Re: [PATCH v2] common/xfs: Add require_xfs_db_write_array function

["Darrick J. Wong" <darrick.wong@oracle.com>]

On Tue, Apr 17, 2018 at 12:03:40PM +0800, xuyang.jy wrote:
> on 2018/4/17 1:22, Darrick J. Wong write:
> 
> >On Mon, Apr 16, 2018 at 05:07:35PM +0800, yang xu wrote:
> >>xfsprogs commit 4222d000ed("db: write via array indexing doesn't
> >>work") fixes a bug that xfs_db write can't support array indexing.
> >>This function will check whether the bug is fixed on the current
> >>xfsprogs.
> >>
> >>xfs/444 applies the function, and skips if this bug exists.
> >>
> >>Signed-off-by: yang xu<xuyang.jy@cn.fujitsu.com>
> >>---
> >>  common/xfs    | 18 ++++++++++++++++++
> >>  tests/xfs/444 |  1 +
> >>  2 files changed, 19 insertions(+)
> >>
> >>diff --git a/common/xfs b/common/xfs
> >>index 3169f87..c4611a1 100644
> >>--- a/common/xfs
> >>+++ b/common/xfs
> >>@@ -701,3 +701,21 @@ _scratch_xfs_set_sb_field()
> >>  {
> >>  	_scratch_xfs_set_metadata_field "$1" "$2" "sb 0"
> >>  }
> >>+
> >>+#Before xfsprogs commit 4222d000ed("db: write via array indexing doesn't work"),
> >>+#xfs_db command to write a specific AGFL index doesn't work.
> >>+_require_xfs_db_write_array()
> >>+{
> >>+	_scratch_xfs_set_metadata_field "bno[32]" "78" "agfl 0">/dev/null 2>&1
> >This is going to fail randomly unless the caller previously called
> >_require_scratch to make sure that there's some kind of xfs fs image
> >sitting on the scratch device (and that there even is a scratch device).
> >Please document that requirement.  Alternately, consider formatting a xfs
> >image file in $TEST_DIR and poking that with xfs_db so you don't have to
> >deal with formatting and cleaning up the scratch device.
> Hi Darrick
> 
> We want to create a xfs image file in /tmp/ and use it, as below:
> 
> [root@RHEL7U5Alpha_CLIENT xfstests-dev]# git diff common/xfs
> diff --git a/common/xfs b/common/xfs
> index 3169f87..39702f7 100644
> --- a/common/xfs
> +++ b/common/xfs
> @@ -701,3 +701,19 @@ _scratch_xfs_set_sb_field()
>  {      	_scratch_xfs_set_metadata_field "$1" "$2" "sb 0"
>  }
> +
> +#Before xfsprogs commit 4222d000ed("db: write via array indexing doesn't work"),
> +##xfs_db command to write a specific AGFL index doesn't work.
> +_require_xfs_db_write_array()
> +{
> +       local supported=0
> +
> +       touch /tmp/$$.img
> +       $MKFS_XFS_PROG -d file,name=/tmp/$$.img,size=512m>/dev/null 2>&1
> +       $XFS_DB_PROG -x -c "agfl 0" -c "write bno[32] 78" /tmp/$$.img \
> +>/dev/null 2>&1
> +       $XFS_DB_PROG -x -c "agfl 0" -c "print bno[32]" /tmp/$$.img \
> +               | grep -q "bno\[32\] = 78"&&  supported=1
> +       rm -f /tmp/$$.img
> +       [ $supported -eq 0 ]&&  _notrun "xfs_db write can't support array"
> +}
> [root@RHEL7U5Alpha_CLIENT xfstests-dev]#
> 
> How about this modification?

$TEST_DIR/$seqres.img, not /tmp/$$.img

Don't assume anything's writable (and big enough for fs images) if it
isn't under $TEST_DIR or $SCRATCH_MNT.

--D

> 
> >>+
> >>+	# Before xfsprogs commit 0ebbf1d58898 ("db: limit AGFL bno
> >>+	# arrayi printing), When asking for a single agfl entry,
> >>+	# the result outputs the entire remainder of the array
> >>+	# starting at the given index.
> >>+	# It is difficult to extract single entry values.
> >>+	# So filter useless information.
> >>+	bno_value=$(_scratch_xfs_get_metadata_field "bno[32]" \
> >>+		"agfl 0" |sed -e 's/ .*$//g' | sed -e 's/^.*://g')
> >>+
> >>+	[ "${bno_value}" != "78" ]&&  _notrun "xfs_db write can't support array"
> >It seems like a bad idea to leave a potentially corrupt scratch fs lying
> >around after this helper exits.
> Agreed.  I will use a temp xfs image file.
> 
> Thanks
> Yang Xu
> 
> >--D
> >
> >>+}
> >>diff --git a/tests/xfs/444 b/tests/xfs/444
> >>index 141be52..9700422 100755
> >>--- a/tests/xfs/444
> >>+++ b/tests/xfs/444
> >>@@ -54,6 +54,7 @@ _supported_os Linux
> >>  _require_check_dmesg
> >>  _require_scratch
> >>  _require_test_program "punch-alternating"
> >>+_require_xfs_db_write_array
> >>
> >>  # This is only a v5 filesystem problem
> >>  _require_scratch_xfs_crc
> >>-- 
> >>1.8.3.1
> >>
> >>
> >>
> >>--
> >>To unsubscribe from this list: send the line "unsubscribe fstests" in
> >>the body of a message tomajordomo@vger.kernel.org
> >>More majordomo info athttp://vger.kernel.org/majordomo-info.html
> >.
> >
> 
> 
>