| Message ID | 20180608061723.5276-1-wqu@suse.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Fri, Jun 08, 2018 at 02:17:23PM +0800, Qu Wenruo wrote: > This is a long existing bug (from 2012) but exposed by a reporter > recently, that when compressed extent without data csum get written to > device-replace target device, the written data is in fact uncompressed data > other than the original compressed data. > > And since btrfs still consider the data is compressed and will try to read it > as compressed, it can cause read error. > > The root cause is located, and one RFC patch already sent to fix it, > titled "[PATCH] btrfs: scrub: Don't use inode pages for device replace". > (The RFC is only for the extra possible way to fix the bug, the fix > itself should work without problem) > > Reported-by: James Harvey <jamespharvey20@gmail.com> > Signed-off-by: Qu Wenruo <wqu@suse.com> > --- > changelog: > v2: > Now the fix patch is no longer RFC. > Remove _require_test as we don't really touch it. > Add comment on the mount cycle. > Add the test to group 'volume'. Thanks for the revision! But again, I'd like to get some explicit reviews from btrfs folks. > --- > tests/btrfs/161 | 91 +++++++++++++++++++++++++++++++++++++++++++++ > tests/btrfs/161.out | 2 + > tests/btrfs/group | 1 + > 3 files changed, 94 insertions(+) > create mode 100755 tests/btrfs/161 > create mode 100644 tests/btrfs/161.out > > diff --git a/tests/btrfs/161 b/tests/btrfs/161 > new file mode 100755 > index 00000000..ce1b0e04 > --- /dev/null > +++ b/tests/btrfs/161 > @@ -0,0 +1,91 @@ > +#! /bin/bash > +# FS QA Test 161 > +# > +# Test if btrfs will corrupt compressed data extent without data csum > +# by replacing it with uncompressed data, when doing replacing device. > +# > +# This could be fixed by the following RFC patch: ^^^^ can be dropped? Thanks, Eryu P.S. *IF* you're going to send v3, could you please follow the new test template (create new test with './new btrfs' would do the work) and rebase against latest master? That'd be easier for me to apply the patch, but I'm also fine with taking it as-is (after we get Reviewed-by tag), I can convert the test and re-number it on commit as always. > +# "[PATCH] btrfs: scrub: Don't use inode pages for device replace" > +# > +#----------------------------------------------------------------------- > +# Copyright (C) 2018 SUSE Linux Products GmbH. 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_scratch_dev_pool_equal_size > + > + > +_scratch_dev_pool_get 1 > +_spare_dev_get > +_scratch_pool_mkfs >> $seqres.full 2>&1 > + > +# Create nodatasum inode > +_scratch_mount "-o nodatasum" > +touch $SCRATCH_MNT/nodatasum_file > +_scratch_remount "datasum,compress" > +_pwrite_byte 0xcd 0 128K $SCRATCH_MNT/nodatasum_file > /dev/null > + > +# Write the compressed data back to disk > +sync > + > +# Replace the device > +_run_btrfs_util_prog replace start -Bf 1 $SPARE_DEV $SCRATCH_MNT > + > +# Unmount to drop all cache so next read will read from disk > +_scratch_unmount > +_mount $SPARE_DEV $SCRATCH_MNT > + > +# Now the EXTENT_DATA item still marks the extent as compressed, > +# but the on-disk data is uncompressed, thus reading it as compressed > +# will definitely cause EIO. > +cat $SCRATCH_MNT/nodatasum_file > /dev/null > + > +_scratch_unmount > +_spare_dev_put > +_scratch_dev_pool_put > + > +echo "Silence is golden" > +# success, all done > +status=0 > +exit > diff --git a/tests/btrfs/161.out b/tests/btrfs/161.out > new file mode 100644 > index 00000000..1752a243 > --- /dev/null > +++ b/tests/btrfs/161.out > @@ -0,0 +1,2 @@ > +QA output created by 161 > +Silence is golden > diff --git a/tests/btrfs/group b/tests/btrfs/group > index f04ee8d5..9195b368 100644 > --- a/tests/btrfs/group > +++ b/tests/btrfs/group > @@ -163,3 +163,4 @@ > 158 auto quick raid scrub > 159 auto quick > 160 auto quick > +161 auto quick replace volume > -- > 2.17.0 > > -- > To unsubscribe from this list: send the line "unsubscribe fstests" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe fstests" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 2018年06月14日 13:30, Eryu Guan wrote: > On Fri, Jun 08, 2018 at 02:17:23PM +0800, Qu Wenruo wrote: >> This is a long existing bug (from 2012) but exposed by a reporter >> recently, that when compressed extent without data csum get written to >> device-replace target device, the written data is in fact uncompressed data >> other than the original compressed data. >> >> And since btrfs still consider the data is compressed and will try to read it >> as compressed, it can cause read error. >> >> The root cause is located, and one RFC patch already sent to fix it, >> titled "[PATCH] btrfs: scrub: Don't use inode pages for device replace". >> (The RFC is only for the extra possible way to fix the bug, the fix >> itself should work without problem) >> >> Reported-by: James Harvey <jamespharvey20@gmail.com> >> Signed-off-by: Qu Wenruo <wqu@suse.com> >> --- >> changelog: >> v2: >> Now the fix patch is no longer RFC. >> Remove _require_test as we don't really touch it. >> Add comment on the mount cycle. >> Add the test to group 'volume'. > > Thanks for the revision! But again, I'd like to get some explicit > reviews from btrfs folks. Something I cannot help here :( > >> --- >> tests/btrfs/161 | 91 +++++++++++++++++++++++++++++++++++++++++++++ >> tests/btrfs/161.out | 2 + >> tests/btrfs/group | 1 + >> 3 files changed, 94 insertions(+) >> create mode 100755 tests/btrfs/161 >> create mode 100644 tests/btrfs/161.out >> >> diff --git a/tests/btrfs/161 b/tests/btrfs/161 >> new file mode 100755 >> index 00000000..ce1b0e04 >> --- /dev/null >> +++ b/tests/btrfs/161 >> @@ -0,0 +1,91 @@ >> +#! /bin/bash >> +# FS QA Test 161 >> +# >> +# Test if btrfs will corrupt compressed data extent without data csum >> +# by replacing it with uncompressed data, when doing replacing device. >> +# >> +# This could be fixed by the following RFC patch: > ^^^^ can be dropped? Yep, that should be dropped. As the fix is already in the latest pull request sent to Linus. > > Thanks, > Eryu > > P.S. > *IF* you're going to send v3, could you please follow the new test > template (create new test with './new btrfs' would do the work) and > rebase against latest master? That'd be easier for me to apply the > patch, but I'm also fine with taking it as-is (after we get Reviewed-by > tag), I can convert the test and re-number it on commit as always. Of course, I'll use the new template in next version, and update the base. Thanks, Qu > >> +# "[PATCH] btrfs: scrub: Don't use inode pages for device replace" >> +# >> +#----------------------------------------------------------------------- >> +# Copyright (C) 2018 SUSE Linux Products GmbH. 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_scratch_dev_pool_equal_size >> + >> + >> +_scratch_dev_pool_get 1 >> +_spare_dev_get >> +_scratch_pool_mkfs >> $seqres.full 2>&1 >> + >> +# Create nodatasum inode >> +_scratch_mount "-o nodatasum" >> +touch $SCRATCH_MNT/nodatasum_file >> +_scratch_remount "datasum,compress" >> +_pwrite_byte 0xcd 0 128K $SCRATCH_MNT/nodatasum_file > /dev/null >> + >> +# Write the compressed data back to disk >> +sync >> + >> +# Replace the device >> +_run_btrfs_util_prog replace start -Bf 1 $SPARE_DEV $SCRATCH_MNT >> + >> +# Unmount to drop all cache so next read will read from disk >> +_scratch_unmount >> +_mount $SPARE_DEV $SCRATCH_MNT >> + >> +# Now the EXTENT_DATA item still marks the extent as compressed, >> +# but the on-disk data is uncompressed, thus reading it as compressed >> +# will definitely cause EIO. >> +cat $SCRATCH_MNT/nodatasum_file > /dev/null >> + >> +_scratch_unmount >> +_spare_dev_put >> +_scratch_dev_pool_put >> + >> +echo "Silence is golden" >> +# success, all done >> +status=0 >> +exit >> diff --git a/tests/btrfs/161.out b/tests/btrfs/161.out >> new file mode 100644 >> index 00000000..1752a243 >> --- /dev/null >> +++ b/tests/btrfs/161.out >> @@ -0,0 +1,2 @@ >> +QA output created by 161 >> +Silence is golden >> diff --git a/tests/btrfs/group b/tests/btrfs/group >> index f04ee8d5..9195b368 100644 >> --- a/tests/btrfs/group >> +++ b/tests/btrfs/group >> @@ -163,3 +163,4 @@ >> 158 auto quick raid scrub >> 159 auto quick >> 160 auto quick >> +161 auto quick replace volume >> -- >> 2.17.0 >> >> -- >> To unsubscribe from this list: send the line "unsubscribe fstests" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html > -- > 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 >
diff --git a/tests/btrfs/161 b/tests/btrfs/161 new file mode 100755 index 00000000..ce1b0e04 --- /dev/null +++ b/tests/btrfs/161 @@ -0,0 +1,91 @@ +#! /bin/bash +# FS QA Test 161 +# +# Test if btrfs will corrupt compressed data extent without data csum +# by replacing it with uncompressed data, when doing replacing device. +# +# This could be fixed by the following RFC patch: +# "[PATCH] btrfs: scrub: Don't use inode pages for device replace" +# +#----------------------------------------------------------------------- +# Copyright (C) 2018 SUSE Linux Products GmbH. 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_scratch_dev_pool_equal_size + + +_scratch_dev_pool_get 1 +_spare_dev_get +_scratch_pool_mkfs >> $seqres.full 2>&1 + +# Create nodatasum inode +_scratch_mount "-o nodatasum" +touch $SCRATCH_MNT/nodatasum_file +_scratch_remount "datasum,compress" +_pwrite_byte 0xcd 0 128K $SCRATCH_MNT/nodatasum_file > /dev/null + +# Write the compressed data back to disk +sync + +# Replace the device +_run_btrfs_util_prog replace start -Bf 1 $SPARE_DEV $SCRATCH_MNT + +# Unmount to drop all cache so next read will read from disk +_scratch_unmount +_mount $SPARE_DEV $SCRATCH_MNT + +# Now the EXTENT_DATA item still marks the extent as compressed, +# but the on-disk data is uncompressed, thus reading it as compressed +# will definitely cause EIO. +cat $SCRATCH_MNT/nodatasum_file > /dev/null + +_scratch_unmount +_spare_dev_put +_scratch_dev_pool_put + +echo "Silence is golden" +# success, all done +status=0 +exit diff --git a/tests/btrfs/161.out b/tests/btrfs/161.out new file mode 100644 index 00000000..1752a243 --- /dev/null +++ b/tests/btrfs/161.out @@ -0,0 +1,2 @@ +QA output created by 161 +Silence is golden diff --git a/tests/btrfs/group b/tests/btrfs/group index f04ee8d5..9195b368 100644 --- a/tests/btrfs/group +++ b/tests/btrfs/group @@ -163,3 +163,4 @@ 158 auto quick raid scrub 159 auto quick 160 auto quick +161 auto quick replace volume
This is a long existing bug (from 2012) but exposed by a reporter recently, that when compressed extent without data csum get written to device-replace target device, the written data is in fact uncompressed data other than the original compressed data. And since btrfs still consider the data is compressed and will try to read it as compressed, it can cause read error. The root cause is located, and one RFC patch already sent to fix it, titled "[PATCH] btrfs: scrub: Don't use inode pages for device replace". (The RFC is only for the extra possible way to fix the bug, the fix itself should work without problem) Reported-by: James Harvey <jamespharvey20@gmail.com> Signed-off-by: Qu Wenruo <wqu@suse.com> --- changelog: v2: Now the fix patch is no longer RFC. Remove _require_test as we don't really touch it. Add comment on the mount cycle. Add the test to group 'volume'. --- tests/btrfs/161 | 91 +++++++++++++++++++++++++++++++++++++++++++++ tests/btrfs/161.out | 2 + tests/btrfs/group | 1 + 3 files changed, 94 insertions(+) create mode 100755 tests/btrfs/161 create mode 100644 tests/btrfs/161.out