* [PATCH v2 1/2] xfs/288: use -d option of xfs_db write command for v5 XFS
@ 2018-07-31 5:41 Zorro Lang
2018-07-31 5:41 ` [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr Zorro Lang
0 siblings, 1 reply; 7+ messages in thread
From: Zorro Lang @ 2018-07-31 5:41 UTC (permalink / raw)
To: fstests; +Cc: linux-xfs, yangx.jy
The commit b3cf8b72334fd35ef961869506e5a72ab398bc82 update xfs/288
to support v5 filesystem testing. That commit thought xfs_db write
command can work well with -d option on V5 XFS. But the truth is
the case doesn't use that option.
So turn to use _scratch_xfs_set_metadata_field, it will help to use
proper options for xfs_db write command.
Signed-off-by: Zorro Lang <zlang@redhat.com>
---
Hi,
This patch has been reviewed by Darrick J. Wong:
https://marc.info/?l=linux-xfs&m=153287978423915&w=2
I sent this patch again only because it's related with next [2/2] patch.
Thanks,
Zorro
tests/xfs/288 | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/tests/xfs/288 b/tests/xfs/288
index bccdb6fc..f4165b6c 100755
--- a/tests/xfs/288
+++ b/tests/xfs/288
@@ -56,9 +56,8 @@ $SETFATTR_PROG -n "user.testattr${seq}" \
_scratch_unmount
# manually corrupt the XFS, by set the header count of attr to 0
-_scratch_xfs_db -x -c "inode $inum" \
- -c "ablock 0" \
- -c "write hdr.count 0" >> $seqres.full
+_scratch_xfs_set_metadata_field "hdr.count" "0" \
+ "inode $inum" "ablock 0" >> $seqres.full
# make sure xfs_repair can find above corruption. If it can't, that
# means we need to fix this bug on current xfs_repair
--
2.14.4
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr
2018-07-31 5:41 [PATCH v2 1/2] xfs/288: use -d option of xfs_db write command for v5 XFS Zorro Lang
@ 2018-07-31 5:41 ` Zorro Lang
2018-08-02 0:46 ` Darrick J. Wong
2018-08-06 0:32 ` Dave Chinner
0 siblings, 2 replies; 7+ messages in thread
From: Zorro Lang @ 2018-07-31 5:41 UTC (permalink / raw)
To: fstests; +Cc: linux-xfs, yangx.jy
Commit b3cf8b72334fd35ef961869506e5a72ab398bc82 help xfs/288 to
support v5 filesystems testing, but there're still some old
distributions don't support xfs_db write '-d' option, or can't
write values into dir/attr of v5 filesystems.
For compatible with older versions, skip this test on v5 xfs
if xfs_db write can't write v5 XFS dir/attr.
Signed-off-by: Zorro Lang <zlang@redhat.com>
---
Hi,
Xiao found v1 can't work well on xfsprogs-4.9, due to 'write -d' can't
write vaules into dir/attr of v5 filesystems before xfsprogs-4.13.
Only check "if write command has -d option" is not enough. So I write
a function _require_xfs_db_write_da() to make sure current xfs_db
can write dir/attr (not only on v5, but especially for v5).
Thanks,
Zorro
common/xfs | 28 ++++++++++++++++++++++++++++
tests/xfs/288 | 1 +
2 files changed, 29 insertions(+)
diff --git a/common/xfs b/common/xfs
index d971b4a8..95dcfd59 100644
--- a/common/xfs
+++ b/common/xfs
@@ -720,6 +720,34 @@ _require_xfs_db_write_array()
[ $supported -eq 0 ] && _notrun "xfs_db write can't support array"
}
+# Before xfsprogs commit 89baf918(xfs_db: write values into dir/attr blocks and
+# recalculate CRCs), xfs_db write command can't write invalid data into dir/attr
+# field of v5 filesystems. For some cases need to write dir/attr (especially crc
+# enabled), use this _require at first.
+_require_xfs_db_write_da()
+{
+ local inum
+ local count
+
+ _require_scratch
+
+ _scratch_mkfs > /dev/null 2>&1
+ _scratch_mount
+ mkdir $SCRATCH_MNT/$seq.dir
+ inum=$(stat -c '%i' $SCRATCH_MNT/$seq.dir)
+ $SETFATTR_PROG -n "user.testda" \
+ -v "$(perl -e "print 'v' x 65536;")" \
+ $SCRATCH_MNT/$seq.dir
+ _scratch_unmount
+ _scratch_xfs_set_metadata_field "hdr.count" "0" \
+ "inode $inum" "ablock 0" >/dev/null 2>&1
+ count=$(_scratch_xfs_get_metadata_field "hdr.count" \
+ "inode $inum" "ablock 0")
+ if [ "$count" != "0" ]; then
+ _notrun "xfs_db write can't write values into dir/attr blocks"
+ fi
+}
+
_require_xfs_spaceman_command()
{
if [ -z "$1" ]; then
diff --git a/tests/xfs/288 b/tests/xfs/288
index f4165b6c..e9589fd5 100755
--- a/tests/xfs/288
+++ b/tests/xfs/288
@@ -35,6 +35,7 @@ _supported_fs xfs
_supported_os Linux
_require_scratch
_require_attrs
+_require_xfs_db_write_da
# get block size ($dbsize) from the mkfs output
_scratch_mkfs_xfs 2>/dev/null | _filter_mkfs 2>$tmp.mkfs >/dev/null
--
2.14.4
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr
2018-07-31 5:41 ` [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr Zorro Lang
@ 2018-08-02 0:46 ` Darrick J. Wong
2018-08-02 1:49 ` Zorro Lang
2018-08-06 0:32 ` Dave Chinner
1 sibling, 1 reply; 7+ messages in thread
From: Darrick J. Wong @ 2018-08-02 0:46 UTC (permalink / raw)
To: Zorro Lang; +Cc: fstests, linux-xfs, yangx.jy
On Tue, Jul 31, 2018 at 01:41:19PM +0800, Zorro Lang wrote:
> Commit b3cf8b72334fd35ef961869506e5a72ab398bc82 help xfs/288 to
> support v5 filesystems testing, but there're still some old
> distributions don't support xfs_db write '-d' option, or can't
> write values into dir/attr of v5 filesystems.
>
> For compatible with older versions, skip this test on v5 xfs
> if xfs_db write can't write v5 XFS dir/attr.
>
> Signed-off-by: Zorro Lang <zlang@redhat.com>
> ---
>
> Hi,
>
> Xiao found v1 can't work well on xfsprogs-4.9, due to 'write -d' can't
> write vaules into dir/attr of v5 filesystems before xfsprogs-4.13.
>
> Only check "if write command has -d option" is not enough. So I write
> a function _require_xfs_db_write_da() to make sure current xfs_db
> can write dir/attr (not only on v5, but especially for v5).
>
> Thanks,
> Zorro
>
> common/xfs | 28 ++++++++++++++++++++++++++++
> tests/xfs/288 | 1 +
> 2 files changed, 29 insertions(+)
>
> diff --git a/common/xfs b/common/xfs
> index d971b4a8..95dcfd59 100644
> --- a/common/xfs
> +++ b/common/xfs
> @@ -720,6 +720,34 @@ _require_xfs_db_write_array()
> [ $supported -eq 0 ] && _notrun "xfs_db write can't support array"
> }
>
> +# Before xfsprogs commit 89baf918(xfs_db: write values into dir/attr blocks and
> +# recalculate CRCs), xfs_db write command can't write invalid data into dir/attr
> +# field of v5 filesystems. For some cases need to write dir/attr (especially crc
> +# enabled), use this _require at first.
> +_require_xfs_db_write_da()
> +{
> + local inum
> + local count
> +
> + _require_scratch
> +
> + _scratch_mkfs > /dev/null 2>&1
> + _scratch_mount
> + mkdir $SCRATCH_MNT/$seq.dir
> + inum=$(stat -c '%i' $SCRATCH_MNT/$seq.dir)
> + $SETFATTR_PROG -n "user.testda" \
> + -v "$(perl -e "print 'v' x 65536;")" \
> + $SCRATCH_MNT/$seq.dir
> + _scratch_unmount
> + _scratch_xfs_set_metadata_field "hdr.count" "0" \
Just FYI an upcoming change to the attr block verifier in 4.19 will
break this because it rejects hdr.count == 0. Can you please set this
to 1 instead of 0?
> + "inode $inum" "ablock 0" >/dev/null 2>&1
> + count=$(_scratch_xfs_get_metadata_field "hdr.count" \
> + "inode $inum" "ablock 0")
> + if [ "$count" != "0" ]; then
And update the test here?
--D
> + _notrun "xfs_db write can't write values into dir/attr blocks"
> + fi
> +}
> +
> _require_xfs_spaceman_command()
> {
> if [ -z "$1" ]; then
> diff --git a/tests/xfs/288 b/tests/xfs/288
> index f4165b6c..e9589fd5 100755
> --- a/tests/xfs/288
> +++ b/tests/xfs/288
> @@ -35,6 +35,7 @@ _supported_fs xfs
> _supported_os Linux
> _require_scratch
> _require_attrs
> +_require_xfs_db_write_da
>
> # get block size ($dbsize) from the mkfs output
> _scratch_mkfs_xfs 2>/dev/null | _filter_mkfs 2>$tmp.mkfs >/dev/null
> --
> 2.14.4
>
> --
> 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
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr
2018-08-02 0:46 ` Darrick J. Wong
@ 2018-08-02 1:49 ` Zorro Lang
2018-08-02 5:14 ` Darrick J. Wong
0 siblings, 1 reply; 7+ messages in thread
From: Zorro Lang @ 2018-08-02 1:49 UTC (permalink / raw)
To: Darrick J. Wong; +Cc: fstests, linux-xfs, yangx.jy
On Wed, Aug 01, 2018 at 05:46:34PM -0700, Darrick J. Wong wrote:
> On Tue, Jul 31, 2018 at 01:41:19PM +0800, Zorro Lang wrote:
> > Commit b3cf8b72334fd35ef961869506e5a72ab398bc82 help xfs/288 to
> > support v5 filesystems testing, but there're still some old
> > distributions don't support xfs_db write '-d' option, or can't
> > write values into dir/attr of v5 filesystems.
> >
> > For compatible with older versions, skip this test on v5 xfs
> > if xfs_db write can't write v5 XFS dir/attr.
> >
> > Signed-off-by: Zorro Lang <zlang@redhat.com>
> > ---
> >
> > Hi,
> >
> > Xiao found v1 can't work well on xfsprogs-4.9, due to 'write -d' can't
> > write vaules into dir/attr of v5 filesystems before xfsprogs-4.13.
> >
> > Only check "if write command has -d option" is not enough. So I write
> > a function _require_xfs_db_write_da() to make sure current xfs_db
> > can write dir/attr (not only on v5, but especially for v5).
> >
> > Thanks,
> > Zorro
> >
> > common/xfs | 28 ++++++++++++++++++++++++++++
> > tests/xfs/288 | 1 +
> > 2 files changed, 29 insertions(+)
> >
> > diff --git a/common/xfs b/common/xfs
> > index d971b4a8..95dcfd59 100644
> > --- a/common/xfs
> > +++ b/common/xfs
> > @@ -720,6 +720,34 @@ _require_xfs_db_write_array()
> > [ $supported -eq 0 ] && _notrun "xfs_db write can't support array"
> > }
> >
> > +# Before xfsprogs commit 89baf918(xfs_db: write values into dir/attr blocks and
> > +# recalculate CRCs), xfs_db write command can't write invalid data into dir/attr
> > +# field of v5 filesystems. For some cases need to write dir/attr (especially crc
> > +# enabled), use this _require at first.
> > +_require_xfs_db_write_da()
> > +{
> > + local inum
> > + local count
> > +
> > + _require_scratch
> > +
> > + _scratch_mkfs > /dev/null 2>&1
> > + _scratch_mount
> > + mkdir $SCRATCH_MNT/$seq.dir
> > + inum=$(stat -c '%i' $SCRATCH_MNT/$seq.dir)
> > + $SETFATTR_PROG -n "user.testda" \
> > + -v "$(perl -e "print 'v' x 65536;")" \
> > + $SCRATCH_MNT/$seq.dir
> > + _scratch_unmount
> > + _scratch_xfs_set_metadata_field "hdr.count" "0" \
>
> Just FYI an upcoming change to the attr block verifier in 4.19 will
> break this because it rejects hdr.count == 0. Can you please set this
> to 1 instead of 0?
>
> > + "inode $inum" "ablock 0" >/dev/null 2>&1
> > + count=$(_scratch_xfs_get_metadata_field "hdr.count" \
> > + "inode $inum" "ablock 0")
> > + if [ "$count" != "0" ]; then
>
> And update the test here?
Sure, I can change the 0 to 1. BTW, does it affect xfs/288? Because xfs/288 set
hdr.count to 0 too.
Thanks,
Zorro
>
> --D
>
> > + _notrun "xfs_db write can't write values into dir/attr blocks"
> > + fi
> > +}
> > +
> > _require_xfs_spaceman_command()
> > {
> > if [ -z "$1" ]; then
> > diff --git a/tests/xfs/288 b/tests/xfs/288
> > index f4165b6c..e9589fd5 100755
> > --- a/tests/xfs/288
> > +++ b/tests/xfs/288
> > @@ -35,6 +35,7 @@ _supported_fs xfs
> > _supported_os Linux
> > _require_scratch
> > _require_attrs
> > +_require_xfs_db_write_da
> >
> > # get block size ($dbsize) from the mkfs output
> > _scratch_mkfs_xfs 2>/dev/null | _filter_mkfs 2>$tmp.mkfs >/dev/null
> > --
> > 2.14.4
> >
> > --
> > 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
> --
> 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
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr
2018-08-02 1:49 ` Zorro Lang
@ 2018-08-02 5:14 ` Darrick J. Wong
2018-08-02 7:10 ` Zorro Lang
0 siblings, 1 reply; 7+ messages in thread
From: Darrick J. Wong @ 2018-08-02 5:14 UTC (permalink / raw)
To: Zorro Lang; +Cc: fstests, linux-xfs, yangx.jy
On Thu, Aug 02, 2018 at 09:49:41AM +0800, Zorro Lang wrote:
> On Wed, Aug 01, 2018 at 05:46:34PM -0700, Darrick J. Wong wrote:
> > On Tue, Jul 31, 2018 at 01:41:19PM +0800, Zorro Lang wrote:
> > > Commit b3cf8b72334fd35ef961869506e5a72ab398bc82 help xfs/288 to
> > > support v5 filesystems testing, but there're still some old
> > > distributions don't support xfs_db write '-d' option, or can't
> > > write values into dir/attr of v5 filesystems.
> > >
> > > For compatible with older versions, skip this test on v5 xfs
> > > if xfs_db write can't write v5 XFS dir/attr.
> > >
> > > Signed-off-by: Zorro Lang <zlang@redhat.com>
> > > ---
> > >
> > > Hi,
> > >
> > > Xiao found v1 can't work well on xfsprogs-4.9, due to 'write -d' can't
> > > write vaules into dir/attr of v5 filesystems before xfsprogs-4.13.
> > >
> > > Only check "if write command has -d option" is not enough. So I write
> > > a function _require_xfs_db_write_da() to make sure current xfs_db
> > > can write dir/attr (not only on v5, but especially for v5).
> > >
> > > Thanks,
> > > Zorro
> > >
> > > common/xfs | 28 ++++++++++++++++++++++++++++
> > > tests/xfs/288 | 1 +
> > > 2 files changed, 29 insertions(+)
> > >
> > > diff --git a/common/xfs b/common/xfs
> > > index d971b4a8..95dcfd59 100644
> > > --- a/common/xfs
> > > +++ b/common/xfs
> > > @@ -720,6 +720,34 @@ _require_xfs_db_write_array()
> > > [ $supported -eq 0 ] && _notrun "xfs_db write can't support array"
> > > }
> > >
> > > +# Before xfsprogs commit 89baf918(xfs_db: write values into dir/attr blocks and
> > > +# recalculate CRCs), xfs_db write command can't write invalid data into dir/attr
> > > +# field of v5 filesystems. For some cases need to write dir/attr (especially crc
> > > +# enabled), use this _require at first.
> > > +_require_xfs_db_write_da()
> > > +{
> > > + local inum
> > > + local count
> > > +
> > > + _require_scratch
> > > +
> > > + _scratch_mkfs > /dev/null 2>&1
> > > + _scratch_mount
> > > + mkdir $SCRATCH_MNT/$seq.dir
> > > + inum=$(stat -c '%i' $SCRATCH_MNT/$seq.dir)
> > > + $SETFATTR_PROG -n "user.testda" \
> > > + -v "$(perl -e "print 'v' x 65536;")" \
> > > + $SCRATCH_MNT/$seq.dir
> > > + _scratch_unmount
> > > + _scratch_xfs_set_metadata_field "hdr.count" "0" \
> >
> > Just FYI an upcoming change to the attr block verifier in 4.19 will
> > break this because it rejects hdr.count == 0. Can you please set this
> > to 1 instead of 0?
> >
> > > + "inode $inum" "ablock 0" >/dev/null 2>&1
> > > + count=$(_scratch_xfs_get_metadata_field "hdr.count" \
> > > + "inode $inum" "ablock 0")
> > > + if [ "$count" != "0" ]; then
> >
> > And update the test here?
>
> Sure, I can change the 0 to 1. BTW, does it affect xfs/288? Because xfs/288 set
> hdr.count to 0 too.
IIUC, xfs/288 deliberately sets hdr.count to 0 to test how the code
handles the (corrupt) attr header, so that use is fine.
--D
> Thanks,
> Zorro
>
> >
> > --D
> >
> > > + _notrun "xfs_db write can't write values into dir/attr blocks"
> > > + fi
> > > +}
> > > +
> > > _require_xfs_spaceman_command()
> > > {
> > > if [ -z "$1" ]; then
> > > diff --git a/tests/xfs/288 b/tests/xfs/288
> > > index f4165b6c..e9589fd5 100755
> > > --- a/tests/xfs/288
> > > +++ b/tests/xfs/288
> > > @@ -35,6 +35,7 @@ _supported_fs xfs
> > > _supported_os Linux
> > > _require_scratch
> > > _require_attrs
> > > +_require_xfs_db_write_da
> > >
> > > # get block size ($dbsize) from the mkfs output
> > > _scratch_mkfs_xfs 2>/dev/null | _filter_mkfs 2>$tmp.mkfs >/dev/null
> > > --
> > > 2.14.4
> > >
> > > --
> > > 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
> > --
> > 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
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr
2018-08-02 5:14 ` Darrick J. Wong
@ 2018-08-02 7:10 ` Zorro Lang
0 siblings, 0 replies; 7+ messages in thread
From: Zorro Lang @ 2018-08-02 7:10 UTC (permalink / raw)
To: Darrick J. Wong; +Cc: fstests, linux-xfs, yangx.jy
On Wed, Aug 01, 2018 at 10:14:52PM -0700, Darrick J. Wong wrote:
> On Thu, Aug 02, 2018 at 09:49:41AM +0800, Zorro Lang wrote:
> > On Wed, Aug 01, 2018 at 05:46:34PM -0700, Darrick J. Wong wrote:
> > > On Tue, Jul 31, 2018 at 01:41:19PM +0800, Zorro Lang wrote:
> > > > Commit b3cf8b72334fd35ef961869506e5a72ab398bc82 help xfs/288 to
> > > > support v5 filesystems testing, but there're still some old
> > > > distributions don't support xfs_db write '-d' option, or can't
> > > > write values into dir/attr of v5 filesystems.
> > > >
> > > > For compatible with older versions, skip this test on v5 xfs
> > > > if xfs_db write can't write v5 XFS dir/attr.
> > > >
> > > > Signed-off-by: Zorro Lang <zlang@redhat.com>
> > > > ---
> > > >
> > > > Hi,
> > > >
> > > > Xiao found v1 can't work well on xfsprogs-4.9, due to 'write -d' can't
> > > > write vaules into dir/attr of v5 filesystems before xfsprogs-4.13.
> > > >
> > > > Only check "if write command has -d option" is not enough. So I write
> > > > a function _require_xfs_db_write_da() to make sure current xfs_db
> > > > can write dir/attr (not only on v5, but especially for v5).
> > > >
> > > > Thanks,
> > > > Zorro
> > > >
> > > > common/xfs | 28 ++++++++++++++++++++++++++++
> > > > tests/xfs/288 | 1 +
> > > > 2 files changed, 29 insertions(+)
> > > >
> > > > diff --git a/common/xfs b/common/xfs
> > > > index d971b4a8..95dcfd59 100644
> > > > --- a/common/xfs
> > > > +++ b/common/xfs
> > > > @@ -720,6 +720,34 @@ _require_xfs_db_write_array()
> > > > [ $supported -eq 0 ] && _notrun "xfs_db write can't support array"
> > > > }
> > > >
> > > > +# Before xfsprogs commit 89baf918(xfs_db: write values into dir/attr blocks and
> > > > +# recalculate CRCs), xfs_db write command can't write invalid data into dir/attr
> > > > +# field of v5 filesystems. For some cases need to write dir/attr (especially crc
> > > > +# enabled), use this _require at first.
> > > > +_require_xfs_db_write_da()
> > > > +{
> > > > + local inum
> > > > + local count
> > > > +
> > > > + _require_scratch
> > > > +
> > > > + _scratch_mkfs > /dev/null 2>&1
> > > > + _scratch_mount
> > > > + mkdir $SCRATCH_MNT/$seq.dir
> > > > + inum=$(stat -c '%i' $SCRATCH_MNT/$seq.dir)
> > > > + $SETFATTR_PROG -n "user.testda" \
> > > > + -v "$(perl -e "print 'v' x 65536;")" \
> > > > + $SCRATCH_MNT/$seq.dir
> > > > + _scratch_unmount
> > > > + _scratch_xfs_set_metadata_field "hdr.count" "0" \
> > >
> > > Just FYI an upcoming change to the attr block verifier in 4.19 will
> > > break this because it rejects hdr.count == 0. Can you please set this
> > > to 1 instead of 0?
> > >
> > > > + "inode $inum" "ablock 0" >/dev/null 2>&1
> > > > + count=$(_scratch_xfs_get_metadata_field "hdr.count" \
> > > > + "inode $inum" "ablock 0")
> > > > + if [ "$count" != "0" ]; then
> > >
> > > And update the test here?
> >
> > Sure, I can change the 0 to 1. BTW, does it affect xfs/288? Because xfs/288 set
> > hdr.count to 0 too.
>
> IIUC, xfs/288 deliberately sets hdr.count to 0 to test how the code
> handles the (corrupt) attr header, so that use is fine.
Hmm... sorry I'm confused now, can you show me the patch which will
"rejects hdr.count == 0"?
xfs/288 trys to check how xfs_repair deal with hdr.count=0. If
"write (-d) hdr.count 0" will be rejected directly, that's another issue.
If it won't affect xfs_db "write", why I can't write 0 in this patch?
Thanks,
Zorro
>
> --D
>
> > Thanks,
> > Zorro
> >
> > >
> > > --D
> > >
> > > > + _notrun "xfs_db write can't write values into dir/attr blocks"
> > > > + fi
> > > > +}
> > > > +
> > > > _require_xfs_spaceman_command()
> > > > {
> > > > if [ -z "$1" ]; then
> > > > diff --git a/tests/xfs/288 b/tests/xfs/288
> > > > index f4165b6c..e9589fd5 100755
> > > > --- a/tests/xfs/288
> > > > +++ b/tests/xfs/288
> > > > @@ -35,6 +35,7 @@ _supported_fs xfs
> > > > _supported_os Linux
> > > > _require_scratch
> > > > _require_attrs
> > > > +_require_xfs_db_write_da
> > > >
> > > > # get block size ($dbsize) from the mkfs output
> > > > _scratch_mkfs_xfs 2>/dev/null | _filter_mkfs 2>$tmp.mkfs >/dev/null
> > > > --
> > > > 2.14.4
> > > >
> > > > --
> > > > 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
> > > --
> > > 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
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> --
> 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
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr
2018-07-31 5:41 ` [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr Zorro Lang
2018-08-02 0:46 ` Darrick J. Wong
@ 2018-08-06 0:32 ` Dave Chinner
1 sibling, 0 replies; 7+ messages in thread
From: Dave Chinner @ 2018-08-06 0:32 UTC (permalink / raw)
To: Zorro Lang; +Cc: fstests, linux-xfs, yangx.jy
On Tue, Jul 31, 2018 at 01:41:19PM +0800, Zorro Lang wrote:
> Commit b3cf8b72334fd35ef961869506e5a72ab398bc82 help xfs/288 to
> support v5 filesystems testing, but there're still some old
> distributions don't support xfs_db write '-d' option, or can't
> write values into dir/attr of v5 filesystems.
>
> For compatible with older versions, skip this test on v5 xfs
> if xfs_db write can't write v5 XFS dir/attr.
>
> Signed-off-by: Zorro Lang <zlang@redhat.com>
> ---
>
> Hi,
>
> Xiao found v1 can't work well on xfsprogs-4.9, due to 'write -d' can't
> write vaules into dir/attr of v5 filesystems before xfsprogs-4.13.
>
> Only check "if write command has -d option" is not enough. So I write
> a function _require_xfs_db_write_da() to make sure current xfs_db
> can write dir/attr (not only on v5, but especially for v5).
>
> Thanks,
> Zorro
>
> common/xfs | 28 ++++++++++++++++++++++++++++
> tests/xfs/288 | 1 +
> 2 files changed, 29 insertions(+)
>
> diff --git a/common/xfs b/common/xfs
> index d971b4a8..95dcfd59 100644
> --- a/common/xfs
> +++ b/common/xfs
> @@ -720,6 +720,34 @@ _require_xfs_db_write_array()
> [ $supported -eq 0 ] && _notrun "xfs_db write can't support array"
> }
>
> +# Before xfsprogs commit 89baf918(xfs_db: write values into dir/attr blocks and
> +# recalculate CRCs), xfs_db write command can't write invalid data into dir/attr
> +# field of v5 filesystems. For some cases need to write dir/attr (especially crc
> +# enabled), use this _require at first.
> +_require_xfs_db_write_da()
> +{
> + local inum
> + local count
> +
> + _require_scratch
Hiding this inside a another requires rule is kinda nasty. If the
test already used _require_scratch_nocheck, this will now cause the
test to check the scratch device at the end of the test.
And vice-versa - if you use _require_scratch_nocheck here, it will
break tests that have already called _require_scratch....
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2018-08-06 2:39 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-07-31 5:41 [PATCH v2 1/2] xfs/288: use -d option of xfs_db write command for v5 XFS Zorro Lang
2018-07-31 5:41 ` [PATCH v2 2/2] xfs/288: _notrun if xfs_db can't write values into dir or attr Zorro Lang
2018-08-02 0:46 ` Darrick J. Wong
2018-08-02 1:49 ` Zorro Lang
2018-08-02 5:14 ` Darrick J. Wong
2018-08-02 7:10 ` Zorro Lang
2018-08-06 0:32 ` Dave Chinner
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.