases too,
as wait for IO while holding a transaction handle should be avoided as
much as possible. After that, for fast fsyncs, we started to wait for
ordered extents to complete before starting to log, which adds some
latency to fsyncs and we even got at least one report about a performance
drop which bisected to that particular change:
https://lore.kernel.org/linux-btrfs/20181109215148.GF23260@techsingularity.net/
This change makes fast fsyncs only wait for writeback to finish before
starting to log the inode, instead of waiting for both the writeback to
finish and for the ordered extents to complete. This brings back part of
the logic we had that extracts checksums from in flight ordered extents,
which are not yet in the checksums tree, and making sure transaction
commits wait for the completion of ordered extents previously logged
(by far most of the time they have already completed by the time a
transaction commit starts, resulting in no wait at all), to avoid any
data loss if an ordered extent completes after the transaction used to
log an inode is committed, followed by a power failure.
When there are no other tasks accessing the checksums and the subvolume
btrees, the ordered extent completion is pretty fast, typically taking
100 to 200 microseconds only in my observations. However when there are
other tasks accessing these btrees, ordered extent completion can take a
lot more time due to lock contention on nodes and leaves of these btrees.
I've seen cases over 2 milliseconds, which starts to be significant. In
particular when we do have concurrent fsyncs against different files there
is a lot of contention on the checksums btree, since we have many tasks
writing the checksums into the btree and other tasks that already started
the logging phase are doing lookups for checksums in the btree.
This change also turns all ranged fsyncs into full ranged fsyncs, which
is something we already did when not using the NO_HOLES features or when
doing a full fsync. This is to guarantee we never miss checksums due to
writeback having been triggered only for a part of an extent, and we end
up logging the full extent but only checksums for the written range, which
results in missing checksums after log replay. Allowing ranged fsyncs to
operate again only in the original range, when using the NO_HOLES feature
and doing a fast fsync is doable but requires some non trivial changes to
the writeback path, which can always be worked on later if needed, but I
don't think they are a very common use case.
Several tests were performed using fio for different numbers of concurrent
jobs, each writing and fsyncing its own file, for both sequential and
random file writes. The tests were run on bare metal, no virtualization,
on a box with 12 cores (Intel i7-8700), 64Gb of RAM and a NVMe device,
with a kernel configuration that is the default of typical distributions
(debian in this case), without debug options enabled (kasan, kmemleak,
slub debug, debug of page allocations, lock debugging, etc).
The following script that calls fio was used:
$ cat test-fsync.sh
#!/bin/bash
DEV=/dev/nvme0n1
MNT=/mnt/btrfs
MOUNT_OPTIONS="-o ssd -o space_cache=v2"
MKFS_OPTIONS="-d single -m single"
if [ $# -ne 5 ]; then
echo "Use $0 NUM_JOBS FILE_SIZE FSYNC_FREQ BLOCK_SIZE [write|randwrite]"
exit 1
fi
NUM_JOBS=$1
FILE_SIZE=$2
FSYNC_FREQ=$3
BLOCK_SIZE=$4
WRITE_MODE=$5
if [ "$WRITE_MODE" != "write" ] && [ "$WRITE_MODE" != "randwrite" ]; then
echo "Invalid WRITE_MODE, must be 'write' or 'randwrite'"
exit 1
fi
cat <<EOF > /tmp/fio-job.ini
[writers]
rw=$WRITE_MODE
fsync=$FSYNC_FREQ
fallocate=none
group_reporting=1
direct=0
bs=$BLOCK_SIZE
ioengine=sync
size=$FILE_SIZE
directory=$MNT
numjobs=$NUM_JOBS
EOF
echo "performance" | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
echo
echo "Using config:"
echo
cat /tmp/fio-job.ini
echo
umount $MNT &> /dev/null
mkfs.btrfs -f $MKFS_OPTIONS $DEV
mount $MOUNT_OPTIONS $DEV $MNT
fio /tmp/fio-job.ini
umount $MNT
The results were the following:
*************************
*** sequential writes ***
*************************
==== 1 job, 8GiB file, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=36.6MiB/s (38.4MB/s), 36.6MiB/s-36.6MiB/s (38.4MB/s-38.4MB/s), io=8192MiB (8590MB), run=223689-223689msec
After patch:
WRITE: bw=40.2MiB/s (42.1MB/s), 40.2MiB/s-40.2MiB/s (42.1MB/s-42.1MB/s), io=8192MiB (8590MB), run=203980-203980msec
(+9.8%, -8.8% runtime)
==== 2 jobs, 4GiB files, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=35.8MiB/s (37.5MB/s), 35.8MiB/s-35.8MiB/s (37.5MB/s-37.5MB/s), io=8192MiB (8590MB), run=228950-228950msec
After patch:
WRITE: bw=43.5MiB/s (45.6MB/s), 43.5MiB/s-43.5MiB/s (45.6MB/s-45.6MB/s), io=8192MiB (8590MB), run=188272-188272msec
(+21.5% throughput, -17.8% runtime)
==== 4 jobs, 2GiB files, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=50.1MiB/s (52.6MB/s), 50.1MiB/s-50.1MiB/s (52.6MB/s-52.6MB/s), io=8192MiB (8590MB), run=163446-163446msec
After patch:
WRITE: bw=64.5MiB/s (67.6MB/s), 64.5MiB/s-64.5MiB/s (67.6MB/s-67.6MB/s), io=8192MiB (8590MB), run=126987-126987msec
(+28.7% throughput, -22.3% runtime)
==== 8 jobs, 1GiB files, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=64.0MiB/s (68.1MB/s), 64.0MiB/s-64.0MiB/s (68.1MB/s-68.1MB/s), io=8192MiB (8590MB), run=126075-126075msec
After patch:
WRITE: bw=86.8MiB/s (91.0MB/s), 86.8MiB/s-86.8MiB/s (91.0MB/s-91.0MB/s), io=8192MiB (8590MB), run=94358-94358msec
(+35.6% throughput, -25.2% runtime)
==== 16 jobs, 512MiB files, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=79.8MiB/s (83.6MB/s), 79.8MiB/s-79.8MiB/s (83.6MB/s-83.6MB/s), io=8192MiB (8590MB), run=102694-102694msec
After patch:
WRITE: bw=107MiB/s (112MB/s), 107MiB/s-107MiB/s (112MB/s-112MB/s), io=8192MiB (8590MB), run=76446-76446msec
(+34.1% throughput, -25.6% runtime)
==== 32 jobs, 512MiB files, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=93.2MiB/s (97.7MB/s), 93.2MiB/s-93.2MiB/s (97.7MB/s-97.7MB/s), io=16.0GiB (17.2GB), run=175836-175836msec
After patch:
WRITE: bw=111MiB/s (117MB/s), 111MiB/s-111MiB/s (117MB/s-117MB/s), io=16.0GiB (17.2GB), run=147001-147001msec
(+19.1% throughput, -16.4% runtime)
==== 64 jobs, 512MiB files, fsync frequency 1, block size 64KiB ====
Before patch:
WRITE: bw=108MiB/s (114MB/s), 108MiB/s-108MiB/s (114MB/s-114MB/s), io=32.0GiB (34.4GB), run=302656-302656msec
After patch:
WRITE: bw=133MiB/s (140MB/s), 133MiB/s-133MiB/s (140MB/s-140MB/s), io=32.0GiB (34.4GB), run=246003-246003msec
(+23.1% throughput, -18.7% runtime)
************************
*** random writes ***
************************
==== 1 job, 8GiB file, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=11.5MiB/s (12.0MB/s), 11.5MiB/s-11.5MiB/s (12.0MB/s-12.0MB/s), io=8192MiB (8590MB), run=714281-714281msec
After patch:
WRITE: bw=11.6MiB/s (12.2MB/s), 11.6MiB/s-11.6MiB/s (12.2MB/s-12.2MB/s), io=8192MiB (8590MB), run=705959-705959msec
(+0.9% throughput, -1.7% runtime)
==== 2 jobs, 4GiB files, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=12.8MiB/s (13.5MB/s), 12.8MiB/s-12.8MiB/s (13.5MB/s-13.5MB/s), io=8192MiB (8590MB), run=638101-638101msec
After patch:
WRITE: bw=13.1MiB/s (13.7MB/s), 13.1MiB/s-13.1MiB/s (13.7MB/s-13.7MB/s), io=8192MiB (8590MB), run=625374-625374msec
(+2.3% throughput, -2.0% runtime)
==== 4 jobs, 2GiB files, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=15.4MiB/s (16.2MB/s), 15.4MiB/s-15.4MiB/s (16.2MB/s-16.2MB/s), io=8192MiB (8590MB), run=531146-531146msec
After patch:
WRITE: bw=17.8MiB/s (18.7MB/s), 17.8MiB/s-17.8MiB/s (18.7MB/s-18.7MB/s), io=8192MiB (8590MB), run=460431-460431msec
(+15.6% throughput, -13.3% runtime)
==== 8 jobs, 1GiB files, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=19.9MiB/s (20.8MB/s), 19.9MiB/s-19.9MiB/s (20.8MB/s-20.8MB/s), io=8192MiB (8590MB), run=412664-412664msec
After patch:
WRITE: bw=22.2MiB/s (23.3MB/s), 22.2MiB/s-22.2MiB/s (23.3MB/s-23.3MB/s), io=8192MiB (8590MB), run=368589-368589msec
(+11.6% throughput, -10.7% runtime)
==== 16 jobs, 512MiB files, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=29.3MiB/s (30.7MB/s), 29.3MiB/s-29.3MiB/s (30.7MB/s-30.7MB/s), io=8192MiB (8590MB), run=279924-279924msec
After patch:
WRITE: bw=30.4MiB/s (31.9MB/s), 30.4MiB/s-30.4MiB/s (31.9MB/s-31.9MB/s), io=8192MiB (8590MB), run=269258-269258msec
(+3.8% throughput, -3.8% runtime)
==== 32 jobs, 512MiB files, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=36.9MiB/s (38.7MB/s), 36.9MiB/s-36.9MiB/s (38.7MB/s-38.7MB/s), io=16.0GiB (17.2GB), run=443581-443581msec
After patch:
WRITE: bw=41.6MiB/s (43.6MB/s), 41.6MiB/s-41.6MiB/s (43.6MB/s-43.6MB/s), io=16.0GiB (17.2GB), run=394114-394114msec
(+12.7% throughput, -11.2% runtime)
==== 64 jobs, 512MiB files, fsync frequency 16, block size 4KiB ====
Before patch:
WRITE: bw=45.9MiB/s (48.1MB/s), 45.9MiB/s-45.9MiB/s (48.1MB/s-48.1MB/s), io=32.0GiB (34.4GB), run=714614-714614msec
After patch:
WRITE: bw=48.8MiB/s (51.1MB/s), 48.8MiB/s-48.8MiB/s (51.1MB/s-51.1MB/s), io=32.0GiB (34.4GB), run=672087-672087msec
(+6.3% throughput, -6.0% runtime)
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Eric reported seeing this message while running generic/475
BTRFS: error (device dm-3) in btrfs_sync_log:3084: errno=-117 Filesystem corrupted
Full stack trace:
BTRFS: error (device dm-0) in btrfs_commit_transaction:2323: errno=-5 IO failure (Error while writing out transaction)
BTRFS info (device dm-0): forced readonly
BTRFS warning (device dm-0): Skipping commit of aborted transaction.
------------[ cut here ]------------
BTRFS: error (device dm-0) in cleanup_transaction:1894: errno=-5 IO failure
BTRFS: Transaction aborted (error -117)
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c6480 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c6488 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c6490 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c6498 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c64a0 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c64a8 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c64b0 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c64b8 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3555 rw 0,0 sector 0x1c64c0 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3572 rw 0,0 sector 0x1b85e8 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3572 rw 0,0 sector 0x1b85f0 len 4096 err no 10
WARNING: CPU: 3 PID: 23985 at fs/btrfs/tree-log.c:3084 btrfs_sync_log+0xbc8/0xd60 [btrfs]
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d4288 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d4290 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d4298 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42a0 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42a8 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42b0 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42b8 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42c0 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42c8 len 4096 err no 10
BTRFS warning (device dm-0): direct IO failed ino 3548 rw 0,0 sector 0x1d42d0 len 4096 err no 10
CPU: 3 PID: 23985 Comm: fsstress Tainted: G W L 5.8.0-rc4-default+ #1181
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-59-gc9ba527-rebuilt.opensuse.org 04/01/2014
RIP: 0010:btrfs_sync_log+0xbc8/0xd60 [btrfs]
RSP: 0018:ffff909a44d17bd0 EFLAGS: 00010286
RAX: 0000000000000000 RBX: 0000000000000001 RCX: 0000000000000001
RDX: ffff8f3be41cb940 RSI: ffffffffb0108d2b RDI: ffffffffb0108ff7
RBP: ffff909a44d17e70 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000037988 R12: ffff8f3bd20e4000
R13: ffff8f3bd20e4428 R14: 00000000ffffff8b R15: ffff909a44d17c70
FS: 00007f6a6ed3fb80(0000) GS:ffff8f3c3dc00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f6a6ed3e000 CR3: 00000000525c0003 CR4: 0000000000160ee0
Call Trace:
? finish_wait+0x90/0x90
? __mutex_unlock_slowpath+0x45/0x2a0
? lock_acquire+0xa3/0x440
? lockref_put_or_lock+0x9/0x30
? dput+0x20/0x4a0
? dput+0x20/0x4a0
? do_raw_spin_unlock+0x4b/0xc0
? _raw_spin_unlock+0x1f/0x30
btrfs_sync_file+0x335/0x490 [btrfs]
do_fsync+0x38/0x70
__x64_sys_fsync+0x10/0x20
do_syscall_64+0x50/0xe0
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x7f6a6ef1b6e3
Code: Bad RIP value.
RSP: 002b:00007ffd01e20038 EFLAGS: 00000246 ORIG_RAX: 000000000000004a
RAX: ffffffffffffffda RBX: 000000000007a120 RCX: 00007f6a6ef1b6e3
RDX: 00007ffd01e1ffa0 RSI: 00007ffd01e1ffa0 RDI: 0000000000000003
RBP: 0000000000000003 R08: 0000000000000001 R09: 00007ffd01e2004c
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000000009f
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
irq event stamp: 0
hardirqs last enabled at (0): [<0000000000000000>] 0x0
hardirqs last disabled at (0): [<ffffffffb007fe0b>] copy_process+0x67b/0x1b00
softirqs last enabled at (0): [<ffffffffb007fe0b>] copy_process+0x67b/0x1b00
softirqs last disabled at (0): [<0000000000000000>] 0x0
---[ end trace af146e0e38433456 ]---
BTRFS: error (device dm-0) in btrfs_sync_log:3084: errno=-117 Filesystem corrupted
This ret came from btrfs_write_marked_extents(). If we get an aborted
transaction via EIO before, we'll see it in btree_write_cache_pages()
and return EUCLEAN, which gets printed as "Filesystem corrupted".
Except we shouldn't be returning EUCLEAN here, we need to be returning
EROFS because EUCLEAN is reserved for actual corruption, not IO errors.
We are inconsistent about our handling of BTRFS_FS_STATE_ERROR
elsewhere, but we want to use EROFS for this particular case. The
original transaction abort has the real error code for why we ended up
with an aborted transaction, all subsequent actions just need to return
EROFS because they may not have a trans handle and have no idea about
the original cause of the abort.
After patch "btrfs: don't WARN if we abort a transaction with EROFS" the
stacktrace will not be dumped either.
Reported-by: Eric Sandeen <esandeen@redhat.com>
CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ add full test stacktrace ]
Signed-off-by: David Sterba <dsterba@suse.com>
commit a514d63882 ("btrfs: qgroup: Commit transaction in advance to
reduce early EDQUOT") tries to reduce the early EDQUOT problems by
checking the qgroup free against threshold and tries to wake up commit
kthread to free some space.
The problem of that mechanism is, it can only free qgroup per-trans
metadata space, can't do anything to data, nor prealloc qgroup space.
Now since we have the ability to flush qgroup space, and implemented
retry-after-EDQUOT behavior, such mechanism can be completely replaced.
So this patch will cleanup such mechanism in favor of
retry-after-EDQUOT.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
When the anonymous block device pool is exhausted, subvolume/snapshot
creation fails with EMFILE (Too many files open). This has been reported
by a user. The allocation happens in the second phase during transaction
commit where it's only way out is to abort the transaction
BTRFS: Transaction aborted (error -24)
WARNING: CPU: 17 PID: 17041 at fs/btrfs/transaction.c:1576 create_pending_snapshot+0xbc4/0xd10 [btrfs]
RIP: 0010:create_pending_snapshot+0xbc4/0xd10 [btrfs]
Call Trace:
create_pending_snapshots+0x82/0xa0 [btrfs]
btrfs_commit_transaction+0x275/0x8c0 [btrfs]
btrfs_mksubvol+0x4b9/0x500 [btrfs]
btrfs_ioctl_snap_create_transid+0x174/0x180 [btrfs]
btrfs_ioctl_snap_create_v2+0x11c/0x180 [btrfs]
btrfs_ioctl+0x11a4/0x2da0 [btrfs]
do_vfs_ioctl+0xa9/0x640
ksys_ioctl+0x67/0x90
__x64_sys_ioctl+0x1a/0x20
do_syscall_64+0x5a/0x110
entry_SYSCALL_64_after_hwframe+0x44/0xa9
---[ end trace 33f2f83f3d5250e9 ]---
BTRFS: error (device sda1) in create_pending_snapshot:1576: errno=-24 unknown
BTRFS info (device sda1): forced readonly
BTRFS warning (device sda1): Skipping commit of aborted transaction.
BTRFS: error (device sda1) in cleanup_transaction:1831: errno=-24 unknown
[CAUSE]
When the global anonymous block device pool is exhausted, the following
call chain will fail, and lead to transaction abort:
btrfs_ioctl_snap_create_v2()
|- btrfs_ioctl_snap_create_transid()
|- btrfs_mksubvol()
|- btrfs_commit_transaction()
|- create_pending_snapshot()
|- btrfs_get_fs_root()
|- btrfs_init_fs_root()
|- get_anon_bdev()
[FIX]
Although we can't enlarge the anonymous block device pool, at least we
can preallocate anon_dev for subvolume/snapshot in the first phase,
outside of transaction context and exactly at the moment the user calls
the creation ioctl.
Reported-by: Greed Rong <greedrong@gmail.com>
Link: https://lore.kernel.org/linux-btrfs/CA+UqX+NTrZ6boGnWHhSeZmEY5J76CTqmYjO2S+=tHJX7nb9DPw@mail.gmail.com/
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The main function to lookup a root by its id btrfs_get_fs_root takes the
whole key, while only using the objectid. The value of offset is preset
to (u64)-1 but not actually used until btrfs_find_root that does the
actual search.
Switch btrfs_get_fs_root to use only objectid and remove all local
variables that existed just for the lookup. The actual key for search is
set up in btrfs_get_fs_root, reusing another key variable.
Signed-off-by: David Sterba <dsterba@suse.com>
The name BTRFS_ROOT_REF_COWS is not very clear about the meaning.
In fact, that bit can only be set to those trees:
- Subvolume roots
- Data reloc root
- Reloc roots for above roots
All other trees won't get this bit set. So just by the result, it is
obvious that, roots with this bit set can have tree blocks shared with
other trees. Either shared by snapshots, or by reloc roots (an special
snapshot created by relocation).
This patch will rename BTRFS_ROOT_REF_COWS to BTRFS_ROOT_SHAREABLE to
make it easier to understand, and update all comment mentioning
"reference counted" to follow the rename.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Back in 2014, commit 04216820fe ("Btrfs: fix race between fs trimming
and block group remove/allocation"), I added the 'trimming' member to the
block group structure. Its purpose was to prevent races between trimming
and block group deletion/allocation by pinning the block group in a way
that prevents its logical address and device extents from being reused
while trimming is in progress for a block group, so that if another task
deletes the block group and then another task allocates a new block group
that gets the same logical address and device extents while the trimming
task is still in progress.
After the previous fix for scrub (patch "btrfs: fix a race between scrub
and block group removal/allocation"), scrub now also has the same needs that
trimming has, so the member name 'trimming' no longer makes sense.
Since there is already a 'pinned' member in the block group that refers
to space reservations (pinned bytes), rename the member to 'frozen',
add a comment on top of it to describe its general purpose and rename
the helpers to increment and decrement the counter as well, to match
the new member name.
The next patch in the series will move the helpers into a more suitable
file (from free-space-cache.c to block-group.c).
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Nikolay noticed a bunch of test failures with my global rsv steal
patches. At first he thought they were introduced by them, but they've
been failing for a while with 64k nodes.
The problem is with 64k nodes we have a global reserve that calculates
out to 13MiB on a freshly made file system, which only has 8MiB of
metadata space. Because of changes I previously made we no longer
account for the global reserve in the overcommit logic, which means we
correctly allow overcommit to happen even though we are already
overcommitted.
However in some corner cases, for example btrfs/170, we will allocate
the entire file system up with data chunks before we have enough space
pressure to allocate a metadata chunk. Then once the fs is full we
ENOSPC out because we cannot overcommit and the global reserve is taking
up all of the available space.
The most ideal way to deal with this is to change our space reservation
stuff to take into account the height of the tree's that we're
modifying, so that our global reserve calculation does not end up so
obscenely large.
However that is a huge undertaking. Instead fix this by forcing a chunk
allocation if the global reserve is larger than the total metadata
space. This gives us essentially the same behavior that happened
before, we get a chunk allocated and these tests can pass.
This is meant to be a stop-gap measure until we can tackle the "tree
height only" project.
Fixes: 0096420adb ("btrfs: do not account global reserve in can_overcommit")
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Tested-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For unlink transactions and block group removal
btrfs_start_transaction_fallback_global_rsv will first try to start an
ordinary transaction and if it fails it will fall back to reserving the
required amount by stealing from the global reserve. This is problematic
because of all the same reasons we had with previous iterations of the
ENOSPC handling, thundering herd. We get a bunch of failures all at
once, everybody tries to allocate from the global reserve, some win and
some lose, we get an ENSOPC.
Fix this behavior by introducing BTRFS_RESERVE_FLUSH_ALL_STEAL. It's
used to mark unlink reservation. To fix this we need to integrate this
logic into the normal ENOSPC infrastructure. We still go through all of
the normal flushing work, and at the moment we begin to fail all the
tickets we try to satisfy any tickets that are allowed to steal by
stealing from the global reserve. If this works we start the flushing
system over again just like we would with a normal ticket satisfaction.
This serializes our global reserve stealing, so we don't have the
thundering herd problem.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Tested-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
One run of btrfs/063 triggered the following lockdep warning:
============================================
WARNING: possible recursive locking detected
5.6.0-rc7-custom+ #48 Not tainted
--------------------------------------------
kworker/u24:0/7 is trying to acquire lock:
ffff88817d3a46e0 (sb_internal#2){.+.+}, at: start_transaction+0x66c/0x890 [btrfs]
but task is already holding lock:
ffff88817d3a46e0 (sb_internal#2){.+.+}, at: start_transaction+0x66c/0x890 [btrfs]
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(sb_internal#2);
lock(sb_internal#2);
*** DEADLOCK ***
May be due to missing lock nesting notation
4 locks held by kworker/u24:0/7:
#0: ffff88817b495948 ((wq_completion)btrfs-endio-write){+.+.}, at: process_one_work+0x557/0xb80
#1: ffff888189ea7db8 ((work_completion)(&work->normal_work)){+.+.}, at: process_one_work+0x557/0xb80
#2: ffff88817d3a46e0 (sb_internal#2){.+.+}, at: start_transaction+0x66c/0x890 [btrfs]
#3: ffff888174ca4da8 (&fs_info->reloc_mutex){+.+.}, at: btrfs_record_root_in_trans+0x83/0xd0 [btrfs]
stack backtrace:
CPU: 0 PID: 7 Comm: kworker/u24:0 Not tainted 5.6.0-rc7-custom+ #48
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
Workqueue: btrfs-endio-write btrfs_work_helper [btrfs]
Call Trace:
dump_stack+0xc2/0x11a
__lock_acquire.cold+0xce/0x214
lock_acquire+0xe6/0x210
__sb_start_write+0x14e/0x290
start_transaction+0x66c/0x890 [btrfs]
btrfs_join_transaction+0x1d/0x20 [btrfs]
find_free_extent+0x1504/0x1a50 [btrfs]
btrfs_reserve_extent+0xd5/0x1f0 [btrfs]
btrfs_alloc_tree_block+0x1ac/0x570 [btrfs]
btrfs_copy_root+0x213/0x580 [btrfs]
create_reloc_root+0x3bd/0x470 [btrfs]
btrfs_init_reloc_root+0x2d2/0x310 [btrfs]
record_root_in_trans+0x191/0x1d0 [btrfs]
btrfs_record_root_in_trans+0x90/0xd0 [btrfs]
start_transaction+0x16e/0x890 [btrfs]
btrfs_join_transaction+0x1d/0x20 [btrfs]
btrfs_finish_ordered_io+0x55d/0xcd0 [btrfs]
finish_ordered_fn+0x15/0x20 [btrfs]
btrfs_work_helper+0x116/0x9a0 [btrfs]
process_one_work+0x632/0xb80
worker_thread+0x80/0x690
kthread+0x1a3/0x1f0
ret_from_fork+0x27/0x50
It's pretty hard to reproduce, only one hit so far.
[CAUSE]
This is because we're calling btrfs_join_transaction() without re-using
the current running one:
btrfs_finish_ordered_io()
|- btrfs_join_transaction() <<< Call #1
|- btrfs_record_root_in_trans()
|- btrfs_reserve_extent()
|- btrfs_join_transaction() <<< Call #2
Normally such btrfs_join_transaction() call should re-use the existing
one, without trying to re-start a transaction.
But the problem is, in btrfs_join_transaction() call #1, we call
btrfs_record_root_in_trans() before initializing current::journal_info.
And in btrfs_join_transaction() call #2, we're relying on
current::journal_info to avoid such deadlock.
[FIX]
Call btrfs_record_root_in_trans() after we have initialized
current::journal_info.
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
At the point we add a root to the dead roots list we have no open inodes
for that root, so we need to hold a ref on that root to keep it from
disappearing.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We are going to make root life be controlled soley by refcounting, and
inodes will be one of the things that hold a ref on the root. This
means we need to handle dropping the ino_cache_inode outside of the root
freeing logic, so move it into btrfs_drop_and_free_fs_root() so it is
cleaned up properly on unmount.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It's no longer used following 30d40577e3 ("btrfs: reloc: Also queue
orphan reloc tree for cleanup to avoid BUG_ON()"), so just remove it.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently the non-prefixed version is a simple wrapper used to hide
the 4th argument of the prefixed version. This doesn't bring much value
in practice and only makes the code harder to follow by adding another
level of indirection. Rectify this by removing the __ prefix and
have only one public function to release bytes from a block reservation.
No semantic changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are new types and helpers that are supposed to be used in new code.
As a preparation to get rid of legacy types and API functions do
the conversion here.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The tree_log_mutex and reloc_mutex locks are properly nested so we can
simplify error handling and add labels for them. This reduces line count
of the function.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit flips the switch to start tracking/processing pinned extents
on a per-transaction basis. It mostly replaces all references from
btrfs_fs_info::(pinned_extents|freed_extents[]) to
btrfs_transaction::pinned_extents.
Two notable modifications that warrant explicit mention are changing
clean_pinned_extents to get a reference to the previously running
transaction. The other one is removal of call to
btrfs_destroy_pinned_extent since transactions are going to be cleaned
in btrfs_cleanup_one_transaction.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The status of aborted transaction can change between calls and it needs
to be accessed by READ_ONCE. Add a helper that also wraps the unlikely
hint.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that all callers of btrfs_get_fs_root are subsequently calling
btrfs_grab_fs_root and handling dropping the ref when they are done
appropriately, go ahead and push btrfs_grab_fs_root up into
btrfs_get_fs_root.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We create the snapshot and then use it for a bunch of things, we need to
hold a ref on it while we're messing with it.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All this does is call btrfs_get_fs_root() with check_ref == true. Just
use btrfs_get_fs_root() so we don't have a bunch of different helpers
that do the same thing.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We clean up the delayed references when we abort a transaction but we
leave the pending qgroup extent records behind, leaking memory.
This patch destroys the extent records when we destroy the delayed refs
and makes sure ensure they're gone before releasing the transaction.
Fixes: 3368d001ba ("btrfs: qgroup: Record possible quota-related extent for qgroup.")
CC: stable@vger.kernel.org # 4.4+
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
[ Rebased to latest upstream, remove to_qgroup() helper, use
rbtree_postorder_for_each_entry_safe() wrapper ]
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If we abort a transaction we have the following sequence
if (!trans->dirty && list_empty(&trans->new_bgs))
return;
WRITE_ONCE(trans->transaction->aborted, err);
The idea being if we didn't modify anything with our trans handle then
we don't really need to abort the whole transaction, maybe the other
trans handles are fine and we can carry on.
However in the case of create_snapshot we add a pending_snapshot object
to our transaction and then commit the transaction. We don't actually
modify anything. sync() behaves the same way, attach to an existing
transaction and commit it. This means that if we have an IO error in
the right places we could abort the committing transaction with our
trans->dirty being not set and thus not set transaction->aborted.
This is a problem because in the create_snapshot() case we depend on
pending->error being set to something, or btrfs_commit_transaction
returning an error.
If we are not the trans handle that gets to commit the transaction, and
we're waiting on the commit to happen we get our return value from
cur_trans->aborted. If this was not set to anything because sync() hit
an error in the transaction commit before it could modify anything then
cur_trans->aborted would be 0. Thus we'd return 0 from
btrfs_commit_transaction() in create_snapshot.
This is a problem because we then try to do things with
pending_snapshot->snap, which will be NULL because we didn't create the
snapshot, and then we'll get a NULL pointer dereference like the
following
"BUG: kernel NULL pointer dereference, address: 00000000000001f0"
RIP: 0010:btrfs_orphan_cleanup+0x2d/0x330
Call Trace:
? btrfs_mksubvol.isra.31+0x3f2/0x510
btrfs_mksubvol.isra.31+0x4bc/0x510
? __sb_start_write+0xfa/0x200
? mnt_want_write_file+0x24/0x50
btrfs_ioctl_snap_create_transid+0x16c/0x1a0
btrfs_ioctl_snap_create_v2+0x11e/0x1a0
btrfs_ioctl+0x1534/0x2c10
? free_debug_processing+0x262/0x2a3
do_vfs_ioctl+0xa6/0x6b0
? do_sys_open+0x188/0x220
? syscall_trace_enter+0x1f8/0x330
ksys_ioctl+0x60/0x90
__x64_sys_ioctl+0x16/0x20
do_syscall_64+0x4a/0x1b0
In order to fix this we need to make sure anybody who calls
commit_transaction has trans->dirty set so that they properly set the
trans->transaction->aborted value properly so any waiters know bad
things happened.
This was found while I was running generic/475 with my modified
fsstress, it reproduced within a few runs. I ran with this patch all
night and didn't see the problem again.
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If we fsync on a subvolume and create a log root for that volume, and
then later delete that subvolume we'll never clean up its log root. Fix
this by making switch_commit_roots free the log for any dropped roots we
encounter. The extra churn is because we need a btrfs_trans_handle, not
the btrfs_transaction.
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The type name is misleading, a single entry is named 'cache' while this
normally means a collection of objects. Rename that everywhere. Also the
identifier was quite long, making function prototypes harder to format.
Suggested-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function is used only during the final phase of freespace cache
writeout. This is necessary since using the plain btrfs_join_transaction
api is deadlock prone. The deadlock looks like:
T1:
btrfs_commit_transaction
commit_cowonly_roots
btrfs_write_dirty_block_groups
btrfs_wait_cache_io
__btrfs_wait_cache_io
btrfs_wait_ordered_range <-- Triggers ordered IO for freespace
inode and blocks transaction commit
until freespace cache writeout
T2: <-- after T1 has triggered the writeout
finish_ordered_fn
btrfs_finish_ordered_io
btrfs_join_transaction <--- this would block waiting for current
transaction to commit, but since trans
commit is waiting for this writeout to
finish
The special purpose functions prevents it by simply skipping the "wait
for writeout" since it's guaranteed the transaction won't proceed until
we are done.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The state was introduced in commit 4a9d8bdee3 ("Btrfs: make the state
of the transaction more readable"), then in commit 302167c50b
("btrfs: don't end the transaction for delayed refs in throttle") the
state is completely removed.
So we can just clean up the state since it's only compared but never
set.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add an overview of the basic btrfs transaction transitions, including
the following states:
- No transaction states
- Transaction N [[TRANS_STATE_RUNNING]]
- Transaction N [[TRANS_STATE_COMMIT_START]]
- Transaction N [[TRANS_STATE_COMMIT_DOING]]
- Transaction N [[TRANS_STATE_UNBLOCKED]]
- Transaction N [[TRANS_STATE_COMPLETED]]
For each state, the comment will include:
- Basic explaination about current state
- How to go next stage
- What will happen if we call various start_transaction() functions
- Relationship to transaction N+1
This doesn't provide tech details, but serves as a cheat sheet for
reader to get into the code a little easier.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It's not used ouside of transaction.c
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
A recent patch to btrfs showed that there was at least 1 case where a
nested transaction was committed. Nested transaction in this case means
a code which has a transaction handle calls some function which in turn
obtains a copy of the same transaction handle. In such cases the correct
thing to do is for the lower callee to call btrfs_end_transaction which
contains appropriate checks so as to not commit the transaction which
will result in stale trans handler for the caller.
To catch such cases add an assert in btrfs_commit_transaction ensuring
btrfs_trans_handle::use_count is always 1.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The file ctree.h serves as a header for everything and has become quite
bloated. Split some helpers that are generic and create a new file that
should be the catch-all for code that's not btrfs-specific.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_calc_trunc_metadata_size differs from trans_metadata_size in that
it doesn't take into account any splitting at the levels, because
truncate will never split nodes. However truncate _and_ changing will
never split nodes, so rename btrfs_calc_trunc_metadata_size to
btrfs_calc_metadata_size. Also btrfs_calc_trans_metadata_size is purely
for inserting items, so rename this to btrfs_calc_insert_metadata_size.
Making these clearer will help when I start using them differently in
upcoming patches.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This is prep work for moving all of the block group cache code into its
own file.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ minor comment updates ]
Signed-off-by: David Sterba <dsterba@suse.com>
The fiemap handler locks a file range that can have unflushed delalloc,
and after locking the range, it tries to attach to a running transaction.
If the running transaction started its commit, that is, it is in state
TRANS_STATE_COMMIT_START, and either the filesystem was mounted with the
flushoncommit option or the transaction is creating a snapshot for the
subvolume that contains the file that fiemap is operating on, we end up
deadlocking. This happens because fiemap is blocked on the transaction,
waiting for it to complete, and the transaction is waiting for the flushed
dealloc to complete, which requires locking the file range that the fiemap
task already locked. The following stack traces serve as an example of
when this deadlock happens:
(...)
[404571.515510] Workqueue: btrfs-endio-write btrfs_endio_write_helper [btrfs]
[404571.515956] Call Trace:
[404571.516360] ? __schedule+0x3ae/0x7b0
[404571.516730] schedule+0x3a/0xb0
[404571.517104] lock_extent_bits+0x1ec/0x2a0 [btrfs]
[404571.517465] ? remove_wait_queue+0x60/0x60
[404571.517832] btrfs_finish_ordered_io+0x292/0x800 [btrfs]
[404571.518202] normal_work_helper+0xea/0x530 [btrfs]
[404571.518566] process_one_work+0x21e/0x5c0
[404571.518990] worker_thread+0x4f/0x3b0
[404571.519413] ? process_one_work+0x5c0/0x5c0
[404571.519829] kthread+0x103/0x140
[404571.520191] ? kthread_create_worker_on_cpu+0x70/0x70
[404571.520565] ret_from_fork+0x3a/0x50
[404571.520915] kworker/u8:6 D 0 31651 2 0x80004000
[404571.521290] Workqueue: btrfs-flush_delalloc btrfs_flush_delalloc_helper [btrfs]
(...)
[404571.537000] fsstress D 0 13117 13115 0x00004000
[404571.537263] Call Trace:
[404571.537524] ? __schedule+0x3ae/0x7b0
[404571.537788] schedule+0x3a/0xb0
[404571.538066] wait_current_trans+0xc8/0x100 [btrfs]
[404571.538349] ? remove_wait_queue+0x60/0x60
[404571.538680] start_transaction+0x33c/0x500 [btrfs]
[404571.539076] btrfs_check_shared+0xa3/0x1f0 [btrfs]
[404571.539513] ? extent_fiemap+0x2ce/0x650 [btrfs]
[404571.539866] extent_fiemap+0x2ce/0x650 [btrfs]
[404571.540170] do_vfs_ioctl+0x526/0x6f0
[404571.540436] ksys_ioctl+0x70/0x80
[404571.540734] __x64_sys_ioctl+0x16/0x20
[404571.540997] do_syscall_64+0x60/0x1d0
[404571.541279] entry_SYSCALL_64_after_hwframe+0x49/0xbe
(...)
[404571.543729] btrfs D 0 14210 14208 0x00004000
[404571.544023] Call Trace:
[404571.544275] ? __schedule+0x3ae/0x7b0
[404571.544526] ? wait_for_completion+0x112/0x1a0
[404571.544795] schedule+0x3a/0xb0
[404571.545064] schedule_timeout+0x1ff/0x390
[404571.545351] ? lock_acquire+0xa6/0x190
[404571.545638] ? wait_for_completion+0x49/0x1a0
[404571.545890] ? wait_for_completion+0x112/0x1a0
[404571.546228] wait_for_completion+0x131/0x1a0
[404571.546503] ? wake_up_q+0x70/0x70
[404571.546775] btrfs_wait_ordered_extents+0x27c/0x400 [btrfs]
[404571.547159] btrfs_commit_transaction+0x3b0/0xae0 [btrfs]
[404571.547449] ? btrfs_mksubvol+0x4a4/0x640 [btrfs]
[404571.547703] ? remove_wait_queue+0x60/0x60
[404571.547969] btrfs_mksubvol+0x605/0x640 [btrfs]
[404571.548226] ? __sb_start_write+0xd4/0x1c0
[404571.548512] ? mnt_want_write_file+0x24/0x50
[404571.548789] btrfs_ioctl_snap_create_transid+0x169/0x1a0 [btrfs]
[404571.549048] btrfs_ioctl_snap_create_v2+0x11d/0x170 [btrfs]
[404571.549307] btrfs_ioctl+0x133f/0x3150 [btrfs]
[404571.549549] ? mem_cgroup_charge_statistics+0x4c/0xd0
[404571.549792] ? mem_cgroup_commit_charge+0x84/0x4b0
[404571.550064] ? __handle_mm_fault+0xe3e/0x11f0
[404571.550306] ? do_raw_spin_unlock+0x49/0xc0
[404571.550608] ? _raw_spin_unlock+0x24/0x30
[404571.550976] ? __handle_mm_fault+0xedf/0x11f0
[404571.551319] ? do_vfs_ioctl+0xa2/0x6f0
[404571.551659] ? btrfs_ioctl_get_supported_features+0x30/0x30 [btrfs]
[404571.552087] do_vfs_ioctl+0xa2/0x6f0
[404571.552355] ksys_ioctl+0x70/0x80
[404571.552621] __x64_sys_ioctl+0x16/0x20
[404571.552864] do_syscall_64+0x60/0x1d0
[404571.553104] entry_SYSCALL_64_after_hwframe+0x49/0xbe
(...)
If we were joining the transaction instead of attaching to it, we would
not risk a deadlock because a join only blocks if the transaction is in a
state greater then or equals to TRANS_STATE_COMMIT_DOING, and the delalloc
flush performed by a transaction is done before it reaches that state,
when it is in the state TRANS_STATE_COMMIT_START. However a transaction
join is intended for use cases where we do modify the filesystem, and
fiemap only needs to peek at delayed references from the current
transaction in order to determine if extents are shared, and, besides
that, when there is no current transaction or when it blocks to wait for
a current committing transaction to complete, it creates a new transaction
without reserving any space. Such unnecessary transactions, besides doing
unnecessary IO, can cause transaction aborts (-ENOSPC) and unnecessary
rotation of the precious backup roots.
So fix this by adding a new transaction join variant, named join_nostart,
which behaves like the regular join, but it does not create a transaction
when none currently exists or after waiting for a committing transaction
to complete.
Fixes: 03628cdbc6 ("Btrfs: do not start a transaction during fiemap")
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When one transaction is finishing its commit, it is possible for another
transaction to start and enter its initial commit phase as well. If the
first ends up getting aborted, we have a small time window where the second
transaction commit does not notice that the previous transaction aborted
and ends up committing, writing a superblock that points to btrees that
reference extent buffers (nodes and leafs) that were not persisted to disk.
The consequence is that after mounting the filesystem again, we will be
unable to load some btree nodes/leafs, either because the content on disk
is either garbage (or just zeroes) or corresponds to the old content of a
previouly COWed or deleted node/leaf, resulting in the well known error
messages "parent transid verify failed on ...".
The following sequence diagram illustrates how this can happen.
CPU 1 CPU 2
<at transaction N>
btrfs_commit_transaction()
(...)
--> sets transaction state to
TRANS_STATE_UNBLOCKED
--> sets fs_info->running_transaction
to NULL
(...)
btrfs_start_transaction()
start_transaction()
wait_current_trans()
--> returns immediately
because
fs_info->running_transaction
is NULL
join_transaction()
--> creates transaction N + 1
--> sets
fs_info->running_transaction
to transaction N + 1
--> adds transaction N + 1 to
the fs_info->trans_list list
--> returns transaction handle
pointing to the new
transaction N + 1
(...)
btrfs_sync_file()
btrfs_start_transaction()
--> returns handle to
transaction N + 1
(...)
btrfs_write_and_wait_transaction()
--> writeback of some extent
buffer fails, returns an
error
btrfs_handle_fs_error()
--> sets BTRFS_FS_STATE_ERROR in
fs_info->fs_state
--> jumps to label "scrub_continue"
cleanup_transaction()
btrfs_abort_transaction(N)
--> sets BTRFS_FS_STATE_TRANS_ABORTED
flag in fs_info->fs_state
--> sets aborted field in the
transaction and transaction
handle structures, for
transaction N only
--> removes transaction from the
list fs_info->trans_list
btrfs_commit_transaction(N + 1)
--> transaction N + 1 was not
aborted, so it proceeds
(...)
--> sets the transaction's state
to TRANS_STATE_COMMIT_START
--> does not find the previous
transaction (N) in the
fs_info->trans_list, so it
doesn't know that transaction
was aborted, and the commit
of transaction N + 1 proceeds
(...)
--> sets transaction N + 1 state
to TRANS_STATE_UNBLOCKED
btrfs_write_and_wait_transaction()
--> succeeds writing all extent
buffers created in the
transaction N + 1
write_all_supers()
--> succeeds
--> we now have a superblock on
disk that points to trees
that refer to at least one
extent buffer that was
never persisted
So fix this by updating the transaction commit path to check if the flag
BTRFS_FS_STATE_TRANS_ABORTED is set on fs_info->fs_state if after setting
the transaction to the TRANS_STATE_COMMIT_START we do not find any previous
transaction in the fs_info->trans_list. If the flag is set, just fail the
transaction commit with -EROFS, as we do in other places. The exact error
code for the previous transaction abort was already logged and reported.
Fixes: 49b25e0540 ("btrfs: enhance transaction abort infrastructure")
CC: stable@vger.kernel.org # 4.4+
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Move this into transaction.c with the rest of the transaction related
code.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The member num_dirty_bgs of struct btrfs_transaction is not used anymore,
it is set and incremented but nothing reads its value anymore. Its last
read use was removed by commit 64403612b7 ("btrfs: rework
btrfs_check_space_for_delayed_refs"). So just remove that member.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The pending chunks list contains chunks that are allocated in the
current transaction but haven't been created yet. The pinned chunks
list contains chunks that are being released in the current transaction.
Both describe chunks that are not reflected on disk as in use but are
unavailable just the same.
The pending chunks list is anchored by the transaction handle, which
means that we need to hold a reference to a transaction when working
with the list.
The way we use them is by iterating over both lists to perform
comparisons on the stripes they describe for each device. This is
backwards and requires that we keep a transaction handle open while
we're trimming.
This patchset adds an extent_io_tree to btrfs_device that maintains
the allocation state of the device. Extents are set dirty when
chunks are first allocated -- when the extent maps are added to the
mapping tree. They're cleared when last removed -- when the extent
maps are removed from the mapping tree. This matches the lifespan
of the pending and pinned chunks list and allows us to do trims
on unallocated space safely without pinning the transaction for what
may be a lengthy operation. We can also use this io tree to mark
which chunks have already been trimmed so we don't repeat the operation.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function is going to be used to clear out the device extent
allocation information. Give it a more generic name and export it. This
is in preparation to replacing the pending/pinned chunk lists with an
extent tree. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We currently overload the pending_chunks list to handle updating
btrfs_device->commit_bytes used. We don't actually care about the
extent mapping or even the device mapping for the chunk - we just need
the device, and we can end up processing it multiple times. The
fs_devices->resized_list does more or less the same thing, but with the
disk size. They are called consecutively during commit and have more or
less the same purpose.
We can combine the two lists into a single list that attaches to the
transaction and contains a list of devices that need updating. Since we
always add the device to a list when we change bytes_used or
disk_total_size, there's no harm in copying both values at once.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Btrfs has the following different extent_io_trees used:
- fs_info::free_extents[2]
- btrfs_inode::io_tree - for both normal inodes and the btree inode
- btrfs_inode::io_failure_tree
- btrfs_transaction::dirty_pages
- btrfs_root::dirty_log_pages
If we want to trace changes in those trees, it will be pretty hard to
distinguish them.
Instead of using hard-to-read pointer address, this patch will introduce
a new member extent_io_tree::owner to track the owner.
This modification needs all the callers of extent_io_tree_init() to
accept a new parameter @owner.
This patch provides the basis for later trace events.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This patch will add a new member fs_info to extent_io_tree.
This provides the basis for later trace events to distinguish the output
between different btrfs filesystems. While this increases the size of
the structure, we want to know the source of the trace events and
passing the fs_info as an argument to all contexts is not possible.
The selftests are now allowed to set it to NULL as they don't use the
tracepoints.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When we are mixing buffered writes with direct IO writes against the same
file and snapshotting is happening concurrently, we can end up with a
corrupt file content in the snapshot. Example:
1) Inode/file is empty.
2) Snapshotting starts.
2) Buffered write at offset 0 length 256Kb. This updates the i_size of the
inode to 256Kb, disk_i_size remains zero. This happens after the task
doing the snapshot flushes all existing delalloc.
3) DIO write at offset 256Kb length 768Kb. Once the ordered extent
completes it sets the inode's disk_i_size to 1Mb (256Kb + 768Kb) and
updates the inode item in the fs tree with a size of 1Mb (which is
the value of disk_i_size).
4) The dealloc for the range [0, 256Kb[ did not start yet.
5) The transaction used in the DIO ordered extent completion, which updated
the inode item, is committed by the snapshotting task.
6) Snapshot creation completes.
7) Dealloc for the range [0, 256Kb[ is flushed.
After that when reading the file from the snapshot we always get zeroes for
the range [0, 256Kb[, the file has a size of 1Mb and the data written by
the direct IO write is found. From an application's point of view this is
a corruption, since in the source subvolume it could never read a version
of the file that included the data from the direct IO write without the
data from the buffered write included as well. In the snapshot's tree,
file extent items are missing for the range [0, 256Kb[.
The issue, obviously, does not happen when using the -o flushoncommit
mount option.
Fix this by flushing delalloc for all the roots that are about to be
snapshotted when committing a transaction. This guarantees total ordering
when updating the disk_i_size of an inode since the flush for dealloc is
done when a transaction is in the TRANS_STATE_COMMIT_START state and wait
is done once no more external writers exist. This is similar to what we
do when using the flushoncommit mount option, but we do it only if the
transaction has snapshots to create and only for the roots of the
subvolumes to be snapshotted. The bulk of the dealloc is flushed in the
snapshot creation ioctl, so the flush work we do inside the transaction
is minimized.
This issue, involving buffered and direct IO writes with snapshotting, is
often triggered by fstest btrfs/078, and got reported by fsck when not
using the NO_HOLES features, for example:
$ cat results/btrfs/078.full
(...)
_check_btrfs_filesystem: filesystem on /dev/sdc is inconsistent
*** fsck.btrfs output ***
[1/7] checking root items
[2/7] checking extents
[3/7] checking free space cache
[4/7] checking fs roots
root 258 inode 264 errors 100, file extent discount
Found file extent holes:
start: 524288, len: 65536
ERROR: errors found in fs roots
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_set_lock_blocking is now only a simple wrapper around
btrfs_set_lock_blocking_write. The name does not bring any semantic
value that could not be inferred from the new function so there's no
point keeping it.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: David Sterba <dsterba@suse.com>
To allow delayed subtree swap rescan, btrfs needs to record per-root
information about which tree blocks get swapped. This patch introduces
the required infrastructure.
The designed workflow will be:
1) Record the subtree root block that gets swapped.
During subtree swap:
O = Old tree blocks
N = New tree blocks
reloc tree subvolume tree X
Root Root
/ \ / \
NA OB OA OB
/ | | \ / | | \
NC ND OE OF OC OD OE OF
In this case, NA and OA are going to be swapped, record (NA, OA) into
subvolume tree X.
2) After subtree swap.
reloc tree subvolume tree X
Root Root
/ \ / \
OA OB NA OB
/ | | \ / | | \
OC OD OE OF NC ND OE OF
3a) COW happens for OB
If we are going to COW tree block OB, we check OB's bytenr against
tree X's swapped_blocks structure.
If it doesn't fit any, nothing will happen.
3b) COW happens for NA
Check NA's bytenr against tree X's swapped_blocks, and get a hit.
Then we do subtree scan on both subtrees OA and NA.
Resulting 6 tree blocks to be scanned (OA, OC, OD, NA, NC, ND).
Then no matter what we do to subvolume tree X, qgroup numbers will
still be correct.
Then NA's record gets removed from X's swapped_blocks.
4) Transaction commit
Any record in X's swapped_blocks gets removed, since there is no
modification to swapped subtrees, no need to trigger heavy qgroup
subtree rescan for them.
This will introduce 128 bytes overhead for each btrfs_root even qgroup
is not enabled. This is to reduce memory allocations and potential
failures.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The first thing we do is loop through the list, this
if (!list_empty())
btrfs_create_pending_block_groups();
thing is just wasted space.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Previously callers to btrfs_end_transaction_throttle() would commit the
transaction if there wasn't enough delayed refs space. This happens in
relocation, and if the fs is relatively empty we'll run out of delayed
refs space basically immediately, so we'll just be stuck in this loop of
committing the transaction over and over again.
This code existed because we didn't have a good feedback mechanism for
running delayed refs, but with the delayed refs rsv we do now. Delete
this throttling code and let the btrfs_start_transaction() in relocation
deal with putting pressure on the delayed refs infrastructure. With
this patch we no longer take 5 minutes to balance a metadata only fs.
Qu has submitted a fstest to catch slow balance or excessive transaction
commits. Steps to reproduce:
* create subvolume
* create many (eg. 16000) inlined files, of size 2KiB
* iteratively snapshot and touch several files to trigger metadata
updates
* start balance -m
Reported-by: Qu Wenruo <wqu@suse.com>
Fixes: 64403612b7 ("btrfs: rework btrfs_check_space_for_delayed_refs")
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[ add tags and steps to reproduce ]
Signed-off-by: David Sterba <dsterba@suse.com>
The typos accumulate over time so once in a while time they get fixed in
a large patch.
Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Over the years we have built up a lot of infrastructure to keep delayed
refs in check, mostly by running them at btrfs_end_transaction() time.
We have a lot of different maths we do to figure out how much, if we
should do it inline or async, etc. This existed because we had no
feedback mechanism to force the flushing of delayed refs when they
became a problem. However with the enospc flushing infrastructure in
place for flushing delayed refs when they put too much pressure on the
enospc system we have this problem solved. Rip out all of this code as
it is no longer needed.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now with the delayed_refs_rsv we can now know exactly how much pending
delayed refs space we need. This means we can drastically simplify
btrfs_check_space_for_delayed_refs by simply checking how much space we
have reserved for the global rsv (which acts as a spill over buffer) and
the delayed refs rsv. If our total size is beyond that amount then we
know it's time to commit the transaction and stop any more delayed refs
from being generated.
With the introduction of dealyed_refs_rsv infrastructure, namely
btrfs_update_delayed_refs_rsv we now know exactly how much pending
delayed refs space is required.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Traditionally we've had voodoo in btrfs to account for the space that
delayed refs may take up by having a global_block_rsv. This works most
of the time, except when it doesn't. We've had issues reported and seen
in production where sometimes the global reserve is exhausted during
transaction commit before we can run all of our delayed refs, resulting
in an aborted transaction. Because of this voodoo we have equally
dubious flushing semantics around throttling delayed refs which we often
get wrong.
So instead give them their own block_rsv. This way we can always know
exactly how much outstanding space we need for delayed refs. This
allows us to make sure we are constantly filling that reservation up
with space, and allows us to put more precise pressure on the enospc
system. Instead of doing math to see if its a good time to throttle,
the normal enospc code will be invoked if we have a lot of delayed refs
pending, and they will be run via the normal flushing mechanism.
For now the delayed_refs_rsv will hold the reservations for the delayed
refs, the block group updates, and deleting csums. We could have a
separate rsv for the block group updates, but the csum deletion stuff is
still handled via the delayed_refs so that will stay there.
Historical background:
The global reserve has grown to cover everything we don't reserve space
explicitly for, and we've grown a lot of weird ad-hoc heuristics to know
if we're running short on space and when it's time to force a commit. A
failure rate of 20-40 file systems when we run hundreds of thousands of
them isn't super high, but cleaning up this code will make things less
ugly and more predictible.
Thus the delayed refs rsv. We always know how many delayed refs we have
outstanding, and although running them generates more we can use the
global reserve for that spill over, which fits better into it's desired
use than a full blown reservation. This first approach is to simply
take how many times we're reserving space for and multiply that by 2 in
order to save enough space for the delayed refs that could be generated.
This is a niave approach and will probably evolve, but for now it works.
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com> # high-level review
[ added background notes from the cover letter ]
Signed-off-by: David Sterba <dsterba@suse.com>
When it was introduced in commit f094ac32ab ("Btrfs: fix NULL pointer
after aborting a transaction"), it was not used.
Signed-off-by: Lu Fengqi <lufq.fnst@cn.fujitsu.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: David Sterba <dsterba@suse.com>
Tracking pending ordered extents per transaction was introduced in commit
50d9aa99bd ("Btrfs: make sure logged extents complete in the current
transaction V3") and later updated in commit 161c3549b4 ("Btrfs: change
how we wait for pending ordered extents").
However now that on fsync we always wait for ordered extents to complete
before logging, done in commit 5636cf7d6d ("btrfs: remove the logged
extents infrastructure"), we no longer need the stuff to track for pending
ordered extents, which was not completely removed in the mentioned commit.
So remove the remaining of the pending ordered extents infrastructure.
Reviewed-by: Liu Bo <bo.liu@linux.alibaba.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This could result in a really bad case where we do something like
evict
evict_refill_and_join
btrfs_commit_transaction
btrfs_run_delayed_iputs
evict
evict_refill_and_join
btrfs_commit_transaction
... forever
We have plenty of other places where we run delayed iputs that are much
safer, let those do the work.
CC: stable@vger.kernel.org # 4.4+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The avg_delayed_ref_runtime can be referenced from the transaction
handle.
Signed-off-by: Lu Fengqi <lufq.fnst@cn.fujitsu.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It can be referenced from the transaction handle.
Signed-off-by: Lu Fengqi <lufq.fnst@cn.fujitsu.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We want to release the unused reservation we have since it refills the
delayed refs reserve, which will make everything go smoother when
running the delayed refs if we're short on our reservation.
CC: stable@vger.kernel.org # 4.4+
Reviewed-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Liu Bo <bo.liu@linux.alibaba.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Too trivial, the purpose can be simply documented in a comment.
Reviewed-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
rb_first_cached() trades an extra pointer "leftmost" for doing the same
job as rb_first() but in O(1).
Functions manipulating href_root need to get the first entry, this
converts href_root to use rb_first_cached().
This patch is first in the sequenct of similar updates to other rbtrees
and this is analysis of the expected behaviour and improvements.
There's a common pattern:
while (node = rb_first) {
entry = rb_entry(node)
next = rb_next(node)
rb_erase(node)
cleanup(entry)
}
rb_first needs to traverse the tree up to logN depth, rb_erase can
completely reshuffle the tree. With the caching we'll skip the traversal
in rb_first. That's a cached memory access vs looped pointer
dereference trade-off that IMHO has a clear winner.
Measurements show there's not much difference in a sample tree with
10000 nodes: 4.5s / rb_first and 4.8s / rb_first_cached. Real effects of
caching and pointer chasing are unpredictable though.
Further optimzations can be done to avoid the expensive rb_erase step.
In some cases it's ok to process the nodes in any order, so the tree can
be traversed in post-order, not rebalancing the children nodes and just
calling free. Care must be taken regarding the next node.
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: Liu Bo <bo.liu@linux.alibaba.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ update changelog from mail discussions ]
Signed-off-by: David Sterba <dsterba@suse.com>
There are two members in struct btrfs_root which indicate root's
objectid: objectid and root_key.objectid.
They are both set to the same value in __setup_root():
static void __setup_root(struct btrfs_root *root,
struct btrfs_fs_info *fs_info,
u64 objectid)
{
...
root->objectid = objectid;
...
root->root_key.objectid = objecitd;
...
}
and not changed to other value after initialization.
grep in btrfs directory shows both are used in many places:
$ grep -rI "root->root_key.objectid" | wc -l
133
$ grep -rI "root->objectid" | wc -l
55
(4.17, inc. some noise)
It is confusing to have two similar variable names and it seems
that there is no rule about which should be used in a certain case.
Since ->root_key itself is needed for tree reloc tree, let's remove
'objecitd' member and unify code to use ->root_key.objectid in all places.
Signed-off-by: Misono Tomohiro <misono.tomohiro@jp.fujitsu.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All callers pass the root tree of dir, we can push that down to the
function itself.
Signed-off-by: Lu Fengqi <lufq.fnst@cn.fujitsu.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
IS_ERR(p) && PTR_ERR(p) == n is a weird way to spell p == ERR_PTR(n).
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Reviewed-by: David Sterba <dsterba@suse.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
[ update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
It can be referenced from the passed transaction handle.
Signed-off-by: Lu Fengqi <lufq.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The transaction times were changed to ktime_get_real_seconds to avoid
the y2038 overflow, but they still have a minor problem when they go
backwards or jump due to settimeofday() or leap seconds.
This changes the transaction handling to instead use ktime_get_seconds(),
which returns a CLOCK_MONOTONIC timestamp that has neither of those
problems.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The get_seconds() function is deprecated as it truncates the timestamp
to 32 bits. Change it to or ktime_get_real_seconds().
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
This is a late set of changes from Deepa Dinamani doing an automated
treewide conversion of the inode and iattr structures from 'timespec'
to 'timespec64', to push the conversion from the VFS layer into the
individual file systems.
There were no conflicts between this and the contents of linux-next
until just before the merge window, when we saw multiple problems:
- A minor conflict with my own y2038 fixes, which I could address
by adding another patch on top here.
- One semantic conflict with late changes to the NFS tree. I addressed
this by merging Deepa's original branch on top of the changes that
now got merged into mainline and making sure the merge commit includes
the necessary changes as produced by coccinelle.
- A trivial conflict against the removal of staging/lustre.
- Multiple conflicts against the VFS changes in the overlayfs tree.
These are still part of linux-next, but apparently this is no longer
intended for 4.18 [1], so I am ignoring that part.
As Deepa writes:
The series aims to switch vfs timestamps to use struct timespec64.
Currently vfs uses struct timespec, which is not y2038 safe.
The series involves the following:
1. Add vfs helper functions for supporting struct timepec64 timestamps.
2. Cast prints of vfs timestamps to avoid warnings after the switch.
3. Simplify code using vfs timestamps so that the actual
replacement becomes easy.
4. Convert vfs timestamps to use struct timespec64 using a script.
This is a flag day patch.
Next steps:
1. Convert APIs that can handle timespec64, instead of converting
timestamps at the boundaries.
2. Update internal data structures to avoid timestamp conversions.
Thomas Gleixner adds:
I think there is no point to drag that out for the next merge window.
The whole thing needs to be done in one go for the core changes which
means that you're going to play that catchup game forever. Let's get
over with it towards the end of the merge window.
[1] https://www.spinics.net/lists/linux-fsdevel/msg128294.html
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQIcBAABAgAGBQJbInZAAAoJEGCrR//JCVInReoQAIlVIIMt5ZX6wmaKbrjy9Itf
MfgbFihQ/djLnuSPVQ3nztcxF0d66BKHZ9puVjz6+mIHqfDvJTRwZs9nU+sOF/T1
g78fRkM1cxq6ZCkGYAbzyjyo5aC4PnSMP/NQLmwqvi0MXqqrbDoq5ZdP9DHJw39h
L9lD8FM/P7T29Fgp9tq/pT5l9X8VU8+s5KQG1uhB5hii4VL6pD6JyLElDita7rg+
Z7/V7jkxIGEUWF7vGaiR1QTFzEtpUA/exDf9cnsf51OGtK/LJfQ0oiZPPuq3oA/E
LSbt8YQQObc+dvfnGxwgxEg1k5WP5ekj/Wdibv/+rQKgGyLOTz6Q4xK6r8F2ahxs
nyZQBdXqHhJYyKr1H1reUH3mrSgQbE5U5R1i3My0xV2dSn+vtK5vgF21v2Ku3A1G
wJratdtF/kVBzSEQUhsYTw14Un+xhBLRWzcq0cELonqxaKvRQK9r92KHLIWNE7/v
c0TmhFbkZA+zR8HdsaL3iYf1+0W/eYy8PcvepyldKNeW2pVk3CyvdTfY2Z87G2XK
tIkK+BUWbG3drEGG3hxZ3757Ln3a9qWyC5ruD3mBVkuug/wekbI8PykYJS7Mx4s/
WNXl0dAL0Eeu1M8uEJejRAe1Q3eXoMWZbvCYZc+wAm92pATfHVcKwPOh8P7NHlfy
A3HkjIBrKW5AgQDxfgvm
=CZX2
-----END PGP SIGNATURE-----
Merge tag 'vfs-timespec64' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/playground
Pull inode timestamps conversion to timespec64 from Arnd Bergmann:
"This is a late set of changes from Deepa Dinamani doing an automated
treewide conversion of the inode and iattr structures from 'timespec'
to 'timespec64', to push the conversion from the VFS layer into the
individual file systems.
As Deepa writes:
'The series aims to switch vfs timestamps to use struct timespec64.
Currently vfs uses struct timespec, which is not y2038 safe.
The series involves the following:
1. Add vfs helper functions for supporting struct timepec64
timestamps.
2. Cast prints of vfs timestamps to avoid warnings after the switch.
3. Simplify code using vfs timestamps so that the actual replacement
becomes easy.
4. Convert vfs timestamps to use struct timespec64 using a script.
This is a flag day patch.
Next steps:
1. Convert APIs that can handle timespec64, instead of converting
timestamps at the boundaries.
2. Update internal data structures to avoid timestamp conversions'
Thomas Gleixner adds:
'I think there is no point to drag that out for the next merge
window. The whole thing needs to be done in one go for the core
changes which means that you're going to play that catchup game
forever. Let's get over with it towards the end of the merge window'"
* tag 'vfs-timespec64' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/playground:
pstore: Remove bogus format string definition
vfs: change inode times to use struct timespec64
pstore: Convert internal records to timespec64
udf: Simplify calls to udf_disk_stamp_to_time
fs: nfs: get rid of memcpys for inode times
ceph: make inode time prints to be long long
lustre: Use long long type to print inode time
fs: add timespec64_truncate()
This function always takes a transaction handle which contains a
reference to the fs_info. Use that and remove the extra argument.
Signed-off-by: Lu Fengqi <lufq.fnst@cn.fujitsu.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that we don't keep long-standing reservations for orphan items,
root->orphan_block_rsv isn't used. We can git rid of it, along with:
- root->orphan_lock, which was used to protect root->orphan_block_rsv
- root->orphan_inodes, which was used as a refcount for root->orphan_block_rsv
- BTRFS_INODE_ORPHAN_META_RESERVED, which was used to track reservations
in root->orphan_block_rsv
- btrfs_orphan_commit_root(), which was the last user of any of these
and does nothing else
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Use the wrappers and reduce the amount of low-level details about the
waitqueue management.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Unlike previous method that tries to commit transaction inside
qgroup_reserve(), this time we will try to commit transaction using
fs_info->transaction_kthread to avoid nested transaction and no need to
worry about locking context.
Since it's an asynchronous function call and we won't wait for
transaction commit, unlike previous method, we must call it before we
hit the qgroup limit.
So this patch will use the ratio and size of qgroup meta_pertrans
reservation as indicator to check if we should trigger a transaction
commit. (meta_prealloc won't be cleaned in transaction committ, it's
useless anyway)
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Remove GPL boilerplate text (long, short, one-line) and keep the rest,
ie. personal, company or original source copyright statements. Add the
SPDX header.
Signed-off-by: David Sterba <dsterba@suse.com>
Btrfs uses 2 different methods to reseve metadata qgroup space.
1) Reserve at btrfs_start_transaction() time
This is quite straightforward, caller will use the trans handler
allocated to modify b-trees.
In this case, reserved metadata should be kept until qgroup numbers
are updated.
2) Reserve by using block_rsv first, and later btrfs_join_transaction()
This is more complicated, caller will reserve space using block_rsv
first, and then later call btrfs_join_transaction() to get a trans
handle.
In this case, before we modify trees, the reserved space can be
modified on demand, and after btrfs_join_transaction(), such reserved
space should also be kept until qgroup numbers are updated.
Since these two types behave differently, split the original "META"
reservation type into 2 sub-types:
META_PERTRANS:
For above case 1)
META_PREALLOC:
For reservations that happened before btrfs_join_transaction() of
case 2)
NOTE: This patch will only convert existing qgroup meta reservation
callers according to its situation, not ensuring all callers are at
correct timing.
Such fix will be added in later patches.
Signed-off-by: Qu Wenruo <wqu@suse.com>
[ update comments ]
Signed-off-by: David Sterba <dsterba@suse.com>
Since userspace transaction have been removed we no longer have use
for this field so delete it.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that the userspace transaction ioctls have been removed,
TRANS_USERSPACE is no longer used hence we can remove it.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When multiple pending snapshots referring to the same source subvolume
are executed, enabled quota will cause root item corruption, where root
items are using old bytenr (no backref in extent tree).
This can be triggered by fstests btrfs/152.
The cause is when source subvolume is still dirty, extra commit
(simplied transaction commit) of qgroup_account_snapshot() can skip
dirty roots not recorded in current transaction, making root item of
source subvolume not updated.
Fix it by forcing recording source subvolume in current transaction
before qgroup sub-transaction commit.
Reported-by: Justin Maggard <jmaggard@netgear.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Even though btrfs_start_dirty_block_groups is fairly in the beginning of
btrfs_commit_transaction outside of the critical section defined by the
transaction states it can only be run by a single comitter. In other
words it defines its own critical section thanks to the
BTRFS_TRANS_DIRTY_BG run flag and ro_block_group_mutex. However, its
error handling is outside of this critical section which is a bit
counter-intuitive. So move the error handling righ after the function is
executed and let the sole runner of dirty block groups handle the return
value. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already pass btrfs_trans_handle which contains a reference to the
fs_info so use that. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already pass the btrfs_transaction which references fs_info so no
need to pass the later as an argument. Also use the opportunity to
shorten transaction->trans. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already pass the trans handle which has a reference to fs_info to
create_pending_snapshot so we can refer to it directly. Doing this
obviates the need to pass the fs_info to create_pending_snapshots as
well. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already have the fs_info from the passed transaction so use it
directly. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The only thing the passed root is used for is:
1. get a reference to the fs_info and to
2. call trace_btrfs_transaction_commit.
We can achieve 1) by simply referring to the fs_info from passed trans
object. As far as 2) is concerned cleanup_transaction is called from
only one place and the 'root' argument passed is the one from the trans
handle. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already pass a transaction handle which refrences the fs_info so
we can grab it from there. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already pass the transaction handle which has a reference to the
fs_info. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already pass the transaction which has a reference to the fs_info,
so use that. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It can be referenced from the passed transaction so no point in passing
it as a function argument. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It can be referenced from the passed transaciton so no point in
passing it as function argument. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All current callers of this function just get a reference to the
trans->fs_info member and pass it as the second argument. Collapse this
into the function itself. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_write_and_wait_transaction is essentially a wrapper of
btrfs_write_and_wait_marked_extents with the addition of calling
clear_btree_io_tree. Having the code split doesn't really bring any
benefit. Open code the later into the former and add proper
documentation header.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ reformat comment ]
Signed-off-by: David Sterba <dsterba@suse.com>
This function is only ever used in __btrfs_end_transaction and
btrfs_commit_transaction so there is no need to export it via header.
Let's move it closer to where it's used, make it static and remove it
from the header. No functional changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Commit 0e8c36a9fd ("Btrfs: fix lots of orphan inodes when the space
is not enough") changed the way transaction reservation is made in
btrfs_evict_node and as a result this function became unused. This has
been the status quo for 5 years in which time no one noticed, so I'd
say it's safe to assume it's unlikely it will ever be used again.
Historical note: there were more attempts to remove the function, the
reasoning was missing and only based on some static analysis tool
reports. Other reason for rejection was that there seemed to be
connection to BTRFS_RESERVE_FLUSH_LIMIT and that would need to be
removeed to. This was not correct so removing the function is all we can
do.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
[ add the note ]
Signed-off-by: David Sterba <dsterba@suse.com>
This reverts commit 3c181c12c4.
The offending patch was merged in 4.16-rc4 and was promptly applied to
stable kernels 4.14.25 and 4.15.8.
The patch causes a corruption in several superblock items on big-endian
machines because of messed up endianity conversions. The damage is
manually repairable. A filesystem cannot be mounted again after it has
been unmounted once.
We do a full revert and not a fixup so stable can pick that patch ASAP.
Fixes: 3c181c12c4 ("btrfs: use proper endianness accessors for super_copy")
Link: https://lkml.kernel.org/r/1521139304@msgid.manchmal.in-ulm.de
CC: stable@vger.kernel.org # 4.14+
Reported-by: Christoph Biedl <linux-kernel.bfrz@manchmal.in-ulm.de>
Signed-off-by: David Sterba <dsterba@suse.com>
The fs_info::super_copy is a byte copy of the on-disk structure and all
members must use the accessor macros/functions to obtain the right
value. This was missing in update_super_roots and in sysfs readers.
Moving between opposite endianness hosts will report bogus numbers in
sysfs, and mount may fail as the root will not be restored correctly. If
the filesystem is always used on a same endian host, this will not be a
problem.
Fix this by using the btrfs_set_super...() functions to set
fs_info::super_copy values, and for the sysfs, use the cached
fs_info::nodesize/sectorsize values.
CC: stable@vger.kernel.org
Fixes: df93589a17 ("btrfs: export more from FS_INFO to sysfs")
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
The use_count is a reference counter, we can use the refcount_t type,
though we don't use the atomicity. This is not a performance critical
code and we could catch the underflows. The type is changed from long,
but the number of references will fit an int.
Signed-off-by: David Sterba <dsterba@suse.com>
All callers use GFP_NOFS, we don't have to pass it as an argument. The
built-in tests pass GFP_KERNEL, but they run only at module load time
and NOFS works there as well.
Signed-off-by: David Sterba <dsterba@suse.com>
We're holding the sb_start_intwrite lock at this point, and doing async
filemap_flush of the inodes will result in a deadlock if we freeze the
fs during this operation. This is because we could do a
btrfs_join_transaction() in the thread we are waiting on which would
block at sb_start_intwrite, and thus deadlock. Using
writeback_inodes_sb() side steps the problem by not introducing all of
these extra locking dependencies.
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We were only doing btrfs_check_space_for_delayed_refs() if the metadata
space was full, ie we couldn't allocate chunks. This assumes we'll be
able to allocate chunks during transaction commit, but since nothing
does a LIMIT flush during the transaction commit this won't actually
happen unless we happen to run shy of actual space. We already take
into account a full fs in btrfs_check_space_for_delayed_refs() so just
kill this extra check to make sure we're ending the transaction when we
need to.
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have started plug in btrfs_write_and_wait_marked_extents() but the
generated IOs actually go to device's schedule IO list where the work
is doing in another task, thus the started plug doesn't make any
sense.
And since we wait for IOs immediately after writing meta blocks, it's
the same case as writing log tree, doing sync submit can merge more
IOs.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Dave Jones hit a WARN_ON(nr < 0) in btrfs_wait_ordered_roots() with
v4.12-rc6. This was because commit 70e7af244 made it possible for
calc_reclaim_items_nr() to return a negative number. It's not really a
bug in that commit, it just didn't go far enough down the stack to find
all the possible 64->32 bit overflows.
This switches calc_reclaim_items_nr() to return a u64 and changes everyone
that uses the results of that math to u64 as well.
Reported-by: Dave Jones <davej@codemonkey.org.uk>
Fixes: 70e7af2 ("Btrfs: fix delalloc accounting leak caused by u32 overflow")
Signed-off-by: Chris Mason <clm@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Quite a lot of qgroup corruption happens due to wrong time of calling
btrfs_qgroup_prepare_account_extents().
Since the safest time is to call it just before
btrfs_qgroup_account_extents(), there is no need to separate these 2
functions.
Merging them will make code cleaner and less bug prone.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
[ changelog and comment adjustments ]
Signed-off-by: David Sterba <dsterba@suse.com>
We can keep the state among the other fs_info flags, there's no reason
why fs_frozen would need to be separate.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Observing the number of slab objects of btrfs_transaction, there's just
one active on an almost quiescent filesystem, and the number of objects
goes to about ten when sync is in progress. Then the nubmer goes down to
1. This matches the expectations of the transaction lifetime.
For such use the separate slab cache is not justified, as we do not
reuse objects frequently. For the shortlived transaction, the generic
slab (size 512) should be ok. We can optimistically expect that the 512
slabs are not all used (fragmentation) and there are free slots to take
when we do the allocation, compared to potentially allocating a whole new
page for the separate slab.
We'll lose the stats about the object use, which could be added later if
we really need them.
Signed-off-by: David Sterba <dsterba@suse.com>
For extent_io tree's we have carried the address_mapping of the inode
around in the io tree in order to pull the inode back out for calling
into various tree ops hooks. This works fine when everything that has
an extent_io_tree has an inode. But we are going to remove the
btree_inode, so we need to change this. Instead just have a generic
void * for private data that we can initialize with, and have all the
tree ops use that instead. This had a lot of cascading changes but
should be relatively straightforward.
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Chandan Rajendra <chandan@linux.vnet.ibm.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ minor reordering of the callback prototypes ]
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
The easist way to reproduce the bug is:
------
# mkfs.btrfs -f $dev -n 16K
# mount $dev $mnt -o inode_cache
# btrfs quota enable $mnt
# btrfs quota rescan -w $mnt
# btrfs qgroup show $mnt
qgroupid rfer excl
-------- ---- ----
0/5 32.00KiB 32.00KiB
^^ Twice the correct value
------
And fstests/btrfs qgroup test group can easily detect them with
inode_cache mount option.
Although some of them are false alerts since old test cases are using
fixed golden output.
While new test cases will use "btrfs check" to detect qgroup mismatch.
[CAUSE]
Inode_cache mount option will make commit_fs_roots() to call
btrfs_save_ino_cache() to update fs/subvol trees, and generate new
delayed refs.
However we call btrfs_qgroup_prepare_account_extents() too early, before
commit_fs_roots().
This makes the "old_roots" for newly generated extents are always NULL.
For freeing extent case, this makes both new_roots and old_roots to be
empty, while correct old_roots should not be empty.
This causing qgroup numbers not decreased correctly.
[FIX]
Modify the timing of calling btrfs_qgroup_prepare_account_extents() to
just before btrfs_qgroup_account_extents(), and add needed delayed_refs
handler.
So qgroup can handle inode_map mount options correctly.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The members have been effectively unused since "Btrfs: rework qgroup
accounting" (fcebe4562d), there's no substitute for
assert_qgroups_uptodate so it's removed as well.
Reviewed-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
refcount_t type and corresponding API should be
used instead of atomic_t when the variable is used as
a reference counter. This allows to avoid accidental
refcounter overflows that might lead to use-after-free
situations.
Signed-off-by: Elena Reshetova <elena.reshetova@intel.com>
Signed-off-by: Hans Liljestrand <ishkamiel@gmail.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: David Windsor <dwindsor@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
write_all_supers and write_ctree_super are almost equal, the parameter
'trans' is unused so we can drop it and have just one helper.
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The quota status used to be tracked as a variable, so the mutex was
needed (until "Btrfs: add a flags field to btrfs_fs_info" afcdd129e0).
Since the status is a bit modified atomically and we don't hold the
mutex beyond the check, we can drop it.
Signed-off-by: David Sterba <dsterba@suse.com>
Once a qgroup limit is exceeded, it's impossible to restore normal
operation to the subvolume without modifying the limit or removing
the subvolume. This is a surprising situation for many users used
to the typical workflow with quotas on other file systems where it's
possible to remove files until the used space is back under the limit.
When we go to unlink a file and start the transaction, we'll hit
the qgroup limit while trying to reserve space for the items we'll
modify while removing the file. We discussed last month how best
to handle this situation and agreed that there is no perfect solution.
The best principle-of-least-surprise solution is to handle it similarly
to how we already handle ENOSPC when unlinking, which is to allow
the operation to succeed with the expectation that it will ultimately
release space under most circumstances.
This patch modifies the transaction start path to select whether to
honor the qgroups limits. btrfs_start_transaction_fallback_global_rsv
is the only caller that skips enforcement. The reservation and tracking
still happens normally -- it just skips the enforcement step.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Reviewed-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently btrfs_ino takes a struct inode and this causes a lot of
internal btrfs functions which consume this ino to take a VFS inode,
rather than btrfs' own struct btrfs_inode. In order to fix this "leak"
of VFS structs into the internals of btrfs first it's necessary to
eliminate all uses of struct inode for the purpose of inode. This patch
does that by using BTRFS_I to convert an inode to btrfs_inode. With
this problem eliminated subsequent patches will start eliminating the
passing of struct inode altogether, eventually resulting in a lot cleaner
code.
Signed-off-by: Nikolay Borisov <n.borisov.lkml@gmail.com>
[ fix btrfs_get_extent tracepoint prototype ]
Signed-off-by: David Sterba <dsterba@suse.com>
This replaces ACCESS_ONCE macro with the corresponding
READ|WRITE macros
Signed-off-by: Seraphime Kirkovski <kirkseraph@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now we only use the root parameter to print the root objectid in
a tracepoint. We can use the root parameter from the transaction
handle for that. It's also used to join the transaction with
async commits, so we remove the comment that it's just for checking.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_write_and_wait_marked_extents and btrfs_sync_log both call
btrfs_wait_marked_extents, which provides a core loop and then handles
errors differently based on whether it's it's a log root or not.
This means that btrfs_write_and_wait_marked_extents needs to take a root
because btrfs_wait_marked_extents requires one, even though it's only
used to determine whether the root is a log root. The log root code
won't ever call into the transaction commit code using a log root, so we
can factor out the core loop and provide the error handling appropriate
to each waiter in new routines. This allows us to eventually remove
the root argument from btrfs_commit_transaction, and as a result,
btrfs_end_transaction.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are loads of functions in btrfs that accept a root parameter
but only use it to obtain an fs_info pointer. Let's convert those to
just accept an fs_info pointer directly.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This results in btrfs_assert_delayed_root_empty and
btrfs_destroy_delayed_inode taking an fs_info instead of a root.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In routines where someptr->fs_info is referenced multiple times, we
introduce a convenience variable. This makes the code considerably
more readable.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We track the node sizes per-root, but they never vary from the values
in the superblock. This patch messes with the 80-column style a bit,
but subsequent patches to factor out root->fs_info into a convenience
variable fix it up again.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are many functions that are always called with the same root
argument. Rather than passing the same root every time, we can
pass an fs_info pointer instead and have the function get the root
pointer itself.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are 11 functions that accept a root parameter and immediately
overwrite it. We can pass those an fs_info pointer instead.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Pull btrfs updates from Chris Mason:
"This is a big variety of fixes and cleanups.
Liu Bo continues to fixup fuzzer related problems, and some of Josef's
cleanups are prep for his bigger extent buffer changes (slated for
v4.10)"
* 'for-linus-4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (39 commits)
Revert "btrfs: let btrfs_delete_unused_bgs() to clean relocated bgs"
Btrfs: remove unnecessary btrfs_mark_buffer_dirty in split_leaf
Btrfs: don't BUG() during drop snapshot
btrfs: fix btrfs_no_printk stub helper
Btrfs: memset to avoid stale content in btree leaf
btrfs: parent_start initialization cleanup
btrfs: Remove already completed TODO comment
btrfs: Do not reassign count in btrfs_run_delayed_refs
btrfs: fix a possible umount deadlock
Btrfs: fix memory leak in do_walk_down
btrfs: btrfs_debug should consume fs_info when DEBUG is not defined
btrfs: convert send's verbose_printk to btrfs_debug
btrfs: convert pr_* to btrfs_* where possible
btrfs: convert printk(KERN_* to use pr_* calls
btrfs: unsplit printed strings
btrfs: clean the old superblocks before freeing the device
Btrfs: kill BUG_ON in run_delayed_tree_ref
Btrfs: don't leak reloc root nodes on error
btrfs: squash lines for simple wrapper functions
Btrfs: improve check_node to avoid reading corrupted nodes
...
current_fs_time() uses struct super_block* as an argument.
As per Linus's suggestion, this is changed to take struct
inode* as a parameter instead. This is because the function
is primarily meant for vfs inode timestamps.
Also the function was renamed as per Arnd's suggestion.
Change all calls to current_fs_time() to use the new
current_time() function instead. current_fs_time() will be
deleted.
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
For many printks, we want to know which file system issued the message.
This patch converts most pr_* calls to use the btrfs_* versions instead.
In some cases, this means adding plumbing to allow call sites access to
an fs_info pointer.
fs/btrfs/check-integrity.c is left alone for another day.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This patch converts printk(KERN_* style messages to use the pr_* versions.
One side effect is that anything that was KERN_DEBUG is now automatically
a dynamic debug message.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
CodingStyle chapter 2:
"[...] never break user-visible strings such as printk messages,
because that breaks the ability to grep for them."
This patch unsplits user-visible strings.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since we could get errors from the concurrent aborted transaction,
the check of this BUG_ON in start_transaction is not true any more.
Say, while flushing free space cache inode's dirty pages,
btrfs_finish_ordered_io
-> btrfs_join_transaction_nolock
(the transaction has been aborted.)
-> BUG_ON(type == TRANS_JOIN_NOLOCK);
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have a lot of random ints in btrfs_fs_info that can be put into flags. This
is mostly equivalent with the exception of how we deal with quota going on or
off, now instead we set a flag when we are turning it on or off and deal with
that appropriately, rather than just having a pending state that the current
quota_enabled gets set to. Thanks,
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When running fstests generic/068, sometimes we got below deadlock:
xfs_io D ffff8800331dbb20 0 6697 6693 0x00000080
ffff8800331dbb20 ffff88007acfc140 ffff880034d895c0 ffff8800331dc000
ffff880032d243e8 fffffffeffffffff ffff880032d24400 0000000000000001
ffff8800331dbb38 ffffffff816a9045 ffff880034d895c0 ffff8800331dbba8
Call Trace:
[<ffffffff816a9045>] schedule+0x35/0x80
[<ffffffff816abab2>] rwsem_down_read_failed+0xf2/0x140
[<ffffffff8118f5e1>] ? __filemap_fdatawrite_range+0xd1/0x100
[<ffffffff8134f978>] call_rwsem_down_read_failed+0x18/0x30
[<ffffffffa06631fc>] ? btrfs_alloc_block_rsv+0x2c/0xb0 [btrfs]
[<ffffffff810d32b5>] percpu_down_read+0x35/0x50
[<ffffffff81217dfc>] __sb_start_write+0x2c/0x40
[<ffffffffa067f5d5>] start_transaction+0x2a5/0x4d0 [btrfs]
[<ffffffffa067f857>] btrfs_join_transaction+0x17/0x20 [btrfs]
[<ffffffffa068ba34>] btrfs_evict_inode+0x3c4/0x5d0 [btrfs]
[<ffffffff81230a1a>] evict+0xba/0x1a0
[<ffffffff812316b6>] iput+0x196/0x200
[<ffffffffa06851d0>] btrfs_run_delayed_iputs+0x70/0xc0 [btrfs]
[<ffffffffa067f1d8>] btrfs_commit_transaction+0x928/0xa80 [btrfs]
[<ffffffffa0646df0>] btrfs_freeze+0x30/0x40 [btrfs]
[<ffffffff81218040>] freeze_super+0xf0/0x190
[<ffffffff81229275>] do_vfs_ioctl+0x4a5/0x5c0
[<ffffffff81003176>] ? do_audit_syscall_entry+0x66/0x70
[<ffffffff810038cf>] ? syscall_trace_enter_phase1+0x11f/0x140
[<ffffffff81229409>] SyS_ioctl+0x79/0x90
[<ffffffff81003c12>] do_syscall_64+0x62/0x110
[<ffffffff816acbe1>] entry_SYSCALL64_slow_path+0x25/0x25
>From this warning, freeze_super() already holds SB_FREEZE_FS, but
btrfs_freeze() will call btrfs_commit_transaction() again, if
btrfs_commit_transaction() finds that it has delayed iputs to handle,
it'll start_transaction(), which will try to get SB_FREEZE_FS lock
again, then deadlock occurs.
The root cause is that in btrfs, sync_filesystem(sb) does not make
sure all metadata is updated. There still maybe some codes adding
delayed iputs, see below sample race window:
CPU1 | CPU2
|-> freeze_super() |
|-> sync_filesystem(sb); |
| |-> cleaner_kthread()
| | |-> btrfs_delete_unused_bgs()
| | |-> btrfs_remove_chunk()
| | |-> btrfs_remove_block_group()
| | |-> btrfs_add_delayed_iput()
| |
|-> sb->s_writers.frozen = SB_FREEZE_FS; |
|-> sb_wait_write(sb, SB_FREEZE_FS); |
| acquire SB_FREEZE_FS lock. |
| |
|-> btrfs_freeze() |
|-> btrfs_commit_transaction() |
|-> btrfs_run_delayed_iputs() |
| will handle delayed iputs, |
| that means start_transaction() |
| will be called, which will try |
| to get SB_FREEZE_FS lock. |
To fix this issue, introduce a "int fs_frozen" to record internally whether
fs has been frozen. If fs has been frozen, we can not handle delayed iputs.
Signed-off-by: Wang Xiaoguang <wangxg.fnst@cn.fujitsu.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ add comment to btrfs_freeze ]
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
__btrfs_abort_transaction doesn't use its root parameter except to
obtain an fs_info pointer. We can obtain that from trans->root->fs_info
for now and from trans->fs_info in a later patch.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_trans_handle->root is documented as for use for confirming
that the root passed in to start the transaction is the same as the
one ending it. It's used in several places when an fs_info pointer
is needed, so let's just add an fs_info pointer directly. Eventually,
the root pointer can be removed.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_test_opt and friends only use the root pointer to access
the fs_info. Let's pass the fs_info directly in preparation to
eliminate similar patterns all over btrfs.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Using the offwakecputime bpf script I noticed most of our time was spent waiting
on the delayed ref throttling. This is what is supposed to happen, but
sometimes the transaction can commit and then we're waiting for throttling that
doesn't matter anymore. So change this stuff to be a little smarter by tracking
the transid we were in when we initiated the throttling. If the transaction we
get is different then we can just bail out. This resulted in a 50% speedup in
my fs_mark test, and reduced the amount of time spent throttling by 60 seconds
over the entire run (which is about 30 minutes). Thanks,
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
Introduced in 2c1984f244 ("btrfs: build fixup for
qgroup_account_snapshot") as temporary bisectability build fixup.
Signed-off-by: David Sterba <dsterba@suse.com>
Before the relocation process of a block group starts, it sets the block
group to readonly mode, then flushes all delalloc writes and then finally
it waits for all ordered extents to complete. This last step includes
waiting for ordered extents destinated at extents allocated in other block
groups, making us waste unecessary time.
So improve this by waiting only for ordered extents that fall into the
block group's range.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
The macro btrfs_std_error got renamed to btrfs_handle_fs_error in an
independent branch for the same merge target (4.7). To make the code
compilable for bisectability reasons, add a temporary stub.
Signed-off-by: David Sterba <dsterba@suse.com>
Current btrfs qgroup design implies a requirement that after calling
btrfs_qgroup_account_extents() there must be a commit root switch.
Normally this is OK, as btrfs_qgroup_accounting_extents() is only called
inside btrfs_commit_transaction() just be commit_cowonly_roots().
However there is a exception at create_pending_snapshot(), which will
call btrfs_qgroup_account_extents() but no any commit root switch.
In case of creating a snapshot whose parent root is itself (create a
snapshot of fs tree), it will corrupt qgroup by the following trace:
(skipped unrelated data)
======
btrfs_qgroup_account_extent: bytenr = 29786112, num_bytes = 16384, nr_old_roots = 0, nr_new_roots = 1
qgroup_update_counters: qgid = 5, cur_old_count = 0, cur_new_count = 1, rfer = 0, excl = 0
qgroup_update_counters: qgid = 5, cur_old_count = 0, cur_new_count = 1, rfer = 16384, excl = 16384
btrfs_qgroup_account_extent: bytenr = 29786112, num_bytes = 16384, nr_old_roots = 0, nr_new_roots = 0
======
The problem here is in first qgroup_account_extent(), the
nr_new_roots of the extent is 1, which means its reference got
increased, and qgroup increased its rfer and excl.
But at second qgroup_account_extent(), its reference got decreased, but
between these two qgroup_account_extent(), there is no switch roots.
This leads to the same nr_old_roots, and this extent just got ignored by
qgroup, which means this extent is wrongly accounted.
Fix it by call commit_cowonly_roots() after qgroup_account_extent() in
create_pending_snapshot(), with needed preparation.
Mark: I added a check at the top of qgroup_account_snapshot() to skip this
code if qgroups are turned off. xfstest btrfs/122 exposes this problem.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Mark Fasheh <mfasheh@suse.de>
Signed-off-by: David Sterba <dsterba@suse.com>
Single caller passes GFP_NOFS. We can get rid of the
gfpflags_allow_blocking checks as NOFS can block but does not recurse to
filesystem through reclaim.
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_std_error() handles errors, puts FS into readonly mode
(as of now). So its good idea to rename it to btrfs_handle_fs_error().
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ edit changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
CURRENT_TIME macro is not appropriate for filesystems as it
doesn't use the right granularity for filesystem timestamps.
Use current_fs_time() instead.
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: Chris Mason <clm@fb.com>
Cc: Josef Bacik <jbacik@fb.com>
Cc: linux-btrfs@vger.kernel.org
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are a few places where we add to trans->bytes_reserved but don't have the
corresponding trace point. With these added my tool no longer sees transaction
leaks.
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We can also preallocate btrfs_path that's used during pending snapshot
creation and avoid another late ENOMEM failure.
Signed-off-by: David Sterba <dsterba@suse.com>
The actual snapshot creation is delayed until transaction commit. If we
cannot get enough memory for the root item there, we have to fail the
whole transaction commit which is bad. So we'll allocate the memory at
the ioctl call and pass it along with the pending_snapshot struct. The
potential ENOMEM will be returned to the caller of snapshot ioctl.
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Alexandru Moise <00moses.alexander00@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When a transaction is aborted, or its commit fails before writing the new
superblock and calling btrfs_finish_extent_commit(), we leak reference
counts on the block groups attached to the transaction's delete_bgs list,
because btrfs_finish_extent_commit() is never called for those two cases.
Fix this by dropping their references at btrfs_put_transaction(), which
is called when transactions are aborted (by making the transaction kthread
commit the transaction) or if their commits fail.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
As of my previous change titled "Btrfs: fix scrub preventing unused block
groups from being deleted", the following warning at
extent-tree.c:btrfs_delete_unused_bgs() can be hit when we mount the a
filesysten with "-o discard":
10263 void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info)
10264 {
(...)
10405 if (trimming) {
10406 WARN_ON(!list_empty(&block_group->bg_list));
10407 spin_lock(&trans->transaction->deleted_bgs_lock);
10408 list_move(&block_group->bg_list,
10409 &trans->transaction->deleted_bgs);
10410 spin_unlock(&trans->transaction->deleted_bgs_lock);
10411 btrfs_get_block_group(block_group);
10412 }
(...)
This happens because scrub can now add back the block group to the list of
unused block groups (fs_info->unused_bgs). This is dangerous because we
are moving the block group from the unused block groups list to the list
of deleted block groups without holding the lock that protects the source
list (fs_info->unused_bgs_lock).
The following diagram illustrates how this happens:
CPU 1 CPU 2
cleaner_kthread()
btrfs_delete_unused_bgs()
sees bg X in list
fs_info->unused_bgs
deletes bg X from list
fs_info->unused_bgs
scrub_enumerate_chunks()
searches device tree using
its commit root
finds device extent for
block group X
gets block group X from the tree
fs_info->block_group_cache_tree
(via btrfs_lookup_block_group())
sets bg X to RO (again)
scrub_chunk(bg X)
sets bg X back to RW mode
adds bg X to the list
fs_info->unused_bgs again,
since it's still unused and
currently not in that list
sets bg X to RO mode
btrfs_remove_chunk(bg X)
--> discard is enabled and bg X
is in the fs_info->unused_bgs
list again so the warning is
triggered
--> we move it from that list into
the transaction's delete_bgs
list, but we can have another
task currently manipulating
the first list (fs_info->unused_bgs)
Fix this by using the same lock (fs_info->unused_bgs_lock) to protect both
the list of unused block groups and the list of deleted block groups. This
makes it safe and there's not much worry for more lock contention, as this
lock is seldom used and only the cleaner kthread adds elements to the list
of deleted block groups. The warning goes away too, as this was previously
an impossible case (and would have been better a BUG_ON/ASSERT) but it's
not impossible anymore.
Reproduced with fstest btrfs/073 (using MOUNT_OPTIONS="-o discard").
Signed-off-by: Filipe Manana <fdmanana@suse.com>
It's possible to reach a state where the cleaner kthread isn't able to
start a transaction to delete an unused block group due to lack of enough
free metadata space and due to lack of unallocated device space to allocate
a new metadata block group as well. If this happens try to use space from
the global block group reserve just like we do for unlink operations, so
that we don't reach a permanent state where starting a transaction for
filesystem operations (file creation, renames, etc) keeps failing with
-ENOSPC. Such an unfortunate state was observed on a machine where over
a dozen unused data block groups existed and the cleaner kthread was
failing to delete them due to ENOSPC error when attempting to start a
transaction, and even running balance with a -dusage=0 filter failed with
ENOSPC as well. Also unmounting and mounting again the filesystem didn't
help. Allowing the cleaner kthread to use the global block reserve to
delete the unused data block groups fixed the problem.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
I want to set some per transaction flags, so instead of adding yet another int
lets just convert the current two int indicators to flags and add a flags field
for future use. Thanks,
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
We have a mechanism to make sure we don't lose updates for ordered extents that
were logged in the transaction that is currently running. We add the ordered
extent to a transaction list and then the transaction waits on all the ordered
extents in that list. However are substantially large file systems this list
can be extremely large, and can give us soft lockups, since the ordered extents
don't remove themselves from the list when they do complete.
To fix this we simply add a counter to the transaction that is incremented any
time we have a logged extent that needs to be completed in the current
transaction. Then when the ordered extent finally completes it decrements the
per transaction counter and wakes up the transaction if we are the last ones.
This will eliminate the softlockup. Thanks,
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
As we have the new metadata reservation functions, use them to replace
the old btrfs_qgroup_reserve() call for metadata.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: Chris Mason <clm@fb.com>
The value of num_items that start_transaction() ultimately
always takes is a small one, so a 64 bit integer is overkill.
Also change num_items for btrfs_start_transaction() and
btrfs_start_transaction_lflush() as well.
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: Alexandru Moise <00moses.alexander00@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Use memset() to null out the btrfs_delayed_ref_root of
btrfs_transaction instead of setting all the members to 0 by hand.
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: Alexandru Moise <00moses.alexander00@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Just call kmem_cache_zalloc() instead of calling kmem_cache_alloc().
We're just initializing most fields to 0, false and NULL later on
_anyway_, so to make the code mode readable and potentially gain
a bit of performance (completely untested claim), we should fill our
btrfs_trans_handle with zeros on allocation then just initialize
those five remaining fields (not counting the list_heads) as normal.
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: Alexandru Moise <00moses.alexander00@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
