Re: [PATCH v2] fstests: regression test for btrfs buffered read's repair

2017-04-28 Thread Liu Bo 
On Fri, Apr 28, 2017 at 10:52:12AM +0100, Filipe Manana wrote:
> On Wed, Apr 26, 2017 at 7:09 PM, Liu Bo  wrote:
> > This case tests whether buffered read can repair the bad copy if we
> > have a good copy.
> >
> > Commit 20a7db8ab3f2 ("btrfs: add dummy callback for readpage_io_failed
> > and drop checks") introduced the regression.
> >
> > The upstream fix is
> > Btrfs: bring back repair during read
> 
> Same issue as reported for the other new test (btrfs/140), the test
> fails on a kernel with the mentioned patch. Seems like it does wrong
> assumptions somewhere (I haven't investigated, just run the test):

Thanks for running the test, what I've missed is that chunk start offset depends
on disk size, so I'll add a filesystem size limit to mkfs.

Thanks,

-liubo

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] fstests: regression test for btrfs buffered read's repair

2017-04-28 Thread Filipe Manana 
On Wed, Apr 26, 2017 at 7:09 PM, Liu Bo  wrote:
> This case tests whether buffered read can repair the bad copy if we
> have a good copy.
>
> Commit 20a7db8ab3f2 ("btrfs: add dummy callback for readpage_io_failed
> and drop checks") introduced the regression.
>
> The upstream fix is
> Btrfs: bring back repair during read

Same issue as reported for the other new test (btrfs/140), the test
fails on a kernel with the mentioned patch. Seems like it does wrong
assumptions somewhere (I haven't investigated, just run the test):

root 08:09:17 /home/fdmanana/git/hub/xfstests (master)> ./check btrfs/141
FSTYP -- btrfs
PLATFORM  -- Linux/x86_64 debian3 4.10.0-rc8-btrfs-next-40+
MKFS_OPTIONS  -- /dev/sdc
MOUNT_OPTIONS -- /dev/sdc /home/fdmanana/btrfs-tests/scratch_1

btrfs/141 - output mismatch (see
/home/fdmanana/git/hub/xfstests/results//btrfs/141.out.bad)
--- tests/btrfs/141.out 2017-04-28 08:09:13.289791597 +0100
+++ /home/fdmanana/git/hub/xfstests/results//btrfs/141.out.bad
2017-04-28 08:09:28.469804304 +0100
@@ -1,39 +1,39 @@
 QA output created by 141
 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 244056064
+wrote 65536/65536 bytes at offset 1103101952
 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-0e8c:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa

...
(Run 'diff -u tests/btrfs/141.out
/home/fdmanana/git/hub/xfstests/results//btrfs/141.out.bad'  to see
the entire diff)
Ran: btrfs/141
Failures: btrfs/141
Failed 1 of 1 tests

root 08:09:29 /home/fdmanana/git/hub/xfstests (master)>
root 08:09:31 /home/fdmanana/git/hub/xfstests (master)> diff -u
tests/btrfs/141.out
/home/fdmanana/git/hub/xfstests/results//btrfs/141.out.bad
--- tests/btrfs/141.out 2017-04-28 08:09:13.289791597 +0100
+++ /home/fdmanana/git/hub/xfstests/results//btrfs/141.out.bad
2017-04-28 08:09:28.469804304 +0100
@@ -1,39 +1,39 @@
 QA output created by 141
 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 244056064
+wrote 65536/65536 bytes at offset 1103101952
 XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-0e8c:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0010:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0020:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0030:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0040:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0050:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0060:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0070:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0080:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0090:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c00a0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c00b0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c00c0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c00d0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c00e0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c00f0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0100:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0110:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0120:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0130:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0140:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0150:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0160:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0170:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0180:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c0190:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c01a0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c01b0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c01c0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c01d0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c01e0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-0e8c01f0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
-read 512/512 bytes at offset 244056064
+41c0:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa  
+41c00010:  aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa

[PATCH v2] fstests: regression test for btrfs buffered read's repair

2017-04-26 Thread Liu Bo 
This case tests whether buffered read can repair the bad copy if we
have a good copy.

Commit 20a7db8ab3f2 ("btrfs: add dummy callback for readpage_io_failed
and drop checks") introduced the regression.

The upstream fix is
Btrfs: bring back repair during read

Signed-off-by: Liu Bo 
---
v2: - Add regression commit and the fix to the description
- Use btrfs inspect-internal dump-tree to get rid of the dependence 
btrfs-map-logical
- Add comments in several places
- Fix typo, dio->buffered.

 tests/btrfs/141 | 169 
 tests/btrfs/141.out |  39 
 tests/btrfs/group   |   1 +
 3 files changed, 209 insertions(+)
 create mode 100755 tests/btrfs/141
 create mode 100644 tests/btrfs/141.out

diff --git a/tests/btrfs/141 b/tests/btrfs/141
new file mode 100755
index 000..f95dd2b
--- /dev/null
+++ b/tests/btrfs/141
@@ -0,0 +1,169 @@
+#! /bin/bash
+# FS QA Test 141
+#
+# Regression test for btrfs buffered read's repair during read.
+#
+# Commit 20a7db8ab3f2 ("btrfs: add dummy callback for readpage_io_failed
+# and drop checks") introduced the regression.
+#
+# The upstream fix is
+#  Btrfs: bring back repair during read
+#
+#---
+# Copyright (c) 2017 Liu Bo.  All Rights Reserved.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation.
+#
+# This program is distributed in the hope that it would be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write the Free Software Foundation,
+# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+#---
+#
+
+seq=`basename $0`
+seqres=$RESULT_DIR/$seq
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1   # failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+_cleanup()
+{
+   cd /
+   rm -f $tmp.*
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+
+# remove previous $seqres.full before test
+rm -f $seqres.full
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs btrfs
+_supported_os Linux
+_require_scratch_dev_pool 2
+
+_require_btrfs_command inspect-internal dump-tree
+_require_command "$FILEFRAG_PROG" filefrag
+
+
+# helpe to convert 'file offset' to btrfs logical offset
+FILEFRAG_FILTER='
+   if (/blocks? of (\d+) bytes/) {
+   $blocksize = $1;
+   next
+   }
+   ($ext, $logical, $physical, $length) =
+   (/^\s*(\d+):\s+(\d+)..\s+\d+:\s+(\d+)..\s+\d+:\s+(\d+):/)
+   or next;
+   ($flags) = /.*:\s*(\S*)$/;
+   print $physical * $blocksize, "#",
+ $length * $blocksize, "#",
+ $logical * $blocksize, "#",
+ $flags, " "'
+
+# this makes filefrag output script readable by using a perl helper.
+# output is one extent per line, with three numbers separated by '#'
+# the numbers are: physical, length, logical (all in bytes)
+# sample output: "1234#10#5678" -> physical 1234, length 10, logical 5678
+_filter_extents()
+{
+   tee -a $seqres.full | $PERL_PROG -ne "$FILEFRAG_FILTER"
+}
+
+_check_file_extents()
+{
+   cmd="filefrag -v $1"
+   echo "# $cmd" >> $seqres.full
+   out=`$cmd | _filter_extents`
+   if [ -z "$out" ]; then
+   return 1
+   fi
+   echo "after filter: $out" >> $seqres.full
+   echo $out
+   return 0
+}
+
+_check_repair()
+{
+   filter=${1:-cat}
+   dmesg | tac | sed -ne "0,\#run fstests $seqnum at $date_time#p" | tac | 
$filter | grep -q -e "csum failed"
+   if [ $? -eq 0 ]; then
+   echo 1
+   else
+   echo 0
+   fi
+}
+
+_get_physical()
+{
+# $1 is logical address
+# print chunk tree and find devid 2 which is $SCRATCH_DEV
+$BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | grep 
$1 -A 6 | awk '($1 ~ /stripe/ && $3 ~ /devid/ && $4 ~ /1/) { print $6 }'
+}
+
+_scratch_dev_pool_get 2
+# step 1, create a raid1 btrfs which contains one 128k file.
+echo "step 1..mkfs.btrfs" >>$seqres.full
+
+mkfs_opts="-d raid1"
+_scratch_pool_mkfs $mkfs_opts >>$seqres.full 2>&1
+
+# -o nospace_cache makes sure data is written to the start position of the data
+# chunk
+_scratch_mount -o nospace_cache
+
+$XFS_IO_PROG -f -d -c "pwrite -S 0xaa -b 128K 0 128K" "$SCRATCH_MNT/foobar" | 
_filter_xfs_io
+
+sync
+
+# step 2, corrupt the first 64k of one copy (on SCRATCH_DEV which is the first
+# one in $SCRATCH_DEV_POOL
+echo "step 2..corrupt