fstests: btrfs/15[78] btrfs/14[23]: Use more accurate devid/phsyical for corruption
cover fstests: btrfs/15[78] btrfs/14[23]: Use more accurate devid/phsyical for corruption
Btrfs-progs commit c501c9e3b816 ("btrfs-progs: mkfs: match devid order to the stripe index") changed the chunk layout, making it more human friendly for manually offset calculation.
However btrfs/14[23] btrfs/15[78] are using hard-coded devid sequence for corruption, thus they can't stand such change and fail.
This patchset will handle such problems in patch 2 and 3, to use more accurate helper to replace the hard-coded one.
And the first patch will address a bad _notrun messages, which is mostly related to btrfs/14[23].
Qu Wenruo (3):
fstests: common: Use more accurate kernel config for
_require_fail_make_request
fstests: btrfs/14[23]: Use proper help to get both devid and physical
offset for corruption.
fstests: btrfs/15[78]: Use proper helper to get both devid and
physical offset for corruption
common/rc | 2 +- tests/btrfs/142 | 50 +++++++++++++++++++++++++++++------------- tests/btrfs/142.out | 2 -- tests/btrfs/143 | 48 +++++++++++++++++++++++++++++----------- tests/btrfs/143.out | 2 -- tests/btrfs/157 | 53 +++++++++++++++++++++++++++++---------------- tests/btrfs/157.out | 4 ---- tests/btrfs/158 | 48 +++++++++++++++++++++++++--------------- tests/btrfs/158.out | 4 ---- 9 files changed, 136 insertions(+), 77 deletions(-)
Signature · 1 line
2.23.0
PATCH 1/3 fstests: common: Use more accurate kernel config for _require_fail_make_request
Just enabling CONFIG_FAIL_MAKE_REQUEST will not fulfill _require_fail_make_request.
It's CONFIG_FAULT_INJECTION_DEBUG_FS.
--- common/rc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
common/rc +1 -1
@@ -2357,7 +2357,7 @@ _require_fail_make_request() { [ -f "$DEBUGFS_MNT/fail_make_request/probability" ] \ || _notrun "$DEBUGFS_MNT/fail_make_request \ - not found. Seems that CONFIG_FAIL_MAKE_REQUEST kernel config option not enabled" + not found. Seems that CONFIG_FAULT_INJECTION_DEBUG_FS kernel config option not enabled" } # Disable extent zeroing for ext4 on the given device
Signature · 1 line
2.23.0
Re: [PATCH 1/3] fstests: common: Use more accurate kernel config for _require_fail_make_request
Quoted 7 lines · Qu Wenruo
Just enabling
CONFIG_FAIL_MAKE_REQUESTwill not fulfill_require_fail_make_request.It's
CONFIG_FAULT_INJECTION_DEBUG_FS.Signed-off-by Qu Wenruo <[email protected]>
Thanks.
Quoted 20 lines
--- common/rc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)common/rc +1
-2@@ -2357,7 +2357,7 @@ _require_fail_make_request() { [ -f "$DEBUGFS_MNT/fail_make_request/probability" ] \ || _notrun "$DEBUGFS_MNT/fail_make_request \ - not found. Seems that CONFIG_FAIL_MAKE_REQUEST kernel config option not enabled" + not found. Seems that CONFIG_FAULT_INJECTION_DEBUG_FS kernel config option not enabled" } # Disable extent zeroing for ext4 on the given device --2.23.0
Signature · 3 lines
Filipe David Manana, “Whether you think you can, or you think you can't — you're right.”
PATCH 2/3 fstests: btrfs/14[23]: Use proper help to get both devid and physical offset for corruption.
In reply to Qu Wenruo[BUG]
When using btrfs-progs v5.4, btrfs/142 and btrfs/143 will fail:
btrfs/142 1s ... - output mismatch (see xfstests/results//btrfs/142.out.bad)
--- tests/btrfs/142.out 2018-09-16 21:30:48.505104287 +0100
+++ xfstests/results//btrfs/142.out.bad
2019-12-10 15:35:40.280392626 +0000
@@ -3,37 +3,37 @@
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
wrote 65536/65536 bytes
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
-XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
-XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
-XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
...
(Run 'diff -u xfstests/tests/btrfs/142.out xfstests/results//btrfs/142.out.bad' to see the entire diff)[CAUSE]
Btrfs/14[23] test whether a read on corrupted stripe will re-silver itself.
Such test by its nature will need to modify on-disk data, thus need to get the btrfs logical -> physical mapping, which is done by near hard-coded lookup function, which rely on certain stripe:devid sequence.
Recent btrfs-progs commit c501c9e3b816 ("btrfs-progs: mkfs: match devid order to the stripe index") changes how we use devices in mkfs.btrfs, this caused a change in chunk layout, and break the hard-coded stripe:devid sequence.
[FIX]
This patch will do full devid and physical offset lookup, instead of old physical offset only lookup.
The only assumption made is, mkfs.btrfs assigns devid sequentially for its devices.
Which means, for "mkfs.btrfs $dev1 $dev2 $dev3", we get devid 1 for $dev1, devid 2 for $dev2, and so on.
This change will allow btrfs/14[23] to handle even future chunk layout change. (Although I hope this will never happen again).
This also addes extra debug output (although less than 10 lines) into $seqres.full, just in case when layout changes and current lookup can't handle it, developer can still pindown the problem easily.
--- tests/btrfs/142 | 50 +++++++++++++++++++++++++++++++-------------- tests/btrfs/142.out | 2 -- tests/btrfs/143 | 48 +++++++++++++++++++++++++++++++------------ tests/btrfs/143.out | 2 -- 4 files changed, 70 insertions(+), 32 deletions(-)
tests/btrfs/142 +35 -15
@@ -47,30 +47,45 @@ _require_command "$FILEFRAG_PROG" filefrag get_physical() { - # $1 is logical address - # print chunk tree and find devid 2 which is $SCRATCH_DEV + local logical=$1 + local stripe=$2 $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep $1 -A 6 | awk '($1 ~ /stripe/ && $3 ~ /devid/ && $4 ~ /1/) { print $6 }' + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }" } +get_devid() +{ + local logical=$1 + local stripe=$2 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }" +} -SYSFS_BDEV=`_sysfs_dev $SCRATCH_DEV` +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" +} start_fail() { + local sysfs_bdev="$1" echo 100 > $DEBUGFS_MNT/fail_make_request/probability echo 2 > $DEBUGFS_MNT/fail_make_request/times echo 1 > $DEBUGFS_MNT/fail_make_request/task-filter echo 0 > $DEBUGFS_MNT/fail_make_request/verbose - echo 1 > $SYSFS_BDEV/make-it-fail + echo 1 > $sysfs_bdev/make-it-fail } stop_fail() { + local sysfs_bdev="$1" echo 0 > $DEBUGFS_MNT/fail_make_request/probability echo 0 > $DEBUGFS_MNT/fail_make_request/times echo 0 > $DEBUGFS_MNT/fail_make_request/task-filter - echo 0 > $SYSFS_BDEV/make-it-fail + echo 0 > $sysfs_bdev/make-it-fail } _scratch_dev_pool_get 2
@@ -87,17 +102,23 @@ _scratch_mount -o nospace_cache,nodatasum $XFS_IO_PROG -f -d -c "pwrite -S 0xaa -b 128K 0 128K" "$SCRATCH_MNT/foobar" |\ _filter_xfs_io_offset -# step 2, corrupt the first 64k of one copy (on SCRATCH_DEV which is the first -# one in $SCRATCH_DEV_POOL +# step 2, corrupt the first 64k of stripe #1 echo "step 2......corrupt file extent" >>$seqres.full ${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar >> $seqres.full logical_in_btrfs=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` -physical_on_scratch=`get_physical ${logical_in_btrfs}` +physical=`get_physical ${logical_in_btrfs} 1` +devid=$(get_devid ${logical_in_btrfs} 1) +target_dev=$(get_device_path $devid) + +SYSFS_BDEV=`_sysfs_dev $target_dev` _scratch_unmount -$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical_on_scratch 64K" $SCRATCH_DEV |\ - _filter_xfs_io_offset +$BTRFS_UTIL_PROG ins dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical_in_btrfs -A 6 >> $seqres.full +echo "Corrupt stripe 1 devid $devid devpath $target_dev physical $physical" \ + >> $seqres.full +$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical 64K" $target_dev > /dev/null _scratch_mount -o nospace_cache
@@ -106,8 +127,7 @@ echo "step 3......repair the bad copy" >>$seqres.full # since raid1 consists of two copies, and the bad copy was put on stripe #1 # while the good copy lies on stripe #0, the bad copy only gets access when the -# reader's pid % 2 == 1 is true -start_fail +start_fail $SYSFS_BDEV while [[ -z ${result} ]]; do # enable task-filter only fails the following dio read so the repair is # supposed to work.
@@ -117,12 +137,12 @@ while [[ -z ${result} ]]; do exec $XFS_IO_PROG -d -c \"pread -b 128K 0 128K\" \"$SCRATCH_MNT/foobar\" fi"); done -stop_fail +stop_fail $SYSFS_BDEV _scratch_unmount # check if the repair works -$XFS_IO_PROG -c "pread -v -b 512 $physical_on_scratch 512" $SCRATCH_DEV |\ +$XFS_IO_PROG -c "pread -v -b 512 $physical 512" $target_dev | \ _filter_xfs_io_offset _scratch_dev_pool_put
tests/btrfs/142.out +0 -2
@@ -1,8 +1,6 @@ QA output created by 142 wrote 131072/131072 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
tests/btrfs/143 +35 -13
100 lines not on this page · raw message
tests/btrfs/143.out +0 -2
@@ -1,8 +1,6 @@ QA output created by 143 wrote 131072/131072 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
Signature · 1 line
2.23.0
Re: [PATCH 2/3] fstests: btrfs/14[23]: Use proper help to get both devid and physical offset for corruption.
Quoted 41 lines · Qu Wenruo
[BUG]
When using btrfs-progs v5.4, btrfs/142 and btrfs/143 will fail:
btrfs/142 1s ... - output mismatch (see xfstests/results//btrfs/142.out.bad)--- tests/btrfs/142.out 2018-09-16 21:30:48.505104287 +0100 +++ xfstests/results//btrfs/142.out.bad 2019-12-10 15:35:40.280392626 +0000 @@ -3,37 +3,37 @@ XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) wrote 65536/65536 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ ... (Run 'diff -u xfstests/tests/btrfs/142.out xfstests/results//btrfs/142.out.bad' to see the entire diff)[CAUSE]
Btrfs/14[23] test whether a read on corrupted stripe will re-silver itself.
Such test by its nature will need to modify on-disk data, thus need to get the btrfs logical -> physical mapping, which is done by near hard-coded lookup function, which rely on certain stripe:devid sequence.Recent btrfs-progs commit
c501c9e3b816("btrfs-progs: mkfs: match devid order to the stripe index") changes how we use devices in mkfs.btrfs, this caused a change in chunk layout, and break the hard-coded stripe:devid sequence.[FIX]
This patch will do full devid and physical offset lookup, instead of old physical offset only lookup.The only assumption made is, mkfs.btrfs assigns devid sequentially for its devices.
Which means, for "mkfs.btrfs $dev1 $dev2 $dev3", we get devid 1 for $dev1, devid 2 for $dev2, and so on.This change will allow btrfs/14[23] to handle even future chunk layout change. (Although I hope this will never happen again).
Quoted 6 lines
This also addes extra debug output (although less than 10 lines) into $seqres.full, just in case when layout changes and current lookup can't handle it, developer can still pindown the problem easily.
Reported-by Filipe Manana <[email protected]>Signed-off-by Qu Wenruo <[email protected]>
Quoted 251 lines
--- tests/btrfs/142 | 50 +++++++++++++++++++++++++++++++-------------- tests/btrfs/142.out | 2 -- tests/btrfs/143 | 48 +++++++++++++++++++++++++++++++------------ tests/btrfs/143.out | 2 -- 4 files changed, 70 insertions(+), 32 deletions(-)tests/btrfs/142 +35
-15@@ -47,30 +47,45 @@ _require_command "$FILEFRAG_PROG" filefrag get_physical() { - # $1 is logical address - # print chunk tree and find devid 2 which is $SCRATCH_DEV + local logical=$1 + local stripe=$2 $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep $1 -A 6 | awk '($1 ~ /stripe/ && $3 ~ /devid/ && $4 ~ /1/) { print $6 }' + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }" } +get_devid() +{ + local logical=$1 + local stripe=$2 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }" +} -SYSFS_BDEV=`_sysfs_dev $SCRATCH_DEV` +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" +} start_fail() { + local sysfs_bdev="$1" echo 100 > $DEBUGFS_MNT/fail_make_request/probability echo 2 > $DEBUGFS_MNT/fail_make_request/times echo 1 > $DEBUGFS_MNT/fail_make_request/task-filter echo 0 > $DEBUGFS_MNT/fail_make_request/verbose - echo 1 > $SYSFS_BDEV/make-it-fail + echo 1 > $sysfs_bdev/make-it-fail } stop_fail() { + local sysfs_bdev="$1" echo 0 > $DEBUGFS_MNT/fail_make_request/probability echo 0 > $DEBUGFS_MNT/fail_make_request/times echo 0 > $DEBUGFS_MNT/fail_make_request/task-filter - echo 0 > $SYSFS_BDEV/make-it-fail + echo 0 > $sysfs_bdev/make-it-fail } _scratch_dev_pool_get 2@@ -87,17 +102,23 @@ _scratch_mount -o nospace_cache,nodatasum $XFS_IO_PROG -f -d -c "pwrite -S 0xaa -b 128K 0 128K" "$SCRATCH_MNT/foobar" |\ _filter_xfs_io_offset -# step 2, corrupt the first 64k of one copy (on SCRATCH_DEV which is the first -# one in $SCRATCH_DEV_POOL +# step 2, corrupt the first 64k of stripe #1 echo "step 2......corrupt file extent" >>$seqres.full ${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar >> $seqres.full logical_in_btrfs=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` -physical_on_scratch=`get_physical ${logical_in_btrfs}` +physical=`get_physical ${logical_in_btrfs} 1` +devid=$(get_devid ${logical_in_btrfs} 1) +target_dev=$(get_device_path $devid) + +SYSFS_BDEV=`_sysfs_dev $target_dev` _scratch_unmount -$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical_on_scratch 64K" $SCRATCH_DEV |\ - _filter_xfs_io_offset +$BTRFS_UTIL_PROG ins dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical_in_btrfs -A 6 >> $seqres.full +echo "Corrupt stripe 1 devid $devid devpath $target_dev physical $physical" \ + >> $seqres.full +$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical 64K" $target_dev > /dev/null _scratch_mount -o nospace_cache@@ -106,8 +127,7 @@ echo "step 3......repair the bad copy" >>$seqres.full # since raid1 consists of two copies, and the bad copy was put on stripe #1 # while the good copy lies on stripe #0, the bad copy only gets access when the -# reader's pid % 2 == 1 is true -start_fail +start_fail $SYSFS_BDEV while [[ -z ${result} ]]; do # enable task-filter only fails the following dio read so the repair is # supposed to work.@@ -117,12 +137,12 @@ while [[ -z ${result} ]]; do exec $XFS_IO_PROG -d -c \"pread -b 128K 0 128K\" \"$SCRATCH_MNT/foobar\" fi"); done -stop_fail +stop_fail $SYSFS_BDEV _scratch_unmount # check if the repair works -$XFS_IO_PROG -c "pread -v -b 512 $physical_on_scratch 512" $SCRATCH_DEV |\ +$XFS_IO_PROG -c "pread -v -b 512 $physical 512" $target_dev | \ _filter_xfs_io_offset _scratch_dev_pool_puttests/btrfs/142.out +0
-2@@ -1,8 +1,6 @@ QA output created by 142 wrote 131072/131072 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................tests/btrfs/143 +35
-13@@ -54,30 +54,48 @@ _require_command "$FILEFRAG_PROG" filefrag get_physical() { - # $1 is logical address - # print chunk tree and find devid 2 which is $SCRATCH_DEV + local logical=$1 + local stripe=$2 $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep $1 -A 6 | awk '($1 ~ /stripe/ && $3 ~ /devid/ && $4 ~ /1/) { print $6 }' + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }" +} + +get_devid() +{ + local logical=$1 + local stripe=$2 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }" +} + +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" } SYSFS_BDEV=`_sysfs_dev $SCRATCH_DEV` start_fail() { + local sysfs_bdev="$1" echo 100 > $DEBUGFS_MNT/fail_make_request/probability # the 1st one fails the first bio which is reading 4k (or more due to # readahead), and the 2nd one fails the retry of validation so that it # triggers read-repair echo 2 > $DEBUGFS_MNT/fail_make_request/times echo 0 > $DEBUGFS_MNT/fail_make_request/verbose - echo 1 > $SYSFS_BDEV/make-it-fail + echo 1 > $sysfs_bdev/make-it-fail } stop_fail() { + local sysfs_bdev="$1" echo 0 > $DEBUGFS_MNT/fail_make_request/probability echo 0 > $DEBUGFS_MNT/fail_make_request/times - echo 0 > $SYSFS_BDEV/make-it-fail + echo 0 > $sysfs_bdev/make-it-fail } _scratch_dev_pool_get 2@@ -94,17 +112,21 @@ _scratch_mount -o nospace_cache,nodatasum $XFS_IO_PROG -f -d -c "pwrite -S 0xaa -b 128K 0 128K" "$SCRATCH_MNT/foobar" |\ _filter_xfs_io_offset -# step 2, corrupt the first 64k of one copy (on SCRATCH_DEV which is the first -# one in $SCRATCH_DEV_POOL +# step 2, corrupt the first 64k of stripe #1 echo "step 2......corrupt file extent" >>$seqres.full ${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar >> $seqres.full logical_in_btrfs=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` -physical_on_scratch=`get_physical ${logical_in_btrfs}` +physical=`get_physical ${logical_in_btrfs} 1` +devid=$(get_devid ${logical_in_btrfs} 1) +target_dev=$(get_device_path $devid) +SYSFS_BDEV=`_sysfs_dev $target_dev` _scratch_unmount -$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical_on_scratch 64K" $SCRATCH_DEV |\ - _filter_xfs_io_offset + +echo "corrupt stripe 1 devid $devid devpath $target_dev physical $physical" \ + >> $seqres.full +$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical 64K" $target_dev > /dev/null _scratch_mount -o nospace_cache@@ -118,18 +140,18 @@ while [[ -z ${result} ]]; do # invalidate the page cache. _scratch_cycle_mount - start_fail + start_fail $SYSFS_BDEV result=$(bash -c " if [[ \$((\$\$ % 2)) -eq 1 ]]; then exec $XFS_IO_PROG -c \"pread 0 4K\" \"$SCRATCH_MNT/foobar\" fi"); - stop_fail + stop_fail $SYSFS_BDEV done _scratch_unmount # check if the repair works -$XFS_IO_PROG -c "pread -v -b 512 $physical_on_scratch 512" $SCRATCH_DEV |\ +$XFS_IO_PROG -c "pread -v -b 512 $physical 512" $target_dev |\ _filter_xfs_io_offset _scratch_dev_pool_puttests/btrfs/143.out +0
-2@@ -1,8 +1,6 @@ QA output created by 143 wrote 131072/131072 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................
Re: [PATCH 2/3] fstests: btrfs/14[23]: Use proper help to get both devid and physical offset for corruption.
In reply to Qu WenruoQuoted 61 lines · Qu Wenruo
[BUG]
When using btrfs-progs v5.4, btrfs/142 and btrfs/143 will fail:
btrfs/142 1s ... - output mismatch (see xfstests/results//btrfs/142.out.bad)--- tests/btrfs/142.out 2018-09-16 21:30:48.505104287 +0100 +++ xfstests/results//btrfs/142.out.bad 2019-12-10 15:35:40.280392626 +0000 @@ -3,37 +3,37 @@ XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) wrote 65536/65536 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ -XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ ... (Run 'diff -u xfstests/tests/btrfs/142.out xfstests/results//btrfs/142.out.bad' to see the entire diff)[CAUSE]
Btrfs/14[23] test whether a read on corrupted stripe will re-silver itself.
Such test by its nature will need to modify on-disk data, thus need to get the btrfs logical -> physical mapping, which is done by near hard-coded lookup function, which rely on certain stripe:devid sequence.Recent btrfs-progs commit
c501c9e3b816("btrfs-progs: mkfs: match devid order to the stripe index") changes how we use devices in mkfs.btrfs, this caused a change in chunk layout, and break the hard-coded stripe:devid sequence.[FIX]
This patch will do full devid and physical offset lookup, instead of old physical offset only lookup.The only assumption made is, mkfs.btrfs assigns devid sequentially for its devices.
Which means, for "mkfs.btrfs $dev1 $dev2 $dev3", we get devid 1 for $dev1, devid 2 for $dev2, and so on.This change will allow btrfs/14[23] to handle even future chunk layout change. (Although I hope this will never happen again).
This also addes extra debug output (although less than 10 lines) into $seqres.full, just in case when layout changes and current lookup can't handle it, developer can still pindown the problem easily.
Reported-by Filipe Manana <[email protected]>Signed-off-by Qu Wenruo <[email protected]>--- tests/btrfs/142 | 50 +++++++++++++++++++++++++++++++-------------- tests/btrfs/142.out | 2 -- tests/btrfs/143 | 48 +++++++++++++++++++++++++++++++------------ tests/btrfs/143.out | 2 -- 4 files changed, 70 insertions(+), 32 deletions(-)tests/btrfs/142 +0
-0@@ -47,30 +47,45 @@ _require_command "$FILEFRAG_PROG" filefrag
tests/btrfs/142 _require_command "$FILEFRAG_PROG" filefrag · line 54
Quoted 10 lines
tests/btrfs/142 +4
-3get_physical() { - # $1 is logical address - # print chunk tree and find devid 2 which is $SCRATCH_DEV + local logical=$1 + local stripe=$2 $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep $1 -A 6 | awk '($1 ~ /stripe/ && $3 ~ /devid/ && $4 ~ /1/) { print $6 }' + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }"
Should use $AWK_PROG instead of direct call to 'awk'.
tests/btrfs/142 _require_command "$FILEFRAG_PROG" filefrag · line 63
Quoted 9 lines
tests/btrfs/142 +7
-0} +get_devid() +{ + local logical=$1 + local stripe=$2 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }"
Same here.
Quoted 104 lines
tests/btrfs/142 +24
-12+} -SYSFS_BDEV=`_sysfs_dev $SCRATCH_DEV` +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" +} start_fail() { + local sysfs_bdev="$1" echo 100 > $DEBUGFS_MNT/fail_make_request/probability echo 2 > $DEBUGFS_MNT/fail_make_request/times echo 1 > $DEBUGFS_MNT/fail_make_request/task-filter echo 0 > $DEBUGFS_MNT/fail_make_request/verbose - echo 1 > $SYSFS_BDEV/make-it-fail + echo 1 > $sysfs_bdev/make-it-fail } stop_fail() { + local sysfs_bdev="$1" echo 0 > $DEBUGFS_MNT/fail_make_request/probability echo 0 > $DEBUGFS_MNT/fail_make_request/times echo 0 > $DEBUGFS_MNT/fail_make_request/task-filter - echo 0 > $SYSFS_BDEV/make-it-fail + echo 0 > $sysfs_bdev/make-it-fail } _scratch_dev_pool_get 2@@ -87,17 +102,23 @@ _scratch_mount -o nospace_cache,nodatasum $XFS_IO_PROG -f -d -c "pwrite -S 0xaa -b 128K 0 128K" "$SCRATCH_MNT/foobar" |\ _filter_xfs_io_offset -# step 2, corrupt the first 64k of one copy (on SCRATCH_DEV which is the first -# one in $SCRATCH_DEV_POOL +# step 2, corrupt the first 64k of stripe #1 echo "step 2......corrupt file extent" >>$seqres.full ${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar >> $seqres.full logical_in_btrfs=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` -physical_on_scratch=`get_physical ${logical_in_btrfs}` +physical=`get_physical ${logical_in_btrfs} 1` +devid=$(get_devid ${logical_in_btrfs} 1) +target_dev=$(get_device_path $devid) + +SYSFS_BDEV=`_sysfs_dev $target_dev` _scratch_unmount -$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical_on_scratch 64K" $SCRATCH_DEV |\ - _filter_xfs_io_offset +$BTRFS_UTIL_PROG ins dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical_in_btrfs -A 6 >> $seqres.full +echo "Corrupt stripe 1 devid $devid devpath $target_dev physical $physical" \ + >> $seqres.full +$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical 64K" $target_dev > /dev/null _scratch_mount -o nospace_cache@@ -106,8 +127,7 @@ echo "step 3......repair the bad copy" >>$seqres.full # since raid1 consists of two copies, and the bad copy was put on stripe #1 # while the good copy lies on stripe #0, the bad copy only gets access when the -# reader's pid % 2 == 1 is true -start_fail +start_fail $SYSFS_BDEV while [[ -z ${result} ]]; do # enable task-filter only fails the following dio read so the repair is # supposed to work.@@ -117,12 +137,12 @@ while [[ -z ${result} ]]; do exec $XFS_IO_PROG -d -c \"pread -b 128K 0 128K\" \"$SCRATCH_MNT/foobar\" fi"); done -stop_fail +stop_fail $SYSFS_BDEV _scratch_unmount # check if the repair works -$XFS_IO_PROG -c "pread -v -b 512 $physical_on_scratch 512" $SCRATCH_DEV |\ +$XFS_IO_PROG -c "pread -v -b 512 $physical 512" $target_dev | \ _filter_xfs_io_offset _scratch_dev_pool_puttests/btrfs/142.out +0
-2@@ -1,8 +1,6 @@ QA output created by 142 wrote 131072/131072 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................tests/btrfs/143 +0
-0@@ -54,30 +54,48 @@ _require_command "$FILEFRAG_PROG" filefrag
tests/btrfs/143 _require_command "$FILEFRAG_PROG" filefrag · line 61
Quoted 10 lines
tests/btrfs/143 +4
-3get_physical() { - # $1 is logical address - # print chunk tree and find devid 2 which is $SCRATCH_DEV + local logical=$1 + local stripe=$2 $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep $1 -A 6 | awk '($1 ~ /stripe/ && $3 ~ /devid/ && $4 ~ /1/) { print $6 }' + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }"
Same here.
tests/btrfs/143 _require_command "$FILEFRAG_PROG" filefrag · line 70
Quoted 9 lines
tests/btrfs/143 +9
-0+} + +get_devid() +{ + local logical=$1 + local stripe=$2 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep $logical -A 6 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }"
Same here.
Quoted 96 lines
tests/btrfs/143 +22
-10+} + +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" } SYSFS_BDEV=`_sysfs_dev $SCRATCH_DEV` start_fail() { + local sysfs_bdev="$1" echo 100 > $DEBUGFS_MNT/fail_make_request/probability # the 1st one fails the first bio which is reading 4k (or more due to # readahead), and the 2nd one fails the retry of validation so that it # triggers read-repair echo 2 > $DEBUGFS_MNT/fail_make_request/times echo 0 > $DEBUGFS_MNT/fail_make_request/verbose - echo 1 > $SYSFS_BDEV/make-it-fail + echo 1 > $sysfs_bdev/make-it-fail } stop_fail() { + local sysfs_bdev="$1" echo 0 > $DEBUGFS_MNT/fail_make_request/probability echo 0 > $DEBUGFS_MNT/fail_make_request/times - echo 0 > $SYSFS_BDEV/make-it-fail + echo 0 > $sysfs_bdev/make-it-fail } _scratch_dev_pool_get 2@@ -94,17 +112,21 @@ _scratch_mount -o nospace_cache,nodatasum $XFS_IO_PROG -f -d -c "pwrite -S 0xaa -b 128K 0 128K" "$SCRATCH_MNT/foobar" |\ _filter_xfs_io_offset -# step 2, corrupt the first 64k of one copy (on SCRATCH_DEV which is the first -# one in $SCRATCH_DEV_POOL +# step 2, corrupt the first 64k of stripe #1 echo "step 2......corrupt file extent" >>$seqres.full ${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar >> $seqres.full logical_in_btrfs=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` -physical_on_scratch=`get_physical ${logical_in_btrfs}` +physical=`get_physical ${logical_in_btrfs} 1` +devid=$(get_devid ${logical_in_btrfs} 1) +target_dev=$(get_device_path $devid) +SYSFS_BDEV=`_sysfs_dev $target_dev` _scratch_unmount -$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical_on_scratch 64K" $SCRATCH_DEV |\ - _filter_xfs_io_offset + +echo "corrupt stripe 1 devid $devid devpath $target_dev physical $physical" \ + >> $seqres.full +$XFS_IO_PROG -d -c "pwrite -S 0xbb -b 64K $physical 64K" $target_dev > /dev/null _scratch_mount -o nospace_cache@@ -118,18 +140,18 @@ while [[ -z ${result} ]]; do # invalidate the page cache. _scratch_cycle_mount - start_fail + start_fail $SYSFS_BDEV result=$(bash -c " if [[ \$((\$\$ % 2)) -eq 1 ]]; then exec $XFS_IO_PROG -c \"pread 0 4K\" \"$SCRATCH_MNT/foobar\" fi"); - stop_fail + stop_fail $SYSFS_BDEV done _scratch_unmount # check if the repair works -$XFS_IO_PROG -c "pread -v -b 512 $physical_on_scratch 512" $SCRATCH_DEV |\ +$XFS_IO_PROG -c "pread -v -b 512 $physical 512" $target_dev |\ _filter_xfs_io_offset _scratch_dev_pool_puttests/btrfs/143.out +0
-3@@ -1,8 +1,6 @@ QA output created by 143 wrote 131072/131072 bytes XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ XXXXXXXX: aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa ................ --
Quoted 2 lines
2.23.0
Other than that (which can probably be fixed at commit time by Eryu) it looks good and it solves the problem.
Thanks.
Signature · 3 lines
Filipe David Manana, “Whether you think you can, or you think you can't — you're right.”
PATCH 3/3 fstests: btrfs/15[78]: Use proper helper to get both devid and physical offset for corruption
In reply to Qu Wenruo[BUG]
When using btrfs-progs v5.4, btrfs/157 and btrfs/158 will fail:
btrfs/157 1s ... - output mismatch (see xfstests/results//btrfs/157.out.bad)
--- tests/btrfs/157.out 2018-09-16 21:30:48.505104287 +0100
+++ xfstests/results//btrfs/157.out.bad
2019-12-10 15:35:43.112390076 +0000
@@ -1,9 +1,9 @@
QA output created by 157
wrote 131072/131072 bytes at offset 0
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-wrote 65536/65536 bytes at offset 9437184
+wrote 65536/65536 bytes at offset 22020096
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-wrote 65536/65536 bytes at offset 9437184
...
(Run 'diff -u xfstests/tests/btrfs/157.out xfstests/results//btrfs/157.out.bad' to see the entire diff)
btrfs/158 2s ... - output mismatch (see xfstests/results//btrfs/158.out.bad)
--- tests/btrfs/158.out 2018-09-16 21:30:48.505104287 +0100
+++ xfstests/results//btrfs/158.out.bad
2019-12-10 15:35:44.844388521 +0000
@@ -1,9 +1,9 @@
QA output created by 158
wrote 131072/131072 bytes at offset 0
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-wrote 65536/65536 bytes at offset 9437184
+wrote 65536/65536 bytes at offset 22020096
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-wrote 65536/65536 bytes at offset 9437184
...
(Run 'diff -u xfstests/tests/btrfs/158.out xfstests/results//btrfs/158.out.bad' to see the entire diff)[CAUSE]
This two tests use physical offset as golden output, while mkfs.btrfs can do whatever it likes to arrange its chunk layout, thus physical offset is never reliable.
And btrfs-progs commit c501c9e3b816 ("btrfs-progs: mkfs: match devid order to the stripe index") just changed the layout.
So the output mismatch and failed.
[FIX]
In fact, that btrfs-progs commit not only changed offset, but also the device sequence.
So we can't just simply remove the physical offset, but also need to use proper helper to get both devid (as its device path) and physical offset for corruption.
As long as mkfs.btrfs still uses sequential devid, these tests should handle future chunk layout change without problem.
--- tests/btrfs/157 | 53 +++++++++++++++++++++++++++++---------------- tests/btrfs/157.out | 4 ---- tests/btrfs/158 | 48 +++++++++++++++++++++++++--------------- tests/btrfs/158.out | 4 ---- 4 files changed, 65 insertions(+), 44 deletions(-)
tests/btrfs/157 +34 -19
@@ -51,22 +51,30 @@ _require_scratch_dev_pool 4 _require_btrfs_command inspect-internal dump-tree _require_btrfs_fs_feature raid56 -get_physical_stripe0() +get_physical() { - $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep " DATA\|RAID6" -A 10 | \ - $AWK_PROG '($1 ~ /stripe/ && $3 ~ /devid/ && $2 ~ /0/) { print $6 }' + local stripe=$1 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep " DATA\|RAID6" -A 10 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }" } -get_physical_stripe1() +get_devid() { - $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep " DATA\|RAID6" -A 10 | \ - $AWK_PROG '($1 ~ /stripe/ && $3 ~ /devid/ && $2 ~ /1/) { print $6 }' + local stripe=$1 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep " DATA\|RAID6" -A 10 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }" +} + +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" } _scratch_dev_pool_get 4 -# step 1: create a raid6 btrfs and create a 4K file +# step 1: create a raid6 btrfs and create a 128K file echo "step 1......mkfs.btrfs" >>$seqres.full mkfs_opts="-d raid6 -b 1G"
@@ -80,18 +88,25 @@ _scratch_mount -o nospace_cache $XFS_IO_PROG -f -d -c "pwrite -S 0xaa 0 128K" -c "fsync" \ "$SCRATCH_MNT/foobar" | _filter_xfs_io +logical=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` _scratch_unmount -stripe_0=`get_physical_stripe0` -stripe_1=`get_physical_stripe1` -dev4=`echo $SCRATCH_DEV_POOL | awk '{print $4}'` -dev3=`echo $SCRATCH_DEV_POOL | awk '{print $3}'` - -# step 2: corrupt the 1st and 2nd stripe (stripe 0 and 1) -echo "step 2......simulate bitrot at offset $stripe_0 of device_4($dev4) and offset $stripe_1 of device_3($dev3)" >>$seqres.full - -$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $stripe_0 64K" $dev4 | _filter_xfs_io -$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $stripe_1 64K" $dev3 | _filter_xfs_io +phy0=$(get_physical 0) +devid0=$(get_devid 0) +devpath0=$(get_device_path $devid0) +phy1=$(get_physical 1) +devid1=$(get_devid 1) +devpath1=$(get_device_path $devid1) + +# step 2: corrupt stripe #0 and #1 +echo "step 2......simulate bitrot at:" >>$seqres.full +echo " ......stripe #0: devid $devid0 devpath $devpath0 phy $phy0" \ + >>$seqres.full +echo " ......stripe #1: devid $devid1 devpath $devpath1 phy $phy1" \ + >>$seqres.full + +$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $phy0 64K" $devpath0 > /dev/null +$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $phy1 64K" $devpath1 > /dev/null # step 3: read foobar to repair the bitrot echo "step 3......repair the bitrot" >> $seqres.full
tests/btrfs/157.out +0 -4
@@ -1,10 +1,6 @@ QA output created by 157 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) 0200000 aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa * 0400000
tests/btrfs/158 +31 -17
69 lines not on this page · raw message
tests/btrfs/158.out +0 -4
@@ -1,10 +1,6 @@ QA output created by 158 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) 0000000 aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa * 0400000
Signature · 1 line
2.23.0
Re: [PATCH 3/3] fstests: btrfs/15[78]: Use proper helper to get both devid and physical offset for corruption
Quoted 50 lines · Qu Wenruo
[BUG]
When using btrfs-progs v5.4, btrfs/157 and btrfs/158 will fail:btrfs/157 1s ... - output mismatch (see xfstests/results//btrfs/157.out.bad)
--- tests/btrfs/157.out 2018-09-16 21:30:48.505104287 +0100 +++ xfstests/results//btrfs/157.out.bad 2019-12-10 15:35:43.112390076 +0000 @@ -1,9 +1,9 @@ QA output created by 157 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 +wrote 65536/65536 bytes at offset 22020096 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 ... (Run 'diff -u xfstests/tests/btrfs/157.out xfstests/results//btrfs/157.out.bad' to see the entire diff) btrfs/158 2s ... - output mismatch (see xfstests/results//btrfs/158.out.bad) --- tests/btrfs/158.out 2018-09-16 21:30:48.505104287 +0100 +++ xfstests/results//btrfs/158.out.bad 2019-12-10 15:35:44.844388521 +0000 @@ -1,9 +1,9 @@ QA output created by 158 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 +wrote 65536/65536 bytes at offset 22020096 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 ... (Run 'diff -u xfstests/tests/btrfs/158.out xfstests/results//btrfs/158.out.bad' to see the entire diff)[CAUSE]
This two tests use physical offset as golden output, while mkfs.btrfs can do whatever it likes to arrange its chunk layout, thus physical offset is never reliable.And btrfs-progs commit
c501c9e3b816("btrfs-progs: mkfs: match devid order to the stripe index") just changed the layout.So the output mismatch and failed.
[FIX]
In fact, that btrfs-progs commit not only changed offset, but also the device sequence.So we can't just simply remove the physical offset, but also need to use proper helper to get both devid (as its device path) and physical offset for corruption.
Quoted 5 lines
As long as mkfs.btrfs still uses sequential devid, these tests should handle future chunk layout change without problem.
Reported-by Filipe Manana <[email protected]>Signed-off-by Qu Wenruo <[email protected]>
Re: [PATCH 3/3] fstests: btrfs/15[78]: Use proper helper to get both devid and physical offset for corruption
In reply to Qu WenruoQuoted 71 lines · Qu Wenruo
[BUG]
When using btrfs-progs v5.4, btrfs/157 and btrfs/158 will fail:btrfs/157 1s ... - output mismatch (see xfstests/results//btrfs/157.out.bad)
--- tests/btrfs/157.out 2018-09-16 21:30:48.505104287 +0100 +++ xfstests/results//btrfs/157.out.bad 2019-12-10 15:35:43.112390076 +0000 @@ -1,9 +1,9 @@ QA output created by 157 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 +wrote 65536/65536 bytes at offset 22020096 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 ... (Run 'diff -u xfstests/tests/btrfs/157.out xfstests/results//btrfs/157.out.bad' to see the entire diff) btrfs/158 2s ... - output mismatch (see xfstests/results//btrfs/158.out.bad) --- tests/btrfs/158.out 2018-09-16 21:30:48.505104287 +0100 +++ xfstests/results//btrfs/158.out.bad 2019-12-10 15:35:44.844388521 +0000 @@ -1,9 +1,9 @@ QA output created by 158 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 +wrote 65536/65536 bytes at offset 22020096 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 ... (Run 'diff -u xfstests/tests/btrfs/158.out xfstests/results//btrfs/158.out.bad' to see the entire diff)[CAUSE]
This two tests use physical offset as golden output, while mkfs.btrfs can do whatever it likes to arrange its chunk layout, thus physical offset is never reliable.And btrfs-progs commit
c501c9e3b816("btrfs-progs: mkfs: match devid order to the stripe index") just changed the layout.So the output mismatch and failed.
[FIX]
In fact, that btrfs-progs commit not only changed offset, but also the device sequence.So we can't just simply remove the physical offset, but also need to use proper helper to get both devid (as its device path) and physical offset for corruption.
As long as mkfs.btrfs still uses sequential devid, these tests should handle future chunk layout change without problem.
Reported-by Filipe Manana <[email protected]>Signed-off-by Qu Wenruo <[email protected]>--- tests/btrfs/157 | 53 +++++++++++++++++++++++++++++---------------- tests/btrfs/157.out | 4 ---- tests/btrfs/158 | 48 +++++++++++++++++++++++++--------------- tests/btrfs/158.out | 4 ---- 4 files changed, 65 insertions(+), 44 deletions(-)tests/btrfs/157 +0
-0@@ -51,22 +51,30 @@ _require_scratch_dev_pool 4 _require_btrfs_command inspect-internal dump-tree _require_btrfs_fs_feature raid56
tests/btrfs/157 _require_scratch_dev_pool 4 · line 59
Quoted 10 lines
tests/btrfs/157 +5
-4-get_physical_stripe0() +get_physical() { - $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep " DATA\|RAID6" -A 10 | \ - $AWK_PROG '($1 ~ /stripe/ && $3 ~ /devid/ && $2 ~ /0/) { print $6 }' + local stripe=$1 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep " DATA\|RAID6" -A 10 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }"
Same as in the other patch, use $AWK_PROG instead.
tests/btrfs/157 _require_scratch_dev_pool 4 · line 67
Quoted 12 lines
tests/btrfs/157 +5
-4} -get_physical_stripe1() +get_devid() { - $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep " DATA\|RAID6" -A 10 | \ - $AWK_PROG '($1 ~ /stripe/ && $3 ~ /devid/ && $2 ~ /1/) { print $6 }' + local stripe=$1 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep " DATA\|RAID6" -A 10 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }"
Same here.
Quoted 73 lines
tests/btrfs/157 +24
-11+} + +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" } _scratch_dev_pool_get 4 -# step 1: create a raid6 btrfs and create a 4K file +# step 1: create a raid6 btrfs and create a 128K file echo "step 1......mkfs.btrfs" >>$seqres.full mkfs_opts="-d raid6 -b 1G"@@ -80,18 +88,25 @@ _scratch_mount -o nospace_cache $XFS_IO_PROG -f -d -c "pwrite -S 0xaa 0 128K" -c "fsync" \ "$SCRATCH_MNT/foobar" | _filter_xfs_io +logical=`${FILEFRAG_PROG} -v $SCRATCH_MNT/foobar | _filter_filefrag | cut -d '#' -f 1` _scratch_unmount -stripe_0=`get_physical_stripe0` -stripe_1=`get_physical_stripe1` -dev4=`echo $SCRATCH_DEV_POOL | awk '{print $4}'` -dev3=`echo $SCRATCH_DEV_POOL | awk '{print $3}'` - -# step 2: corrupt the 1st and 2nd stripe (stripe 0 and 1) -echo "step 2......simulate bitrot at offset $stripe_0 of device_4($dev4) and offset $stripe_1 of device_3($dev3)" >>$seqres.full - -$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $stripe_0 64K" $dev4 | _filter_xfs_io -$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $stripe_1 64K" $dev3 | _filter_xfs_io +phy0=$(get_physical 0) +devid0=$(get_devid 0) +devpath0=$(get_device_path $devid0) +phy1=$(get_physical 1) +devid1=$(get_devid 1) +devpath1=$(get_device_path $devid1) + +# step 2: corrupt stripe #0 and #1 +echo "step 2......simulate bitrot at:" >>$seqres.full +echo " ......stripe #0: devid $devid0 devpath $devpath0 phy $phy0" \ + >>$seqres.full +echo " ......stripe #1: devid $devid1 devpath $devpath1 phy $phy1" \ + >>$seqres.full + +$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $phy0 64K" $devpath0 > /dev/null +$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $phy1 64K" $devpath1 > /dev/null # step 3: read foobar to repair the bitrot echo "step 3......repair the bitrot" >> $seqres.fulltests/btrfs/157.out +0
-4@@ -1,10 +1,6 @@ QA output created by 157 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) 0200000 aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa * 0400000tests/btrfs/158 +0
-0@@ -43,22 +43,30 @@ _require_scratch_dev_pool 4 _require_btrfs_command inspect-internal dump-tree _require_btrfs_fs_feature raid56
tests/btrfs/158 _require_scratch_dev_pool 4 · line 51
Quoted 10 lines
tests/btrfs/158 +5
-4-get_physical_stripe0() +get_physical() { - $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep " DATA\|RAID6" -A 10 | \ - $AWK_PROG '($1 ~ /stripe/ && $3 ~ /devid/ && $2 ~ /0/) { print $6 }' + local stripe=$1 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep " DATA\|RAID6" -A 10 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }"
Same here.
tests/btrfs/158 _require_scratch_dev_pool 4 · line 59
Quoted 12 lines
tests/btrfs/158 +5
-4} -get_physical_stripe1() +get_devid() { - $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ - grep " DATA\|RAID6" -A 10 | \ - $AWK_PROG '($1 ~ /stripe/ && $3 ~ /devid/ && $2 ~ /1/) { print $6 }' + local stripe=$1 + $BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \ + grep " DATA\|RAID6" -A 10 | \ + awk "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }"
Same here.
Quoted 61 lines
tests/btrfs/158 +21
-9+} + +get_device_path() +{ + local devid=$1 + echo "$SCRATCH_DEV_POOL" | awk "{print \$$devid}" } _scratch_dev_pool_get 4 -# step 1: create a raid6 btrfs and create a 4K file +# step 1: create a raid6 btrfs and create a 128K file echo "step 1......mkfs.btrfs" >>$seqres.full mkfs_opts="-d raid6 -b 1G"@@ -74,16 +82,22 @@ $XFS_IO_PROG -f -d -c "pwrite -S 0xaa 0 128K" -c "fsync" \ _scratch_unmount -stripe_0=`get_physical_stripe0` -stripe_1=`get_physical_stripe1` -dev4=`echo $SCRATCH_DEV_POOL | awk '{print $4}'` -dev3=`echo $SCRATCH_DEV_POOL | awk '{print $3}'` +phy0=$(get_physical 0) +devid0=$(get_devid 0) +devpath0=$(get_device_path $devid0) +phy1=$(get_physical 1) +devid1=$(get_devid 1) +devpath1=$(get_device_path $devid1) # step 2: corrupt the 1st and 2nd stripe (stripe 0 and 1) -echo "step 2......simulate bitrot at offset $stripe_0 of device_4($dev4) and offset $stripe_1 of device_3($dev3)" >>$seqres.full - -$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $stripe_0 64K" $dev4 | _filter_xfs_io -$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $stripe_1 64K" $dev3 | _filter_xfs_io +echo "step 2......simulate bitrot at:" >>$seqres.full +echo " ......stripe #0: devid $devid0 devpath $devpath0 phy $phy0" \ + >>$seqres.full +echo " ......stripe #1: devid $devid1 devpath $devpath1 phy $phy1" \ + >>$seqres.full + +$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $phy0 64K" $devpath0 > /dev/null +$XFS_IO_PROG -f -d -c "pwrite -S 0xbb $phy1 64K" $devpath1 > /dev/null # step 3: scrub filesystem to repair the bitrot echo "step 3......repair the bitrot" >> $seqres.fulltests/btrfs/158.out +0
-5@@ -1,10 +1,6 @@ QA output created by 158 wrote 131072/131072 bytes at offset 0 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -wrote 65536/65536 bytes at offset 9437184 -XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) 0000000 aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa * 0400000 --
Quoted 2 lines
2.23.0
Other than that (which can probably be fixed at commit time by Eryu) it looks good and it solves the problem.
Thanks.
Signature · 3 lines
Filipe David Manana, “Whether you think you can, or you think you can't — you're right.”