There is a bug report that fstrim crashed, and that crash is eventually
pinned down to a missing device which re-appeared and screwed up callers
that only checks BTRFS_DEV_STATE_MISSING, but not
BTRFS_DEV_STATE_WRITEABLE nor device->bdev.
A missing device re-appearing can be very tricky, as for now it will
result in a device without WRITEABLE or MISSING flag, and still no bdev
pointer.
As the first step to enhance handling of such re-appearing missing
devices, add a dmesg output when a missing device re-appeared.
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
There is a bug report that btrfs/242 can randomly fail with the
following NULL pointer dereference:
run fstests btrfs/242 at 2026-06-01 10:25:08
BTRFS: device fsid d4d7f234-487c-4787-88e4-47a8b68c9874 devid 1 transid 9 /dev/sdc (8:32) scanned by mount (122609)
BTRFS info (device sdc): first mount of filesystem d4d7f234-487c-4787-88e4-47a8b68c9874
BTRFS info (device sdc): using crc32c checksum algorithm
BTRFS warning (device sdc): devid 2 uuid fbe72d72-3272-482d-80fb-ab88ed398192 is missing
BTRFS warning (device sdc): devid 2 uuid fbe72d72-3272-482d-80fb-ab88ed398192 is missing
BTRFS info (device sdc): allowing degraded mounts
BTRFS info (device sdc): turning on async discard
BTRFS info (device sdc): enabling free space tree
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000018
user pgtable: 4k pages, 48-bit VAs, pgdp=000000013fd6b000
CPU: 4 UID: 0 PID: 122625 Comm: fstrim Not tainted 7.0.10-2-default #1 PREEMPT(full) openSUSE Tumbleweed e9a5f6b24978fba3bf015a992f865837fdfff3dd
Hardware name: QEMU KVM Virtual Machine, BIOS edk2-20250812-19.fc42 08/12/2025
pstate: 01400005 (nzcv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=--)
pc : btrfs_trim_fs+0x34c/0xa00 [btrfs]
lr : btrfs_trim_fs+0x1f0/0xa00 [btrfs]
Call trace:
btrfs_trim_fs+0x34c/0xa00 [btrfs f02c1d570ceea621c69d302ba75dd61868083840] (P)
btrfs_ioctl_fitrim+0xe8/0x178 [btrfs f02c1d570ceea621c69d302ba75dd61868083840]
btrfs_ioctl+0xdd4/0x2bd8 [btrfs f02c1d570ceea621c69d302ba75dd61868083840]
__arm64_sys_ioctl+0xac/0x108
invoke_syscall.constprop.0+0x5c/0xd0
el0_svc_common.constprop.0+0x40/0xf0
do_el0_svc+0x24/0x40
el0_svc+0x40/0x1d0
el0t_64_sync_handler+0xa0/0xe8
el0t_64_sync+0x1b0/0x1b8
Code: 17ffff83 f94017e0 f9002be0 f9402ea0 (f9400c00)
---[ end trace 0000000000000000 ]---
Also the reporter is very kind to test the following ASSERT() added to
btrfs_trim_free_extents_throttle():
ASSERT(device->bdev,
"devid=%llu path=%s dev_state=0x%lx\n",
device->devid, btrfs_dev_name(device), device->dev_state);
And it shows the following output:
assertion failed: device->bdev, in extent-tree.c:6630 (devid=2 path=/dev/sdd dev_state=0x82)
Which means the device->bdev is NULL, and the dev_state is
BTRFS_DEV_STATE_IN_FS_METADATA | BTRFS_DEV_STATE_ITEM_FOUND, without
BTRFS_DEV_STATE_WRITEABLE flag set.
[CAUSE]
The pc points to the following call chain:
btrfs_trim_fs()
|- btrfs_trim_free_extents()
|- btrfs_trim_free_extents_throttle()
|- bdev_max_discard_sectors(device->bdev)
So the NULL pointer dereference is caused by device->bdev being NULL.
This looks impossible by a quick glance, as just before calling
btrfs_trim_free_extents_throttle(), we have skipped any device that has
BTRFS_DEV_STATE_MISSING flag set.
However in this particular case, there is a window where the missing
device is later re-scanned, causing btrfs to remove the
BTRFS_DEV_STATE_MISSING flag:
btrfs_control_ioctl()
|- btrfs_scan_one_device()
|- device_list_add()
|- rcu_assign_pointer(device->name, name);
| This updates the missing device's path to the new good path.
|
|- clear_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state)
This removes the BTRFS_DEV_STATE_MISSING flag.
This allows the missing device to re-appear and clear the
BTRFS_DEV_STATE_MISSING flag. However the device still does not have
the BTRFS_DEV_STATE_WRITEABLE flag set, nor is its bdev pointer updated.
The bdev pointer remains NULL, triggering the crash later.
[FIX]
This is a big de-synchronization between BTRFS_DEV_STATE_MISSING and
device->bdev pointer, and shows a gap in btrfs's re-appearing-device
handling.
The proper handling of re-appearing device will need quite some extra
work, which is out of the context of this small fix.
Thankfully the regular bbio submission path has already handled it well
by checking if the device->bdev is NULL before submitting.
So here we just fix the crash by checking if the device is writeable and
has a bdev pointer before calling bdev_max_discard_sectors().
Reported-by: Su Yue <glass.su@suse.com>
Link: https://lore.kernel.org/linux-btrfs/wlwir19t.fsf@damenly.org/
Fixes: 499f377f49 ("btrfs: iterate over unused chunk space in FITRIM")
CC: stable@vger.kernel.org # 5.10+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If we fail to lookup the dir item, we are always returning -ENOENT but
that may not be the reason for the failure, as btrfs_lookup_dir_item() can
return many different errors, such as -EIO or -ENOMEM for example.
Fix this by returning the real error, and also fixup the silly error
message, including the id of the directory and the error.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There's no need to exclusively lock the mapping, shared locking is enough
to protect from a concurrent set block size operation (BLKBSZSET ioctl).
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Under heavy memcg-driven slab reclaim with many memcgs and CPUs,
shrink_slab_memcg() invokes the per-superblock count callback once per
(memcg, NUMA node) tuple. For btrfs that callback reaches
percpu_counter_sum_positive() on fs_info->evictable_extent_maps, which
takes the percpu_counter's raw spinlock with IRQs disabled and walks
every online CPU. With hundreds of memcgs driving reclaim on a host with
dozens of CPUs, this counter lock becomes a global serialization point:
profiles show CPU pinned in the spin_lock_irqsave acquire under
__percpu_counter_sum, with cross-CPU IPIs hitting csd_lock_wait_toolong
while waiting for spinning vCPUs.
The shrinker count is advisory -- super_cache_count() already notes
"counts can change between super_cache_count and super_cache_scan, so we
really don't need locks here." Use percpu_counter_read_positive(), which
is lockless. Worst-case skew is bounded by batch * num_online_cpus (a
few thousand), negligible compared to the millions of extent maps a busy
filesystem accumulates and well within the noise that the shrinker
already tolerates.
Tested-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Shakeel Butt <shakeel.butt@linux.dev>
Signed-off-by: Ben Maurer <bmaurer@meta.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Convert more multiplications of sectorsize or nodesize to use the
shifts. The remaining cases are multiplications by constants that
compiler can optimize by itself, and in tests.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
Extent buffer pages allocated by alloc_extent_buffer() are attached to
btree_inode->i_mapping (the buffer_tree path), reach the LRU, and are
served by the btree_migrate_folio aops in fs/btrfs/disk-io.c. They are
migratable in practice once their owning extent buffer hits refs == 1,
which happens naturally. The buddy allocator classifies them by GFP,
however, and bare GFP_NOFS lands them in MIGRATE_UNMOVABLE pageblocks.
The result: every btree_inode page we read in pins an unmovable pageblock
from the page-superblock allocator's perspective, even though the page
itself can be moved.
Have each caller of btrfs_alloc_page_array, btrfs_alloc_folio_array,
and alloc_eb_folio_array pass in the full GFP mask directly, instead
of having the functions calculate it from boolean flags.
The alloc_extent_buffer call site passes GFP_NOFS | __GFP_NOFAIL |
__GFP_MOVABLE. All other call sites pass plain GFP_NOFS.
Three categories of caller stay on bare GFP_NOFS, deliberately:
- alloc_dummy_extent_buffer / btrfs_clone_extent_buffer: the
resulting eb is EXTENT_BUFFER_UNMAPPED, folio->mapping stays NULL,
the folios never enter LRU, never get migrate_folio aops. Tagging
them __GFP_MOVABLE would violate the page allocator's migrability
contract and they would defeat compaction in MOVABLE pageblocks
where isolate_migratepages_block skips non-LRU non-movable_ops
pages outright.
- btrfs_alloc_page_array callers in fs/btrfs/raid56.c (stripe
pages), fs/btrfs/inode.c (encoded reads), fs/btrfs/ioctl.c (io_uring
encoded reads), fs/btrfs/relocation.c (relocation buffers): same
contract violation. raid56 stripe_pages additionally persist in
the stripe cache (RBIO_CACHE_SIZE=1024) well beyond a single I/O,
so they are not transient enough to hand-wave the contract.
- btrfs_alloc_folio_array caller in fs/btrfs/scrub.c (stripe
folios): same -- stripe->folios[] are private buffers freed via
folio_put in release_scrub_stripe.
This change targets the dominant fragmentation source observed on the
page-superblock series: ~28 GB of btree_inode pages parked across
many tainted superpageblocks on a 250 GB test system with btrfs root,
preventing 1 GiB hugepage allocation from those regions. With the
movable hint, those pages now land in MOVABLE pageblocks where the
existing background defragger drains them through the standard
PB_has_movable gate, no LRU-sample fallback needed.
Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Rik van Riel <riel@surriel.com>
Signed-off-by: David Sterba <dsterba@suse.com>
On 64-bit kernels with 32-bit userspace, struct btrfs_ioctl_timespec is
laid out as 16 bytes (8B sec + 4B nsec + 4B trailing padding) instead of
the 12 bytes a 32-bit userspace expects, because the surrounding struct
is not packed. As a result, struct btrfs_ioctl_get_subvol_info_args has
a different size and layout in 32-bit userspace than in the 64-bit
kernel, and BTRFS_IOC_GET_SUBVOL_INFO returns garbage to 32-bit callers.
Mirror what was done for BTRFS_IOC_SET_RECEIVED_SUBVOL: add a packed
btrfs_ioctl_get_subvol_info_args_32 with btrfs_ioctl_timespec_32 fields,
define BTRFS_IOC_GET_SUBVOL_INFO_32 with that struct as the size
argument, factor the existing handler into a shared _btrfs_ioctl_get_
subvol_info() helper, and add btrfs_ioctl_get_subvol_info_32() which
fills the kernel struct and translates field-by-field into the 32-bit
struct before copy_to_user().
Signed-off-by: Daan De Meyer <daan@amutable.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The f_fsid was originally derived from fs_devices->fsid and the
subvolume root ID. However, when temp_fsid is active, fs_devices->fsid
is randomized, making the standard derivation inconsistent.
Since metadata_uuid is optional, it is not a reliable alternative. This
patch instead retrieves the on-disk UUID from fs_info->super_copy->fsid.
To prevent f_fsid collisions between original and cloned filesystems,
this implementation hashes the dev_t for single-device btrfs filesystems
to ensure uniqueness. This is limited to single-device filesystems as
cloned mounts are currently only supported for that configuration. Note
that f_fsid will change if the device is replaced.
Additionally, since the kernel cannot distinguish between the original
and the cloned filesystem, this new f_fsid derivation is applied to
both.
Link: https://lore.kernel.org/linux-btrfs/cover.1772095546.git.asj@kernel.org/
Link: https://lore.kernel.org/linux-btrfs/cover.1774092915.git.asj@kernel.org/
Signed-off-by: Anand Jain <asj@kernel.org>
Signed-off-by: David Sterba <dsterba@suse.com>
When mounting a cloned filesystem with a temporary fsuuid (temp_fsid),
layered modules like overlayfs require a persistent identifier.
While internal in-memory fs_devices->fsid must remain unique to
the kernel module, let s_uuid carry the original on-disk UUID.
Signed-off-by: Anand Jain <asj@kernel.org>
Signed-off-by: David Sterba <dsterba@suse.com>
[MINOR PROBLEM]
When mounting a filesystem with a valid DEV_STATS item, we will always
update the DEV_STATS again in the next transaction commit, even if there
is no change the values.
[CAUSE]
During the mount, btrfs_device_init_dev_stats() will read out the
on-disk DEV_STATS item for each device.
Then it calls btrfs_dev_stat_set() to update the in-memory structure.
However btrfs_dev_stat_set() does not only set the dev stats value, but
also increase device->dev_stats_ccnt.
That member determines if we should update the device item at the next
transaction commit. Since we have called btrfs_dev_stat_set() for each
dev status member, dev_stats_ccnt will be non-zero and we will update
the dev stats item even it doesn't change at all.
[FIX]
Instead of using btrfs_dev_stat_set() for valid on-disk DEV_STATUS
values, directly call atomic_set() to set the in-memory values.
For other call sites, we still want to use btrfs_dev_stat_set() so that
we will force updating/creating the dev stats item.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[MINOR PROBLEM]
When adding a new btrfs device, the corresponding DEV_STATS item creation
can only triggered by a mount cycle if there is no other error
triggered:
# mkfs.btrfs -f $dev1 $mnt
# mount $dev1 $mnt
# btrfs dev add $dev2 $mnt
# sync
# btrfs ins dump-tree -t dev $dev1
device tree key (DEV_TREE ROOT_ITEM 0)
leaf 30588928 items 6 free space 15853 generation 9 owner DEV_TREE
item 0 key (DEV_STATS PERSISTENT_ITEM 1) itemoff 16243 itemsize 40 <<<
persistent item objectid DEV_STATS offset 1
device stats
write_errs 0 read_errs 0 flush_errs 0 corruption_errs 0 generation 0
item 1 key (1 DEV_EXTENT 13631488) itemoff 16195 itemsize 48
Only after a mount cycle and a new transaction, the DEV_STATS for devid
2 can show up:
# umount $mnt
# mount $dev1 $mnt
# touch $mnt
# sync
# btrfs ins dump-tree -t dev $dev1
device tree key (DEV_TREE ROOT_ITEM 0)
leaf 30605312 items 7 free space 15788 generation 10 owner DEV_TREE
item 0 key (DEV_STATS PERSISTENT_ITEM 1) itemoff 16243 itemsize 40
persistent item objectid DEV_STATS offset 1
device stats
write_errs 0 read_errs 0 flush_errs 0 corruption_errs 0 generation 0
item 1 key (DEV_STATS PERSISTENT_ITEM 2) itemoff 16203 itemsize 40
persistent item objectid DEV_STATS offset 2
device stats
write_errs 0 read_errs 0 flush_errs 0 corruption_errs 0 generation 0
[CAUSE]
Btrfs only updates the DEV_STATS item when the device->dev_stats_ccnt
counter is not 0.
This is to reduce COW for the device tree. However that dev_stats_ccnt is
only increased at the following call sites:
- btrfs_dev_stat_inc()
This happens when some IO error happened.
- btrfs_dev_stat_read_and_reset()
This happens for GET_DEV_STATS ioctl with BTRFS_DEV_STATS_RESET flag.
- btrfs_dev_stat_set()
This happens inside btrfs_device_init_dev_stats().
So when a new device is added, its dev_stats_ccnt is just initialized to
0, and btrfs won't create nor update the corresponding DEV_STATS item at
all.
[ENHANCEMENT]
When a new device is added, also increase the dev_stats_ccnt by one.
This includes both device add ioctl and dev-replace.
This will force btrfs to create a new DEV_STATS item or update the
existing one with the correct values.
This not only makes the DEV_STATS creation early, but also prevents
old DEV_STATS left from older kernels to cause false alerts for the
newly added device.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[MINOR BUG]
The following script will cause DEV_STATS item to be left after the
corresponding device is removed:
# mkfs.btrfs -f $dev1
# mount $dev1 $mnt
# btrfs dev add $dev2 $mnt
# umount $mnt
## Without real errors, only at mount time btrfs will update
## dev->dev_stats_ccnt, thus we need a mount cycle to create the
## DEV_STATS item for the new device.
# mount $dev1 $mnt
# touch $mnt/foobar
# sync
# btrfs dev remove $dev2 $mnt
# umount $mnt
This will result the DEV_STATS item for devid 2 still left in device
tree:
device tree key (DEV_TREE ROOT_ITEM 0)
leaf 31064064 items 7 free space 15788 generation 18 owner DEV_TREE
leaf 31064064 flags 0x1(WRITTEN) backref revision 1
fs uuid 4bd853ed-f6ef-45fd-bbf1-1c3a2d9987cb
chunk uuid b496eab1-ec23-46b5-81c1-2f1b3503ca07
item 0 key (DEV_STATS PERSISTENT_ITEM 1) itemoff 16243 itemsize 40
persistent item objectid DEV_STATS offset 1
device stats
write_errs 0 read_errs 0 flush_errs 0 corruption_errs 0 generation 0
item 1 key (DEV_STATS PERSISTENT_ITEM 2) itemoff 16203 itemsize 40
persistent item objectid DEV_STATS offset 2
device stats
write_errs 0 read_errs 0 flush_errs 0 corruption_errs 0 generation 0
This is not a huge problem, but if the existing DEV_STATS contains
errors, and a new device is added into the fs taking the old devid, then
after a mount cycle, the new device will suddenly inherit old errors
which can give false alerts.
[CAUSE]
Btrfs never has the ability to delete DEV_STATS items.
It either create a new one through update_dev_stat_item(), or read an
existing one through btrfs_device_init_dev_stats().
However update_dev_stat_item() is only called lazily, if a new device is
created and no new update to dev stats, then it will skip the update of
the on-disk item.
So if the old DEV_STATS item exists and a new device is added, and no
errors during the remaining operations, the old DEV_STATS will not be
updated.
Then at the next mount cycle, btrfs_device_init_dev_stats() is called at
mount time, which will read out the old records, causing false alerts to
the newly added device.
[FIX]
Manually remove the DEV_STATS item during btrfs_rm_device().
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[MINOR PROBLEM]
When a running dev-replace hits some error for the target device (devid
0), there will be a DEV_STATS with error records created at the next
transaction commit.
Unfortunately that item will never to be deleted.
This means at the next dev-replace, if the replace is interrupted, then
at the next mount, the target device will suddenly inherit the old error
records from that DEV_STATS item, which can give some false alerts on
that device.
This shouldn't affect end users that much, as it requires all the
following conditions to be met, which is pretty rare:
- The initial dev-replace hits some error on the target device
E.g. write errors, but those errors itself is already a big problem
for a running replace.
This is required to create the DEV_STATS item in the first place.
- The next replace is interrupted
This is required to allow btrfs to read from the old records.
[CAUSE]
Btrfs just never deletes the DEV_STATS after a replace is finished.
[FIX]
Remove the DEV_STATS item for devid 0 after the replace is finished.
This is not going to completely fix the error, as we still have other
error paths, e.g. by somehow the fs flips RO and can not start a new
transaction for the DEV_STATS item removal.
But those corner cases will be addressed by later patches which provide
a more generic fix to DEV_STATS related problems.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
get_new_location() uses BUG_ON() to crash the kernel if the file extent
item it looks up has any of offset, compression, encryption, or
other_encoding set non-zero. The data reloc inode is only written by
relocation's own paths and the four fields are always 0 in what the
kernel writes:
- insert_prealloc_file_extent() memsets the stack item to zero and
only fills in type, disk_bytenr, disk_num_bytes and num_bytes, so
offset/compression/encryption/other_encoding stay 0.
- insert_ordered_extent_file_extent() copies oe->compress_type into
the file extent's compression field, but the data reloc inode is
created with BTRFS_INODE_NOCOMPRESS so compress_type is always 0;
encryption and other_encoding are reserved-and-zero in btrfs.
A non-zero value here means the leaf decoded from disk does not match
what the kernel wrote, i.e. on-disk corruption. A malformed image
reaches this code via balance and panics the kernel.
A previous attempt to enforce all four constraints in tree-checker's
check_extent_data_item() was merged as commit 7d0ee95979e9 ("btrfs:
validate data reloc tree file extent item members in tree-checker")
and then reverted by commit 1c034697fcaa after btrfs/061 produced
false positives on arm64 with 64K pages. The reason: relocation
writeback legitimately produces REG file_extent_items with offset != 0
in the data reloc tree. When an ordered extent covers only the back
portion of an underlying PREALLOC (num_bytes < ram_bytes on the input
file_extent), insert_ordered_extent_file_extent() inserts a REG with
offset = oe->offset
num_bytes = oe->num_bytes
ram_bytes preserved from the original PREALLOC,
and this item can reach disk if a transaction commit fires while it
is present in the leaf.
The four fields belong in different layers:
- compression, encryption and other_encoding are universal
invariants for every item in the data reloc tree, regardless of
cluster geometry. Enforce them in tree-checker's
check_extent_data_item() so a corrupt leaf is rejected at read
time.
- offset is only an invariant at the cluster-boundary keys that
get_new_location() searches (the key is computed as
src_disk_bytenr - reloc_block_group_start). Partial-PREALLOC
writebacks legitimately place REG items at non-boundary keys with
offset != 0; tree-checker cannot reject these. The cluster-
boundary item is always written by either
insert_prealloc_file_extent() (offset=0 by memset) or by the
front portion of a partial writeback (offset=0 by construction),
so a non-zero offset there is corruption.
Enforce the universal invariants in check_extent_data_item() with a
file_extent_err() rejection. Convert the BUG_ON() in
get_new_location() to a -EUCLEAN return paired with btrfs_print_leaf()
and btrfs_err() so the offending leaf is logged. The caller in
replace_file_extents() already handles non-zero returns from
get_new_location() by breaking out of the loop without aborting the
transaction.
Suggested-by: Qu Wenruo <wqu@suse.com>
Suggested-by: David Sterba <dsterba@suse.com>
Reported-by: syzbot+3e20d8f3d41bac5dc9a2@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=3e20d8f3d41bac5dc9a2
Signed-off-by: Teng Liu <27rabbitlt@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
should_nocow() reads inode->defrag_bytes without holding inode->lock,
while btrfs_set_delalloc_extent() and btrfs_clear_delalloc_extent()
update it under that spinlock.
This is a data race. The read is a quick check used to decide whether
to fall back to COW for a NOCOW inode: if defrag_bytes is non-zero and
the range is tagged EXTENT_DEFRAG, we force COW so that defragmentation
can rewrite the extent. Reading a stale value is harmless because:
- A missed increment may skip COW once, but the defrag pass will
redo the extent later.
- A stale non-zero may force an unnecessary COW, which is a minor
efficiency loss, not a correctness issue.
On 64-bit platforms an aligned u64 load is naturally atomic so tearing
cannot happen. On 32-bit platforms u64 may tear, but we only test for
zero vs non-zero, so the heuristic stays correct regardless. Use
data_race() annotation.
Fixes: 47059d930f ("Btrfs: make defragment work with nodatacow option")
Signed-off-by: Cen Zhang <zzzccc427@gmail.com>
[ Use data_race() instead of READ_ONCXE() ]
Signed-off-by: David Sterba <dsterba@suse.com>
The fs_path can use the auto freeing pattern and it's completely
contained in send. Define the freeing wrapper and add the cleanup
attributes.
Almost all conversions are straightforward, replacing goto with direct
return.
Signed-off-by: David Sterba <dsterba@suse.com>
The qgroupid has a specific format, add common format specifier, similar
to what we have for checksums and keys.
Reviewed-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
When a block device does not report a maximum number of open or active
zones, currently assign BTRFS_DEFAULT_MAX_ACTIVE_ZONES (128) to
the internal limit, if the device has more than
BTRFS_DEFAULT_MAX_ACTIVE_ZONES zones.
But if the device has less than BTRFS_DEFAULT_MAX_ACTIVE_ZONES the
internal max_active_zones limit will stay at 0, even if the device has
zone resource limits. Furthermore, if the device has a total number of
zones that is less than BTRFS_DEFAULT_MAX_ACTIVE_ZONE, max_active_zones
should be set to at most the number of zones.
Also move the max_active_zone calculation and setting into a dedicated
helper, to shrink btrfs_get_dev_zone_info().
Fixes: 04147d8394 ("btrfs: zoned: limit active zones to max_open_zones")
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This is open-coded bvec_phys(), also remove direct use of bv_page.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Boris Burkov <boris@bur.io>
Tested-by: Boris Burkov <boris@bur.io>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: David Sterba <dsterba@suse.com>
Calling __free_page() on folio_page() happens to work today, but
won't always. Besides, it's far simpler to call folio_put().
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Tested-by: Boris Burkov <boris@bur.io>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: David Sterba <dsterba@suse.com>
It seems that af566bdaff was tested against a tree which did not
contain commit 12851bd921 ("fs: Turn page_offset() into a wrapper
around folio_pos()). Unfortunately it has a bug of its own; on 32-bit
systems, shifting by PAGE_SHIFT will overflow on files larger than 4GiB.
Since page_offset() is now fixed, just revert af566bdaff.
Fixes: af566bdaff (btrfs: fix the file offset calculation inside btrfs_decompress_buf2page())
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Boris Burkov <boris@bur.io>
Tested-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
When performing data relocation on a zoned filesystem, BTRFS can deadlock
in handle_reserve_tickets(). The relocation process is waiting on a space
reservation ticket that can never be fulfilled, because the relocation
itself is the operation responsible for freeing up that space.
Fix this by introducing a new flush state,
BTRFS_RESERVE_FLUSH_ZONED_RELOCATION, specifically for data chunk
allocation during zoned relocation. Like
BTRFS_RESERVE_FLUSH_FREE_SPACE_INODE, this state uses
priority_reclaim_data_space() instead of the normal flushing path, which
avoids re-entering the relocation code and breaking the deadlock cycle.
In btrfs_alloc_data_chunk_ondemand(), select this new flush state when the
inode belongs to a data relocation root on a zoned filesystem.
Fixes: e2a7fd2237 ("btrfs: zoned: add zone reclaim flush state for DATA space_info")
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Don't account the free space in a data relocation space-info sub-group as
usable free space in statfs.
This is misleading as no user allocations can be made in this space-info
sub-group. It is only a target for relocation.
Fixes: f92ee31e03 ("btrfs: introduce btrfs_space_info sub-group")
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When searching for a data relocation block-group on mount,
btrfs_zoned_reserve_data_reloc_bg() is looking for the first empty DATA
block-group. But it first checks if the block-group is empty and if yes
continues the search, and then checks if it is the first DATA block-group.
There is actually no point in looking for the second empty DATA block
group as new DATA allocations will just allocate a new chunk for it. Pick
the first DATA block-group without any allocations done and set it as
relocation block-group.
At first, the commit 694ce5e143 ("btrfs: zoned: reserve data_reloc
block group on mount") introduced the functionality. At that time, we
took second unused (used == 0) block group, as the first one might be a
block group used for normal data. Later, commit daa0fde322 ("btrfs:
zoned: fix data relocation block group reservation") switched to look
for an empty block group (alloc_offset == 0). At this point, there is no
reason taking the second one anymore. So, this commit is fixing an issue
in commit daa0fde322.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Decode the 'RECLAIM_ZONES' state in tracepoints, as of now only the
numerical state is shown in the tracepoints.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Document the purpose of the RECLAIM_ZONES flush state.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
With all the previous preparations, it's finally time to enable the
huge folio support.
- The max folio size
Here we define BTRFS_MAX_FOLIO_SIZE, which is fixed at 2MiB.
This will ensure we have a large enough but not too large folio for
btrfs. This limit applies to all systems regardless of page size.
Then we also define BTRFS_MAX_BLOCKS_PER_FOLIO, which depends on
CONFIG_BTRFS_EXPERIMENTAL.
If it's an experimental build, BTRFS_MAX_BLOCKS_PER_FOLIO is 512,
otherwise it's BITS_PER_LONG.
The filemap max order will be calculated using both
BTRFS_MAX_FOLIO_SIZE and BTRFS_MAX_BLOCKS_PER_FOLIO.
E.g. for 64K page size with 64K fs block size, the limit will be
BTRFS_MAX_FOLIO_SIZE (2M), which limits the filemap max order to 5.
This will be lower than the old order (6), but folios larger than 2M
are rarely any better for IO performance. Meanwhile excessively large
folios can cause other problems like stalling the IO pipeline for too
long.
For 4K page size and 4K fs block size, the limit will be increased to
2M from the old 256K.
This new size is constrained by both BTRFS_MAX_FOLIO_SIZE (2M) and
BTRFS_MAX_BLOCKS_PER_FOLIO (512 * 4K), allowing x86_64 to achieve huge
folio support, and the filemap max order will be 9.
- btrfs_bio_ctrl::submit_bitmap
This will be enlarged to contain BTRFS_MAX_BLOCKS_PER_FOLIO bits, and
this will be on-stack memory.
This will increase on-stack memory usage by 56 bytes compared to the
baseline (before the first patch in the series).
- Local @delalloc_bitmap inside writepage_delalloc()
Unfortunately we cannot afford to handle an allocation error here, thus
again we use on-stack memory.
Thus this will increase on-stack memory usage by 56 bytes again.
So unfortunately this means during the delalloc window, the writeback path
will have +112 bytes on-stack memory usage, and for other cases the
writeback path will have +56 bytes on-stack memory usage.
The +56 bytes (btrfs_bio_ctrl::submit_bitmap) can be removed
after we have reworked the compression submission, so the current
on-stack submit_bitmap is mostly a workaround until then.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[CURRENT LIMIT]
Btrfs currently only supports sub-bitmaps (e.g. dirty bitmap) no larger
than BITS_PER_LONG.
One call site that utilizes this limit is btrfs_bio_ctrl::submit_bitmap,
which makes it very simple and straightforward to just grab an unsigned
long value and assign it to submit_bitmap.
Unfortunately that limit prevents us from supporting huge folios.
For 4K page size and block size, a huge folio (order 9) means 512 blocks
inside a 2M folio.
[ENHANCEMENT]
Instead of using a fixed unsigned long value, change
btrfs_bio_ctrl::submit_bitmap to an unsigned long pointer.
And for cases where an unsigned long can hold the whole bitmap,
introduce @submit_bitmap_value, and just point that pointer to that
unsigned long.
Then update all direct users of bio_ctrl->submit_bitmap to use the
pointer version.
There are several call sites that get extra changes:
- @range_bitmap inside extent_writepage_io()
Which is only utilized to truncate the bitmap.
Since we do not want to allocate new memory just for such temporary
usage, change the original bitmap_set() and bitmap_and() into
bitmap_clear() for the ranges outside of the target range.
- Getting dirty subpage bitmap inside writepage_delalloc()
Since we're passing an unsigned long pointer now, we need to go with
different handling (bs == ps, blocks_per_folio <= BITS_PER_LONG,
blocks_per_folio > BITS_PER_LONG).
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[CURRENT LIMIT]
Btrfs currently only supports sub-bitmaps (e.g. dirty bitmap) no larger
than BITS_PER_LONG.
That limit allows us to easily grab an unsigned long without the need to
properly allocate memory for a larger bitmap.
Unfortunately that limit prevents us from supporting huge folios.
For 4K page size and block size, a huge folio (order 9) means 512 blocks
inside a 2M folio.
[ENHANCEMENT]
To allow direct bitmap operations without allocating new memory,
introduce two different ways to access the subpage bitmaps:
- Return an unsigned long value
This only happens if blocks_per_folio <= BITS_PER_LONG.
We read out the sub-bitmap into an unsigned long, and return the
value.
This is the old existing method.
This involves get_bitmap_value_##name() helper functions.
And this time the helper functions are defined as inline functions
instead of macros to provide better type checks.
- Return a pointer where the sub-bitmap starts
This only happens if blocks_per_folio >= BITS_PER_LONG.
This is the new method for sub-bitmaps larger than BITS_PER_LONG.
Since the sizes of sub-bitmaps are all aligned to BITS_PER_LONG, we
can directly access the start word of the sub-bitmap.
This involves get_bitmap_pointer_##name() helper functions.
Then change the existing sub-bitmaps users to use the new helpers:
- Bitmap dumping
Switch between get_bitmap_value_##name() and
get_bitmap_pointer_##name() depending on the sub-bitmap size.
- btrfs_get_subpage_dirty_bitmap()
Rename it to btrfs_get_subpage_dirty_bitmap_value() to follow the new
value/pointer naming.
Since we do not support huge folios yet, there is no pointer version
for the dirty bitmap.
Furthermore, add the support for block size == page size cases for
btrfs_get_subpage_dirty_bitmap_value(), so that the caller no longer
needs to check if the folio needs subpage handling.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The comments at the beginning of subpage.c are out-of-date, a lot of the
limitations have been already resolved.
Update them to reflect the latest status.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Optimize the btrfs_abort_transaction() for size as it (by our
convention) must be put right after the error condition is detected.
The exact file:line is reported so there's a portion that must be
inlined. As this is cold code it bloats functions. In previous patch
"btrfs: move transaction abort message to __btrfs_abort_transaction()"
the error message was moved to the common helper, saving like 20KiB of
btrfs.ko and several instructions per call site and some stack space.
There's little left to be optimized, we need to keep the atomic
test_and_set_bit() and to convey that as 'first hit' to
__btrfs_abort_transaction().
Right now it's a bool, which takes 8 bytes on stack for each call but
it's 1 bit of information. We can encode that to some of the other
parameters.
For that let's use the 'error' parameter, by convention it's negative
errno so we can reliably detect if it's the first hit or a later error.
Also the negation is usually implemented by a single instruction (NEG on
x86_64) so the resulting object code is kept short.
This reduces btrfs.ko by 8K and stack in several functions by 8 bytes.
Cumulative effect with the other commit is -30K of btrfs.ko. While the
encoding is an implementation detail, it's contained within the API.
Making the transaction abort calls very light is desired.
Signed-off-by: David Sterba <dsterba@suse.com>
In preparation to encode more information to the error value add a step
that verifies if the value is valid (i.e. < 0). This works for
compile-time and runtime (in debugging mode).
The compile-time check recognizes direct constants and defines an array
type. An invalid condition leads to negative array size which is caught
by compiler.
The runtime check constructs the array type from the condition and only
verifies the correct size, as we don't need to tweak the size to be
negative.
The sizeof() expressions do not generate any code. In the debugging
config the warning adds about 9KiB of btrfs.ko code size.
The array size trick is needed as we can't use static_array(), not even
with __builtin_constant_p().
Sample error message:
In file included from inode.c:40:
inode.c: In function ‘__cow_file_range_inline’:
transaction.h:261:26: error: size of unnamed array is negative
261 | (void)sizeof(char[-!(__builtin_constant_p(error) ? (error) < 0 : 1)]); \
| ^
transaction.h:275:9: note: in expansion of macro ‘VERIFY_NEGATIVE_ERROR’
275 | VERIFY_NEGATIVE_ERROR(error); \
| ^~~~~~~~~~~~~~~~~~~~~
inode.c:665:17: note: in expansion of macro ‘btrfs_abort_transaction’
665 | btrfs_abort_transaction(trans, 17);
| ^~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: David Sterba <dsterba@suse.com>
In the beginning of the loop, we try to obtain a locked delayed ref head,
if 'locked_ref' is currently NULL, by calling btrfs_select_ref_head(),
which can return an error pointer. If the error pointer is -EAGAIN we do
a continue and go back to the beginning of the loop, which will not try
again to call btrfs_select_ref_head() since 'locked_ref' is no longer
NULL but it's ERR_PTR(-EAGAIN), and then we do:
spin_lock(&locked_ref->lock);
against a ERR_PTR(-EAGAIN) value, generating an invalid pointer
dereference.
Fix this by ensuring that 'locked_ref' is set to NULL when
btrfs_select_ref_head() returns ERR_PTR(-EAGAIN) and incrementing 'count'
as well, to prevent infinite looping. We do this by doing a goto to the
bottom of the loop that already sets 'locked_ref' to NULL and does a
cond_resched(), with an increment to 'count' right before the goto.
These measures were in place before the refactoring in commit 0110a4c434
("btrfs: refactor __btrfs_run_delayed_refs loop") but were unintentionally
lost afterwards.
Reported-by: Dan Carpenter <error27@gmail.com>
Link: https://lore.kernel.org/linux-btrfs/ag8ARRwykv8bpJ87@stanley.mountain/
Fixes: 0110a4c434 ("btrfs: refactor __btrfs_run_delayed_refs loop")
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
sb_write_pointer() reads the super block from the block device page cache
using read_cache_page_gfp(). This has the same race with BLKBSZSET as the
one fixed by commit 3f29d661e5 ("btrfs: sync read disk super and set
block size").
Take the mapping invalidate lock around read_cache_page_gfp() to
serialize the read against block size changes.
Signed-off-by: KangNing Liao <lkangn.kernel@gmail.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Print the type of the inode (directory, regular file, symlink, etc) to
facilitate debugging.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_sync_log() is one of the main functions called during a fsync.
Add trace events for when entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_log_new_name() is an important function that affects inode logging
and is called during link and rename operations. Add trace events for when
entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_record_new_subvolume() is an important operation that affects
inode logging and is called during subvolume creation. Add a trace event
for it to help debug issues.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_record_snapshot_destroy() is an important operation that affects
inode logging and is called during subvolume/snapshot deletion as well as
during rmdir. Add a trace event for it to help debug issues.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_record_unlink_dir() is an important operation that affects inode
logging and is called during unlink and rename operations. Add a trace
event for it to help debug issues.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
log_new_delayed_dentries() is an important step called during a fsync, as
well as during rename and link operations on inodes that were previously
logged. Add trace events for when entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In overwrite_item():
There's no point in printing the root's ID if the assertion fails, since
it can only be BTRFS_TREE_LOG_OBJECTID if it fails.
In log_new_delayed_dentries():
There's no point in using a verbose assertion to print the value of
ctx->logging_new_delayed_dentries because it's a boolean, so if the
assertion fails we know its value is true (1).
So convert them to simpler assertion to make the code less verbose.
It also slightly reduces the object size, at least on x86_64 using
Debian's gcc 14.2.0-19 (if CONFIG_BTRFS_ASSERT is enabled in the kernel
config, which is the case for SUSE distributions for example).
Before:
$ size fs/btrfs/btrfs.ko
text data bss dec hex filename
2028244 197176 15624 2241044 223214 fs/btrfs/btrfs.ko
After:
$ size fs/btrfs/btrfs.ko
text data bss dec hex filename
2028228 197176 15624 2241028 223204 fs/btrfs/btrfs.ko
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
log_conflicting_inodes() is an important step called during a fsync, as
well as during rename and link operations on inodes that were previously
logged. Add trace events for when entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
add_conflicting_inode() is an important step called during a fsync, as
well as during rename and link operations on inodes that were previously
logged. Add trace events for when entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
log_new_dir_dentries() is an important step called during a fsync, as
well as during rename and link operations on inodes that were previously
logged. Add trace events for when entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
log_all_new_ancestors() is an important step called during a fsync, as
well as during rename and link operations on inodes that were previously
logged. Add trace events for when entering and exiting that function.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
