| Message ID | 20180428230627.16791-1-jlayton@kernel.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Sat, Apr 28, 2018 at 07:06:27PM -0400, Jeff Layton wrote: > From: Jeff Layton <jlayton@redhat.com> > > This adds a regression test for the following kernel patch: > > errseq: Always report a writeback error once > > This is motivated by some rather odd behavior done by the PostgreSQL > project. The main database writers will offload the fsync calls to a > separate process, which can open files after a writeback error has > already occurred. > > This used to work with older kernels that reported the error to only > one fd, but with the errseq_t changes we lost the ability to see > errors that occurred before the open. The above patch restores that > behavior. > > Signed-off-by: Jeff Layton <jlayton@redhat.com> > Reviewed-by: Amir Goldstein <amir73il@gmail.com> > --- > tests/generic/999 | 95 +++++++++++++++++++++++++++++++++++++++++++++++++++ > tests/generic/999.out | 5 +++ > tests/generic/group | 1 + > 3 files changed, 101 insertions(+) > create mode 100755 tests/generic/999 > create mode 100644 tests/generic/999.out > > diff --git a/tests/generic/999 b/tests/generic/999 > new file mode 100755 > index 000000000000..b816fd3bb423 > --- /dev/null > +++ b/tests/generic/999 > @@ -0,0 +1,95 @@ > +#! /bin/bash > +# FS QA Test No. XXX ^^^ Need a test number here :) > +# > +# Open a file several times, write to it, fsync on all fds and make sure that > +# they all return 0. Change the device to start throwing errors. Write again > +# on all fds and fsync on all fds. Ensure that we get errors on all of them. > +# Then fsync on all one last time and verify that all return 0. > +# > +#----------------------------------------------------------------------- > +# Copyright (c) 2018, Jeff Layton <jlayton@redhat.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" > + > +here=`pwd` > +tmp=/tmp/$$ > +status=1 # failure is the default! > +trap "_cleanup; exit \$status" 0 1 2 3 15 > + > +_cleanup() > +{ > + cd / > + rm -rf $tmp.* $testdir > + _dmerror_cleanup > +} > + > +# get standard environment, filters and checks > +. ./common/rc > +. ./common/filter > +. ./common/dmerror > + > +# real QA test starts here > +_supported_os Linux > +_require_scratch_nocheck > +# This test uses "dm" without taking into account the data could be on > +# realtime subvolume, thus the test will fail with rtinherit=1 > +_require_no_rtinherit > + > +_require_dm_target error > + > +rm -f $seqres.full > + > +echo "Format and mount" > +_scratch_mkfs > $seqres.full 2>&1 > +_dmerror_init > +_dmerror_mount > + > +datalen=65536 > +_require_fs_space $SCRATCH_MNT $datalen > + > +# use fd 5 to hold file open > +testfile=$SCRATCH_MNT/fsync-open-after-err > +exec 5>$testfile > + > +# write some data to file and fsync it out > +$XFS_IO_PROG -c "pwrite -q 0 $datalen" -c fsync $testfile > + > +# flip device to non-working mode > +_dmerror_load_error_table > + > +# rewrite the data, call sync to ensure it's written back w/o scraping error > +$XFS_IO_PROG -c "pwrite -q 0 $datalen" -c sync $testfile > + > +# heal the device error > +_dmerror_load_working_table > + > +# open again and call fsync > +echo "The following fsync should fail with EIO:" > +$XFS_IO_PROG -c fsync $testfile > +echo "done" I built latest Linus tree, which should contain the mentioned fix, and I saw different results on xfs, ext4 and btrfs. XFS fails the test as: -fsync: Input/output error +/mnt/scratch/fsync-open-after-err: Input/output error While btrfs fails as: -fsync: Input/output error +/mnt/scratch/fsync-open-after-err: Read-only file system And both XFS and btrfs behave in the same way no matter I have the fix applied or not (tested on v4.16 kernel and v4.17-rc3+ kernel). Only ext4 passes the test with v4.17-rc3+ kernel, and fails on v4.16 kernel as expected: -fsync: Input/output error Could you please take a look? Thanks, Eryu > + > +# close file > +exec 5>&- > + > +# success, all done > +_dmerror_unmount > +_dmerror_cleanup > + > +status=0 > +exit > diff --git a/tests/generic/999.out b/tests/generic/999.out > new file mode 100644 > index 000000000000..38d2d7f6495f > --- /dev/null > +++ b/tests/generic/999.out > @@ -0,0 +1,5 @@ > +QA output created by 999 > +Format and mount > +The following fsync should fail with EIO: > +fsync: Input/output error > +done > diff --git a/tests/generic/group b/tests/generic/group > index ea8e51b35e79..48f491a5c32b 100644 > --- a/tests/generic/group > +++ b/tests/generic/group > @@ -486,3 +486,4 @@ > 481 auto quick log metadata > 482 auto metadata replay > 483 auto quick log metadata > +999 auto quick > -- > 2.14.3 > > -- > 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 Wed, 2018-05-02 at 13:50 +0800, Eryu Guan wrote: > On Sat, Apr 28, 2018 at 07:06:27PM -0400, Jeff Layton wrote: > > From: Jeff Layton <jlayton@redhat.com> > > > > This adds a regression test for the following kernel patch: > > > > errseq: Always report a writeback error once > > > > This is motivated by some rather odd behavior done by the PostgreSQL > > project. The main database writers will offload the fsync calls to a > > separate process, which can open files after a writeback error has > > already occurred. > > > > This used to work with older kernels that reported the error to only > > one fd, but with the errseq_t changes we lost the ability to see > > errors that occurred before the open. The above patch restores that > > behavior. > > > > Signed-off-by: Jeff Layton <jlayton@redhat.com> > > Reviewed-by: Amir Goldstein <amir73il@gmail.com> > > --- > > tests/generic/999 | 95 +++++++++++++++++++++++++++++++++++++++++++++++++++ > > tests/generic/999.out | 5 +++ > > tests/generic/group | 1 + > > 3 files changed, 101 insertions(+) > > create mode 100755 tests/generic/999 > > create mode 100644 tests/generic/999.out > > > > diff --git a/tests/generic/999 b/tests/generic/999 > > new file mode 100755 > > index 000000000000..b816fd3bb423 > > --- /dev/null > > +++ b/tests/generic/999 > > @@ -0,0 +1,95 @@ > > +#! /bin/bash > > +# FS QA Test No. XXX > > ^^^ Need a test number here :) > > +# > > +# Open a file several times, write to it, fsync on all fds and make sure that > > +# they all return 0. Change the device to start throwing errors. Write again > > +# on all fds and fsync on all fds. Ensure that we get errors on all of them. > > +# Then fsync on all one last time and verify that all return 0. > > +# > > +#----------------------------------------------------------------------- > > +# Copyright (c) 2018, Jeff Layton <jlayton@redhat.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" > > + > > +here=`pwd` > > +tmp=/tmp/$$ > > +status=1 # failure is the default! > > +trap "_cleanup; exit \$status" 0 1 2 3 15 > > + > > +_cleanup() > > +{ > > + cd / > > + rm -rf $tmp.* $testdir > > + _dmerror_cleanup > > +} > > + > > +# get standard environment, filters and checks > > +. ./common/rc > > +. ./common/filter > > +. ./common/dmerror > > + > > +# real QA test starts here > > +_supported_os Linux > > +_require_scratch_nocheck > > +# This test uses "dm" without taking into account the data could be on > > +# realtime subvolume, thus the test will fail with rtinherit=1 > > +_require_no_rtinherit > > + > > +_require_dm_target error > > + > > +rm -f $seqres.full > > + > > +echo "Format and mount" > > +_scratch_mkfs > $seqres.full 2>&1 > > +_dmerror_init > > +_dmerror_mount > > + > > +datalen=65536 > > +_require_fs_space $SCRATCH_MNT $datalen > > + > > +# use fd 5 to hold file open > > +testfile=$SCRATCH_MNT/fsync-open-after-err > > +exec 5>$testfile > > + > > +# write some data to file and fsync it out > > +$XFS_IO_PROG -c "pwrite -q 0 $datalen" -c fsync $testfile > > + > > +# flip device to non-working mode > > +_dmerror_load_error_table > > + > > +# rewrite the data, call sync to ensure it's written back w/o scraping error > > +$XFS_IO_PROG -c "pwrite -q 0 $datalen" -c sync $testfile > > + > > +# heal the device error > > +_dmerror_load_working_table > > + > > +# open again and call fsync > > +echo "The following fsync should fail with EIO:" > > +$XFS_IO_PROG -c fsync $testfile > > +echo "done" > > > I built latest Linus tree, which should contain the mentioned fix, and I > saw different results on xfs, ext4 and btrfs. > > XFS fails the test as: > -fsync: Input/output error > +/mnt/scratch/fsync-open-after-err: Input/output error I think I see the problem. I was testing this with a logdev. I think we'll need a _require_logdev directive in the test. I'll send a v4 in a bit with that and some other corrections. > > > While btrfs fails as: > -fsync: Input/output error > +/mnt/scratch/fsync-open-after-err: Read-only file system > I think btrfs will require a specialized testcase (like we did for some of the earlier tests in this area). I'll send a separate patch for that one. > > > And both XFS and btrfs behave in the same way no matter I have the fix > applied or not (tested on v4.16 kernel and v4.17-rc3+ kernel). > > Only ext4 passes the test with v4.17-rc3+ kernel, and fails on v4.16 > kernel as expected: > -fsync: Input/output error > > Could you please take a look? > > Thanks, > Eryu > > > + > > +# close file > > +exec 5>&- > > + > > +# success, all done > > +_dmerror_unmount > > +_dmerror_cleanup > > + > > +status=0 > > +exit > > diff --git a/tests/generic/999.out b/tests/generic/999.out > > new file mode 100644 > > index 000000000000..38d2d7f6495f > > --- /dev/null > > +++ b/tests/generic/999.out > > @@ -0,0 +1,5 @@ > > +QA output created by 999 > > +Format and mount > > +The following fsync should fail with EIO: > > +fsync: Input/output error > > +done > > diff --git a/tests/generic/group b/tests/generic/group > > index ea8e51b35e79..48f491a5c32b 100644 > > --- a/tests/generic/group > > +++ b/tests/generic/group > > @@ -486,3 +486,4 @@ > > 481 auto quick log metadata > > 482 auto metadata replay > > 483 auto quick log metadata > > +999 auto quick > > -- > > 2.14.3 > > > > -- > > 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
diff --git a/tests/generic/999 b/tests/generic/999 new file mode 100755 index 000000000000..b816fd3bb423 --- /dev/null +++ b/tests/generic/999 @@ -0,0 +1,95 @@ +#! /bin/bash +# FS QA Test No. XXX +# +# Open a file several times, write to it, fsync on all fds and make sure that +# they all return 0. Change the device to start throwing errors. Write again +# on all fds and fsync on all fds. Ensure that we get errors on all of them. +# Then fsync on all one last time and verify that all return 0. +# +#----------------------------------------------------------------------- +# Copyright (c) 2018, Jeff Layton <jlayton@redhat.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" + +here=`pwd` +tmp=/tmp/$$ +status=1 # failure is the default! +trap "_cleanup; exit \$status" 0 1 2 3 15 + +_cleanup() +{ + cd / + rm -rf $tmp.* $testdir + _dmerror_cleanup +} + +# get standard environment, filters and checks +. ./common/rc +. ./common/filter +. ./common/dmerror + +# real QA test starts here +_supported_os Linux +_require_scratch_nocheck +# This test uses "dm" without taking into account the data could be on +# realtime subvolume, thus the test will fail with rtinherit=1 +_require_no_rtinherit + +_require_dm_target error + +rm -f $seqres.full + +echo "Format and mount" +_scratch_mkfs > $seqres.full 2>&1 +_dmerror_init +_dmerror_mount + +datalen=65536 +_require_fs_space $SCRATCH_MNT $datalen + +# use fd 5 to hold file open +testfile=$SCRATCH_MNT/fsync-open-after-err +exec 5>$testfile + +# write some data to file and fsync it out +$XFS_IO_PROG -c "pwrite -q 0 $datalen" -c fsync $testfile + +# flip device to non-working mode +_dmerror_load_error_table + +# rewrite the data, call sync to ensure it's written back w/o scraping error +$XFS_IO_PROG -c "pwrite -q 0 $datalen" -c sync $testfile + +# heal the device error +_dmerror_load_working_table + +# open again and call fsync +echo "The following fsync should fail with EIO:" +$XFS_IO_PROG -c fsync $testfile +echo "done" + +# close file +exec 5>&- + +# success, all done +_dmerror_unmount +_dmerror_cleanup + +status=0 +exit diff --git a/tests/generic/999.out b/tests/generic/999.out new file mode 100644 index 000000000000..38d2d7f6495f --- /dev/null +++ b/tests/generic/999.out @@ -0,0 +1,5 @@ +QA output created by 999 +Format and mount +The following fsync should fail with EIO: +fsync: Input/output error +done diff --git a/tests/generic/group b/tests/generic/group index ea8e51b35e79..48f491a5c32b 100644 --- a/tests/generic/group +++ b/tests/generic/group @@ -486,3 +486,4 @@ 481 auto quick log metadata 482 auto metadata replay 483 auto quick log metadata +999 auto quick