[PATCH 1/2] xfs/055: add missing 055.cfg

[PATCH 1/2] xfs/055: add missing 055.cfg

[Xiao Yang]

_link_output_file sets output file suffix to null because 055.cfg is
missing, so add it.

Signed-off-by: Xiao Yang <yangx.jy@cn.fujitsu.com>
---
 tests/xfs/055.cfg | 2 ++
 1 file changed, 2 insertions(+)
 create mode 100644 tests/xfs/055.cfg

diff --git a/tests/xfs/055.cfg b/tests/xfs/055.cfg
new file mode 100644
index 0000000..7ffdfc0
--- /dev/null
+++ b/tests/xfs/055.cfg
@@ -0,0 +1,2 @@
+irix: irix
+linux: linux
-- 
1.8.3.1

Re: [PATCH 1/2] xfs/055: add missing 055.cfg

[Eryu Guan]

On Wed, May 31, 2017 at 05:10:26PM +0800, Xiao Yang wrote:
> _link_output_file sets output file suffix to null because 055.cfg is
> missing, so add it.
> 
> Signed-off-by: Xiao Yang <yangx.jy@cn.fujitsu.com>

Thanks for the patch! But we eventually want to remove IRIX support, I
don't think this is needed at this stage, especially that Eric Biggers
has posted patches to remove IRIX support.

BTW, your patch has '1/2' prefix, but I didn't see the second patch, do
you have a second patch that didn't hit the list?

Thanks,
Eryu

> ---
>  tests/xfs/055.cfg | 2 ++
>  1 file changed, 2 insertions(+)
>  create mode 100644 tests/xfs/055.cfg
> 
> diff --git a/tests/xfs/055.cfg b/tests/xfs/055.cfg
> new file mode 100644
> index 0000000..7ffdfc0
> --- /dev/null
> +++ b/tests/xfs/055.cfg
> @@ -0,0 +1,2 @@
> +irix: irix
> +linux: linux
> -- 
> 1.8.3.1
> 
> 
> 
> --
> 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

Re: [PATCH 1/2] xfs/055: add missing 055.cfg

[Xiao Yang]

On 2017/06/01 10:53, Eryu Guan wrote:
> On Wed, May 31, 2017 at 05:10:26PM +0800, Xiao Yang wrote:
>> _link_output_file sets output file suffix to null because 055.cfg is
>> missing, so add it.
>>
>> Signed-off-by: Xiao Yang<yangx.jy@cn.fujitsu.com>
> Thanks for the patch! But we eventually want to remove IRIX support, I
> don't think this is needed at this stage, especially that Eric Biggers
> has posted patches to remove IRIX support.
>
> BTW, your patch has '1/2' prefix, but I didn't see the second patch, do
> you have a second patch that didn't hit the list?
Hi Eryu

Yeah, i will send the second patch again. :-)

Thanks,
Xiao Yang
> Thanks,
> Eryu
>
>> ---
>>   tests/xfs/055.cfg | 2 ++
>>   1 file changed, 2 insertions(+)
>>   create mode 100644 tests/xfs/055.cfg
>>
>> diff --git a/tests/xfs/055.cfg b/tests/xfs/055.cfg
>> new file mode 100644
>> index 0000000..7ffdfc0
>> --- /dev/null
>> +++ b/tests/xfs/055.cfg
>> @@ -0,0 +1,2 @@
>> +irix: irix
>> +linux: linux
>> -- 
>> 1.8.3.1
>>
>>
>>
>> --
>> 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
>
>
>

Re: [PATCH 2/2] xfs: fix installation for extended output files

[Eryu Guan]

On Thu, Jun 01, 2017 at 11:01:13AM +0800, Xiao Yang wrote:
> The installation of these extended output files are skipped
> because they aren't matched by xfstests.  So we rename them
> to adapt building rules.
> 
> Signed-off-by: Xiao Yang <yangx.jy@cn.fujitsu.com>

This patch is 2.4M in size, no wonder it didn't hit the list..

(cc linux-xfs list for these xfs specific tests)

> ---
>  tests/xfs/018                         |    8 +-
>  tests/xfs/018.op.irix                 | 5628 ---------------------------------
>  tests/xfs/018.op.linux                | 4714 ---------------------------
>  tests/xfs/018.out.irix                | 5628 +++++++++++++++++++++++++++++++++
>  tests/xfs/018.out.linux               | 4714 +++++++++++++++++++++++++++
>  tests/xfs/018.out.trans_buf           | 1821 +++++++++++
>  tests/xfs/018.out.trans_inode         | 2609 +++++++++++++++
>  tests/xfs/018.trans_buf               | 1821 -----------
>  tests/xfs/018.trans_inode             | 2609 ---------------

xfs/018 081 and 082 are all marked as 'deprecated', I'm wondering if we
can just remove them.

>  tests/xfs/040                         |    4 +-
>  tests/xfs/040.good                    |   16 -
>  tests/xfs/040.out.good                |   16 +

xfs/040 is doing source code diff between userspace vs kernel space, I'm
not sure if this is still needed. But xfs/040 requires KWORKAREA points
to a directory that contains both userspace and kernel code, I suspect
it's rarely run.

>  tests/xfs/081                         |    2 +-
>  tests/xfs/081.out.ugquota.trans_inode | 3091 ++++++++++++++++++
>  tests/xfs/081.ugquota.trans_inode     | 3091 ------------------
>  tests/xfs/082                         |    8 +-
>  tests/xfs/082.op.irix                 | 5628 ---------------------------------
>  tests/xfs/082.op.linux                | 4714 ---------------------------
>  tests/xfs/082.out.irix                | 5628 +++++++++++++++++++++++++++++++++
>  tests/xfs/082.out.linux               | 4714 +++++++++++++++++++++++++++
>  tests/xfs/082.out.trans_buf           | 1821 +++++++++++
>  tests/xfs/082.out.trans_inode         | 2609 +++++++++++++++
>  tests/xfs/082.trans_buf               | 1821 -----------
>  tests/xfs/082.trans_inode             | 2609 ---------------
>  24 files changed, 32662 insertions(+), 32662 deletions(-)

Thanks,
Eryu

Re: [PATCH 2/2] xfs: fix installation for extended output files

[Darrick J. Wong]

On Thu, Jun 01, 2017 at 12:07:33PM +0800, Eryu Guan wrote:
> On Thu, Jun 01, 2017 at 11:01:13AM +0800, Xiao Yang wrote:
> > The installation of these extended output files are skipped
> > because they aren't matched by xfstests.  So we rename them
> > to adapt building rules.
> > 
> > Signed-off-by: Xiao Yang <yangx.jy@cn.fujitsu.com>
> 
> This patch is 2.4M in size, no wonder it didn't hit the list..
> 
> (cc linux-xfs list for these xfs specific tests)
> 
> > ---
> >  tests/xfs/018                         |    8 +-
> >  tests/xfs/018.op.irix                 | 5628 ---------------------------------
> >  tests/xfs/018.op.linux                | 4714 ---------------------------
> >  tests/xfs/018.out.irix                | 5628 +++++++++++++++++++++++++++++++++
> >  tests/xfs/018.out.linux               | 4714 +++++++++++++++++++++++++++
> >  tests/xfs/018.out.trans_buf           | 1821 +++++++++++
> >  tests/xfs/018.out.trans_inode         | 2609 +++++++++++++++
> >  tests/xfs/018.trans_buf               | 1821 -----------
> >  tests/xfs/018.trans_inode             | 2609 ---------------
> 
> xfs/018 081 and 082 are all marked as 'deprecated', I'm wondering if we
> can just remove them.

I have little good to say about these -- diffing logprint outputs will
break every time we add new log item types (which we have) so these are
brittle.

> >  tests/xfs/040                         |    4 +-
> >  tests/xfs/040.good                    |   16 -
> >  tests/xfs/040.out.good                |   16 +
> 
> xfs/040 is doing source code diff between userspace vs kernel space, I'm
> not sure if this is still needed. But xfs/040 requires KWORKAREA points
> to a directory that contains both userspace and kernel code, I suspect
> it's rarely run.

It's /never/ run, because it requires dmapi.h, which doesn't exist on
Linux, and it appears to assume the pre-libxfs-split directory layout.
Seeing as xfsprogs ships its own "compare the libxfs code" tool now,
I think it better to go fix it.  Patch for x/040 coming soon.

> >  tests/xfs/081                         |    2 +-
> >  tests/xfs/081.out.ugquota.trans_inode | 3091 ++++++++++++++++++
> >  tests/xfs/081.ugquota.trans_inode     | 3091 ------------------
> >  tests/xfs/082                         |    8 +-
> >  tests/xfs/082.op.irix                 | 5628 ---------------------------------
> >  tests/xfs/082.op.linux                | 4714 ---------------------------
> >  tests/xfs/082.out.irix                | 5628 +++++++++++++++++++++++++++++++++
> >  tests/xfs/082.out.linux               | 4714 +++++++++++++++++++++++++++
> >  tests/xfs/082.out.trans_buf           | 1821 +++++++++++
> >  tests/xfs/082.out.trans_inode         | 2609 +++++++++++++++
> >  tests/xfs/082.trans_buf               | 1821 -----------
> >  tests/xfs/082.trans_inode             | 2609 ---------------
> >  24 files changed, 32662 insertions(+), 32662 deletions(-)

Ye $deities!

--D

> 
> Thanks,
> Eryu
> --
> 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

Re: [PATCH 2/2] xfs: fix installation for extended output files

[Xiao Yang]

On 2017/06/01 23:48, Darrick J. Wong wrote:
> On Thu, Jun 01, 2017 at 12:07:33PM +0800, Eryu Guan wrote:
>> On Thu, Jun 01, 2017 at 11:01:13AM +0800, Xiao Yang wrote:
>>> The installation of these extended output files are skipped
>>> because they aren't matched by xfstests.  So we rename them
>>> to adapt building rules.
>>>
>>> Signed-off-by: Xiao Yang<yangx.jy@cn.fujitsu.com>
>> This patch is 2.4M in size, no wonder it didn't hit the list..
>>
>> (cc linux-xfs list for these xfs specific tests)
>>
>>> ---
>>>   tests/xfs/018                         |    8 +-
>>>   tests/xfs/018.op.irix                 | 5628 ---------------------------------
>>>   tests/xfs/018.op.linux                | 4714 ---------------------------
>>>   tests/xfs/018.out.irix                | 5628 +++++++++++++++++++++++++++++++++
>>>   tests/xfs/018.out.linux               | 4714 +++++++++++++++++++++++++++
>>>   tests/xfs/018.out.trans_buf           | 1821 +++++++++++
>>>   tests/xfs/018.out.trans_inode         | 2609 +++++++++++++++
>>>   tests/xfs/018.trans_buf               | 1821 -----------
>>>   tests/xfs/018.trans_inode             | 2609 ---------------
>> xfs/018 081 and 082 are all marked as 'deprecated', I'm wondering if we
>> can just remove them.
> I have little good to say about these -- diffing logprint outputs will
> break every time we add new log item types (which we have) so these are
> brittle.
>
>>>   tests/xfs/040                         |    4 +-
>>>   tests/xfs/040.good                    |   16 -
>>>   tests/xfs/040.out.good                |   16 +
>> xfs/040 is doing source code diff between userspace vs kernel space, I'm
>> not sure if this is still needed. But xfs/040 requires KWORKAREA points
>> to a directory that contains both userspace and kernel code, I suspect
>> it's rarely run.
> It's /never/ run, because it requires dmapi.h, which doesn't exist on
> Linux, and it appears to assume the pre-libxfs-split directory layout.
> Seeing as xfsprogs ships its own "compare the libxfs code" tool now,
> I think it better to go fix it.  Patch for x/040 coming soon.
Hi Darrick

The libxfs code is being updated regularly, so I think the difference is 
not immutable.
Is it necessary to run this case?

Thanks,
Xiao Yang
>>>   tests/xfs/081                         |    2 +-
>>>   tests/xfs/081.out.ugquota.trans_inode | 3091 ++++++++++++++++++
>>>   tests/xfs/081.ugquota.trans_inode     | 3091 ------------------
>>>   tests/xfs/082                         |    8 +-
>>>   tests/xfs/082.op.irix                 | 5628 ---------------------------------
>>>   tests/xfs/082.op.linux                | 4714 ---------------------------
>>>   tests/xfs/082.out.irix                | 5628 +++++++++++++++++++++++++++++++++
>>>   tests/xfs/082.out.linux               | 4714 +++++++++++++++++++++++++++
>>>   tests/xfs/082.out.trans_buf           | 1821 +++++++++++
>>>   tests/xfs/082.out.trans_inode         | 2609 +++++++++++++++
>>>   tests/xfs/082.trans_buf               | 1821 -----------
>>>   tests/xfs/082.trans_inode             | 2609 ---------------
>>>   24 files changed, 32662 insertions(+), 32662 deletions(-)
> Ye $deities!
>
> --D
>
>> Thanks,
>> Eryu
>> --
>> 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
>
> .
>

Re: [PATCH 2/2] xfs: fix installation for extended output files

[Darrick J. Wong]

On Fri, Jun 02, 2017 at 10:40:39AM +0800, Xiao Yang wrote:
> On 2017/06/01 23:48, Darrick J. Wong wrote:
> >On Thu, Jun 01, 2017 at 12:07:33PM +0800, Eryu Guan wrote:
> >>On Thu, Jun 01, 2017 at 11:01:13AM +0800, Xiao Yang wrote:
> >>>The installation of these extended output files are skipped
> >>>because they aren't matched by xfstests.  So we rename them
> >>>to adapt building rules.
> >>>
> >>>Signed-off-by: Xiao Yang<yangx.jy@cn.fujitsu.com>
> >>This patch is 2.4M in size, no wonder it didn't hit the list..
> >>
> >>(cc linux-xfs list for these xfs specific tests)
> >>
> >>>---
> >>>  tests/xfs/018                         |    8 +-
> >>>  tests/xfs/018.op.irix                 | 5628 ---------------------------------
> >>>  tests/xfs/018.op.linux                | 4714 ---------------------------
> >>>  tests/xfs/018.out.irix                | 5628 +++++++++++++++++++++++++++++++++
> >>>  tests/xfs/018.out.linux               | 4714 +++++++++++++++++++++++++++
> >>>  tests/xfs/018.out.trans_buf           | 1821 +++++++++++
> >>>  tests/xfs/018.out.trans_inode         | 2609 +++++++++++++++
> >>>  tests/xfs/018.trans_buf               | 1821 -----------
> >>>  tests/xfs/018.trans_inode             | 2609 ---------------
> >>xfs/018 081 and 082 are all marked as 'deprecated', I'm wondering if we
> >>can just remove them.
> >I have little good to say about these -- diffing logprint outputs will
> >break every time we add new log item types (which we have) so these are
> >brittle.
> >
> >>>  tests/xfs/040                         |    4 +-
> >>>  tests/xfs/040.good                    |   16 -
> >>>  tests/xfs/040.out.good                |   16 +
> >>xfs/040 is doing source code diff between userspace vs kernel space, I'm
> >>not sure if this is still needed. But xfs/040 requires KWORKAREA points
> >>to a directory that contains both userspace and kernel code, I suspect
> >>it's rarely run.
> >It's /never/ run, because it requires dmapi.h, which doesn't exist on
> >Linux, and it appears to assume the pre-libxfs-split directory layout.
> >Seeing as xfsprogs ships its own "compare the libxfs code" tool now,
> >I think it better to go fix it.  Patch for x/040 coming soon.
> Hi Darrick
> 
> The libxfs code is being updated regularly, so I think the difference is not
> immutable.
> Is it necessary to run this case?

For the upstream maintainers, yes, because we want to keep the libxfs/
code synchronized between, say, Linux 4.12 and xfsprogs 4.12.  Right now
we more or less just run it manually, but I think there's value in
enabling everyone else to keep us honest by providing everyone a way to
look for missing pieces in an automated fashion. :)

--D

> 
> Thanks,
> Xiao Yang
> >>>  tests/xfs/081                         |    2 +-
> >>>  tests/xfs/081.out.ugquota.trans_inode | 3091 ++++++++++++++++++
> >>>  tests/xfs/081.ugquota.trans_inode     | 3091 ------------------
> >>>  tests/xfs/082                         |    8 +-
> >>>  tests/xfs/082.op.irix                 | 5628 ---------------------------------
> >>>  tests/xfs/082.op.linux                | 4714 ---------------------------
> >>>  tests/xfs/082.out.irix                | 5628 +++++++++++++++++++++++++++++++++
> >>>  tests/xfs/082.out.linux               | 4714 +++++++++++++++++++++++++++
> >>>  tests/xfs/082.out.trans_buf           | 1821 +++++++++++
> >>>  tests/xfs/082.out.trans_inode         | 2609 +++++++++++++++
> >>>  tests/xfs/082.trans_buf               | 1821 -----------
> >>>  tests/xfs/082.trans_inode             | 2609 ---------------
> >>>  24 files changed, 32662 insertions(+), 32662 deletions(-)
> >Ye $deities!
> >
> >--D
> >
> >>Thanks,
> >>Eryu
> >>--
> >>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 linux-xfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 2/2] xfs: fix installation for extended output files

[Eryu Guan]

On Thu, Jun 01, 2017 at 10:45:39PM -0700, Darrick J. Wong wrote:
> On Fri, Jun 02, 2017 at 10:40:39AM +0800, Xiao Yang wrote:
> > On 2017/06/01 23:48, Darrick J. Wong wrote:
> > >On Thu, Jun 01, 2017 at 12:07:33PM +0800, Eryu Guan wrote:
> > >>On Thu, Jun 01, 2017 at 11:01:13AM +0800, Xiao Yang wrote:
> > >>>The installation of these extended output files are skipped
> > >>>because they aren't matched by xfstests.  So we rename them
> > >>>to adapt building rules.
> > >>>
> > >>>Signed-off-by: Xiao Yang<yangx.jy@cn.fujitsu.com>
> > >>This patch is 2.4M in size, no wonder it didn't hit the list..
> > >>
> > >>(cc linux-xfs list for these xfs specific tests)
> > >>
> > >>>---
> > >>>  tests/xfs/018                         |    8 +-
> > >>>  tests/xfs/018.op.irix                 | 5628 ---------------------------------
> > >>>  tests/xfs/018.op.linux                | 4714 ---------------------------
> > >>>  tests/xfs/018.out.irix                | 5628 +++++++++++++++++++++++++++++++++
> > >>>  tests/xfs/018.out.linux               | 4714 +++++++++++++++++++++++++++
> > >>>  tests/xfs/018.out.trans_buf           | 1821 +++++++++++
> > >>>  tests/xfs/018.out.trans_inode         | 2609 +++++++++++++++
> > >>>  tests/xfs/018.trans_buf               | 1821 -----------
> > >>>  tests/xfs/018.trans_inode             | 2609 ---------------
> > >>xfs/018 081 and 082 are all marked as 'deprecated', I'm wondering if we
> > >>can just remove them.
> > >I have little good to say about these -- diffing logprint outputs will
> > >break every time we add new log item types (which we have) so these are
> > >brittle.
> > >
> > >>>  tests/xfs/040                         |    4 +-
> > >>>  tests/xfs/040.good                    |   16 -
> > >>>  tests/xfs/040.out.good                |   16 +
> > >>xfs/040 is doing source code diff between userspace vs kernel space, I'm
> > >>not sure if this is still needed. But xfs/040 requires KWORKAREA points
> > >>to a directory that contains both userspace and kernel code, I suspect
> > >>it's rarely run.
> > >It's /never/ run, because it requires dmapi.h, which doesn't exist on
> > >Linux, and it appears to assume the pre-libxfs-split directory layout.
> > >Seeing as xfsprogs ships its own "compare the libxfs code" tool now,
> > >I think it better to go fix it.  Patch for x/040 coming soon.
> > Hi Darrick
> > 
> > The libxfs code is being updated regularly, so I think the difference is not
> > immutable.
> > Is it necessary to run this case?
> 
> For the upstream maintainers, yes, because we want to keep the libxfs/
> code synchronized between, say, Linux 4.12 and xfsprogs 4.12.  Right now
> we more or less just run it manually, but I think there's value in
> enabling everyone else to keep us honest by providing everyone a way to
> look for missing pieces in an automated fashion. :)

Agreed. The only concerns I have are a) it only passes with latest
upstream kernel & xfsprogs code, and is a false alert for distribution
testers; and b) 040.out files needs update regularly, which seems like a
maintaince burden.

For the first problem, I think we could at least remove it from auto
group. But I don't have a better thought on the second problem right
now..

Thanks,
Eryu

> 
> --D
> 
> > 
> > Thanks,
> > Xiao Yang
> > >>>  tests/xfs/081                         |    2 +-
> > >>>  tests/xfs/081.out.ugquota.trans_inode | 3091 ++++++++++++++++++
> > >>>  tests/xfs/081.ugquota.trans_inode     | 3091 ------------------
> > >>>  tests/xfs/082                         |    8 +-
> > >>>  tests/xfs/082.op.irix                 | 5628 ---------------------------------
> > >>>  tests/xfs/082.op.linux                | 4714 ---------------------------
> > >>>  tests/xfs/082.out.irix                | 5628 +++++++++++++++++++++++++++++++++
> > >>>  tests/xfs/082.out.linux               | 4714 +++++++++++++++++++++++++++
> > >>>  tests/xfs/082.out.trans_buf           | 1821 +++++++++++
> > >>>  tests/xfs/082.out.trans_inode         | 2609 +++++++++++++++
> > >>>  tests/xfs/082.trans_buf               | 1821 -----------
> > >>>  tests/xfs/082.trans_inode             | 2609 ---------------
> > >>>  24 files changed, 32662 insertions(+), 32662 deletions(-)
> > >Ye $deities!
> > >
> > >--D
> > >
> > >>Thanks,
> > >>Eryu
> > >>--
> > >>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 linux-xfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html