The following kernel commit introduced the bug: 51f395ad btrfs: Use right extent length when inserting overlap extent map. When btrfs commit race with btrfs_get_extent(), merge_extent_mapping() may build up a new extent which length overflows and cause extent map insert fail, causing the caller get a -EEXIST error. This regression is fixed by the following patches: btrfs: Fix and enhance merge_extent_mapping() to insert best fitted extent map btrfs: Fix the wrong condition judgment about subset extent map Cc: Filipe Manana <fdmanana@suse.com> Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com> --- tests/btrfs/078 | 128 ++++++++++++++++++++++++++++++++++++++++++++++++++++ tests/btrfs/078.out | 2 + tests/btrfs/group | 1 + 3 files changed, 131 insertions(+) create mode 100755 tests/btrfs/078 create mode 100644 tests/btrfs/078.out diff --git a/tests/btrfs/078 b/tests/btrfs/078 new file mode 100755 index 0000000..bdc20af --- /dev/null +++ b/tests/btrfs/078 @@ -0,0 +1,128 @@ +#! /bin/bash +# FS QA Test No. btrfs/078 +# +# Do write along with fiemap ioctl. +# Regression test for the kernel comit: +# 51f395ad btrfs: Use right extent length when inserting overlap extent map. +# +# When calling fiemap(without SYNC flag) and btrfs fs is commiting, +# it will cause race condition and cause btrfs to generate a wrong extent +# whose len is overflow and fail to insert into the extent map tree, +# returning -EEXIST. +# +# Fixed by the following patches (not merged in mainline yet): +# btrfs: Fix and enhance merge_extent_mapping() to insert best fitted extent map +# btrfs: Fix the wrong condition judgment about subset extent map +# +#----------------------------------------------------------------------- +# Copyright (C) 2014 Fujitsu All Rights Reserved. +# Author: Qu Wenruo <quwenruo@cn.fujitsu.com> +# +# 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" + +tmp=`mktemp -d` +status=1 # failure is the default! +trap "_cleanup; exit \$status" 0 1 2 3 15 + +_cleanup() +{ + kill $dd_pid &> /dev/null + kill $fiemap_pid &> /dev/null + wait + rm -fr $test_file + rm -fr $tmp +} + +# get standard environment, filters and checks +. ./common/rc +. ./common/filter + +# real QA test starts here +_supported_fs btrfs +_supported_os Linux +_require_scratch +_need_to_be_root +# Since xfs_io's fiemap always use SYNC flag and can't be unset, +# we must use filefrag to call fiemap without SYNC flag. +_require_command "/usr/sbin/filefrag" +_require_xfs_io_command "falloc" + +test_file=$SCRATCH_MNT/testfile +filesize=$((10 * 1024 * 1024 * 1024)) #10G size +buffersize=$((1024 * 1024)) # 1M bs for dd +count=$(($filesize / $buffersize)) +testfile=$SCRATCH_MNT/testfile + +rm -f $seqres.full +rm -f $test_file + +_scratch_mkfs >>$seqres.full 2>&1 +_scratch_mount +$XFS_IO_PROG -f -c "falloc 0 $filesize" $testfile || \ + _fail 'need device larger than 10G' + +dd_work() { + out=$1 + dd if=/dev/zero of=$out bs=$buffersize count=$count \ + conv=notrunc &> /dev/null +} + +# There is a bug for e2fsprogs, at least in version 1.42.9, filefrag will +# leak the return value, so we can't judge return value only, +# but also to filter the output +_filter_error() { + # when filefrag fails FIEMAP ioctl, it will fall back to FIBMAP, + # which is not supported by btrfs and will report "FIBMAP: strerr()" + # However due to other btrfs commits, kernel will return -ENOTTY + # or -EINVAL, so only grep for "FIBMAP" for max compatibility. + grep "FIBMAP" +} + +fiemap_work() { + filename=$1 + while true; do + filefrag $filename 2> $tmp/output 1> /dev/null + ret=$? + err=`cat $tmp/output | _filter_error` + if [ $ret -ne 0 -o -n "$err" ]; then + kill $dd_pid + return 1 + fi + done +} + +dd_work $testfile & +dd_pid=$! +fiemap_work $testfile & +fiemap_pid=$! +wait $dd_pid +ddret=$? +kill $fiemap_pid &> /dev/null +wait $fiemap_pid + +if [ $ddret -ne 0 ]; then + echo "Extent merge bug detected" + status=1 + exit +else + echo "Silence is golden" + status=0 + exit +fi diff --git a/tests/btrfs/078.out b/tests/btrfs/078.out new file mode 100644 index 0000000..b8acea8 --- /dev/null +++ b/tests/btrfs/078.out @@ -0,0 +1,2 @@ +QA output created by 078 +Silence is golden diff --git a/tests/btrfs/group b/tests/btrfs/group index 9adf862..40e7430 100644 --- a/tests/btrfs/group +++ b/tests/btrfs/group @@ -80,3 +80,4 @@ 075 auto quick subvol 076 auto quick 077 auto quick +078 auto -- 2.1.2 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html