Although
commit e6c4efd87ab0 ("btrfs: Fix and enhance merge_extent_mapping() to insert best fitted extent map")
fixed up the negetive em->len, it has introduced several regressions, several has been fixed by
commit 32be3a1ac6d0 ("btrfs: Fix the wrong condition judgment about subset extent map"),
commit 8dff9c853410 ("Btrfs: deal with duplciates during extent_map insertion in btrfs_get_extent") and
commit 8e2bd3b7fac9 ("Btrfs: deal with existing encompassing extent map in btrfs_get_extent()").
Unfortunately, there is one more regression which is caught recently in
our test farm, more details are explained in patch 7.
While debugging the above issue, I found that all of these bugs are caused
by some racy situations, which can be very tricky to reproduce, so I made
several extent map specific test cases in btrfs's selftest framework.
Patch 1-2 are some preparatory work.
Patch 3-5 are regression tests for handling EEXIST from adding extent map.
Patch 6-7 are fixing two bugs which can be reproduced by the above test cases.
Patch 8-10 are debugging wise, so that we can know what happened easily.
Liu Bo (10):
Btrfs: add helper for em merge logic
Btrfs: move extent map specific code to extent_map.c
Btrfs: add extent map selftests
Btrfs: extent map selftest: buffered write vs dio read
Btrfs: extent map selftest: dio write vs dio read
Btrfs: fix incorrect block_len in merge_extent_mapping
Btrfs: fix unexpected EEXIST from btrfs_get_extent
Btrfs: add WARN_ONCE to detect unexpected error from
merge_extent_mapping
Btrfs: add tracepoint for em's EEXIST case
Btrfs: noinline merge_extent_mapping
fs/btrfs/Makefile | 2 +-
fs/btrfs/extent_map.c | 120 +++++++++++++
fs/btrfs/extent_map.h | 2 +
fs/btrfs/inode.c | 108 +-----------
fs/btrfs/tests/btrfs-tests.c | 3 +
fs/btrfs/tests/btrfs-tests.h | 1 +
fs/btrfs/tests/extent-map-tests.c | 363 ++++++++++++++++++++++++++++++++++++++
include/trace/events/btrfs.h | 35 ++++
8 files changed, 526 insertions(+), 108 deletions(-)
create mode 100644 fs/btrfs/tests/extent-map-tests.c
--
2.9.4
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
