Re: [PATCH] xfs: test changing UUID on V5 sueprblock

[Brian Foster <bfoster@redhat.com>]

On Thu, May 14, 2015 at 04:50:35PM -0500, Eric Sandeen wrote:
> Tests xfs_db's ability to change & restore UUIDs on
> V5 filesystems, and tests xfs_copy's ability to
> change the UUID on the copy.
> 
> Update to _filter_uuid is so that it will catch the
> UUID output from xfs_admin -u, which is slightly different
> than the regexp it was expecting.
> 
> This requires new userspace which knows how to change the
> UUID on a V5 superblock.
> 
> Signed-off-by: Eric Sandeen <sandeen@redhat.com>
> ---
> 
> (I dunno, is 60s "quick?" - I could cut down the fsstress
> ops, just wanted to populate the fs w/ metadata).
> 

Seems a bit on the long side to me, but not that bad. Might be worth a
try to cut the fsstress activity in half or so?

> diff --git a/common/filter b/common/filter
> index 05dbae6..8db1b3e 100644
> --- a/common/filter
> +++ b/common/filter
> @@ -275,9 +275,9 @@ _filter_uuid()
>  {
>  	if [ ! -z $1 ]; then
>  		UUID=$1
> -		sed -e "s/\(uuid:\) $UUID/\1 <EXACTUUID>/i"
> +		sed -e "s/\(uuid[ :=]\+\) $UUID/\1 <EXACTUUID>/i"
>  	else
> -		sed -e "s/\(uuid:\) *[0-9a-f-][0-9a-f-]*/\1 <UUID>/i"
> +		sed -e "s/\(uuid[ :=]\+\) *[0-9a-f-][0-9a-f-]*/\1 <UUID>/ig"
>  	fi
>  }
>  
> diff --git a/tests/xfs/074 b/tests/xfs/074
> new file mode 100755
> index 0000000..738050e
> --- /dev/null
> +++ b/tests/xfs/074
> @@ -0,0 +1,139 @@
> +#! /bin/bash
> +# FS QA Test 074
> +#
> +# test UUID modification of CRC-enabled filesystems
> +#
> +# CRC-enabled / V5 superblock filesystems have a UUID stamped into
> +# every piece of metadata, and a mechanism was added later to allow
> +# changing the user-visible UUID by copying the original UUID (which
> +# matches all the existing metadata) to a new superblock location.
> +# Exercise some of that behavior.
> +#
> +#-----------------------------------------------------------------------
> +# 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"
> +
> +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
> +
> +# real QA test starts here
> +
> +_supported_fs xfs
> +_supported_os Linux
> +_require_scratch
> +_require_xfs_crc
> +
> +_fs_has_META_UUID()
> +{
> +	FS=$1
> +	$XFS_DB_PROG -r -c version $FS | grep -q META_UUID
> +}
> +
> +rm -f $seqres.full
> +
> +_scratch_mkfs_xfs -m crc=1 >> $seqres.full 2>&1 || _fail "mkfs failed"
> +
> +xfs_admin -U restore $SCRATCH_DEV 2>&1 | \
> +	grep -q "invalid UUID\|supported on V5 fs" && \
> +		_notrun "Userspace does not support V5 super UUID modifications"
> +

Could we use $XFS_DB_PROG throughout the test so it's possible to modify
the test environment?

> +ORIG_UUID=`xfs_admin -u $SCRATCH_DEV | awk '{print $NF}'`
> +
> +_scratch_mount
> +# Put some stuff on the fs
> +$FSSTRESS_PROG -d $SCRATCH_MNT -n 1000 -p 4 >> $seqres.full 2>&1
> +_scratch_unmount
> +
> +# Can xfs_db change it?
> +
> +echo "== Generate new UUID"
> +xfs_admin -U generate $SCRATCH_DEV | _filter_uuid 
> +NEW_UUID=`xfs_admin -u $SCRATCH_DEV | awk '{print $NF}'`
> +[ "$NEW_UUID" == "$ORIG_UUID" ] && _fail "Failed to change UUID"
> +_check_scratch_fs
> +_scratch_mount || _notrun "Mount failed; no META_UUID kernel support?"
> +_scratch_unmount

Could we define a helper somewhere in common/ that determines whether
this works (e.g., change the uuid and mount)? Then this check can be a
less ambiguous failure and we wouldn't run part of the test in the case
where it's not supported.

> +_fs_has_META_UUID $SCRATCH_DEV || _fail "META_UUID feature not set"
> +
> +# This should be a no-op
> +echo "== Rewrite UUID"
> +xfs_admin -U rewrite $SCRATCH_DEV | _filter_uuid $NEW_UUID
> +_check_scratch_fs
> +_scratch_mount || _fail "Mount failed after UUID rewrite"
> +_scratch_unmount
> +_fs_has_META_UUID $SCRATCH_DEV || _fail "META_UUID feature not set"
> +

I also wonder if these hunks could be genericized a bit into a helper
function above..? E.g., something invoked like:

echo "== Rewrite UUID"
_test_uuid "rewrite" $UUID_FILTER
_fs_has_META_UUID $SCRATCH_DEV || _fail "META_UUID feature not set"

... or maybe it could also return the current post-command uuid to
support the explicit uuid checks as well. Just a thought, it might or
might not be a worthwhile cleanup...

> +# Can we change it back?
> +echo "== Restore old UUID"
> +xfs_admin -U restore $SCRATCH_DEV | _filter_uuid $ORIG_UUID
> +NEW_UUID=`xfs_admin -u $SCRATCH_DEV | awk '{print $NF}'`
> +[ "$NEW_UUID" != "$ORIG_UUID" ] && _fail "Failed to restore UUID"
> +_check_scratch_fs
> +_scratch_mount || _fail "Mount failed after UUID restore"
> +_scratch_unmount
> +_fs_has_META_UUID $SCRATCH_DEV &&  _fail "META_UUID feature should not be not set"
> +
> +# This should be a no-op too.
> +echo "== Rewrite UUID"
> +xfs_admin -U rewrite $SCRATCH_DEV | _filter_uuid  $ORIG_UUID
> +_check_scratch_fs
> +_scratch_mount || _fail "Mount failed after UUID rewrite"
> +_scratch_unmount
> +_fs_has_META_UUID $SCRATCH_DEV && _fail "META_UUID feature should not be not set"
> +
> +# Ok, now what does xfs_copy do; it changes UUID by default
> +
> +IMGFILE=$TEST_DIR/$seq.copy.img
> +rm -f $IMGFILE
> +
> +# xfs_copy changes the UUID by default
> +echo "== xfs_copy with new UUID"
> +$XFS_COPY_PROG $SCRATCH_DEV $IMGFILE 2>&1 >> $seqres.full || \
> +	_fail "xfs_copy (new UUID) failed"
> +_check_xfs_filesystem $IMGFILE none none || _fail "Copy looks corrupted"
> +# The copy should have META_UUID feature set
> +_fs_has_META_UUID $IMGFILE || _fail "META_UUID feature not set on copy"
> +

I suppose repair might be sufficient, but a mount check of the copy
target couldn't hurt here as well.

Brian

> +rm -f $IMGFILE
> +
> +# duplicating the UUID should be fine too
> +echo "== xfs_copy with duplicate UUID"
> +$XFS_COPY_PROG -d $SCRATCH_DEV $IMGFILE 2>&1 >> $seqres.full || \
> +	_fail "xfs_copy (duplicate) failed"
> +_check_xfs_filesystem $IMGFILE none none || _fail "Duplicate copy looks corrupted"
> +# The copy should not have META_UUID feature set
> +_fs_has_META_UUID $IMGFILE && _fail "META_UUID feature should not be set on copy"
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/xfs/074.out b/tests/xfs/074.out
> new file mode 100644
> index 0000000..b511107
> --- /dev/null
> +++ b/tests/xfs/074.out
> @@ -0,0 +1,21 @@
> +QA output created by 074
> +== Generate new UUID
> +Clearing log and setting UUID
> +writing all SBs
> +new UUID =  <UUID>
> +== Rewrite UUID
> +old UUID = <EXACTUUID>
> +Clearing log and setting UUID
> +writing all SBs
> +new UUID = <EXACTUUID>
> +== Restore old UUID
> +Clearing log and setting UUID
> +writing all SBs
> +new UUID = <EXACTUUID>
> +== Rewrite UUID
> +old UUID = <EXACTUUID>
> +Clearing log and setting UUID
> +writing all SBs
> +new UUID = <EXACTUUID>
> +== xfs_copy with new UUID
> +== xfs_copy with duplicate UUID
> diff --git a/tests/xfs/group b/tests/xfs/group
> index 58144d2..d1f6121 100644
> --- a/tests/xfs/group
> +++ b/tests/xfs/group
> @@ -70,6 +71,7 @@
>  071 rw auto
>  072 rw auto prealloc quick
>  073 copy auto
> +074 auto quick copy
>  078 growfs auto quick
>  080 rw ioctl
>  081 deprecated # log logprint quota
> 
> --
> 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