On 07/17/2017 02:48 PM, Bill O'Donnell wrote:
> On Mon, Jul 17, 2017 at 02:41:04PM -0500, Eric Sandeen wrote:
>>
>>
>> On 07/17/2017 01:53 PM, Bill O'Donnell wrote:
>>> xfs_db should take type size into account when setting type.
>>> If type size isn't updated whenever type is set, a false crc
>>> error can occur due to the stale size. This test checks for
>>> that false crc error.
>>>
>>> Signed-off-by: Bill O'Donnell <billodo@xxxxxxxxxx>
>>> ---
>>> v2: Clarify commit message and test comments. Add a few more test cases.
>>> v3: Add more test cases. Use _scratch_xfs_db instead of $XFS_DB_PROG to
>>> accomodate fses with external logs. Remove superfluous daddr and type
>>> commands.
>>>
>>> tests/xfs/424 | 93 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>> tests/xfs/424.out | 1 +
>>> tests/xfs/group | 1 +
>>> 3 files changed, 95 insertions(+)
>>> create mode 100755 tests/xfs/424
>>> create mode 100644 tests/xfs/424.out
>>>
>>> diff --git a/tests/xfs/424 b/tests/xfs/424
>>> new file mode 100755
>>> index 00000000..8415ecd1
>>> --- /dev/null
>>> +++ b/tests/xfs/424
>>> @@ -0,0 +1,93 @@
>>> +#! /bin/bash
>>> +# FS QA Test 424
>>> +#
>>> +# xfs_db should take type size into account when setting type.
>>> +# If type size isn't updated whenever type is set, a false crc
>>> +# error can occur due to the stale size. This test checks for
>>> +# that false crc error.
>>> +#
>>> +#-----------------------------------------------------------------------
>>> +# Copyright (c) 2017 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.*
>>> +}
>>> +
>>> +_filter_dbval()
>>> +{
>>> + awk '{ print $4 }'
>>> +}
>>> +
>>> +# get standard environment, filters and checks
>>> +. ./common/rc
>>> +. ./common/filter
>>> +
>>> +# remove previous $seqres.full before test
>>> +rm -f $seqres.full
>>> +
>>> +# Modify as appropriate
>>> +_supported_os Linux
>>> +_supported_fs xfs
>>> +_require_scratch
>>> +
>>> +# real QA test starts here
>>> +_scratch_unmount >> $seqres.full 2>&1
>>> +
>>> +# for different sector sizes, ensure no CRC errors are falsely reported.
>>> +
>>> +# Supported types include: agf, agfl, agi, attr3, bmapbta,
>>> +# bmapbtd, bnobt, cntbt, data, dir3, dqblk, inobt, inodata,
>>> +# inode, log, rtbitmap, rtsummary, sb, symlink, text, finobt.
>>> +# For various sector sizes, test some types that involve type size.
>>> +for SECTOR_SIZE in 512 1024 2048 4096; do
>>> + $MKFS_XFS_PROG -f -s size=$SECTOR_SIZE $SCRATCH_DEV > /dev/null
>>> +
>>> + for TYPE in agf agi agfl sb; do
>>> + DADDR=`_scratch_xfs_db -c "$TYPE" -c "daddr" | _filter_dbval`
>>> + _scratch_xfs_db -c "daddr $DADDR" -c "type $TYPE"
>>> + done
>>> +
>>> + for TYPE in inodata data log text rtbitmap rtsummary; do
>>> + DADDR=`_scratch_xfs_db -c "sb" -c "addr rootino" -c "daddr" | _filter_dbval`
>>> + _scratch_xfs_db -c "daddr $DADDR" -c "type $TYPE"
>>> + done
>>
>> Ok, I don't understand this part. For each of those types, you get the address of
>> the root inode, then set it to the type - including types which are... not inodes?
>
> For this case, where addr set to rootino, the types included indeed work fine, but
> "type inode" yields crc error(s). :/
My [PATCH] xfs_db: properly set inode type should fix that... but I got hung up on what
to do if the current disk address isn't even an inode-aligned sector.
I ... think we should let it proceed, it's a debug tool after all.
Sorry, I forgot this wasn't in.
> It might make more sense to just leave the inodata data log text rtbitmap rtsummary
> types out of the test.
Or just set them up properly:
+ DADDR=`_scratch_xfs_db -c "sb" -c "addr rbmino" -c "daddr" | _filter_dbval`
+ _scratch_xfs_db -c "daddr $DADDR" -c "type rtbitmap"
+ DADDR=`_scratch_xfs_db -c "sb" -c "addr rsumino" -c "daddr" | _filter_dbval`
+ _scratch_xfs_db -c "daddr $DADDR" -c "type rtsummary"
+ DADDR=`_scratch_xfs_db -c "sb" -c "addr logstart" -c "daddr" | _filter_dbval`
+ _scratch_xfs_db -c "daddr $DADDR" -c "type log"
but i'm a little confused by some of these types. if you go to the sb and then
to logstart, xfs_db does indeed set the 'log' type. but if you go to rbmino,
it does /not/ set rtbitmap. Given that I don't know what these types are for,
I guess it might be best to leave them out of the test for now.
testing inode would be good though - i'll try to get that patch to a point where
I can merge it for 4.12, yet.
(and types data & text are trivial enough to do - not much to go wrong there,
but may as well exercise them while we're at it)
Might be worth it to add a comment about which ones don't get tested at this point.
-Eric
>
> fs_db> sb
> xfs_db> addr rootino
> xfs_db> daddr
> current daddr is 96
> xfs_db> type inode
> Metadata corruption detected at xfs_inode block 0x60/0x200
> Metadata corruption detected at xfs_inode block 0x60/0x200
>
>
>>
>> (sure, type text & type data can be anything, but the others are specific data
>> types...)
>>
>>> +
>>> + DADDR=`_scratch_xfs_db -c "agf" -c "addr bnoroot" -c "daddr" | _filter_dbval`
>>> + _scratch_xfs_db -c "daddr $DADDR" -c "type bnobt"
>>> + DADDR=`_scratch_xfs_db -c "agf" -c "addr cntroot" -c "daddr" | _filter_dbval`
>>> + _scratch_xfs_db -c "daddr $DADDR" -c "type cntbt"
>>> + DADDR=`_scratch_xfs_db -c "agi" -c "addr root" -c "daddr" | _filter_dbval`
>>> + _scratch_xfs_db -c "daddr $DADDR" -c "type inobt"
>>> + DADDR=`_scratch_xfs_db -c "agi" -c "addr free_root" -c "daddr" | _filter_dbval`
>>> + _scratch_xfs_db -c "daddr $DADDR" -c "type finobt"
>>
>> to test the generic data & text types, you could just do:
>>
>> + _scratch_xfs_db -c "daddr $DADDR" -c "type text"
>> + _scratch_xfs_db -c "daddr $DADDR" -c "type data"
>>
>> here. They don't hae CRCs anyway so if the test is truly to "test for crc errors"
>> then maybe it's not needed, but it can't hurt to actually iterate through each type
>> that's possible.
>>
>>> +done
>>> +
>>> +# success, all done
>>> +status=0
>>> +exit
>>> diff --git a/tests/xfs/424.out b/tests/xfs/424.out
>>> new file mode 100644
>>> index 00000000..d879a949
>>> --- /dev/null
>>> +++ b/tests/xfs/424.out
>>> @@ -0,0 +1 @@
>>> +QA output created by 424
>>> diff --git a/tests/xfs/group b/tests/xfs/group
>>> index ffdb0615..75c8280c 100644
>>> --- a/tests/xfs/group
>>> +++ b/tests/xfs/group
>>> @@ -421,3 +421,4 @@
>>> 421 auto quick clone dedupe
>>> 422 dangerous_scrub dangerous_online_repair
>>> 423 dangerous_scrub
>>> +424 auto quick db
>>>
>
--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
