the relock, and a forced stuck delalloc bit makes both lock loops return
-EUCLEAN with the message.
-Patches 1-2 carry Fixes: 05a5a7621ce6, patch 3 Fixes: c2832898126f.
+Patches 1-2 carry Fixes: 05a5a7621ce6 ("Btrfs: implement full reflink
+support for inline extents", v5.7), patch 3 Fixes: c2832898126f ("btrfs:
+make relocate_one_page() handle subpage case", v5.15).
Zygo Blaxell (3):
btrfs: fix flushoncommit deadlock when cloning an inline extent inside
Reviewed-by: Qu Wenruo, Reviewed-by: Filipe Manana, Tested-by: Zygo
Blaxell), as posted; v2 changed only its commit message and tags. A split bio's error was dropped when the split
that failed was not the last to complete and the last one succeeded.
-Not raid56-specific. Fixes: d48e1dea3931. It replaces our own patch
-for the same bug (identical code, different comment); it is carried
-here until upstream takes it, and "git am --skip" drops it then.
+Not raid56-specific. Fixes: d48e1dea3931 ("btrfs: fix error propagation
+of split bios", v6.12). It replaces our own patch for the same bug
+(identical code, different comment); it is carried here until upstream
+takes it, and "git am --skip" drops it then.
Patch 4 drops cached rbios when their stripes' extents are freed. A
cached rbio's pages can be stolen by a later sub-stripe write to the
writeback. A writer that does not fsync loses that data silently, and
the allocator hands the same free space out again. Read repair already
walks P, Q and P+Q; the RMW now retries in the same order.
-Fixes: 7a3150723061.
+Fixes: 7a3150723061 ("btrfs: raid56: do data csum verification during
+RMW cycle", v6.2).
Patch 9 counts raid56 member I/O errors in the device stats. raid56
submits its stripe bios itself and their end_io only marked the rbio's
tolerance, so the operation succeeded) was never counted in
write_io_errs or read_io_errs: a failing member looked healthy in
`btrfs device stats`. The device is found by the bio's bdev, so writes
-to a replace target are charged to the target. Fixes: 53b381b3abeb.
+to a replace target are charged to the target. Fixes: 53b381b3abeb
+("Btrfs: RAID5 and RAID6", v3.9).
Patch 10 makes a sub-stripe write write back the sectors its RMW read
repaired. The read phase verifies data sectors against their csums and
with an error bit, and nodatasum sectors rebuilt from P+Q together with
a verified csummed sector of the same row are written back; a missing
device never is, and a failed write-back does not count against the
-write's tolerance. Fixes: 7a3150723061.
+write's tolerance. Fixes: 7a3150723061 ("btrfs: raid56: do data csum
+verification during RMW cycle", v6.2).
Josef Bacik (1):
btrfs: don't drop the error of the last split bio
scrub rewrote P/Q that disagreed with the data, on a read-only mount
too. The mismatches are still verified and counted. It comes before
patch 9 so that no point in the series rebuilds data in a read-only
-scrub. Fixes: 5a6ac9eacb49.
+scrub. Fixes: 5a6ac9eacb49 ("Btrfs, raid56: support parity scrub on
+raid56", v3.19).
Patch 9 lets raid56 parity scrub repair a raid6 data strip that has
no csum. When the scrubbed P and Q both disagree with the data in a
of Q does not repeat it. RAID5 has no such redundancy and keeps
today's behaviour. After an abort the rebuilt data is not written
either; a running dev-replace still copies parity to its target.
-Fixes: 5a6ac9eacb49.
+Fixes: 5a6ac9eacb49 ("Btrfs, raid56: support parity scrub on raid56",
+v3.19).
Patch 10 counts parity mismatches found by raid56 parity scrub as
corruption errors: in the device stats of the data member patch 9
"fixed up error / checksum error ... on dev X" lines mapped the sector
with a non-READ op, which on raid56 returns the full-stripe write map,
so they named the device of the full stripe's first data strip: damage
-on one device was reported on most of them. Fixes: 0096580713ff.
+on one device was reported on most of them. Fixes: 0096580713ff
+("btrfs: scrub: introduce error reporting functionality for
+scrub_stripe", v6.4).
Patch 12 counts the data repairs a raid56 parity scrub makes. The
parity scrub of the P/Q holder reads and repairs the full stripe's data
when that member was scrubbed last. Repairs are now counted once, by
the parity scrub only when the data device is not being scrubbed at the
same time, and not at all in a read-only scrub or after an abort.
-Fixes: 1009254bf22a.
+Fixes: 1009254bf22a ("btrfs: scrub: use scrub_stripe to implement RAID56
+P/Q scrub", v6.4).
Testing: a raid5 filesystem filled with block-location-identifying data
in partially filled stripes, two devices corrupted in chosen rows, every
was reviewed, and the clone/dedupe regression tests run clean with both
patches.
-Patch 1 Fixes: 3115deb381e9, patch 2 Fixes: 05a5a7621ce6.
+Patch 1 Fixes: 3115deb381e9 ("btrfs: reflink: make copy_inline_to_page()
+to be subpage compatible", v5.14), patch 2 Fixes: 05a5a7621ce6 ("Btrfs:
+implement full reflink support for inline extents", v5.7).
Zygo Blaxell (2):
btrfs: reflink: zero the right part of the folio after copying inline