Skip to contentSkip to conversation

xfstests: generic xattr enospc cleanup test

PATCH Superseded Brian Foster 2 participants 5 messages Jun 23, 2015 – Jun 26, 2015

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 xfstests: generic xattr enospc cleanup test

XFS had a regression where inode reclaim in the unlink codepath would not correctly tear down extended attribute forks where no xattr extents are present. Add a generic test to create this condition.

The test sets extended attributes on a series of files under ENOSPC conditions and then verifies that the files can be removed without syslog warnings or errors.

---
 tests/generic/103     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/103.out |  2 ++
 tests/generic/group   |  1 +
 3 files changed, 94 insertions(+)
 create mode 100755 tests/generic/103
 create mode 100644 tests/generic/103.out
tests/generic/103 new file +91 -0
new file mode 100755
index 0000000..373cd9d
@@ -0,0 +1,91 @@
+#! /bin/bash
+# FSQA Test No. 103
+#
+# Test attribute fork teardown. This test is inspired by a regression in XFS
+# that resulted in problematic removal of inodes with remote attribute forks
+# without attribute extents. The attribute fork condition is created by
+# attempting to set larger attribute values on a filesystem that is at or near
+# ENOSPC.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2015 Red Hat, Inc.  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"
+
+_cleanup()
+{
+        cd /
+        rm -f $tmp.*
+}
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15 25
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/attr
+
+# real QA test starts here
+_supported_fs generic
+_supported_os Linux
+_require_scratch
+_require_attrs
+
+rm -f $seqres.full
+
+_consume_freesp()
+{
+	file=$1
+
+	# consume nearly all available space (leave ~512kB)
+	avail=`_get_available_space $SCRATCH_MNT`
+	filesizekb=$((avail / 1024 - 512))
+	$XFS_IO_PROG -fc "falloc 0 ${filesizekb}k" $file
+}
+
+_scratch_mkfs >> $seqres.full 2>&1
+_scratch_mount
+
+for i in $(seq 0 63); do
+	touch $SCRATCH_MNT/$seq.$i
+done
+
+# Generate a large attribute value and consume the rest of the space in the
+# filesystem.
+$XFS_IO_PROG -fc "pwrite 0 64k" $SCRATCH_MNT/attrval > /dev/null 2>&1
+_consume_freesp $SCRATCH_MNT/spc
+
+# Set attributes on the test files. These should start to hit ENOSPC.
+for i in $(seq 0 63); do
+	$SETFATTR_PROG -n user.test -v "`cat $SCRATCH_MNT/attrval`" \
+		$SCRATCH_MNT/$seq.$i >> $seqres.full 2>&1
+done
+
+# Remove the files with attributes to test attribute fork teardown. Problems
+# result in dmesg output.
+rm -f $SCRATCH_MNT/$seq.*
+
+echo Silence is golden.
+
+status=0
+exit
tests/generic/103.out new file +2 -0
new file mode 100644
index 0000000..ce229bf
@@ -0,0 +1,2 @@
+QA output created by 103
+Silence is golden.
tests/generic/group +1 -0
index 0c8964c..41f3039 100644
@@ -98,6 +98,7 @@
 097 udf auto
 099 udf auto
 100 udf auto
+103 auto enospc quick
 105 acl auto quick
 112 rw aio auto quick
 113 rw aio auto quick
Signature · 1 line
1.9.3
#1 lore raw
Quoted 74 lines · Brian Foster

XFS had a regression where inode reclaim in the unlink codepath would not correctly tear down extended attribute forks where no xattr extents are present. Add a generic test to create this condition.

The test sets extended attributes on a series of files under ENOSPC conditions and then verifies that the files can be removed without syslog warnings or errors.

---
 tests/generic/103     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/103.out |  2 ++
 tests/generic/group   |  1 +
 3 files changed, 94 insertions(+)
 create mode 100755 tests/generic/103
 create mode 100644 tests/generic/103.out
tests/generic/103 new file +51 -0
new file mode 100755
index 0000000..373cd9d
@@ -0,0 +1,91 @@
+#! /bin/bash
+# FSQA Test No. 103
+#
+# Test attribute fork teardown. This test is inspired by a regression in XFS
+# that resulted in problematic removal of inodes with remote attribute forks
+# without attribute extents. The attribute fork condition is created by
+# attempting to set larger attribute values on a filesystem that is at or near
+# ENOSPC.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2015 Red Hat, Inc.  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"
+
+_cleanup()
+{
+        cd /
+        rm -f $tmp.*
+}
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15 25
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/attr
+
+# real QA test starts here
+_supported_fs generic
+_supported_os Linux
+_require_scratch

tests/generic/103 line 61

Quoted 10 lines
tests/generic/103 new file +10 -0
+_require_attrs
+
+rm -f $seqres.full
+
+_consume_freesp()
+{
+	file=$1
+
+	# consume nearly all available space (leave ~512kB)
+	avail=`_get_available_space $SCRATCH_MNT`

Seems you missed "_get_available_space" function definition in common/rc? I got "command not found" error.

Thanks,
Eryu

Quoted 57 lines
tests/generic/103 new file +30 -0
+	filesizekb=$((avail / 1024 - 512))
+	$XFS_IO_PROG -fc "falloc 0 ${filesizekb}k" $file
+}
+
+_scratch_mkfs >> $seqres.full 2>&1
+_scratch_mount
+
+for i in $(seq 0 63); do
+	touch $SCRATCH_MNT/$seq.$i
+done
+
+# Generate a large attribute value and consume the rest of the space in the
+# filesystem.
+$XFS_IO_PROG -fc "pwrite 0 64k" $SCRATCH_MNT/attrval > /dev/null 2>&1
+_consume_freesp $SCRATCH_MNT/spc
+
+# Set attributes on the test files. These should start to hit ENOSPC.
+for i in $(seq 0 63); do
+	$SETFATTR_PROG -n user.test -v "`cat $SCRATCH_MNT/attrval`" \
+		$SCRATCH_MNT/$seq.$i >> $seqres.full 2>&1
+done
+
+# Remove the files with attributes to test attribute fork teardown. Problems
+# result in dmesg output.
+rm -f $SCRATCH_MNT/$seq.*
+
+echo Silence is golden.
+
+status=0
+exit
tests/generic/103.out new file +2 -0
new file mode 100644
index 0000000..ce229bf
@@ -0,0 +1,2 @@
+QA output created by 103
+Silence is golden.
tests/generic/group +1 -0
index 0c8964c..41f3039 100644
@@ -98,6 +98,7 @@
 097 udf auto
 099 udf auto
 100 udf auto
+103 auto enospc quick
 105 acl auto quick
 112 rw aio auto quick
 113 rw aio auto quick
Signature · 6 lines
1.9.3

--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#2 lore raw
Quoted 80 lines · Eryu Guan
Quoted 79 lines · Brian Foster

XFS had a regression where inode reclaim in the unlink codepath would not correctly tear down extended attribute forks where no xattr extents are present. Add a generic test to create this condition.

The test sets extended attributes on a series of files under ENOSPC conditions and then verifies that the files can be removed without syslog warnings or errors.

---
 tests/generic/103     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/103.out |  2 ++
 tests/generic/group   |  1 +
 3 files changed, 94 insertions(+)
 create mode 100755 tests/generic/103
 create mode 100644 tests/generic/103.out
tests/generic/103 new file +56 -0
new file mode 100755
index 0000000..373cd9d
@@ -0,0 +1,91 @@
+#! /bin/bash
+# FSQA Test No. 103
+#
+# Test attribute fork teardown. This test is inspired by a regression in XFS
+# that resulted in problematic removal of inodes with remote attribute forks
+# without attribute extents. The attribute fork condition is created by
+# attempting to set larger attribute values on a filesystem that is at or near
+# ENOSPC.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2015 Red Hat, Inc.  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"
+
+_cleanup()
+{
+        cd /
+        rm -f $tmp.*
+}
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15 25
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/attr
+
+# real QA test starts here
+_supported_fs generic
+_supported_os Linux
+_require_scratch
+_require_attrs
+
+rm -f $seqres.full
+
+_consume_freesp()
Quoted 9 lines

tests/generic/103 line 61

Quoted 5 lines
tests/generic/103 new file +5 -0
+{
+	file=$1
+
+	# consume nearly all available space (leave ~512kB)
+	avail=`_get_available_space $SCRATCH_MNT`

Seems you missed "_get_available_space" function definition in common/rc? I got "command not found" error.

Ah, sorry... that is actually part of a previous patch. I meant to point out that this depends on the following patch due to the use of the new helper:

http://thread.gmane.org/gmane.comp.file-systems.fstests/834

Sorry for the confusion.

Brian

Quoted 59 lines

Thanks,
Eryu

Quoted 57 lines
tests/generic/103 new file +30 -0
+	filesizekb=$((avail / 1024 - 512))
+	$XFS_IO_PROG -fc "falloc 0 ${filesizekb}k" $file
+}
+
+_scratch_mkfs >> $seqres.full 2>&1
+_scratch_mount
+
+for i in $(seq 0 63); do
+	touch $SCRATCH_MNT/$seq.$i
+done
+
+# Generate a large attribute value and consume the rest of the space in the
+# filesystem.
+$XFS_IO_PROG -fc "pwrite 0 64k" $SCRATCH_MNT/attrval > /dev/null 2>&1
+_consume_freesp $SCRATCH_MNT/spc
+
+# Set attributes on the test files. These should start to hit ENOSPC.
+for i in $(seq 0 63); do
+	$SETFATTR_PROG -n user.test -v "`cat $SCRATCH_MNT/attrval`" \
+		$SCRATCH_MNT/$seq.$i >> $seqres.full 2>&1
+done
+
+# Remove the files with attributes to test attribute fork teardown. Problems
+# result in dmesg output.
+rm -f $SCRATCH_MNT/$seq.*
+
+echo Silence is golden.
+
+status=0
+exit
tests/generic/103.out new file +2 -0
new file mode 100644
index 0000000..ce229bf
@@ -0,0 +1,2 @@
+QA output created by 103
+Silence is golden.
tests/generic/group +1 -0
index 0c8964c..41f3039 100644
@@ -98,6 +98,7 @@
 097 udf auto
 099 udf auto
 100 udf auto
+103 auto enospc quick
 105 acl auto quick
 112 rw aio auto quick
 113 rw aio auto quick
Signature · 6 lines
1.9.3

--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#3 lore raw
In reply to Brian Foster
Quoted 76 lines · Brian Foster

XFS had a regression where inode reclaim in the unlink codepath would not correctly tear down extended attribute forks where no xattr extents are present. Add a generic test to create this condition.

The test sets extended attributes on a series of files under ENOSPC conditions and then verifies that the files can be removed without syslog warnings or errors.

---
 tests/generic/103     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/103.out |  2 ++
 tests/generic/group   |  1 +
 3 files changed, 94 insertions(+)
 create mode 100755 tests/generic/103
 create mode 100644 tests/generic/103.out
tests/generic/103 new file +53 -0
new file mode 100755
index 0000000..373cd9d
@@ -0,0 +1,91 @@
+#! /bin/bash
+# FSQA Test No. 103
+#
+# Test attribute fork teardown. This test is inspired by a regression in XFS
+# that resulted in problematic removal of inodes with remote attribute forks
+# without attribute extents. The attribute fork condition is created by
+# attempting to set larger attribute values on a filesystem that is at or near
+# ENOSPC.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2015 Red Hat, Inc.  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"
+
+_cleanup()
+{
+        cd /
+        rm -f $tmp.*
+}
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15 25
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/attr
+
+# real QA test starts here
+_supported_fs generic
+_supported_os Linux
+_require_scratch
+_require_attrs
+

tests/generic/103 line 63

Quoted 10 lines
tests/generic/103 new file +10 -0
+rm -f $seqres.full
+
+_consume_freesp()
+{
+	file=$1
+
+	# consume nearly all available space (leave ~512kB)
+	avail=`_get_available_space $SCRATCH_MNT`
+	filesizekb=$((avail / 1024 - 512))
+	$XFS_IO_PROG -fc "falloc 0 ${filesizekb}k" $file

This fails ext2/3 which don't have fallocate(2) support, need _require_xfs_io_command "falloc".

This blocks ext2/3 from testing but extN just doesn't support 64k xattr value, we don't lose much test coverage. On the other hand, changing falloc to pwrite extends test time significantly (2s -> 30+s) and doesn't gain much benefit.

Quoted 13 lines
tests/generic/103 new file +13 -0
+}
+
+_scratch_mkfs >> $seqres.full 2>&1
+_scratch_mount
+
+for i in $(seq 0 63); do
+	touch $SCRATCH_MNT/$seq.$i
+done
+
+# Generate a large attribute value and consume the rest of the space in the
+# filesystem.
+$XFS_IO_PROG -fc "pwrite 0 64k" $SCRATCH_MNT/attrval > /dev/null 2>&1
+_consume_freesp $SCRATCH_MNT/spc

tests/generic/103 line 86

Quoted 10 lines
tests/generic/103 new file +10 -0
+
+# Set attributes on the test files. These should start to hit ENOSPC.
+for i in $(seq 0 63); do
+	$SETFATTR_PROG -n user.test -v "`cat $SCRATCH_MNT/attrval`" \
+		$SCRATCH_MNT/$seq.$i >> $seqres.full 2>&1
+done
+
+# Remove the files with attributes to test attribute fork teardown. Problems
+# result in dmesg output.
+rm -f $SCRATCH_MNT/$seq.*

I see xfs corruption after the test, I think that's expected.

But I just want to point out here that _check_dmesg is unable to detect the call traces in dmesg now, as the call trace doesn't have "WARNING", "kernel BUG" etc. keywords that _check_dmesg searches for, we rely on _check_fimesystems to fail the test.

Thanks,
Eryu

Quoted 32 lines
tests/generic/103 new file +5 -0
+
+echo Silence is golden.
+
+status=0
+exit
tests/generic/103.out new file +2 -0
new file mode 100644
index 0000000..ce229bf
@@ -0,0 +1,2 @@
+QA output created by 103
+Silence is golden.
tests/generic/group +1 -0
index 0c8964c..41f3039 100644
@@ -98,6 +98,7 @@
 097 udf auto
 099 udf auto
 100 udf auto
+103 auto enospc quick
 105 acl auto quick
 112 rw aio auto quick
 113 rw aio auto quick
Signature · 6 lines
1.9.3

--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#4 lore raw
Quoted 82 lines · Eryu Guan
Quoted 81 lines · Brian Foster

XFS had a regression where inode reclaim in the unlink codepath would not correctly tear down extended attribute forks where no xattr extents are present. Add a generic test to create this condition.

The test sets extended attributes on a series of files under ENOSPC conditions and then verifies that the files can be removed without syslog warnings or errors.

---
 tests/generic/103     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/103.out |  2 ++
 tests/generic/group   |  1 +
 3 files changed, 94 insertions(+)
 create mode 100755 tests/generic/103
 create mode 100644 tests/generic/103.out
tests/generic/103 new file +58 -0
new file mode 100755
index 0000000..373cd9d
@@ -0,0 +1,91 @@
+#! /bin/bash
+# FSQA Test No. 103
+#
+# Test attribute fork teardown. This test is inspired by a regression in XFS
+# that resulted in problematic removal of inodes with remote attribute forks
+# without attribute extents. The attribute fork condition is created by
+# attempting to set larger attribute values on a filesystem that is at or near
+# ENOSPC.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2015 Red Hat, Inc.  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"
+
+_cleanup()
+{
+        cd /
+        rm -f $tmp.*
+}
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15 25
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/attr
+
+# real QA test starts here
+_supported_fs generic
+_supported_os Linux
+_require_scratch
+_require_attrs
+
+rm -f $seqres.full
+
+_consume_freesp()
+{
+	file=$1
Quoted 9 lines

tests/generic/103 line 63

Quoted 5 lines
tests/generic/103 new file +5 -0
+
+	# consume nearly all available space (leave ~512kB)
+	avail=`_get_available_space $SCRATCH_MNT`
+	filesizekb=$((avail / 1024 - 512))
+	$XFS_IO_PROG -fc "falloc 0 ${filesizekb}k" $file

This fails ext2/3 which don't have fallocate(2) support, need _require_xfs_io_command "falloc".

Ok.

Quoted 5 lines

This blocks ext2/3 from testing but extN just doesn't support 64k xattr value, we don't lose much test coverage. On the other hand, changing falloc to pwrite extends test time significantly (2s -> 30+s) and doesn't gain much benefit.

Yeah, I don't think it's worth that. It runs really quickly now and the reason behind it is an XFS regression.

Quoted 17 lines
Quoted 17 lines
tests/generic/103 new file +17 -0
+}
+
+_scratch_mkfs >> $seqres.full 2>&1
+_scratch_mount
+
+for i in $(seq 0 63); do
+	touch $SCRATCH_MNT/$seq.$i
+done
+
+# Generate a large attribute value and consume the rest of the space in the
+# filesystem.
+$XFS_IO_PROG -fc "pwrite 0 64k" $SCRATCH_MNT/attrval > /dev/null 2>&1
+_consume_freesp $SCRATCH_MNT/spc
+
+# Set attributes on the test files. These should start to hit ENOSPC.
+for i in $(seq 0 63); do
+	$SETFATTR_PROG -n user.test -v "`cat $SCRATCH_MNT/attrval`" \
Quoted 9 lines

tests/generic/103 line 86

Quoted 6 lines
tests/generic/103 new file +6 -0
+		$SCRATCH_MNT/$seq.$i >> $seqres.full 2>&1
+done
+
+# Remove the files with attributes to test attribute fork teardown. Problems
+# result in dmesg output.
+rm -f $SCRATCH_MNT/$seq.*

I see xfs corruption after the test, I think that's expected.

The problem should be fixed upstream, though 4.1 might still fail. I also need to backport the fix now that it's merged. ;)

Quoted 5 lines

But I just want to point out here that _check_dmesg is unable to detect the call traces in dmesg now, as the call trace doesn't have "WARNING", "kernel BUG" etc. keywords that _check_dmesg searches for, we rely on _check_fimesystems to fail the test.

Good catch, thanks. The failure here is an XFS error as opposed to a BUG() or warning. I've updated _check_dmesg() to catch "Internal error" as well.

Brian

Quoted 34 lines

Thanks,
Eryu

Quoted 32 lines
tests/generic/103 new file +5 -0
+
+echo Silence is golden.
+
+status=0
+exit
tests/generic/103.out new file +2 -0
new file mode 100644
index 0000000..ce229bf
@@ -0,0 +1,2 @@
+QA output created by 103
+Silence is golden.
tests/generic/group +1 -0
index 0c8964c..41f3039 100644
@@ -98,6 +98,7 @@
 097 udf auto
 099 udf auto
 100 udf auto
+103 auto enospc quick
 105 acl auto quick
 112 rw aio auto quick
 113 rw aio auto quick
Signature · 6 lines
1.9.3

--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html