[BUG]
The following script can lead to tons of beyond device boundary access:
mkfs.btrfs -f $dev -b 10G
mount $dev $mnt
trimfs $mnt
btrfs filesystem resize 1:-1G $mnt
trimfs $mnt
[CAUSE]
Since commit 929be17a9b ("btrfs: Switch btrfs_trim_free_extents to
find_first_clear_extent_bit"), we try to avoid trimming ranges that's
already trimmed.
So we check device->alloc_state by finding the first range which doesn't
have CHUNK_TRIMMED and CHUNK_ALLOCATED not set.
But if we shrunk the device, that bits are not cleared, thus we could
easily got a range starts beyond the shrunk device size.
This results the returned @start and @end are all beyond device size,
then we call "end = min(end, device->total_bytes -1);" making @end
smaller than device size.
Then finally we goes "len = end - start + 1", totally underflow the
result, and lead to the beyond-device-boundary access.
[FIX]
This patch will fix the problem in two ways:
- Clear CHUNK_TRIMMED | CHUNK_ALLOCATED bits when shrinking device
This is the root fix
- Add extra safety check when trimming free device extents
We check and warn if the returned range is already beyond current
device.
Link: https://github.com/kdave/btrfs-progs/issues/282
Fixes: 929be17a9b ("btrfs: Switch btrfs_trim_free_extents to find_first_clear_extent_bit")
CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
There is a bug report about bad signal timing could lead to read-only
fs during balance:
BTRFS info (device xvdb): balance: start -d -m -s
BTRFS info (device xvdb): relocating block group 73001861120 flags metadata
BTRFS info (device xvdb): found 12236 extents, stage: move data extents
BTRFS info (device xvdb): relocating block group 71928119296 flags data
BTRFS info (device xvdb): found 3 extents, stage: move data extents
BTRFS info (device xvdb): found 3 extents, stage: update data pointers
BTRFS info (device xvdb): relocating block group 60922265600 flags metadata
BTRFS: error (device xvdb) in btrfs_drop_snapshot:5505: errno=-4 unknown
BTRFS info (device xvdb): forced readonly
BTRFS info (device xvdb): balance: ended with status: -4
[CAUSE]
The direct cause is the -EINTR from the following call chain when a
fatal signal is pending:
relocate_block_group()
|- clean_dirty_subvols()
|- btrfs_drop_snapshot()
|- btrfs_start_transaction()
|- btrfs_delayed_refs_rsv_refill()
|- btrfs_reserve_metadata_bytes()
|- __reserve_metadata_bytes()
|- wait_reserve_ticket()
|- prepare_to_wait_event();
|- ticket->error = -EINTR;
Normally this behavior is fine for most btrfs_start_transaction()
callers, as they need to catch any other error, same for the signal, and
exit ASAP.
However for balance, especially for the clean_dirty_subvols() case, we're
already doing cleanup works, getting -EINTR from btrfs_drop_snapshot()
could cause a lot of unexpected problems.
From the mentioned forced read-only report, to later balance error due
to half dropped reloc trees.
[FIX]
Fix this problem by using btrfs_join_transaction() if
btrfs_drop_snapshot() is called from relocation context.
Since btrfs_join_transaction() won't get interrupted by signal, we can
continue the cleanup.
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>3
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
Sometime fsstress could lead to qgroup warning for case like
generic/013:
BTRFS warning (device dm-3): qgroup 0/259 has unreleased space, type 1 rsv 81920
------------[ cut here ]------------
WARNING: CPU: 9 PID: 24535 at fs/btrfs/disk-io.c:4142 close_ctree+0x1dc/0x323 [btrfs]
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
RIP: 0010:close_ctree+0x1dc/0x323 [btrfs]
Call Trace:
btrfs_put_super+0x15/0x17 [btrfs]
generic_shutdown_super+0x72/0x110
kill_anon_super+0x18/0x30
btrfs_kill_super+0x17/0x30 [btrfs]
deactivate_locked_super+0x3b/0xa0
deactivate_super+0x40/0x50
cleanup_mnt+0x135/0x190
__cleanup_mnt+0x12/0x20
task_work_run+0x64/0xb0
__prepare_exit_to_usermode+0x1bc/0x1c0
__syscall_return_slowpath+0x47/0x230
do_syscall_64+0x64/0xb0
entry_SYSCALL_64_after_hwframe+0x44/0xa9
---[ end trace 6c341cdf9b6cc3c1 ]---
BTRFS error (device dm-3): qgroup reserved space leaked
While that subvolume 259 is no longer in that filesystem.
[CAUSE]
Normally per-trans qgroup reserved space is freed when a transaction is
committed, in commit_fs_roots().
However for completely dropped subvolume, that subvolume is completely
gone, thus is no longer in the fs_roots_radix, and its per-trans
reserved qgroup will never be freed.
Since the subvolume is already gone, leaked per-trans space won't cause
any trouble for end users.
[FIX]
Just call btrfs_qgroup_free_meta_all_pertrans() before a subvolume is
completely dropped.
Signed-off-by: Qu Wenruo <wqu@suse.com>
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>
Deleting a subvolume on a full filesystem leads to ENOSPC followed by a
forced read-only. This is not a transaction abort and the filesystem is
otherwise ok, so the error should be just propagated to the callers.
This is caused by unnecessary call to btrfs_handle_fs_error for all
errors, except EAGAIN. This does not make sense as the standard
transaction abort mechanism is in btrfs_drop_snapshot so all relevant
failures are handled.
Originally in commit cb1b69f450 ("Btrfs: forced readonly when
btrfs_drop_snapshot() fails") there was no return value at all, so the
btrfs_std_error made some sense but once the error handling and
propagation has been implemented we don't need it anymore.
Signed-off-by: David Sterba <dsterba@suse.com>
Sparse reports a warning at btrfs_lock_cluster()
warning: context imbalance in btrfs_lock_cluster()
- wrong count
The root cause is the missing annotation at btrfs_lock_cluster()
Add the missing __acquires(&cluster->refill_lock) annotation.
Signed-off-by: Jules Irenge <jbi.octave@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Readahead will generate a lot of extra reads for adjacent nodes, but
when running delayed refs we have no idea if the next ref is going to be
adjacent or not, so this potentially just generates a lot of extra IO.
To make matters worse each ref is truly just looking for one item, it
doesn't generally search forward, so we simply don't need it here.
Reviewed-by: Qu Wenruo <wqu@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>
There are a few different ways to free roots, either you allocated them
yourself and you just do
free_extent_buffer(root->node);
free_extent_buffer(root->commit_node);
btrfs_put_root(root);
Which is the pattern for log roots. Or for snapshots/subvolumes that
are being dropped you simply call btrfs_free_fs_root() which does all
the cleanup for you.
Unify this all into btrfs_put_root(), so that we don't free up things
associated with the root until the last reference is dropped. This
makes the root freeing code much more significant.
The only caveat is at close_ctree() time we have to free the extent
buffers for all of our main roots (extent_root, chunk_root, etc) because
we have to drop the btree_inode and we'll run into issues if we hold
onto those nodes until ->kill_sb() time. This will be addressed in the
future when we kill the btree_inode.
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>
No need to add a level of indirection for hiding a simple 'if'. Open
code insert_extent_backref in its sole caller. No functional changes.
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>
This function finally factor out prepare_allocation() form
find_free_extent(). This function is called before the allocation loop
and a specific allocator function like prepare_allocation_clustered()
should initialize their private information and can set proper hint_byte
to indicate where to start the allocation with.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
LOOP_NO_EMPTY_SIZE is solely dedicated for clustered allocation. So, we
can skip this stage and give up the allocation.
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Factor out chunk_allocation_failed() from
find_free_extent_update_loop(). This function is called when it failed
to allocate a chunk. The function can modify "ffe_ctl->loop" and return
0 to continue with the next stage. Or, it can return -ENOSPC to give up
here.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that, we don't use last_ptr and use_cluster in the function. Drop
these arguments from it.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Factor out found_extent() from find_free_extent_update_loop(). This
function is called when a proper extent is found and before returning
from find_free_extent(). Hook functions like found_extent_clustered()
should save information for a next allocation.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Factor out release_block_group() from find_free_extent(). This function
is called when it gives up an allocation from a block group. Each
allocation policy should reset its information for an allocation in
the next block group.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that, find_free_extent_clustered() and find_free_extent_unclustered()
can access "last_ptr" from the "clustered" variable, we can drop it from
the arguments.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Factor out do_allocation() from find_free_extent(). This function do an
actual extent allocation in a given block group. The ffe_ctl->policy is
used to determine the actual allocator function to use.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Move "last_ptr" and "use_cluster" into struct find_free_extent_ctl, so
that hook functions for clustered allocator can use these variables.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit moves hint_byte into find_free_extent_ctl, so that we can
modify the hint_byte in the other functions. This will help us split
find_free_extent further. This commit also renames the function argument
"hint_byte" to "hint_byte_orig" to avoid misuse.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit introduces extent allocation policy for btrfs. This policy
controls how btrfs allocate an extents from block groups. There is no
functional change introduced with this commit.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
While the "full_search" variable defined in find_free_extent() is bool,
but the full_search argument of find_free_extent_update_loop() is
defined as int. Let's trivially fix the argument type.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All callers pass extent buffer start and length so the extent buffer
itself should work fine.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This patch removes all haphazard code implementing nocow writers
exclusion from pending snapshot creation and switches to using the drew
lock to ensure this invariant still holds.
'Readers' are snapshot creators from create_snapshot and 'writers' are
nocow writers from buffered write path or btrfs_setsize. This locking
scheme allows for multiple snapshots to happen while any nocow writers
are blocked, since writes to page cache in the nocow path will make
snapshots inconsistent.
So for performance reasons we'd like to have the ability to run multiple
concurrent snapshots and also favors readers in this case. And in case
there aren't pending snapshots (which will be the majority of the cases)
we rely on the percpu's writers counter to avoid cacheline contention.
The main gain from using the drew lock is it's now a lot easier to
reason about the guarantees of the locking scheme and whether there is
some silent breakage lurking.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If we're allocating a logged extent we attempt to insert an extent
record for the file extent directly. We increase
space_info->bytes_reserved, because the extent entry addition will call
btrfs_update_block_group(), which will convert the ->bytes_reserved to
->bytes_used. However if we fail at any point while inserting the
extent entry we will bail and leave space on ->bytes_reserved, which
will trigger a WARN_ON() on umount. Fix this by pinning the space if we
fail to insert, which is what happens in every other failure case that
involves adding the extent entry.
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@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 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>
In preparation to making pinned extents per-transaction ensure that log
such extents are always excluded from caching. To achieve this in
addition to marking them via btrfs_pin_extent_for_log_replay they also
need to be marked with btrfs_add_excluded_extent to prevent log tree
extent buffer being loaded by the free space caching thread. That's
required since log tree blocks are not recorded in the extent tree, hence
they always look free.
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>
All callers have a reference to a transaction handle so pass it to
pin_down_extent. This is the final step before switching pinned extent
tracking to a per-transaction basis.
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>
Preparation for refactoring pinned extents tracking.
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>
btrfs_pin_reserved_extent is now only called with a valid transaction so
exploit the fact to take a transaction. This is preparation for tracking
pinned extents on a per-transaction basis.
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>
Preparation for switching pinned extent tracking to a per-transaction
basis.
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>
We are now using these for all roots, rename them to btrfs_put_root()
and btrfs_grab_root();
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>
If we're allocating a logged extent we attempt to insert an extent
record for the file extent directly. We increase
space_info->bytes_reserved, because the extent entry addition will call
btrfs_update_block_group(), which will convert the ->bytes_reserved to
->bytes_used. However if we fail at any point while inserting the
extent entry we will bail and leave space on ->bytes_reserved, which
will trigger a WARN_ON() on umount. Fix this by pinning the space if we
fail to insert, which is what happens in every other failure case that
involves adding the extent entry.
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@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>
An earlier patch keeps track of discardable_extents. These are
undiscarded extents managed by the free space cache. Here, we will use
this to dynamically calculate the discard delay interval.
There are 3 rate to consider. The first is the target convergence rate,
the rate to discard all discardable_extents over the
BTRFS_DISCARD_TARGET_MSEC time frame. This is clamped by the lower
limit, the iops limit or BTRFS_DISCARD_MIN_DELAY (1ms), and the upper
limit, BTRFS_DISCARD_MAX_DELAY (1s). We reevaluate this delay every
transaction commit.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Dennis Zhou <dennis@kernel.org>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When discard is enabled, everytime a pinned extent is released back to
the block_group's free space cache, a discard is issued for the extent.
This is an overeager approach when it comes to discarding and helping
the SSD maintain enough free space to prevent severe garbage collection
situations.
This adds the beginning of async discard. Instead of issuing a discard
prior to returning it to the free space, it is just marked as untrimmed.
The block_group is then added to a LRU which then feeds into a workqueue
to issue discards at a much slower rate. Full discarding of unused block
groups is still done and will be addressed in a future patch of the
series.
For now, we don't persist the discard state of extents and bitmaps.
Therefore, our failure recovery mode will be to consider extents
untrimmed. This lets us handle failure and unmounting as one in the
same.
On a number of Facebook webservers, I collected data every minute
accounting the time we spent in btrfs_finish_extent_commit() (col. 1)
and in btrfs_commit_transaction() (col. 2). btrfs_finish_extent_commit()
is where we discard extents synchronously before returning them to the
free space cache.
discard=sync:
p99 total per minute p99 total per minute
Drive | extent_commit() (ms) | commit_trans() (ms)
---------------------------------------------------------------
Drive A | 434 | 1170
Drive B | 880 | 2330
Drive C | 2943 | 3920
Drive D | 4763 | 5701
discard=async:
p99 total per minute p99 total per minute
Drive | extent_commit() (ms) | commit_trans() (ms)
--------------------------------------------------------------
Drive A | 134 | 956
Drive B | 64 | 1972
Drive C | 59 | 1032
Drive D | 62 | 1200
While it's not great that the stats are cumulative over 1m, all of these
servers are running the same workload and and the delta between the two
are substantial. We are spending significantly less time in
btrfs_finish_extent_commit() which is responsible for discarding.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Dennis Zhou <dennis@kernel.org>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This series introduces async discard which will use the flag
DISCARD_ASYNC, so rename the original flag to DISCARD_SYNC as it is
synchronously done in transaction commit.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Dennis Zhou <dennis@kernel.org>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This hasn't been used since it was first introduced in commit
b4bd745d12 ("btrfs: Introduce find_free_extent_ctl structure for later
rework"). Passing that to btrfs_add_reserved_bytes in find_free_extent
is not strictly necessary and using the local ram_bytes instead seems
cleaner.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
__btrfs_free_reserved_extent now performs the actions of
btrfs_free_and_pin_reserved_extent. But this name is a bit of a
misnomer, since the extent is not really freed but just pinned. Reflect
this in the new name. No semantics changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
__btrfs_free_reserved_extent performs 2 entirely different operations
depending on whether its 'pin' argument is true or false. This patch
lifts the 2nd case (pin is false) into it's sole caller
btrfs_free_reserved_extent. No semantics changes.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All callers of btrfs_free_reserved_extent (respectively
__btrfs_free_reserved_extent with in set to 0) pass in extents which
have only been reserved but not yet written to. Namely,
* in cow_file_range that function is called only if create_io_em fails
or btrfs_add_ordered_extent fail, both of which happen _before_ any IO
is submitted to the newly reserved range
* in submit_compressed_extents the code flow is similar -
out_free_reserve can be called only before
btrfs_submit_compressed_write which is where any writes to the range
could occur
* btrfs_new_extent_direct also calls btrfs_free_reserved_extent only
if extent_map fails, before any IO is issued
* __btrfs_prealloc_file_range also calls btrfs_free_reserved_extent
in case insertion of the metadata fails
* btrfs_alloc_tree_block again can only be called in case in-memory
operations fail, before any IO is submitted
* btrfs_finish_ordered_io - this is the only caller where discarding
the extent could have a material effect, since it can be called for
an extent which was partially written.
With this change the submission of discards is optimised since discards
are now not being created for extents which are known to not have been
touched on disk.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When logging a file that has shared extents (reflinked with other files or
with itself), we can end up logging multiple checksum items that cover
overlapping ranges. This confuses the search for checksums at log replay
time causing some checksums to never be added to the fs/subvolume tree.
Consider the following example of a file that shares the same extent at
offsets 0 and 256Kb:
[ bytenr 13893632, offset 64Kb, len 64Kb ]
0 64Kb
[ bytenr 13631488, offset 64Kb, len 192Kb ]
64Kb 256Kb
[ bytenr 13893632, offset 0, len 256Kb ]
256Kb 512Kb
When logging the inode, at tree-log.c:copy_items(), when processing the
file extent item at offset 0, we log a checksum item covering the range
13959168 to 14024704, which corresponds to 13893632 + 64Kb and 13893632 +
64Kb + 64Kb, respectively.
Later when processing the extent item at offset 256K, we log the checksums
for the range from 13893632 to 14155776 (which corresponds to 13893632 +
256Kb). These checksums get merged with the checksum item for the range
from 13631488 to 13893632 (13631488 + 256Kb), logged by a previous fsync.
So after this we get the two following checksum items in the log tree:
(...)
item 6 key (EXTENT_CSUM EXTENT_CSUM 13631488) itemoff 3095 itemsize 512
range start 13631488 end 14155776 length 524288
item 7 key (EXTENT_CSUM EXTENT_CSUM 13959168) itemoff 3031 itemsize 64
range start 13959168 end 14024704 length 65536
The first one covers the range from the second one, they overlap.
So far this does not cause a problem after replaying the log, because
when replaying the file extent item for offset 256K, we copy all the
checksums for the extent 13893632 from the log tree to the fs/subvolume
tree, since searching for an checksum item for bytenr 13893632 leaves us
at the first checksum item, which covers the whole range of the extent.
However if we write 64Kb to file offset 256Kb for example, we will
not be able to find and copy the checksums for the last 128Kb of the
extent at bytenr 13893632, referenced by the file range 384Kb to 512Kb.
After writing 64Kb into file offset 256Kb we get the following extent
layout for our file:
[ bytenr 13893632, offset 64K, len 64Kb ]
0 64Kb
[ bytenr 13631488, offset 64Kb, len 192Kb ]
64Kb 256Kb
[ bytenr 14155776, offset 0, len 64Kb ]
256Kb 320Kb
[ bytenr 13893632, offset 64Kb, len 192Kb ]
320Kb 512Kb
After fsync'ing the file, if we have a power failure and then mount
the filesystem to replay the log, the following happens:
1) When replaying the file extent item for file offset 320Kb, we
lookup for the checksums for the extent range from 13959168
(13893632 + 64Kb) to 14155776 (13893632 + 256Kb), through a call
to btrfs_lookup_csums_range();
2) btrfs_lookup_csums_range() finds the checksum item that starts
precisely at offset 13959168 (item 7 in the log tree, shown before);
3) However that checksum item only covers 64Kb of data, and not 192Kb
of data;
4) As a result only the checksums for the first 64Kb of data referenced
by the file extent item are found and copied to the fs/subvolume tree.
The remaining 128Kb of data, file range 384Kb to 512Kb, doesn't get
the corresponding data checksums found and copied to the fs/subvolume
tree.
5) After replaying the log userspace will not be able to read the file
range from 384Kb to 512Kb, because the checksums are missing and
resulting in an -EIO error.
The following steps reproduce this scenario:
$ mkfs.btrfs -f /dev/sdc
$ mount /dev/sdc /mnt/sdc
$ xfs_io -f -c "pwrite -S 0xa3 0 256K" /mnt/sdc/foobar
$ xfs_io -c "fsync" /mnt/sdc/foobar
$ xfs_io -c "pwrite -S 0xc7 256K 256K" /mnt/sdc/foobar
$ xfs_io -c "reflink /mnt/sdc/foobar 320K 0 64K" /mnt/sdc/foobar
$ xfs_io -c "fsync" /mnt/sdc/foobar
$ xfs_io -c "pwrite -S 0xe5 256K 64K" /mnt/sdc/foobar
$ xfs_io -c "fsync" /mnt/sdc/foobar
<power failure>
$ mount /dev/sdc /mnt/sdc
$ md5sum /mnt/sdc/foobar
md5sum: /mnt/sdc/foobar: Input/output error
$ dmesg | tail
[165305.003464] BTRFS info (device sdc): no csum found for inode 257 start 401408
[165305.004014] BTRFS info (device sdc): no csum found for inode 257 start 405504
[165305.004559] BTRFS info (device sdc): no csum found for inode 257 start 409600
[165305.005101] BTRFS info (device sdc): no csum found for inode 257 start 413696
[165305.005627] BTRFS info (device sdc): no csum found for inode 257 start 417792
[165305.006134] BTRFS info (device sdc): no csum found for inode 257 start 421888
[165305.006625] BTRFS info (device sdc): no csum found for inode 257 start 425984
[165305.007278] BTRFS info (device sdc): no csum found for inode 257 start 430080
[165305.008248] BTRFS warning (device sdc): csum failed root 5 ino 257 off 393216 csum 0x1337385e expected csum 0x00000000 mirror 1
[165305.009550] BTRFS warning (device sdc): csum failed root 5 ino 257 off 393216 csum 0x1337385e expected csum 0x00000000 mirror 1
Fix this simply by deleting first any checksums, from the log tree, for the
range of the extent we are logging at copy_items(). This ensures we do not
get checksum items in the log tree that have overlapping ranges.
This is a long time issue that has been present since we have the clone
(and deduplication) ioctl, and can happen both when an extent is shared
between different files and within the same file.
A test case for fstests follows soon.
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have a BUG_ON(ret < 0) in find_free_extent from
btrfs_cache_block_group. If we fail to allocate our ctl we'll just
panic, which is not good. Instead just go on to another block group.
If we fail to find a block group we don't want to return ENOSPC, because
really we got a ENOMEM and that's the root of the problem. Save our
return from btrfs_cache_block_group(), and then if we still fail to make
our allocation return that ret so we get the right error back.
Tested with inject-error.py from bcc.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.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>
[BUG]
When deleting large files (which cross block group boundary) with
discard mount option, we find some btrfs_discard_extent() calls only
trimmed part of its space, not the whole range:
btrfs_discard_extent: type=0x1 start=19626196992 len=2144530432 trimmed=1073741824 ratio=50%
type: bbio->map_type, in above case, it's SINGLE DATA.
start: Logical address of this trim
len: Logical length of this trim
trimmed: Physically trimmed bytes
ratio: trimmed / len
Thus leaving some unused space not discarded.
[CAUSE]
When discard mount option is specified, after a transaction is fully
committed (super block written to disk), we begin to cleanup pinned
extents in the following call chain:
btrfs_commit_transaction()
|- btrfs_finish_extent_commit()
|- find_first_extent_bit(unpin, 0, &start, &end, EXTENT_DIRTY);
|- btrfs_discard_extent()
However, pinned extents are recorded in an extent_io_tree, which can
merge adjacent extent states.
When a large file gets deleted and it has adjacent file extents across
block group boundary, we will get a large merged range like this:
|<--- BG1 --->|<--- BG2 --->|
|//////|<-- Range to discard --->|/////|
To discard that range, we have the following calls:
btrfs_discard_extent()
|- btrfs_map_block()
| Returned bbio will end at BG1's end. As btrfs_map_block()
| never returns result across block group boundary.
|- btrfs_issuse_discard()
Issue discard for each stripe.
So we will only discard the range in BG1, not the remaining part in BG2.
Furthermore, this bug is not that reliably observed, for above case, if
there is no other extent in BG2, BG2 will be empty and btrfs will trim
all space of BG2, covering up the bug.
[FIX]
- Allow __btrfs_map_block_for_discard() to modify @length parameter
btrfs_map_block() uses its @length paramter to notify the caller how
many bytes are mapped in current call.
With __btrfs_map_block_for_discard() also modifing the @length,
btrfs_discard_extent() now understands when to do extra trim.
- Call btrfs_map_block() in a loop until we hit the range end Since we
now know how many bytes are mapped each time, we can iterate through
each block group boundary and issue correct trim for each range.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Tested-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The on-disk format of block group item makes use of the key that stores
the offset and length. This is further used in the code, although this
makes thing harder to understand. The key is also packed so the
offset/length is not properly aligned as u64.
Add start (key.objectid) and length (key.offset) members to block group
and remove the embedded key. When the item is searched or written, a
local variable for key is used.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For unknown reasons, the member 'used' in the block group struct is
stored in the b-tree item and accessed everywhere using the special
accessor helper. Let's unify it and make it a regular member and only
update the item before writing it to the tree.
The item is still being used for flags and chunk_objectid, there's some
duplication until the item is removed in following patches.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Using an ASSERT in btrfs_pin_extent allows to more stringently observe
whether the function is called under a transaction or not.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Qu Wenruo
Filipe Manana
YueHaibing
David Sterba
Jules Irenge
Josef Bacik
Nikolay Borisov
Naohiro Aota
Dennis Zhou
Omar Sandoval