Currently we are only documenting two uses of the bg_list member of a
block group, but there two more:
1) To track deleted block groups for discard purposes, introduced in
commit e33e17ee10 ("btrfs: add missing discards when unpinning
extents with -o discard");
2) To track block groups for automatic reclaim, introduced more recently
by commit 18bb8bbf13 ("btrfs: zoned: automatically reclaim zones")
So document those two other use cases.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The old scrub code has different entrance to verify the content, and
since we have removed the writeback path, now we can start removing the
re-check part, including:
- scrub_recover structure
- scrub_sector::recover member
- function scrub_setup_recheck_block()
- function scrub_recheck_block()
- function scrub_recheck_block_checksum()
- function scrub_repair_block_group_good_copy()
- function scrub_repair_sector_from_good_copy()
- function scrub_is_page_on_raid56()
- function full_stripe_lock()
- function search_full_stripe_lock()
- function get_full_stripe_logical()
- function insert_full_stripe_lock()
- function lock_full_stripe()
- function unlock_full_stripe()
- btrfs_block_group::full_stripe_locks_root member
- btrfs_full_stripe_locks_tree structure
This infrastructure is to ensure RAID56 scrub is properly handling
recovery and P/Q scrub correctly.
This is no longer needed, before P/Q scrub we will wait for all
the involved data stripes to be scrubbed first, and RAID56 code has
internal lock to ensure no race in the same full stripe.
- function scrub_print_warning()
- function scrub_get_recover()
- function scrub_put_recover()
- function scrub_handle_errored_block()
- function scrub_setup_recheck_block()
- function scrub_bio_wait_endio()
- function scrub_submit_raid56_bio_wait()
- function scrub_recheck_block_on_raid56()
- function scrub_recheck_block()
- function scrub_recheck_block_checksum()
- function scrub_repair_block_from_good_copy()
- function scrub_repair_sector_from_good_copy()
And two more functions exported temporarily for later cleanup:
- alloc_scrub_sector()
- alloc_scrub_block()
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The only caller of btrfs_make_block_group() always passes 0 as the value
for the bytes_used argument, so remove it.
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>
The only user in the zoned remap code is gone now, so remove the argument.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When a file system has ZNS devices which are constrained by a maximum
number of active block groups, then not being able to use all the block
groups for every allocation is not ideal, and could cause us to loop a
ton with mixed size allocations.
In general, since zoned doesn't write into gaps behind where block
groups are writing, it is not susceptible to the same sort of
fragmentation that size classes are designed to solve, so we can skip
size classes for zoned file systems in general, even though there would
probably be no harm for SMR devices.
Signed-off-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
The aim of this patch is to reduce the fragmentation of block groups
under certain unhappy workloads. It is particularly effective when the
size of extents correlates with their lifetime, which is something we
have observed causing fragmentation in the fleet at Meta.
This patch categorizes extents into size classes:
- x < 128KiB: "small"
- 128KiB < x < 8MiB: "medium"
- x > 8MiB: "large"
and as much as possible reduces allocations of extents into block groups
that don't match the size class. This takes advantage of any (possible)
correlation between size and lifetime and also leaves behind predictable
re-usable gaps when extents are freed; small writes don't gum up bigger
holes.
Size classes are implemented in the following way:
- Mark each new block group with a size class of the first allocation
that goes into it.
- Add two new passes to ffe: "unset size class" and "wrong size class".
First, try only matching block groups, then try unset ones, then allow
allocation of new ones, and finally allow mismatched block groups.
- Filtering is done just by skipping inappropriate ones, there is no
special size class indexing.
Other solutions I considered were:
- A best fit allocator with an rb-tree. This worked well, as small
writes didn't leak big holes from large freed extents, but led to
regressions in ffe and write performance due to lock contention on
the rb-tree with every allocation possibly updating it in parallel.
Perhaps something clever could be done to do the updates in the
background while being "right enough".
- A fixed size "working set". This prevents freeing an extent
drastically changing where writes currently land, and seems like a
good option too. Doesn't take advantage of size in any way.
- The same size class idea, but implemented with xarray marks. This
turned out to be slower than looping the linked list and skipping
wrong block groups, and is also less flexible since we must have only
3 size classes (max #marks). With the current approach we can have as
many as we like.
Performance testing was done via: https://github.com/josefbacik/fsperf
Of particular relevance are the new fragmentation specific tests.
A brief summary of the testing results:
- Neutral results on existing tests. There are some minor regressions
and improvements here and there, but nothing that truly stands out as
notable.
- Improvement on new tests where size class and extent lifetime are
correlated. Fragmentation in these cases is completely eliminated
and write performance is generally a little better. There is also
significant improvement where extent sizes are just a bit larger than
the size class boundaries.
- Regression on one new tests: where the allocations are sized
intentionally a hair under the borders of the size classes. Results
are neutral on the test that intentionally attacks this new scheme by
mixing extent size and lifetime.
The full dump of the performance results can be found here:
https://bur.io/fsperf/size-class-2022-11-15.txt
(there are ANSI escape codes, so best to curl and view in terminal)
Here is a snippet from the full results for a new test which mixes
buffered writes appending to a long lived set of files and large short
lived fallocates:
bufferedappendvsfallocate results
metric baseline current stdev diff
======================================================================================
avg_commit_ms 31.13 29.20 2.67 -6.22%
bg_count 14 15.60 0 11.43%
commits 11.10 12.20 0.32 9.91%
elapsed 27.30 26.40 2.98 -3.30%
end_state_mount_ns 11122551.90 10635118.90 851143.04 -4.38%
end_state_umount_ns 1.36e+09 1.35e+09 12248056.65 -1.07%
find_free_extent_calls 116244.30 114354.30 964.56 -1.63%
find_free_extent_ns_max 599507.20 1047168.20 103337.08 74.67%
find_free_extent_ns_mean 3607.19 3672.11 101.20 1.80%
find_free_extent_ns_min 500 512 6.67 2.40%
find_free_extent_ns_p50 2848 2876 37.65 0.98%
find_free_extent_ns_p95 4916 5000 75.45 1.71%
find_free_extent_ns_p99 20734.49 20920.48 1670.93 0.90%
frag_pct_max 61.67 0 8.05 -100.00%
frag_pct_mean 43.59 0 6.10 -100.00%
frag_pct_min 25.91 0 16.60 -100.00%
frag_pct_p50 42.53 0 7.25 -100.00%
frag_pct_p95 61.67 0 8.05 -100.00%
frag_pct_p99 61.67 0 8.05 -100.00%
fragmented_bg_count 6.10 0 1.45 -100.00%
max_commit_ms 49.80 46 5.37 -7.63%
sys_cpu 2.59 2.62 0.29 1.39%
write_bw_bytes 1.62e+08 1.68e+08 17975843.50 3.23%
write_clat_ns_mean 57426.39 54475.95 2292.72 -5.14%
write_clat_ns_p50 46950.40 42905.60 2101.35 -8.62%
write_clat_ns_p99 148070.40 143769.60 2115.17 -2.90%
write_io_kbytes 4194304 4194304 0 0.00%
write_iops 2476.15 2556.10 274.29 3.23%
write_lat_ns_max 2101667.60 2251129.50 370556.59 7.11%
write_lat_ns_mean 59374.91 55682.00 2523.09 -6.22%
write_lat_ns_min 17353.10 16250 1646.08 -6.36%
There are some mixed improvements/regressions in most metrics along with
an elimination of fragmentation in this workload.
On the balance, the drastic 1->0 improvement in the happy cases seems
worth the mix of regressions and improvements we do observe.
Some considerations for future work:
- Experimenting with more size classes
- More hinting/search ordering work to approximate a best-fit allocator
Signed-off-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
In zoned mode the sequential status of zone can be also tracked in the
runtime flags of block group.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We already have flags in block group to track various status bits,
convert needs_free_space as well and reduce size of btrfs_block_group.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BACKGROUND]
When committing a transaction, we will update block group items for all
dirty block groups.
But in fact, dirty block groups don't always need to update their block
group items.
It's pretty common to have a metadata block group which experienced
several COW operations, but still have the same amount of used bytes.
In that case, we may unnecessarily COW a tree block doing nothing.
[ENHANCEMENT]
This patch will introduce btrfs_block_group::commit_used member to
remember the last used bytes, and use that new member to skip
unnecessary block group item update.
This would be more common for large filesystems, where metadata block
group can be as large as 1GiB, containing at most 64K metadata items.
In that case, if COW added and then deleted one metadata item near the
end of the block group, then it's completely possible we don't need to
touch the block group item at all.
[BENCHMARK]
The change itself can have quite a high chance (20~80%) to skip block
group item updates in lot of workloads.
As a result, it would result shorter time spent on
btrfs_write_dirty_block_groups(), and overall reduce the execution time
of the critical section of btrfs_commit_transaction().
Here comes a fio command, which will do random writes in 4K block size,
causing a very heavy metadata updates.
fio --filename=$mnt/file --size=512M --rw=randwrite --direct=1 --bs=4k \
--ioengine=libaio --iodepth=64 --runtime=300 --numjobs=4 \
--name=random_write --fallocate=none --time_based --fsync_on_close=1
The file size (512M) and number of threads (4) means 2GiB file size in
total, but during the full 300s run time, my dedicated SATA SSD is able
to write around 20~25GiB, which is over 10 times the file size.
Thus after we fill the initial 2G, we should not cause much block group
item updates.
Please note, the fio numbers by themselves don't have much change, but
if we look deeper, there is some reduced execution time, especially for
the critical section of btrfs_commit_transaction().
I added extra trace_printk() to measure the following per-transaction
execution time:
- Critical section of btrfs_commit_transaction()
By re-using the existing update_commit_stats() function, which
has already calculated the interval correctly.
- The while() loop for btrfs_write_dirty_block_groups()
Although this includes the execution time of btrfs_run_delayed_refs(),
it should still be representative overall.
Both result involves transid 7~30, the same amount of transaction
committed.
The result looks like this:
| Before | After | Diff
----------------------+-------------------+----------------+--------
Transaction interval | 229247198.5 | 215016933.6 | -6.2%
Block group interval | 23133.33333 | 18970.83333 | -18.0%
The change in block group item updates is more obvious, as skipped block
group item updates also mean less delayed refs.
And the overall execution time for that block group update loop is
pretty small, thus we can assume the extent tree is already mostly
cached. If we can skip an uncached tree block, it would cause more
obvious change.
Unfortunately the overall reduction in commit transaction critical
section is much smaller, as the block group item updates loop is not
really the major part, at least not for the above fio script.
But still we have a observable reduction in the critical section.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This function uses functions that are not defined in block-group.h, move
it into block-group.c in order to keep the header clean.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.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 actually embedded in struct btrfs_block_group, so move this
definition to block-group.h, and then open-code the init of the tree
where we init the rest of the block group instead of using a helper.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.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 a block group related definition, move it into block-group.h.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
struct btrfs_caching_ctl::progress and struct
btrfs_block_group::last_byte_to_unpin were previously needed to ensure
that unpin_extent_range() didn't return a range to the free space cache
before the caching thread had a chance to cache that range. However, the
commit "btrfs: fix space cache corruption and potential double
allocations" made it so that we always synchronously cache the block
group at the time that we pin the extent, so this machinery is no longer
necessary.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We used to use this in a few spots, but now we only use it directly
inside of block-group.c, so remove the helper and just open code where
we were using it.
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 used mostly to determine if we need to look at the caching ctl
list and clean up any references to this block group. However we never
clear this flag, specifically because we need to know if we have to
remove a caching ctl we have for this block group still. This is in the
remove block group path which isn't a fast path, so the optimization
doesn't really matter, simplify this logic and remove the flag.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We use a bit field in the btrfs_block_group for different flags, however
this is awkward because we have to hold the block_group->lock for any
modification of any of these fields, and makes the code clunky for a few
of these flags. Convert these to a properly flags setup so we can
utilize the bit helpers.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When testing space_cache v2 on a large set of machines, we encountered a
few symptoms:
1. "unable to add free space :-17" (EEXIST) errors.
2. Missing free space info items, sometimes caught with a "missing free
space info for X" error.
3. Double-accounted space: ranges that were allocated in the extent tree
and also marked as free in the free space tree, ranges that were
marked as allocated twice in the extent tree, or ranges that were
marked as free twice in the free space tree. If the latter made it
onto disk, the next reboot would hit the BUG_ON() in
add_new_free_space().
4. On some hosts with no on-disk corruption or error messages, the
in-memory space cache (dumped with drgn) disagreed with the free
space tree.
All of these symptoms have the same underlying cause: a race between
caching the free space for a block group and returning free space to the
in-memory space cache for pinned extents causes us to double-add a free
range to the space cache. This race exists when free space is cached
from the free space tree (space_cache=v2) or the extent tree
(nospace_cache, or space_cache=v1 if the cache needs to be regenerated).
struct btrfs_block_group::last_byte_to_unpin and struct
btrfs_block_group::progress are supposed to protect against this race,
but commit d0c2f4fa55 ("btrfs: make concurrent fsyncs wait less when
waiting for a transaction commit") subtly broke this by allowing
multiple transactions to be unpinning extents at the same time.
Specifically, the race is as follows:
1. An extent is deleted from an uncached block group in transaction A.
2. btrfs_commit_transaction() is called for transaction A.
3. btrfs_run_delayed_refs() -> __btrfs_free_extent() runs the delayed
ref for the deleted extent.
4. __btrfs_free_extent() -> do_free_extent_accounting() ->
add_to_free_space_tree() adds the deleted extent back to the free
space tree.
5. do_free_extent_accounting() -> btrfs_update_block_group() ->
btrfs_cache_block_group() queues up the block group to get cached.
block_group->progress is set to block_group->start.
6. btrfs_commit_transaction() for transaction A calls
switch_commit_roots(). It sets block_group->last_byte_to_unpin to
block_group->progress, which is block_group->start because the block
group hasn't been cached yet.
7. The caching thread gets to our block group. Since the commit roots
were already switched, load_free_space_tree() sees the deleted extent
as free and adds it to the space cache. It finishes caching and sets
block_group->progress to U64_MAX.
8. btrfs_commit_transaction() advances transaction A to
TRANS_STATE_SUPER_COMMITTED.
9. fsync calls btrfs_commit_transaction() for transaction B. Since
transaction A is already in TRANS_STATE_SUPER_COMMITTED and the
commit is for fsync, it advances.
10. btrfs_commit_transaction() for transaction B calls
switch_commit_roots(). This time, the block group has already been
cached, so it sets block_group->last_byte_to_unpin to U64_MAX.
11. btrfs_commit_transaction() for transaction A calls
btrfs_finish_extent_commit(), which calls unpin_extent_range() for
the deleted extent. It sees last_byte_to_unpin set to U64_MAX (by
transaction B!), so it adds the deleted extent to the space cache
again!
This explains all of our symptoms above:
* If the sequence of events is exactly as described above, when the free
space is re-added in step 11, it will fail with EEXIST.
* If another thread reallocates the deleted extent in between steps 7
and 11, then step 11 will silently re-add that space to the space
cache as free even though it is actually allocated. Then, if that
space is allocated *again*, the free space tree will be corrupted
(namely, the wrong item will be deleted).
* If we don't catch this free space tree corruption, it will continue
to get worse as extents are deleted and reallocated.
The v1 space_cache is synchronously loaded when an extent is deleted
(btrfs_update_block_group() with alloc=0 calls btrfs_cache_block_group()
with load_cache_only=1), so it is not normally affected by this bug.
However, as noted above, if we fail to load the space cache, we will
fall back to caching from the extent tree and may hit this bug.
The easiest fix for this race is to also make caching from the free
space tree or extent tree synchronous. Josef tested this and found no
performance regressions.
A few extra changes fall out of this change. Namely, this fix does the
following, with step 2 being the crucial fix:
1. Factor btrfs_caching_ctl_wait_done() out of
btrfs_wait_block_group_cache_done() to allow waiting on a caching_ctl
that we already hold a reference to.
2. Change the call in btrfs_cache_block_group() of
btrfs_wait_space_cache_v1_finished() to
btrfs_caching_ctl_wait_done(), which makes us wait regardless of the
space_cache option.
3. Delete the now unused btrfs_wait_space_cache_v1_finished() and
space_cache_v1_done().
4. Change btrfs_cache_block_group()'s `int load_cache_only` parameter to
`bool wait` to more accurately describe its new meaning.
5. Change a few callers which had a separate call to
btrfs_wait_block_group_cache_done() to use wait = true instead.
6. Make btrfs_wait_block_group_cache_done() static now that it's not
used outside of block-group.c anymore.
Fixes: d0c2f4fa55 ("btrfs: make concurrent fsyncs wait less when waiting for a transaction commit")
CC: stable@vger.kernel.org # 5.12+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
After commit 5f0addf7b8 ("btrfs: zoned: use dedicated lock for data
relocation"), we observe IO errors on e.g, btrfs/232 like below.
[09.0][T4038707] WARNING: CPU: 3 PID: 4038707 at fs/btrfs/extent-tree.c:2381 btrfs_cross_ref_exist+0xfc/0x120 [btrfs]
<snip>
[09.9][T4038707] Call Trace:
[09.5][T4038707] <TASK>
[09.3][T4038707] run_delalloc_nocow+0x7f1/0x11a0 [btrfs]
[09.6][T4038707] ? test_range_bit+0x174/0x320 [btrfs]
[09.2][T4038707] ? fallback_to_cow+0x980/0x980 [btrfs]
[09.3][T4038707] ? find_lock_delalloc_range+0x33e/0x3e0 [btrfs]
[09.5][T4038707] btrfs_run_delalloc_range+0x445/0x1320 [btrfs]
[09.2][T4038707] ? test_range_bit+0x320/0x320 [btrfs]
[09.4][T4038707] ? lock_downgrade+0x6a0/0x6a0
[09.2][T4038707] ? orc_find.part.0+0x1ed/0x300
[09.5][T4038707] ? __module_address.part.0+0x25/0x300
[09.0][T4038707] writepage_delalloc+0x159/0x310 [btrfs]
<snip>
[09.4][ C3] sd 10:0:1:0: [sde] tag#2620 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
[09.5][ C3] sd 10:0:1:0: [sde] tag#2620 Sense Key : Illegal Request [current]
[09.9][ C3] sd 10:0:1:0: [sde] tag#2620 Add. Sense: Unaligned write command
[09.5][ C3] sd 10:0:1:0: [sde] tag#2620 CDB: Write(16) 8a 00 00 00 00 00 02 f3 63 87 00 00 00 2c 00 00
[09.4][ C3] critical target error, dev sde, sector 396041272 op 0x1:(WRITE) flags 0x800 phys_seg 3 prio class 0
[09.9][ C3] BTRFS error (device dm-1): bdev /dev/mapper/dml_102_2 errs: wr 1, rd 0, flush 0, corrupt 0, gen 0
The IO errors occur when we allocate a regular extent in previous data
relocation block group.
On zoned btrfs, we use a dedicated block group to relocate a data
extent. Thus, we allocate relocating data extents (pre-alloc) only from
the dedicated block group and vice versa. Once the free space in the
dedicated block group gets tight, a relocating extent may not fit into
the block group. In that case, we need to switch the dedicated block
group to the next one. Then, the previous one is now freed up for
allocating a regular extent. The BG is already not enough to allocate
the relocating extent, but there is still room to allocate a smaller
extent. Now the problem happens. By allocating a regular extent while
nocow IOs for the relocation is still on-going, we will issue WRITE IOs
(for relocation) and ZONE APPEND IOs (for the regular writes) at the
same time. That mixed IOs confuses the write pointer and arises the
unaligned write errors.
This commit introduces a new bit 'zoned_data_reloc_ongoing' to the
btrfs_block_group. We set this bit before releasing the dedicated block
group, and no extent are allocated from a block group having this bit
set. This bit is similar to setting block_group->ro, but is different from
it by allowing nocow writes to start.
Once all the nocow IO for relocation is done (hooked from
btrfs_finish_ordered_io), we reset the bit to release the block group for
further allocation.
Fixes: c2707a2556 ("btrfs: zoned: add a dedicated data relocation block group")
CC: stable@vger.kernel.org # 5.16+
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Commit be1a1d7a5d ("btrfs: zoned: finish fully written block group")
introduced zone finishing code both for data and metadata end_io path.
However, the metadata side is not working as it should. First, it
compares logical address (eb->start + eb->len) with offset within a
block group (cache->zone_capacity) in submit_eb_page(). That essentially
disabled zone finishing on metadata end_io path.
Furthermore, fixing the issue above revealed we cannot call
btrfs_zone_finish_endio() in end_extent_buffer_writeback(). We cannot
call btrfs_lookup_block_group() which require spin lock inside end_io
context.
Introduce btrfs_schedule_zone_finish_bg() to wait for the extent buffer
writeback and do the zone finish IO in a workqueue.
Also, drop EXTENT_BUFFER_ZONE_FINISH as it is no longer used.
Fixes: be1a1d7a5d ("btrfs: zoned: finish fully written block group")
CC: stable@vger.kernel.org # 5.16+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When doing a NOCOW write, either through direct IO or buffered IO, we do
two lookups for the block group that contains the target extent: once
when we call btrfs_inc_nocow_writers() and then later again when we call
btrfs_dec_nocow_writers() after creating the ordered extent.
The lookups require taking a lock and navigating the red black tree used
to track all block groups, which can take a non-negligible amount of time
for a large filesystem with thousands of block groups, as well as lock
contention and cache line bouncing.
Improve on this by having a single block group search: making
btrfs_inc_nocow_writers() return the block group to its caller and then
have the caller pass that block group to btrfs_dec_nocow_writers().
This is part of a patchset comprised of the following patches:
btrfs: remove search start argument from first_logical_byte()
btrfs: use rbtree with leftmost node cached for tracking lowest block group
btrfs: use a read/write lock for protecting the block groups tree
btrfs: return block group directly at btrfs_next_block_group()
btrfs: avoid double search for block group during NOCOW writes
The following test was used to test these changes from a performance
perspective:
$ cat test.sh
#!/bin/bash
modprobe null_blk nr_devices=0
NULL_DEV_PATH=/sys/kernel/config/nullb/nullb0
mkdir $NULL_DEV_PATH
if [ $? -ne 0 ]; then
echo "Failed to create nullb0 directory."
exit 1
fi
echo 2 > $NULL_DEV_PATH/submit_queues
echo 16384 > $NULL_DEV_PATH/size # 16G
echo 1 > $NULL_DEV_PATH/memory_backed
echo 1 > $NULL_DEV_PATH/power
DEV=/dev/nullb0
MNT=/mnt/nullb0
LOOP_MNT="$MNT/loop"
MOUNT_OPTIONS="-o ssd -o nodatacow"
MKFS_OPTIONS="-R free-space-tree -O no-holes"
cat <<EOF > /tmp/fio-job.ini
[io_uring_writes]
rw=randwrite
fsync=0
fallocate=posix
group_reporting=1
direct=1
ioengine=io_uring
iodepth=64
bs=64k
filesize=1g
runtime=300
time_based
directory=$LOOP_MNT
numjobs=8
thread
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 &> /dev/null
mount $MOUNT_OPTIONS $DEV $MNT
mkdir $LOOP_MNT
truncate -s 4T $MNT/loopfile
mkfs.btrfs -f $MKFS_OPTIONS $MNT/loopfile &> /dev/null
mount $MOUNT_OPTIONS $MNT/loopfile $LOOP_MNT
# Trigger the allocation of about 3500 data block groups, without
# actually consuming space on underlying filesystem, just to make
# the tree of block group large.
fallocate -l 3500G $LOOP_MNT/filler
fio /tmp/fio-job.ini
umount $LOOP_MNT
umount $MNT
echo 0 > $NULL_DEV_PATH/power
rmdir $NULL_DEV_PATH
The test was run on a non-debug kernel (Debian's default kernel config),
the result were the following.
Before patchset:
WRITE: bw=1455MiB/s (1526MB/s), 1455MiB/s-1455MiB/s (1526MB/s-1526MB/s), io=426GiB (458GB), run=300006-300006msec
After patchset:
WRITE: bw=1503MiB/s (1577MB/s), 1503MiB/s-1503MiB/s (1577MB/s-1577MB/s), io=440GiB (473GB), run=300006-300006msec
+3.3% write throughput and +3.3% IO done in the same time period.
The test has somewhat limited coverage scope, as with only NOCOW writes
we get less contention on the red black tree of block groups, since we
don't have the extra contention caused by COW writes, namely when
allocating data extents, pinning and unpinning data extents, but on the
hand there's access to tree in the NOCOW path, when incrementing a block
group's number of NOCOW writers.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In btrfs_make_block_group(), we activate the allocated block group,
expecting that the block group is soon used for allocation. However, the
chunk allocation from flush_space() context broke the assumption. There
can be a large time gap between the chunk allocation time and the extent
allocation time from the chunk.
Activating the empty block groups pre-allocated from flush_space()
context can exhaust the active zone counter of a device. Once we use all
the active zone counts for empty pre-allocated block groups, we cannot
activate new block group for the other things: metadata, tree-log, or
data relocation block group. That failure results in a fake -ENOSPC.
This patch introduces CHUNK_ALLOC_FORCE_FOR_EXTENT to distinguish the
chunk allocation from find_free_extent(). Now, the new block group is
activated only in that context.
Fixes: eb66a010d5 ("btrfs: zoned: activate new block group")
CC: stable@vger.kernel.org # 5.16+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Tested-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
With extent tree v2 you will be able to create multiple csum, extent,
and free space trees. They will be used based on the block group, which
will now use the block_group_item->chunk_objectid to point to the set of
global roots that it will use. When allocating new block groups we'll
simply mod the gigabyte offset of the block group against the number of
global roots we have and that will be the block groups global id.
>From there we can take the bytenr that we're modifying in the respective
tree, look up the block group and get that block groups corresponding
global root id. From there we can get to the appropriate global root
for that bytenr.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When a task is doing some modification to the chunk btree and it is not in
the context of a chunk allocation or a chunk removal, it can deadlock with
another task that is currently allocating a new data or metadata chunk.
These contexts are the following:
* When relocating a system chunk, when we need to COW the extent buffers
that belong to the chunk btree;
* When adding a new device (ioctl), where we need to add a new device item
to the chunk btree;
* When removing a device (ioctl), where we need to remove a device item
from the chunk btree;
* When resizing a device (ioctl), where we need to update a device item in
the chunk btree and may need to relocate a system chunk that lies beyond
the new device size when shrinking a device.
The problem happens due to a sequence of steps like the following:
1) Task A starts a data or metadata chunk allocation and it locks the
chunk mutex;
2) Task B is relocating a system chunk, and when it needs to COW an extent
buffer of the chunk btree, it has locked both that extent buffer as
well as its parent extent buffer;
3) Since there is not enough available system space, either because none
of the existing system block groups have enough free space or because
the only one with enough free space is in RO mode due to the relocation,
task B triggers a new system chunk allocation. It blocks when trying to
acquire the chunk mutex, currently held by task A;
4) Task A enters btrfs_chunk_alloc_add_chunk_item(), in order to insert
the new chunk item into the chunk btree and update the existing device
items there. But in order to do that, it has to lock the extent buffer
that task B locked at step 2, or its parent extent buffer, but task B
is waiting on the chunk mutex, which is currently locked by task A,
therefore resulting in a deadlock.
One example report when the deadlock happens with system chunk relocation:
INFO: task kworker/u9:5:546 blocked for more than 143 seconds.
Not tainted 5.15.0-rc3+ #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:kworker/u9:5 state:D stack:25936 pid: 546 ppid: 2 flags:0x00004000
Workqueue: events_unbound btrfs_async_reclaim_metadata_space
Call Trace:
context_switch kernel/sched/core.c:4940 [inline]
__schedule+0xcd9/0x2530 kernel/sched/core.c:6287
schedule+0xd3/0x270 kernel/sched/core.c:6366
rwsem_down_read_slowpath+0x4ee/0x9d0 kernel/locking/rwsem.c:993
__down_read_common kernel/locking/rwsem.c:1214 [inline]
__down_read kernel/locking/rwsem.c:1223 [inline]
down_read_nested+0xe6/0x440 kernel/locking/rwsem.c:1590
__btrfs_tree_read_lock+0x31/0x350 fs/btrfs/locking.c:47
btrfs_tree_read_lock fs/btrfs/locking.c:54 [inline]
btrfs_read_lock_root_node+0x8a/0x320 fs/btrfs/locking.c:191
btrfs_search_slot_get_root fs/btrfs/ctree.c:1623 [inline]
btrfs_search_slot+0x13b4/0x2140 fs/btrfs/ctree.c:1728
btrfs_update_device+0x11f/0x500 fs/btrfs/volumes.c:2794
btrfs_chunk_alloc_add_chunk_item+0x34d/0xea0 fs/btrfs/volumes.c:5504
do_chunk_alloc fs/btrfs/block-group.c:3408 [inline]
btrfs_chunk_alloc+0x84d/0xf50 fs/btrfs/block-group.c:3653
flush_space+0x54e/0xd80 fs/btrfs/space-info.c:670
btrfs_async_reclaim_metadata_space+0x396/0xa90 fs/btrfs/space-info.c:953
process_one_work+0x9df/0x16d0 kernel/workqueue.c:2297
worker_thread+0x90/0xed0 kernel/workqueue.c:2444
kthread+0x3e5/0x4d0 kernel/kthread.c:319
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295
INFO: task syz-executor:9107 blocked for more than 143 seconds.
Not tainted 5.15.0-rc3+ #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:syz-executor state:D stack:23200 pid: 9107 ppid: 7792 flags:0x00004004
Call Trace:
context_switch kernel/sched/core.c:4940 [inline]
__schedule+0xcd9/0x2530 kernel/sched/core.c:6287
schedule+0xd3/0x270 kernel/sched/core.c:6366
schedule_preempt_disabled+0xf/0x20 kernel/sched/core.c:6425
__mutex_lock_common kernel/locking/mutex.c:669 [inline]
__mutex_lock+0xc96/0x1680 kernel/locking/mutex.c:729
btrfs_chunk_alloc+0x31a/0xf50 fs/btrfs/block-group.c:3631
find_free_extent_update_loop fs/btrfs/extent-tree.c:3986 [inline]
find_free_extent+0x25cb/0x3a30 fs/btrfs/extent-tree.c:4335
btrfs_reserve_extent+0x1f1/0x500 fs/btrfs/extent-tree.c:4415
btrfs_alloc_tree_block+0x203/0x1120 fs/btrfs/extent-tree.c:4813
__btrfs_cow_block+0x412/0x1620 fs/btrfs/ctree.c:415
btrfs_cow_block+0x2f6/0x8c0 fs/btrfs/ctree.c:570
btrfs_search_slot+0x1094/0x2140 fs/btrfs/ctree.c:1768
relocate_tree_block fs/btrfs/relocation.c:2694 [inline]
relocate_tree_blocks+0xf73/0x1770 fs/btrfs/relocation.c:2757
relocate_block_group+0x47e/0xc70 fs/btrfs/relocation.c:3673
btrfs_relocate_block_group+0x48a/0xc60 fs/btrfs/relocation.c:4070
btrfs_relocate_chunk+0x96/0x280 fs/btrfs/volumes.c:3181
__btrfs_balance fs/btrfs/volumes.c:3911 [inline]
btrfs_balance+0x1f03/0x3cd0 fs/btrfs/volumes.c:4301
btrfs_ioctl_balance+0x61e/0x800 fs/btrfs/ioctl.c:4137
btrfs_ioctl+0x39ea/0x7b70 fs/btrfs/ioctl.c:4949
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:874 [inline]
__se_sys_ioctl fs/ioctl.c:860 [inline]
__x64_sys_ioctl+0x193/0x200 fs/ioctl.c:860
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae
So fix this by making sure that whenever we try to modify the chunk btree
and we are neither in a chunk allocation context nor in a chunk remove
context, we reserve system space before modifying the chunk btree.
Reported-by: Hao Sun <sunhao.th@gmail.com>
Link: https://lore.kernel.org/linux-btrfs/CACkBjsax51i4mu6C0C3vJqQN3NR_iVuucoeG3U1HXjrgzn5FFQ@mail.gmail.com/
Fixes: 79bd37120b ("btrfs: rework chunk allocation to avoid exhaustion of the system chunk array")
CC: stable@vger.kernel.org # 5.14+
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_update_block_group() accounts for the number of bytes allocated or
freed. Argument @alloc specifies whether the call is for alloc or free.
Convert the argument @alloc type from int to bool.
Reviewed-by: Su Yue <l@damenly.su>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add zone_is_active flag to btrfs_block_group. This flag indicates the
underlying zones are all active. Such zone active block groups are tracked
by fs_info->active_bg_list.
btrfs_dev_{set,clear}_active_zone() take responsibility for the underlying
device part. They set/clear the bitmap to indicate zone activeness and
count the number of zones we can activate left.
btrfs_zone_{activate,finish}() take responsibility for the logical part and
the list management. In addition, btrfs_zone_finish() wait for any writes
on it and send REQ_OP_ZONE_FINISH to the zone.
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We will use a block group's physical location to track active zones and
finish fully written zones in the following commits. Since the zone
activation is done in the extent allocation context which already holding
the tree locks, we can't query the chunk tree for the physical locations.
So, copy the location info into a block group and use it for activation.
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The ZNS specification introduces the concept of a Zone Capacity. A zone
capacity is an additional per-zone attribute that indicates the number of
usable logical blocks within each zone, starting from the first logical
block of each zone. It is always smaller or equal to the zone size.
With the SINGLE profile, we can set a block group's "capacity" as the same
as the underlying zone's Zone Capacity. We will limit the allocation not
to exceed in a following commit.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Commit eafa4fd0ad ("btrfs: fix exhaustion of the system chunk array
due to concurrent allocations") fixed a problem that resulted in
exhausting the system chunk array in the superblock when there are many
tasks allocating chunks in parallel. Basically too many tasks enter the
first phase of chunk allocation without previous tasks having finished
their second phase of allocation, resulting in too many system chunks
being allocated. That was originally observed when running the fallocate
tests of stress-ng on a PowerPC machine, using a node size of 64K.
However that commit also introduced a deadlock where a task in phase 1 of
the chunk allocation waited for another task that had allocated a system
chunk to finish its phase 2, but that other task was waiting on an extent
buffer lock held by the first task, therefore resulting in both tasks not
making any progress. That change was later reverted by a patch with the
subject "btrfs: fix deadlock with concurrent chunk allocations involving
system chunks", since there is no simple and short solution to address it
and the deadlock is relatively easy to trigger on zoned filesystems, while
the system chunk array exhaustion is not so common.
This change reworks the chunk allocation to avoid the system chunk array
exhaustion. It accomplishes that by making the first phase of chunk
allocation do the updates of the device items in the chunk btree and the
insertion of the new chunk item in the chunk btree. This is done while
under the protection of the chunk mutex (fs_info->chunk_mutex), in the
same critical section that checks for available system space, allocates
a new system chunk if needed and reserves system chunk space. This way
we do not have chunk space reserved until the second phase completes.
The same logic is applied to chunk removal as well, since it keeps
reserved system space long after it is done updating the chunk btree.
For direct allocation of system chunks, the previous behaviour remains,
because otherwise we would deadlock on extent buffers of the chunk btree.
Changes to the chunk btree are by large done by chunk allocation and chunk
removal, which first reserve chunk system space and then later do changes
to the chunk btree. The other remaining cases are uncommon and correspond
to adding a device, removing a device and resizing a device. All these
other cases do not pre-reserve system space, they modify the chunk btree
right away, so they don't hold reserved space for a long period like chunk
allocation and chunk removal do.
The diff of this change is huge, but more than half of it is just addition
of comments describing both how things work regarding chunk allocation and
removal, including both the new behavior and the parts of the old behavior
that did not change.
CC: stable@vger.kernel.org # 5.12+
Tested-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com>
Tested-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Tested-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When a file gets deleted on a zoned file system, the space freed is not
returned back into the block group's free space, but is migrated to
zone_unusable.
As this zone_unusable space is behind the current write pointer it is not
possible to use it for new allocations. In the current implementation a
zone is reset once all of the block group's space is accounted as zone
unusable.
This behaviour can lead to premature ENOSPC errors on a busy file system.
Instead of only reclaiming the zone once it is completely unusable,
kick off a reclaim job once the amount of unusable bytes exceeds a user
configurable threshold between 51% and 100%. It can be set per mounted
filesystem via the sysfs tunable bg_reclaim_threshold which is set to 75%
by default.
Similar to reclaiming unused block groups, these dirty block groups are
added to a to_reclaim list and then on a transaction commit, the reclaim
process is triggered but after we deleted unused block groups, which will
free space for the relocation process.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When we active a swap file, at btrfs_swap_activate(), we acquire the
exclusive operation lock to prevent the physical location of the swap
file extents to be changed by operations such as balance and device
replace/resize/remove. We also call there can_nocow_extent() which,
among other things, checks if the block group of a swap file extent is
currently RO, and if it is we can not use the extent, since a write
into it would result in COWing the extent.
However we have no protection against a scrub operation running after we
activate the swap file, which can result in the swap file extents to be
COWed while the scrub is running and operating on the respective block
group, because scrub turns a block group into RO before it processes it
and then back again to RW mode after processing it. That means an attempt
to write into a swap file extent while scrub is processing the respective
block group, will result in COWing the extent, changing its physical
location on disk.
Fix this by making sure that block groups that have extents that are used
by active swap files can not be turned into RO mode, therefore making it
not possible for a scrub to turn them into RO mode. When a scrub finds a
block group that can not be turned to RO due to the existence of extents
used by swap files, it proceeds to the next block group and logs a warning
message that mentions the block group was skipped due to active swap
files - this is the same approach we currently use for balance.
Fixes: ed46ff3d42 ("Btrfs: support swap files")
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When a bad checksum is found and if the filesystem has a mirror of the
damaged data, we read the correct data from the mirror and writes it to
damaged blocks. This however, violates the sequential write constraints
of a zoned block device.
We can consider three methods to repair an IO failure in zoned filesystems:
(1) Reset and rewrite the damaged zone
(2) Allocate new device extent and replace the damaged device extent to
the new extent
(3) Relocate the corresponding block group
Method (1) is most similar to a behavior done with regular devices.
However, it also wipes non-damaged data in the same device extent, and
so it unnecessary degrades non-damaged data.
Method (2) is much like device replacing but done in the same device. It
is safe because it keeps the device extent until the replacing finish.
However, extending device replacing is non-trivial. It assumes
"src_dev->physical == dst_dev->physical". Also, the extent mapping
replacing function should be extended to support replacing device extent
position in one device.
Method (3) invokes relocation of the damaged block group and is
straightforward to implement. It relocates all the mirrored device
extents, so it potentially is a more costly operation than method (1) or
(2). But it relocates only used extents which reduce the total IO size.
Let's apply method (3) for now. In the future, we can extend device-replace
and apply method (2).
For protecting a block group gets relocated multiple time with multiple
IO errors, this commit introduces "relocating_repair" bit to show it's
now relocating to repair IO failures. Also it uses a new kthread
"btrfs-relocating-repair", not to block IO path with relocating process.
This commit also supports repairing in the scrub process.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This is the 1/4 patch to support device-replace on zoned filesystems.
We have two types of IOs during the device replace process. One is an IO
to "copy" (by the scrub functions) all the device extents from the source
device to the destination device. The other one is an IO to "clone" (by
handle_ops_on_dev_replace()) new incoming write IOs from users to the
source device into the target device.
Cloning incoming IOs can break the sequential write rule in on target
device. When a write is mapped in the middle of a block group, the IO is
directed to the middle of a target device zone, which breaks the
sequential write requirement.
However, the cloning function cannot be disabled since incoming IOs
targeting already copied device extents must be cloned so that the IO is
executed on the target device.
We cannot use dev_replace->cursor_{left,right} to determine whether a bio
is going to a not yet copied region. Since we have a time gap between
finishing btrfs_scrub_dev() and rewriting the mapping tree in
btrfs_dev_replace_finishing(), we can have a newly allocated device extent
which is never cloned nor copied.
So the point is to copy only already existing device extents. This patch
introduces mark_block_group_to_copy() to mark existing block groups as a
target of copying. Then, handle_ops_on_dev_replace() and dev-replace can
check the flag to do their job.
Also, btrfs_finish_block_group_to_copy() will check if the copied stripe
is the last stripe in the block group. With the last stripe copied,
the to_copy flag is finally disabled. Afterwards we can safely clone
incoming IOs on this block group.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We cannot use zone append for writing metadata, because the B-tree nodes
have references to each other using logical address. Without knowing
the address in advance, we cannot construct the tree in the first place.
So we need to serialize write IOs for metadata.
We cannot add a mutex around allocation and submission because metadata
blocks are allocated in an earlier stage to build up B-trees.
Add a zoned_meta_io_lock and hold it during metadata IO submission in
btree_write_cache_pages() to serialize IOs.
Furthermore, this adds a per-block group metadata IO submission pointer
"meta_write_pointer" to ensure sequential writing, which can break when
attempting to write back blocks in an unfinished transaction. If the
writing out failed because of a hole and the write out is for data
integrity (WB_SYNC_ALL), it returns EAGAIN.
A caller like fsync() code should handle this properly e.g. by falling
back to a full transaction commit.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
On a zoned filesystem, cache if a block group is on a sequential write
only zone.
On sequential write only zones, we can use REQ_OP_ZONE_APPEND for
writing data, therefore provide btrfs_use_zone_append() to figure out if
IO is targeting a sequential write only zone and we can use
REQ_OP_ZONE_APPEND for data writing.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_rmap_block currently reverse-maps the physical addresses on all
devices to the corresponding logical addresses.
Extend the function to match to a specified device. The old functionality
of querying all devices is left intact by specifying NULL as target
device.
A block_device instead of a btrfs_device is passed into btrfs_rmap_block,
as this function is intended to reverse-map the result of a bio, which
only has a block_device.
Also export the function for later use.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In a zoned filesystem a once written then freed region is not usable
until the underlying zone has been reset. So we need to distinguish such
unusable space from usable free space.
Therefore we need to introduce the "zone_unusable" field to the block
group structure, and "bytes_zone_unusable" to the space_info structure
to track the unusable space.
Pinned bytes are always reclaimed to the unusable space. But, when an
allocated region is returned before using e.g., the block group becomes
read-only between allocation time and reservation time, we can safely
return the region to the block group. For the situation, this commit
introduces "btrfs_add_free_space_unused". This behaves the same as
btrfs_add_free_space() on regular filesystem. On zoned filesystems, it
rewinds the allocation offset.
Because the read-only bytes tracks free but unusable bytes when the block
group is read-only, we need to migrate the zone_unusable bytes to
read-only bytes when a block group is marked read-only.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
A zoned filesystem must allocate blocks at the zones' write pointer. The
device's write pointer position can be mapped to a logical address within
a block group. To facilitate this, add an "alloc_offset" to the
block-group to track the logical addresses of the write pointer.
This logical address is populated in btrfs_load_block_group_zone_info()
from the write pointers of corresponding zones.
For now, zoned filesystems the single profile. Supporting non-single
profile with zone append writing is not trivial. For example, in the DUP
profile, we send a zone append writing IO to two zones on a device. The
device reply with written LBAs for the IOs. If the offsets of the
returned addresses from the beginning of the zone are different, then it
results in different logical addresses.
We need fine-grained logical to physical mapping to support such separated
physical address issue. Since it should require additional metadata type,
disable non-single profiles for now.
This commit supports the case all the zones in a block group are
sequential. The next patch will handle the case having a conventional
zone.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Anand Jain <anand.jain@oracle.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 documenting the usage of the commit_root_sem, I noticed that we do
not actually take the commit_root_sem in the case of the free space
cache. This is problematic because we're supposed to hold that sem
while we're reading the commit roots, which is what we do for the free
space cache.
The reason I did it inline when I originally wrote the code was because
there's the case of unpinning where we need to make sure that the free
space cache is loaded if we're going to use the free space cache. But
we can accomplish the same thing by simply waiting for the cache to be
loaded.
Rework this code to load the free space cache asynchronously. This
allows us to greatly cleanup the caching code because now it's all
shared by the various caching methods. We also are now in a position to
have the commit_root semaphore held while we're loading the free space
cache. And finally our modification of ->last_byte_to_unpin is removed
because it can be handled in the proper way on commit.
Some care must be taken when replaying the log, when we expect that the
free space cache will be read entirely before we start excluding space
to replay. This could lead to overwriting space during replay.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have refcount_t now with the associated library to handle refcounts,
which gives us extra debugging around reference count mistakes that may
be made. For example it'll warn on any transition from 0->1 or 0->-1,
which is handy for noticing cases where we've messed up reference
counting. Convert the block group ref counting from an atomic_t to
refcount_t and use the appropriate helpers.
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 helpers btrfs_freeze_block_group() and btrfs_unfreeze_block_group()
used to be named btrfs_get_block_group_trimming() and
btrfs_put_block_group_trimming() respectively.
At the time they were added to free-space-cache.c, by commit e33e17ee10
("btrfs: add missing discards when unpinning extents with -o discard")
because all the trimming related functions were in free-space-cache.c.
Now that the helpers were renamed and are used in scrub context as well,
move them to block-group.c, a much more logical location for them.
Signed-off-by: Filipe Manana <fdmanana@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>
It's used only during initial block group reading to map physical
address of super block to a list of logical ones. Make it private to
block-group.c, add proper kernel doc and ensure it's exported only for
tests.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
As mentioned earlier, discarding data can be done either by issuing an
explicit discard or implicitly by reusing the LBA. Metadata block_groups
see much more frequent reuse due to well it being metadata. So instead
of explicitly discarding metadata block_groups, just leave them be and
let the latter implicit discarding be done for them.
For mixed block_groups, block_groups which contain both metadata and
data, we let them be as higher fragmentation is expected.
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>
The prior two patches added discarding via a background workqueue. This
just piggybacked off of the fstrim code to trim the whole block at once.
Well inevitably this is worse performance wise and will aggressively
overtrim. But it was nice to plumb the other infrastructure to keep the
patches easier to review.
This adds the real goal of this series which is discarding slowly (ie. a
slow long running fstrim). The discarding is split into two phases,
extents and then bitmaps. The reason for this is two fold. First, the
bitmap regions overlap the extent regions. Second, discarding the
extents first will let the newly trimmed bitmaps have the highest chance
of coalescing when being readded to the free space cache.
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>
[BUG]
When running btrfs/072 with only one online CPU, it has a pretty high
chance to fail:
btrfs/072 12s ... _check_dmesg: something found in dmesg (see xfstests-dev/results//btrfs/072.dmesg)
- output mismatch (see xfstests-dev/results//btrfs/072.out.bad)
--- tests/btrfs/072.out 2019-10-22 15:18:14.008965340 +0800
+++ /xfstests-dev/results//btrfs/072.out.bad 2019-11-14 15:56:45.877152240 +0800
@@ -1,2 +1,3 @@
QA output created by 072
Silence is golden
+Scrub find errors in "-m dup -d single" test
...
And with the following call trace:
BTRFS info (device dm-5): scrub: started on devid 1
------------[ cut here ]------------
BTRFS: Transaction aborted (error -27)
WARNING: CPU: 0 PID: 55087 at fs/btrfs/block-group.c:1890 btrfs_create_pending_block_groups+0x3e6/0x470 [btrfs]
CPU: 0 PID: 55087 Comm: btrfs Tainted: G W O 5.4.0-rc1-custom+ #13
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
RIP: 0010:btrfs_create_pending_block_groups+0x3e6/0x470 [btrfs]
Call Trace:
__btrfs_end_transaction+0xdb/0x310 [btrfs]
btrfs_end_transaction+0x10/0x20 [btrfs]
btrfs_inc_block_group_ro+0x1c9/0x210 [btrfs]
scrub_enumerate_chunks+0x264/0x940 [btrfs]
btrfs_scrub_dev+0x45c/0x8f0 [btrfs]
btrfs_ioctl+0x31a1/0x3fb0 [btrfs]
do_vfs_ioctl+0x636/0xaa0
ksys_ioctl+0x67/0x90
__x64_sys_ioctl+0x43/0x50
do_syscall_64+0x79/0xe0
entry_SYSCALL_64_after_hwframe+0x49/0xbe
---[ end trace 166c865cec7688e7 ]---
[CAUSE]
The error number -27 is -EFBIG, returned from the following call chain:
btrfs_end_transaction()
|- __btrfs_end_transaction()
|- btrfs_create_pending_block_groups()
|- btrfs_finish_chunk_alloc()
|- btrfs_add_system_chunk()
This happens because we have used up all space of
btrfs_super_block::sys_chunk_array.
The root cause is, we have the following bad loop of creating tons of
system chunks:
1. The only SYSTEM chunk is being scrubbed
It's very common to have only one SYSTEM chunk.
2. New SYSTEM bg will be allocated
As btrfs_inc_block_group_ro() will check if we have enough space
after marking current bg RO. If not, then allocate a new chunk.
3. New SYSTEM bg is still empty, will be reclaimed
During the reclaim, we will mark it RO again.
4. That newly allocated empty SYSTEM bg get scrubbed
We go back to step 2, as the bg is already mark RO but still not
cleaned up yet.
If the cleaner kthread doesn't get executed fast enough (e.g. only one
CPU), then we will get more and more empty SYSTEM chunks, using up all
the space of btrfs_super_block::sys_chunk_array.
[FIX]
Since scrub/dev-replace doesn't always need to allocate new extent,
especially chunk tree extent, so we don't really need to do chunk
pre-allocation.
To break above spiral, here we introduce a new parameter to
btrfs_inc_block_group(), @do_chunk_alloc, which indicates whether we
need extra chunk pre-allocation.
For relocation, we pass @do_chunk_alloc=true, while for scrub, we pass
@do_chunk_alloc=false.
This should keep unnecessary empty chunks from popping up for scrub.
Also, since there are two parameters for btrfs_inc_block_group_ro(),
add more comment for it.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@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>
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>
The members ::used and ::flags are now in the block group cache
structure, the last one is chunk_objectid, but that's set to a fixed
value and otherwise unused. The item is constructed from a local
variable before write, so we can remove the embedded one from block
group.
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>
Filipe Manana