Open for contribution towards xfstests for btrfs

Open for contribution towards xfstests for btrfs

[Kiran Patil]

Hello,

We have a team of 5 students who would like to contribute to btrfs
filesystem testing using xfstests.

Is there space for them to contribute?

If yes, to whom do they need to keep in touch for guidance.

Thank you,
Kiran Patil (Director)
Green Turtles Technologies, INDIA.

Re: Open for contribution towards xfstests for btrfs

[Liu Bo]

On Mon, Jan 07, 2013 at 03:48:43PM +0530, Kiran Patil wrote:
> Hello,
> 
> We have a team of 5 students who would like to contribute to btrfs
> filesystem testing using xfstests.
> 
> Is there space for them to contribute?

xfstests is lacking of testcases targetting for btrfs send/recieve,
you may be interested in it.

thanks,
liubo

> 
> If yes, to whom do they need to keep in touch for guidance.
> 
> Thank you,
> Kiran Patil (Director)
> Green Turtles Technologies, INDIA.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Open for contribution towards xfstests for btrfs

[Arne Jansen]

On 07.01.2013 14:01, Liu Bo wrote:
> On Mon, Jan 07, 2013 at 03:48:43PM +0530, Kiran Patil wrote:
>> Hello,
>>
>> We have a team of 5 students who would like to contribute to btrfs
>> filesystem testing using xfstests.
>>
>> Is there space for them to contribute?
> 
> xfstests is lacking of testcases targetting for btrfs send/recieve,
> you may be interested in it.

There is even a basis for that already, we have a testsuite for the
zfs send -> btrfs receive feature at

git://git.kernel.org/pub/scm/linux/kernel/git/arne/far-progs.git

see test.pl. It can also already be used to test btrfs send ->
btrfs receive.

-Arne

> 
> thanks,
> liubo
> 
>>
>> If yes, to whom do they need to keep in touch for guidance.
>>
>> Thank you,
>> Kiran Patil (Director)
>> Green Turtles Technologies, INDIA.
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" 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-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Open for contribution towards xfstests for btrfs

[David Sterba]

On Mon, Jan 07, 2013 at 03:48:43PM +0530, Kiran Patil wrote:
> We have a team of 5 students who would like to contribute to btrfs
> filesystem testing using xfstests.

By any chance, are these the same people behind
http://sourceforge.net/projects/btrfstest/ ?

> Is there space for them to contribute?

Sure. As an example of where to start, look at a some patches from Miao
Xie with reproducers for several bugs. This IMHO is a smooth path to
start with xfstests, the task is to embody the test itself into the
teitesinfrastructure.

The more advanced tasks:
* look at commits in btrfs and try to write a specific test for them
  (depends on the nature of the bug, the difficulty ranges from easy to
  impossible)
* write test coverage for a particular feature (eg. for Jeff Liu's label
  setting patches)
* several btrfs specific xfstests were sent to the mailinglists over
  time, I'm sure not all of them were merged

> If yes, to whom do they need to keep in touch for guidance.

You can learn a lot from existing xfstests and review comments of
proposed tests.  Feel free to ask in the respective mailinglists or IRC
channels.


david

Re: Open for contribution towards xfstests for btrfs

[Eric Sandeen]

On 1/7/13 4:18 AM, Kiran Patil wrote:
> Hello,
> 
> We have a team of 5 students who would like to contribute to btrfs
> filesystem testing using xfstests.
> 
> Is there space for them to contribute?
> 
> If yes, to whom do they need to keep in touch for guidance.

(cc: xfs list)

Are you thinking of writing new tests, or doing more formal testing
with the existing testsuite?  Both would be useful.

In particular, any time there's a bugfix, a regression test is helpful.
Generic new tests of btrfs functionality would be good, too.

I would suggest starting with a couple new tests, and cc: the xfs list
for review.  In the course of the review you'll get a sense of the various
conventions used in xfstests, etc.

Simply running xfstests over various btrfs configurations would also
be useful, to be sure any regressions caught by existing tests don't go
unnoticed.

thanks,
-Eric

> Thank you,
> Kiran Patil (Director)
> Green Turtles Technologies, INDIA.

Re: Open for contribution towards xfstests for btrfs

[Kiran Patil]

On Mon, Jan 7, 2013 at 9:28 PM, Eric Sandeen <sandeen@redhat.com> wrote:
> On 1/7/13 4:18 AM, Kiran Patil wrote:
>> Hello,
>>
>> We have a team of 5 students who would like to contribute to btrfs
>> filesystem testing using xfstests.
>>
>> Is there space for them to contribute?
>>
>> If yes, to whom do they need to keep in touch for guidance.
>
> (cc: xfs list)
>
> Are you thinking of writing new tests, or doing more formal testing
> with the existing testsuite?  Both would be useful.
>
> In particular, any time there's a bugfix, a regression test is helpful.
> Generic new tests of btrfs functionality would be good, too.
>
> I would suggest starting with a couple new tests, and cc: the xfs list
> for review.  In the course of the review you'll get a sense of the various
> conventions used in xfstests, etc.
>
> Simply running xfstests over various btrfs configurations would also
> be useful, to be sure any regressions caught by existing tests don't go
> unnoticed.
>
> thanks,
> -Eric
>
Dear Eric and David,

Thanks a lot for your inputs.

We will work on them and soon we will start our work.

Thank you,

-Kiran

Re: Open for contribution towards xfstests for btrfs

[Kiran Patil]

On Mon, Jan 7, 2013 at 10:31 PM, Kiran Patil <kirantpatil@gmail.com> wrote:
> On Mon, Jan 7, 2013 at 9:28 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>> On 1/7/13 4:18 AM, Kiran Patil wrote:
>>> Hello,
>>>
>>> We have a team of 5 students who would like to contribute to btrfs
>>> filesystem testing using xfstests.
>>>
>>> Is there space for them to contribute?
>>>
>>> If yes, to whom do they need to keep in touch for guidance.
>>
>> (cc: xfs list)
>>
>> Are you thinking of writing new tests, or doing more formal testing
>> with the existing testsuite?  Both would be useful.
>>
>> In particular, any time there's a bugfix, a regression test is helpful.
>> Generic new tests of btrfs functionality would be good, too.
>>
>> I would suggest starting with a couple new tests, and cc: the xfs list
>> for review.  In the course of the review you'll get a sense of the various
>> conventions used in xfstests, etc.
>>
>> Simply running xfstests over various btrfs configurations would also
>> be useful, to be sure any regressions caught by existing tests don't go
>> unnoticed.
>>
>> thanks,
>> -Eric
>>
> Dear Eric and David,
>
> Thanks a lot for your inputs.
>
> We will work on them and soon we will start our work.
>
> Thank you,
>
> -Kiran

Sorry, I forgot to say thanks to Liu and Arne.

Thank you Liu and Arne for your suggestions.

-Kiran

Re: Open for contribution towards xfstests for btrfs

[Eric Sandeen]

On 1/7/13 9:31 AM, David Sterba wrote:
> On Mon, Jan 07, 2013 at 03:48:43PM +0530, Kiran Patil wrote:
>> We have a team of 5 students who would like to contribute to btrfs
>> filesystem testing using xfstests.
> 
> By any chance, are these the same people behind
> http://sourceforge.net/projects/btrfstest/ ?

If so, you might want to consider that

http://www.tuxera.com/community/posix-test-suite/

probably already covers some of the tests that "btrfstest" is
(re)writing.  i.e.:

+/* Creating Link on non-existing prefix path should return ENOTDIR

and

+/* Creating link on ENAMETOOLONG if either file name exceeds maximum characters

are almost certainly covered in the above posix test suite.

FWIW, that testsuite has a few failures on btrfs (not all unique to btrfs, IIRC):

Test Summary Report
-------------------
/root/pjd-fstest/pjd-fstest-20090130-RC/tests/chown/00.t   (Wstat: 0 Tests: 171 Failed: 6)
  Failed tests:  84, 88, 141, 145, 149, 153
/root/pjd-fstest/pjd-fstest-20090130-RC/tests/truncate/00.t (Wstat: 0 Tests: 21 Failed: 1)
  Failed test:  15
/root/pjd-fstest/pjd-fstest-20090130-RC/tests/xacl/00.t    (Wstat: 0 Tests: 45 Failed: 1)
  Failed test:  45

-Eric

Re: Open for contribution towards xfstests for btrfs

[Kiran Patil]

Hi,

>>
>> By any chance, are these the same people behind
>> http://sourceforge.net/projects/btrfstest/ ?

No, they are not the same people.

>
> If so, you might want to consider that
>
> http://www.tuxera.com/community/posix-test-suite/
>
> probably already covers some of the tests that "btrfstest" is
> (re)writing.  i.e.:
>
> +/* Creating Link on non-existing prefix path should return ENOTDIR
>
> and
>
> +/* Creating link on ENAMETOOLONG if either file name exceeds maximum characters
>
> are almost certainly covered in the above posix test suite.
>
> FWIW, that testsuite has a few failures on btrfs (not all unique to btrfs, IIRC):

Sure, we will look at them.

> Test Summary Report
> -------------------
> /root/pjd-fstest/pjd-fstest-20090130-RC/tests/chown/00.t   (Wstat: 0 Tests: 171 Failed: 6)
>   Failed tests:  84, 88, 141, 145, 149, 153
> /root/pjd-fstest/pjd-fstest-20090130-RC/tests/truncate/00.t (Wstat: 0 Tests: 21 Failed: 1)
>   Failed test:  15
> /root/pjd-fstest/pjd-fstest-20090130-RC/tests/xacl/00.t    (Wstat: 0 Tests: 45 Failed: 1)
>   Failed test:  45
>
> -Eric
>
Thanks,
Kiran.