-----BEGIN PGP SIGNATURE-----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=4OkE
-----END PGP SIGNATURE-----
Merge tag 'for-5.6-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull btrfs updates from David Sterba:
"Features, highlights:
- async discard
- "mount -o discard=async" to enable it
- freed extents are not discarded immediatelly, but grouped
together and trimmed later, with IO rate limiting
- the "sync" mode submits short extents that could have been
ignored completely by the device, for SATA prior to 3.1 the
requests are unqueued and have a big impact on performance
- the actual discard IO requests have been moved out of
transaction commit to a worker thread, improving commit latency
- IO rate and request size can be tuned by sysfs files, for now
enabled only with CONFIG_BTRFS_DEBUG as we might need to
add/delete the files and don't have a stable-ish ABI for
general use, defaults are conservative
- export device state info in sysfs, eg. missing, writeable
- no discard of extents known to be untouched on disk (eg. after
reservation)
- device stats reset is logged with process name and PID that called
the ioctl
Fixes:
- fix missing hole after hole punching and fsync when using NO_HOLES
- writeback: range cyclic mode could miss some dirty pages and lead
to OOM
- two more corner cases for metadata_uuid change after power loss
during the change
- fix infinite loop during fsync after mix of rename operations
Core changes:
- qgroup assign returns ENOTCONN when quotas not enabled, used to
return EINVAL that was confusing
- device closing does not need to allocate memory anymore
- snapshot aware code got removed, disabled for years due to
performance problems, reimplmentation will allow to select wheter
defrag breaks or does not break COW on shared extents
- tree-checker:
- check leaf chunk item size, cross check against number of
stripes
- verify location keys for DIR_ITEM, DIR_INDEX and XATTR items
- new self test for physical -> logical mapping code, used for super
block range exclusion
- assertion helpers/macros updated to avoid objtool "unreachable
code" reports on older compilers or config option combinations"
* tag 'for-5.6-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (84 commits)
btrfs: free block groups after free'ing fs trees
btrfs: Fix split-brain handling when changing FSID to metadata uuid
btrfs: Handle another split brain scenario with metadata uuid feature
btrfs: Factor out metadata_uuid code from find_fsid.
btrfs: Call find_fsid from find_fsid_inprogress
Btrfs: fix infinite loop during fsync after rename operations
btrfs: set trans->drity in btrfs_commit_transaction
btrfs: drop log root for dropped roots
btrfs: sysfs, add devid/dev_state kobject and device attributes
btrfs: Refactor btrfs_rmap_block to improve readability
btrfs: Add self-tests for btrfs_rmap_block
btrfs: selftests: Add support for dummy devices
btrfs: Move and unexport btrfs_rmap_block
btrfs: separate definition of assertion failure handlers
btrfs: device stats, log when stats are zeroed
btrfs: fix improper setting of scanned for range cyclic write cache pages
btrfs: safely advance counter when looking up bio csums
btrfs: remove unused member btrfs_device::work
btrfs: remove unnecessary wrapper get_alloc_profile
btrfs: add correction to handle -1 edge case in async discard
...
[BUG]
For dev-replace test cases with fsstress, like btrfs/06[45] btrfs/071,
looped runs can lead to random failure, where scrub finds csum error.
The possibility is not high, around 1/20 to 1/100, but it's causing data
corruption.
The bug is observable after commit b12de52896 ("btrfs: scrub: Don't
check free space before marking a block group RO")
[CAUSE]
Dev-replace has two source of writes:
- Write duplication
All writes to source device will also be duplicated to target device.
Content: Not yet persisted data/meta
- Scrub copy
Dev-replace reused scrub code to iterate through existing extents, and
copy the verified data to target device.
Content: Previously persisted data and metadata
The difference in contents makes the following race possible:
Regular Writer | Dev-replace
-----------------------------------------------------------------
^ |
| Preallocate one data extent |
| at bytenr X, len 1M |
v |
^ Commit transaction |
| Now extent [X, X+1M) is in |
v commit root |
================== Dev replace starts =========================
| ^
| | Scrub extent [X, X+1M)
| | Read [X, X+1M)
| | (The content are mostly garbage
| | since it's preallocated)
^ | v
| Write back happens for |
| extent [X, X+512K) |
| New data writes to both |
| source and target dev. |
v |
| ^
| | Scrub writes back extent [X, X+1M)
| | to target device.
| | This will over write the new data in
| | [X, X+512K)
| v
This race can only happen for nocow writes. Thus metadata and data cow
writes are safe, as COW will never overwrite extents of previous
transaction (in commit root).
This behavior can be confirmed by disabling all fallocate related calls
in fsstress (*), then all related tests can pass a 2000 run loop.
*: FSSTRESS_AVOID="-f fallocate=0 -f allocsp=0 -f zero=0 -f insert=0 \
-f collapse=0 -f punch=0 -f resvsp=0"
I didn't expect resvsp ioctl will fallback to fallocate in VFS...
[FIX]
Make dev-replace to require mandatory block group RO, and wait for current
nocow writes before calling scrub_chunk().
This patch will mostly revert commit 76a8efa171 ("btrfs: Continue replace
when set_block_ro failed") for dev-replace path.
The side effect is, dev-replace can be more strict on avaialble space, but
definitely worth to avoid data corruption.
Reported-by: Filipe Manana <fdmanana@suse.com>
Fixes: 76a8efa171 ("btrfs: Continue replace when set_block_ro failed")
Fixes: b12de52896 ("btrfs: scrub: Don't check free space before marking a block group RO")
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
block_group removal is a little tricky. It can race with the extent
allocator, the cleaner thread, and balancing. The current path is for a
block_group to be added to the unused_bgs list. Then, when the cleaner
thread comes around, it starts a transaction and then proceeds with
removing the block_group. Extents that are pinned are subsequently
removed from the pinned trees and then eventually a discard is issued
for the entire block_group.
Async discard introduces another player into the game, the discard
workqueue. While it has none of the racing issues, the new problem is
ensuring we don't leave free space untrimmed prior to forgetting the
block_group. This is handled by placing fully free block_groups on a
separate discard queue. This is necessary to maintain discarding order
as in the future we will slowly trim even fully free block_groups. The
ordering helps us make progress on the same block_group rather than say
the last fully freed block_group or needing to search through the fully
freed block groups at the beginning of a list and insert after.
The new order of events is a fully freed block group gets placed on the
unused discard queue first. Once it's processed, it will be placed on
the unusued_bgs list and then the original sequence of events will
happen, just without the final whole block_group discard.
The mount flags can change when processing unused_bgs, so when flipping
from DISCARD to DISCARD_ASYNC, the unused_bgs must be punted to the
discard_list to be trimmed. If we flip off DISCARD_ASYNC, we punt
free block groups on the discard_list to the unused_bg queue which will
do the final discard for us.
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 doing a device replace, while at scrub.c:scrub_enumerate_chunks(), we
set the block group to RO mode and then wait for any ongoing writes into
extents of the block group to complete. While doing that wait we overwrite
the value of the variable 'ret' and can break out of the loop if an error
happens without turning the block group back into RW mode. So what happens
is the following:
1) btrfs_inc_block_group_ro() returns 0, meaning it set the block group
to RO mode (its ->ro field set to 1 or incremented to some value > 1);
2) Then btrfs_wait_ordered_roots() returns a value > 0;
3) Then if either joining or committing the transaction fails, we break
out of the loop wihtout calling btrfs_dec_block_group_ro(), leaving
the block group in RO mode forever.
To fix this, just remove the code that waits for ongoing writes to extents
of the block group, since it's not needed because in the initial setup
phase of a device replace operation, before starting to find all chunks
and their extents, we set the target device for replace while holding
fs_info->dev_replace->rwsem, which ensures that after releasing that
semaphore, any writes into the source device are made to the target device
as well (__btrfs_map_block() guarantees that). So while at
scrub_enumerate_chunks() we only need to worry about finding and copying
extents (from the source device to the target device) that were written
before we started the device replace operation.
Fixes: f0e9b7d640 ("Btrfs: fix race setting block group readonly during device replace")
Signed-off-by: Filipe Manana <fdmanana@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 "&fs_info->dev_replace.rwsem" and "&dev_replace->rwsem" refer to
the same lock but Smatch is not clever enough to figure that out so it
leads to static checker warnings. It's better to use it consistently
anyway.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The on-disk format of block group item makes use of the key that stores
the offset and length. This is further used in the code, although this
makes thing harder to understand. The key is also packed so the
offset/length is not properly aligned as u64.
Add start (key.objectid) and length (key.offset) members to block group
and remove the embedded key. When the item is searched or written, a
local variable for key is used.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For unknown reasons, the member 'used' in the block group struct is
stored in the b-tree item and accessed everywhere using the special
accessor helper. Let's unify it and make it a regular member and only
update the item before writing it to the tree.
The item is still being used for flags and chunk_objectid, there's some
duplication until the item is removed in following patches.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Commit 9e0af23764 ("Btrfs: fix task hang under heavy compressed
write") worked around the issue that a recycled work item could get a
false dependency on the original work item due to how the workqueue code
guarantees non-reentrancy. It did so by giving different work functions
to different types of work.
However, the fixes in the previous few patches are more complete, as
they prevent a work item from being recycled at all (except for a tiny
window that the kernel workqueue code handles for us). This obsoletes
the previous fix, so we don't need the unique helpers for correctness.
The only other reason to keep them would be so they show up in stack
traces, but they always seem to be optimized to a tail call, so they
don't show up anyways. So, let's just get rid of the extra indirection.
While we're here, rename normal_work_helper() to the more informative
btrfs_work_helper().
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently, scrub_missing_raid56_worker() puts and potentially frees
sblock (which embeds the work item) and then submits a bio through
scrub_wr_submit(). This is another potential instance of the bug in
"btrfs: don't prematurely free work in run_ordered_work()". Fix it by
dropping the reference after we submit the bio.
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This is prep work for moving all of the block group cache code into its
own file.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ minor comment updates ]
Signed-off-by: David Sterba <dsterba@suse.com>
Preparatory patch for additional RAID1 profiles with more copies. The
mask will contain 3-copy and 4-copy, most of the checks for plain RAID1
work the same for the other profiles.
Signed-off-by: David Sterba <dsterba@suse.com>
Currently btrfs_csum_data() relied on the crc32c() wrapper around the
crypto framework for calculating the CRCs.
As we have our own crypto_shash structure in the fs_info now, we can
directly call into the crypto framework without going trough the wrapper.
This way we can even remove the btrfs_csum_data() and btrfs_csum_final()
wrappers.
The module dependency on crc32c is preserved via MODULE_SOFTDEP("pre:
crc32c"), which was previously provided by LIBCRC32C config option doing
the same.
Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
BTRFS has the implicit assumption that a checksum in btrfs_orderd_sums
is 4 bytes. While this is true for CRC32C, it is not for any other
checksum.
Change the data type to be a byte array and adjust loop index
calculation accordingly.
This includes moving the adjustment of 'index' by 'ins_size' in
btrfs_csum_file_blocks() before dividing 'ins_size' by the checksum
size, because before this patch the 'sums' member of 'struct
btrfs_ordered_sum' was 4 Bytes in size and afterwards it is only one
byte.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
fs_info::mapping_tree is the physical<->logical mapping tree and uses
the same underlying structure as extents, but is embedded to another
structure. There are no other members and this indirection is useless.
No functional change.
Signed-off-by: David Sterba <dsterba@suse.com>
The error code used here is wrong as it's not invalid to try to start
scrub when umount has begun. Returning EAGAIN is more user friendly as
it's recoverable.
Signed-off-by: David Sterba <dsterba@suse.com>
The scrub_ctx csum_list member must be initialized before scrub_free_ctx
is called. If the csum_list is not initialized beforehand, the
list_empty call in scrub_free_csums will result in a null deref if the
allocation fails in the for loop.
Fixes: a2de733c78 ("btrfs: scrub")
CC: stable@vger.kernel.org # 3.0+
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Dan Robertson <dan@dlrobertson.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The scrub worker pointers are not NULL iff the scrub is running, so
reset them back once the last reference is dropped. Add assertions to
the initial phase of scrub to verify that.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Use the refcount_t for fs_info::scrub_workers_refcnt instead of int so
we get the extra checks. All reference changes are still done under
scrub_lock.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
scrub_workers_refcnt is protected by scrub_lock, add lockdep_assert_held()
in scrub_workers_get().
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Suggested-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The kernel log messages help debugging and audit, add them for scrub
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Both btrfs_find_device() and find_device() does the same thing except
that the latter does not take the seed device onto account in the device
scanning context. We can merge them.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_find_device() accepts fs_info as an argument and retrieves
fs_devices from fs_info.
Instead use fs_devices, so that this function can be used in non-mount
(during device scanning) context as well.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The typos accumulate over time so once in a while time they get fixed in
a large patch.
Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since scrub workers only do memory allocation with GFP_KERNEL when they
need to perform repair, we can move the recent setup of the nofs context
up to scrub_handle_errored_block() instead of setting it up down the call
chain at insert_full_stripe_lock() and scrub_add_page_to_wr_bio(),
removing some duplicate code and comment. So the only paths for which a
scrub worker can do memory allocations using GFP_KERNEL are the following:
scrub_bio_end_io_worker()
scrub_block_complete()
scrub_handle_errored_block()
lock_full_stripe()
insert_full_stripe_lock()
-> kmalloc with GFP_KERNEL
scrub_bio_end_io_worker()
scrub_block_complete()
scrub_handle_errored_block()
scrub_write_page_to_dev_replace()
scrub_add_page_to_wr_bio()
-> kzalloc with GFP_KERNEL
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The scrub context is allocated with GFP_KERNEL and called from
btrfs_scrub_dev under the fs_info::device_list_mutex. This is not safe
regarding reclaim that could try to flush filesystem data in order to
get the memory. And the device_list_mutex is held during superblock
commit, so this would cause a lockup.
Move the alocation and initialization before any changes that require
the mutex.
Signed-off-by: David Sterba <dsterba@suse.com>
We can pass fs_info directly as this is the only member of btrfs_device
that's bing used inside scrub_setup_ctx.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The dev-replace locking functions are now trivial wrappers around rw
semaphore that can be used directly everywhere. No functional change.
Signed-off-by: David Sterba <dsterba@suse.com>
When a transaction commit starts, it attempts to pause scrub and it blocks
until the scrub is paused. So while the transaction is blocked waiting for
scrub to pause, we can not do memory allocation with GFP_KERNEL from scrub,
otherwise we risk getting into a deadlock with reclaim.
Checking for scrub pause requests is done early at the beginning of the
while loop of scrub_stripe() and later in the loop, scrub_extent() and
scrub_raid56_parity() are called, which in turn call scrub_pages() and
scrub_pages_for_parity() respectively. These last two functions do memory
allocations using GFP_KERNEL. Same problem could happen while scrubbing
the super blocks, since it calls scrub_pages().
We also can not have any of the worker tasks, created by the scrub task,
doing GFP_KERNEL allocations, because before pausing, the scrub task waits
for all the worker tasks to complete (also done at scrub_stripe()).
So make sure GFP_NOFS is used for the memory allocations because at any
time a scrub pause request can happen from another task that started to
commit a transaction.
Fixes: 58c4e17384 ("btrfs: scrub: use GFP_KERNEL on the submission path")
CC: stable@vger.kernel.org # 4.6+
Reviewed-by: Nikolay Borisov <nborisov@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>
The wrapper is too trivial, open coding does not make it less readable.
Reviewed-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
struct scrub_ctx has an ->is_dev_replace member, so there's no point in
passing around is_dev_replace where sctx is available.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Remove the remaining code that misused the page cache pages during
device replace and could cause data corruption for compressed nodatasum
extents. Such files do not normally exist but there's a bug that allows
this combination and the corruption was exposed by device replace fixup
code.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Introduce a small helper, btrfs_mark_bg_unused(), to acquire locks and
add a block group to unused_bgs list.
No functional modification, and only 3 callers are involved.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It can be referenced from the passed bg cache.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since commit ac0b4145d6 ("btrfs: scrub: Don't use inode pages
for device replace") the function is not used and we can remove all
functions down the call chain.
There was an optimization that reused inode pages to speed up device
replace, but broke when there was nodatasum and compressed page. The
potential performance gain is small so we don't loose much by removing
it and using scrub_pages same as the other pages.
Signed-off-by: Qu Wenruo <wqu@suse.com>
[ update changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
In commit ac0b4145d6 ("btrfs: scrub: Don't use inode pages for device
replace") we removed the branch of copy_nocow_pages() to avoid
corruption for compressed nodatasum extents.
However above commit only solves the problem in scrub_extent(), if
during scrub_pages() we failed to read some pages,
sctx->no_io_error_seen will be non-zero and we go to fixup function
scrub_handle_errored_block().
In scrub_handle_errored_block(), for sctx without csum (no matter if
we're doing replace or scrub) we go to scrub_fixup_nodatasum() routine,
which does the similar thing with copy_nocow_pages(), but does it
without the extra check in copy_nocow_pages() routine.
So for test cases like btrfs/100, where we emulate read errors during
replace/scrub, we could corrupt compressed extent data again.
This patch will fix it just by avoiding any "optimization" for
nodatasum, just falls back to the normal fixup routine by try read from
any good copy.
This also solves WARN_ON() or dead lock caused by lame backref iteration
in scrub_fixup_nodatasum() routine.
The deadlock or WARN_ON() won't be triggered before commit ac0b4145d6
("btrfs: scrub: Don't use inode pages for device replace") since
copy_nocow_pages() have better locking and extra check for data extent,
and it's already doing the fixup work by try to read data from any good
copy, so it won't go scrub_fixup_nodatasum() anyway.
This patch disables the faulty code and will be removed completely in a
followup patch.
Fixes: ac0b4145d6 ("btrfs: scrub: Don't use inode pages for device replace")
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
Btrfs can create compressed extent without checksum (even though it
shouldn't), and if we then try to replace device containing such extent,
the result device will contain all the uncompressed data instead of the
compressed one.
Test case already submitted to fstests:
https://patchwork.kernel.org/patch/10442353/
[CAUSE]
When handling compressed extent without checksum, device replace will
goe into copy_nocow_pages() function.
In that function, btrfs will get all inodes referring to this data
extents and then use find_or_create_page() to get pages direct from that
inode.
The problem here is, pages directly from inode are always uncompressed.
And for compressed data extent, they mismatch with on-disk data.
Thus this leads to corrupted compressed data extent written to replace
device.
[FIX]
In this attempt, we could just remove the "optimization" branch, and let
unified scrub_pages() to handle it.
Although scrub_pages() won't bother reusing page cache, it will be a
little slower, but it does the correct csum checking and won't cause
such data corruption caused by "optimization".
Note about the fix: this is the minimal fix that can be backported to
older stable trees without conflicts. The whole callchain from
copy_nocow_pages() can be deleted, and will be in followup patches.
Fixes: ff023aac31 ("Btrfs: add code to scrub to copy read data to another disk")
CC: stable@vger.kernel.org # 4.4+
Reported-by: James Harvey <jamespharvey20@gmail.com>
Reviewed-by: James Harvey <jamespharvey20@gmail.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
[ remove code removal, add note why ]
Signed-off-by: David Sterba <dsterba@suse.com>
This patch will add the following trace events:
1) btrfs_remove_block_group
For btrfs_remove_block_group() function.
Triggered when a block group is really removed.
2) btrfs_add_unused_block_group
Triggered which block group is added to unused_bgs list.
3) btrfs_skip_unused_block_group
Triggered which unused block group is not deleted.
These trace events is pretty handy to debug case related to block group
auto remove.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Remove GPL boilerplate text (long, short, one-line) and keep the rest,
ie. personal, company or original source copyright statements. Add the
SPDX header.
Signed-off-by: David Sterba <dsterba@suse.com>
The current calls are unclear in what way btrfs_dev_replace_lock takes
the locks, so drop the argument, split the helpers and use similar
naming as for read and write locks.
Signed-off-by: David Sterba <dsterba@suse.com>
Added in b5d67f64f9 ("Btrfs: change scrub to support big blocks") but
rendered redundant by be50a8ddaa ("Btrfs: Simplify
scrub_setup_recheck_block()'s argument").
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In case of raid56, writes and rebuilds always take BTRFS_STRIPE_LEN(64K)
as unit, however, scrub_extent() sets blocksize as unit, so rebuild
process may be triggered on every block on a same stripe.
A typical example would be that when we're replacing a disappeared disk,
all reads on the disks get -EIO, every block (size is 4K if blocksize is
4K) would go thru these,
scrub_handle_errored_block
scrub_recheck_block # re-read pages one by one
scrub_recheck_block # rebuild by calling raid56_parity_recover()
page by page
Although with raid56 stripe cache most of reads during rebuild can be
avoided, the parity recover calculation(xor or raid6 algorithms) needs to
be done $(BTRFS_STRIPE_LEN / blocksize) times.
This makes it smarter by doing raid56 scrub/replace on stripe length.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In the last step of scrub_handle_error_block, we try to combine good
copies on all possible mirrors, this works fine for raid1 and raid10,
but not for raid56 as it's doing parity rebuild.
If parity rebuild doesn't get back with correct data which matches its
checksum, in case of replace we'd rather write what is stored in the
source device than the data calculuated from parity.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
