[PATCH 0/4] xfsqa: make some tests work again, add xfs_fsr test

[PATCH 0/4] xfsqa: make some tests work again, add xfs_fsr test

[Dave Chinner]

Enable the growfs test 104 and filter out the useless output, remove
hacks from 073 to do with loop devices and filter the log size out
in 206 to prevent failures with new mkfs binaries that allow large
logs.

Also, add a simple xfs_fsr test to improve coverage in the test
suite.  That's not hard given we have only one test (042) that
defragments only one file.

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

[PATCH 1/4] xfsqa: filter the fsstress seed output in 104

[Dave Chinner]

Otherwise it fails to match the golden output. Also add
the test to the auto group so that it is run regularly now
that the growfs problems are fixed.

Signed-off-by: Dave Chinner <david@fromorbit.com>
---
 104   |    2 +-
 group |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/104 b/104
index 6dea1d3..5f18d08 100755
--- a/104
+++ b/104
@@ -60,7 +60,7 @@ _stress_scratch()
 	procs=3
 	nops=1000
 	# -w ensures that the only ops are ones which cause write I/O
-	$FSSTRESS_PROG -d $SCRATCH_MNT -w -p $procs -n $nops $FSSTRESS_AVOID &
+	$FSSTRESS_PROG -d $SCRATCH_MNT -w -p $procs -n $nops $FSSTRESS_AVOID > /dev/null &
 }
 
 # real QA test starts here
diff --git a/group b/group
index 8d055a2..2c2e293 100644
--- a/group
+++ b/group
@@ -210,7 +210,7 @@ prealloc
 101 udf
 102 udf
 103 metadata dir ioctl auto quick
-104 growfs ioctl prealloc
+104 growfs ioctl prealloc auto
 105 acl auto quick
 106 quota
 107 quota
-- 
1.6.5

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

[PATCH 2/4] xfsqa: make unmounting loop devices in 073 work again

[Dave Chinner]

Loop devices are now created and destroyed on demand, and
the existing hacks to do this no longer work. Remove them.

Signed-off-by: Dave Chinner <david@fromorbit.com>
---
 073 |   13 +------------
 1 files changed, 1 insertions(+), 12 deletions(-)

diff --git a/073 b/073
index 5059448..ac4c64f 100755
--- a/073
+++ b/073
@@ -3,15 +3,6 @@
 #
 # Test xfs_copy
 #
-# HACK WARNING:
-#
-# Due to the severe brokenness of mount's handling of loopback devices, we
-# hardcode the loop devices we use for this test. This enables us to clean up
-# the pieces when we remount the loop device because mount loses all trace of
-# the fact this is a loop device. Hence to enable us to unmount the hosting
-# filesystem, we need to manually tear down the relevant loop device. If
-# mount ever gets fixed then this hack can be removed.
-#
 #-----------------------------------------------------------------------
 # Copyright (c) 2000-2003,2008 Silicon Graphics, Inc.  All Rights Reserved.
 #
@@ -127,9 +118,7 @@ _verify_copy()
 	diff -u $tmp.geometry1 $tmp.geometry2
 
 	echo unmounting and removing new image
-	loop=`mount | grep $target | grep -o -e 'loop=.*[^),]' | grep -o -e '/.*$'`
-	umount $source $target
-	losetup -d $loop > /dev/null 2>&1
+	umount $source_dir $target_dir
 	rm -f $target
 }
 
-- 
1.6.5

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

[PATCH 3/4] xfsqa: filter log size in 206

[Dave Chinner]

With the log size now going past 128MB, test 206 creates a filesystem
with a much larger log than 128MB. The golden output expects at
128MB log size, so the test breaks. Filter the log size out of the
mkfs output and golden output so that the test works on both older and
newer mkfs binaries.

Signed-off-by: Dave Chinner <david@fromorbit.com>
---
 206     |    9 ++++++---
 206.out |    6 +++---
 2 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/206 b/206
index b1b2286..0935335 100755
--- a/206
+++ b/206
@@ -75,19 +75,22 @@ dd if=/dev/zero of=$tmpfile bs=1 seek=19998630180864 count=1 >/dev/null 2>&1 \
 # mkfs slightly smaller than that
 echo "=== mkfs.xfs ==="
 mkfs.xfs -f -bsize=4096 -dagsize=76288719b,size=3905982455b -llazy-count=0 $tmpfile \
-	| sed -e "s,^meta-data=.*isize,meta-data=FILE                   isize,g"
+	| sed -e "s,^meta-data=.*isize,meta-data=FILE                   isize,g" \
+		-e "s/\(^log.*blocks=\)\([0-9]*,\)/\1XXXXX,/"
 
 mount -o loop $tmpfile $tmpdir || _fail "!!! failed to loopback mount"
 
 # see what happens when we growfs it
 echo "=== xfs_growfs ==="
 xfs_growfs $tmpdir \
-	| sed -e "s,^meta-data=.*isize,meta-data=FILE                   isize,g"
+	| sed -e "s,^meta-data=.*isize,meta-data=FILE                   isize,g" \
+		-e "s/\(^log.*blocks=\)\([0-9]*,\)/\1XXXXX,/"
 
 # and double-check the new geometry
 echo "=== xfs_info ==="
 xfs_info $tmpdir \
-	| sed -e "s,^meta-data=.*isize,meta-data=FILE                   isize,g"
+	| sed -e "s,^meta-data=.*isize,meta-data=FILE                   isize,g" \
+		-e "s/\(^log.*blocks=\)\([0-9]*,\)/\1XXXXX,/"
 
 # _cleanup cleans up for us
 
diff --git a/206.out b/206.out
index 57d4f4e..da70d01 100644
--- a/206.out
+++ b/206.out
@@ -6,7 +6,7 @@ meta-data=FILE                   isize=256    agcount=52, agsize=76288719 blks
 data     =                       bsize=4096   blocks=3905982455, imaxpct=5
          =                       sunit=0      swidth=0 blks
 naming   =version 2              bsize=4096   ascii-ci=0
-log      =internal log           bsize=4096   blocks=32768, version=2
+log      =internal log           bsize=4096   blocks=XXXXX, version=2
          =                       sectsz=512   sunit=0 blks, lazy-count=0
 realtime =none                   extsz=4096   blocks=0, rtextents=0
 === xfs_growfs ===
@@ -15,7 +15,7 @@ meta-data=FILE                   isize=256    agcount=52, agsize=76288719 blks
 data     =                       bsize=4096   blocks=3905982455, imaxpct=5
          =                       sunit=0      swidth=0 blks
 naming   =version 2              bsize=4096   ascii-ci=0
-log      =internal               bsize=4096   blocks=32768, version=2
+log      =internal               bsize=4096   blocks=XXXXX, version=2
          =                       sectsz=512   sunit=0 blks, lazy-count=0
 realtime =none                   extsz=4096   blocks=0, rtextents=0
 data blocks changed from 3905982455 to 4882478016
@@ -25,6 +25,6 @@ meta-data=FILE                   isize=256    agcount=64, agsize=76288719 blks
 data     =                       bsize=4096   blocks=4882478016, imaxpct=5
          =                       sunit=0      swidth=0 blks
 naming   =version 2              bsize=4096   ascii-ci=0
-log      =internal               bsize=4096   blocks=32768, version=2
+log      =internal               bsize=4096   blocks=XXXXX, version=2
          =                       sectsz=512   sunit=0 blks, lazy-count=0
 realtime =none                   extsz=4096   blocks=0, rtextents=0
-- 
1.6.5

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

[PATCH 4/4] xfsqa: add simple xfs_fsr test

[Dave Chinner]

We have very little fsr test coverage in the test suite right now.
042 is the only test that runs it, and it only defragments a single
file.

Improve the coverage by simply running xfs_fsr of the test
filesystem. This gives xfs_fsr a wide variety of inodes to try to
defragment. A simple test like this would have tripped over the
attr2 fork offset problems in the swap extents ioctl long ago
thanks to the varied nature of inodes the test suite leaves around
in this filesystem.

Signed-off-by: Dave Chinner <david@fromorbit.com>
---
 221     |   57 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 221.out |    2 ++
 group   |    1 +
 3 files changed, 60 insertions(+), 0 deletions(-)
 create mode 100644 221
 create mode 100644 221.out

diff --git a/221 b/221
new file mode 100644
index 0000000..5fe5583
--- /dev/null
+++ b/221
@@ -0,0 +1,57 @@
+#! /bin/sh
+# FS QA Test No. 221
+#
+# xfs_fsr QA tests
+# run xfs_fsr over the test filesystem to give it a wide and varied set of
+# inodes to try to defragment. This is effectively a crash/assert failure
+# test looking for corruption induced by xfs_fsr runs.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2010 Dave Chinner.  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
+#
+#-----------------------------------------------------------------------
+#
+# creator
+owner=david@fromorbit.com
+
+seq=`basename $0`
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+
+_cleanup()
+{
+    rm -f $tmp.*
+}
+
+trap "_cleanup ; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+
+# real QA test starts here
+_supported_fs xfs
+_supported_os Linux
+
+[ "$XFS_FSR_PROG" = "" ] && _notrun "xfs_fsr not found"
+
+xfs_fsr $TEST_MNT > /dev/null 2>&1
+
+echo "--- silence is golden ---"
+status=0 ; exit
diff --git a/221.out b/221.out
new file mode 100644
index 0000000..34b5d1b
--- /dev/null
+++ b/221.out
@@ -0,0 +1,2 @@
+QA output created by 221
+--- silence is golden ---
diff --git a/group b/group
index 2c2e293..528f5e9 100644
--- a/group
+++ b/group
@@ -330,3 +330,4 @@ prealloc
 218 auto fsr quick
 219 auto quota quick
 220 auto quota quick
+221 fsr ioctl auto quick
-- 
1.6.5

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH 1/4] xfsqa: filter the fsstress seed output in 104

[Christoph Hellwig]

On Tue, Jan 12, 2010 at 03:41:06PM +1100, Dave Chinner wrote:
> Otherwise it fails to match the golden output. Also add
> the test to the auto group so that it is run regularly now
> that the growfs problems are fixed.

Well, they're not actually fixed yet, Alex still needs to push out your
patches to the xfs tree :)

anyway, patch looks good,


Reviewed-by: Christoph Hellwig <hch@lst.de>

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH 3/4] xfsqa: filter log size in 206

[Christoph Hellwig]

On Tue, Jan 12, 2010 at 03:41:08PM +1100, Dave Chinner wrote:
> With the log size now going past 128MB, test 206 creates a filesystem
> with a much larger log than 128MB. The golden output expects at
> 128MB log size, so the test breaks. Filter the log size out of the
> mkfs output and golden output so that the test works on both older and
> newer mkfs binaries.

I guess you'll need a large enough test filesystem to hit this so I
don't see it, but it looks good,


Reviewed-by: Christoph Hellwig <hch@lst.de>

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH 4/4] xfsqa: add simple xfs_fsr test

[Christoph Hellwig]

On Tue, Jan 12, 2010 at 03:41:09PM +1100, Dave Chinner wrote:
> We have very little fsr test coverage in the test suite right now.
> 042 is the only test that runs it, and it only defragments a single
> file.
> 
> Improve the coverage by simply running xfs_fsr of the test
> filesystem. This gives xfs_fsr a wide variety of inodes to try to
> defragment. A simple test like this would have tripped over the
> attr2 fork offset problems in the swap extents ioctl long ago
> thanks to the varied nature of inodes the test suite leaves around
> in this filesystem.

Looks good,

and having it after 200+ testcases also means it should have at least a
bit of work to do.


Reviewed-by: Christoph Hellwig <hch@lst.de>

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH 2/4] xfsqa: make unmounting loop devices in 073 work again

[Christoph Hellwig]

On Tue, Jan 12, 2010 at 03:41:07PM +1100, Dave Chinner wrote:
> Loop devices are now created and destroyed on demand, and
> the existing hacks to do this no longer work. Remove them.

Just noticed that I had not been running this test at all for a long
time as it doesn't work with external logs.

Seems like the current code indeed fails reliably for me and this
patch fixes it.

The only problem is that the loop device autodestruct features is only
available since Linux 2.6.25 and also needs a recent util-linux.  Is
this old enough to simply go with the auto destruct version, or do
we need to support either one?

What about the patch below?  This adds the -d argument to the umount
calls for the loop images, which destroys the loop images in older
kernels and is a no-op for modern kernels that have the loop
auto-destruct feature:


Loop devices are now created and destroyed on demand, and
the existing hacks to do this no longer work. Remove them.

Signed-off-by: Dave Chinner <david@fromorbit.com>
---
 073 |   13 +------------
 1 files changed, 1 insertions(+), 12 deletions(-)

Index: xfstests-dev/073
===================================================================
--- xfstests-dev.orig/073	2009-05-28 18:29:58.000000000 +0000
+++ xfstests-dev/073	2010-01-12 09:53:37.000000000 +0000
@@ -3,15 +3,6 @@
 #
 # Test xfs_copy
 #
-# HACK WARNING:
-#
-# Due to the severe brokenness of mount's handling of loopback devices, we
-# hardcode the loop devices we use for this test. This enables us to clean up
-# the pieces when we remount the loop device because mount loses all trace of
-# the fact this is a loop device. Hence to enable us to unmount the hosting
-# filesystem, we need to manually tear down the relevant loop device. If
-# mount ever gets fixed then this hack can be removed.
-#
 #-----------------------------------------------------------------------
 # Copyright (c) 2000-2003,2008 Silicon Graphics, Inc.  All Rights Reserved.
 #
@@ -51,7 +42,7 @@ _cleanup()
 {
 	cd /
 	umount $SCRATCH_MNT 2>/dev/null
-	umount $imgs.loop 2>/dev/null
+	umount -d $imgs.loop 2>/dev/null
 	[ -d $imgs.loop ] && rmdir $imgs.loop
 	[ -d $imgs.source_dir ] && rm -rf $imgs.source_dir
 	rm -f $imgs.* $tmp.* /var/tmp/xfs_copy.log.*
@@ -127,9 +118,7 @@ _verify_copy()
 	diff -u $tmp.geometry1 $tmp.geometry2
 
 	echo unmounting and removing new image
-	loop=`mount | grep $target | grep -o -e 'loop=.*[^),]' | grep -o -e '/.*$'`
-	umount $source $target
-	losetup -d $loop > /dev/null 2>&1
+	umount -d $source_dir $target_dir
 	rm -f $target
 }
 

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH 2/4] xfsqa: make unmounting loop devices in 073 work again

[Dave Chinner]

On Tue, Jan 12, 2010 at 05:00:31AM -0500, Christoph Hellwig wrote:
> On Tue, Jan 12, 2010 at 03:41:07PM +1100, Dave Chinner wrote:
> > Loop devices are now created and destroyed on demand, and
> > the existing hacks to do this no longer work. Remove them.
> 
> Just noticed that I had not been running this test at all for a long
> time as it doesn't work with external logs.
> 
> Seems like the current code indeed fails reliably for me and this
> patch fixes it.
> 
> The only problem is that the loop device autodestruct features is only
> available since Linux 2.6.25 and also needs a recent util-linux.  Is
> this old enough to simply go with the auto destruct version, or do
> we need to support either one?
> 
> What about the patch below?  This adds the -d argument to the umount
> calls for the loop images, which destroys the loop images in older
> kernels and is a no-op for modern kernels that have the loop
> auto-destruct feature:

Yes, that looks like a better way to solve the problem.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs