Bug#909530: Please help debugging (Was: Bug#909530: t-coffee: autopkgtest regression)

2018-10-21 Thread Andreas Tille
Hi Liubov,

thanks a lot for your patch.  I've tested it and uploaded the package.

Happy that you persist in the team

  Andreas. 

-- 
http://fam-tille.de



Bug#909530: Please help debugging (Was: Bug#909530: t-coffee: autopkgtest regression)

2018-09-30 Thread Andreas Tille
Control: tags -1 help

Hi,

On Fri, Sep 28, 2018 at 02:38:19PM +0200, Andreas Tille wrote:
> > Strange.  I'm able to reproduce this here even on my local machine in a
> > shell where I did some `unset HOME`.  I wonder why this error was not
> > produced with version 11.00.8cbe486-6 before.  Anyway, now since the
> > issue can be easily reproduced I'll find a fix soonish.
> 
> While that is now really fixed[1] the other issues seem to be caused
> by something else.  This is really only happening in the package in
> unstable.  Back to gdb ... :-(

I've tried to track down the issue using gdb + fprintf[2] to fix the
coredump issue.  Unfortunately I did not finalised and will not have any
time for the next seven days.

Any help would be welcome

   Andreas.

 
> [1] 
> https://salsa.debian.org/med-team/t-coffee/commit/9cffa995e4603e396054062c6f3dbff1568bf243
[2] 
https://salsa.debian.org/med-team/t-coffee/blob/master/debian/patches/debug.patch

-- 
http://fam-tille.de



Bug#909530: t-coffee: autopkgtest regression

2018-09-28 Thread Andreas Tille
Hi again,

On Fri, Sep 28, 2018 at 12:12:58PM +0200, Andreas Tille wrote:
> > also for me) I ran t_coffee:
> > root@autopkgtest-lxc-vfquea:/tmp/autopkgtest-lxc.3n021a13/downtmp/build.S6P/src#
> > t_coffee
> > ERROR: Could not set a HOME directory.
> > Set any of the following environement variables to some suitable
> > location: HOME, HOME_4_TCOFFEE, TMP or TEMP [FATAL:T-COFFEE]
> > Segmentation fault
> 
> Strange.  I'm able to reproduce this here even on my local machine in a
> shell where I did some `unset HOME`.  I wonder why this error was not
> produced with version 11.00.8cbe486-6 before.  Anyway, now since the
> issue can be easily reproduced I'll find a fix soonish.

While that is now really fixed[1] the other issues seem to be caused
by something else.  This is really only happening in the package in
unstable.  Back to gdb ... :-(
  
> > Paul
> > PS: also there is a typo in the error message:
> > environement -> environment
> 
> Will be fixed as well.

Is also fixed in[1].

Kind regards

Andreas.


[1] 
https://salsa.debian.org/med-team/t-coffee/commit/9cffa995e4603e396054062c6f3dbff1568bf243

-- 
http://fam-tille.de



Bug#909530: t-coffee: autopkgtest regression

2018-09-28 Thread Andreas Tille
Hi Paul,

On Thu, Sep 27, 2018 at 02:05:19PM +0200, Paul Gevers wrote:
> 
> I ran autopkgtest locally with the --shell option. After it failed (yes,
> also for me) I ran t_coffee:
> root@autopkgtest-lxc-vfquea:/tmp/autopkgtest-lxc.3n021a13/downtmp/build.S6P/src#
> t_coffee
> ERROR: Could not set a HOME directory.
> Set any of the following environement variables to some suitable
> location: HOME, HOME_4_TCOFFEE, TMP or TEMP [FATAL:T-COFFEE]
> Segmentation fault

Strange.  I'm able to reproduce this here even on my local machine in a
shell where I did some `unset HOME`.  I wonder why this error was not
produced with version 11.00.8cbe486-6 before.  Anyway, now since the
issue can be easily reproduced I'll find a fix soonish.
 
> Paul
> PS: also there is a typo in the error message:
> environement -> environment

Will be fixed as well.

Thanks a lot for all your QA work

 Andreas.

-- 
http://fam-tille.de



Bug#909530: t-coffee: autopkgtest regression

2018-09-27 Thread Paul Gevers
Hi Andreas,

On 24-09-18 22:05, Andreas Tille wrote:
> I admit I have no clue why this fails.  I've build the package with
> autopkg hook in pbuilder successfully and the unit test also runs
> on my local machine.
> 
> I suspect that using a more modern compiler than that was used for the
> recent package compared to the previous with successful tests had
> uncovered some hidden issue but it only happens under certain
> circumstances.
> 
> Any idea why the test fails on ci infrastructure is welcome.

I ran autopkgtest locally with the --shell option. After it failed (yes,
also for me) I ran t_coffee:
root@autopkgtest-lxc-vfquea:/tmp/autopkgtest-lxc.3n021a13/downtmp/build.S6P/src#
t_coffee
ERROR: Could not set a HOME directory.
Set any of the following environement variables to some suitable
location: HOME, HOME_4_TCOFFEE, TMP or TEMP [FATAL:T-COFFEE]
Segmentation fault

Paul
PS: also there is a typo in the error message:
environement -> environment



signature.asc
Description: OpenPGP digital signature


Bug#909530: t-coffee: autopkgtest regression

2018-09-24 Thread Andreas Tille
Control: tags -1 unreproducible
Control: tags -1 help

On Mon, Sep 24, 2018 at 09:12:16PM +0200, Paul Gevers wrote:
> 
> https://ci.debian.net/data/autopkgtest/testing/amd64/t/t-coffee/1045746/log.gz
> 
> # Command Line: t_coffee sample_seq1.fasta  [PROGRAM:T-COFFEE]
> 
> # TERMINATION STATUS: FAILURE [PROGRAM: T-COFFEE pid 1242 ppid 1131
> #CL: t_coffee sample_seq1.fasta

I admit I have no clue why this fails.  I've build the package with
autopkg hook in pbuilder successfully and the unit test also runs
on my local machine.

I suspect that using a more modern compiler than that was used for the
recent package compared to the previous with successful tests had
uncovered some hidden issue but it only happens under certain
circumstances.

Any idea why the test fails on ci infrastructure is welcome.

Kind regards

  Andreas.

-- 
http://fam-tille.de



Bug#909530: t-coffee: autopkgtest regression

2018-09-24 Thread Paul Gevers
Source: t-coffee
Version: 11.00.8cbe486-7
X-Debbugs-CC: debian...@lists.debian.org
User: debian...@lists.debian.org
Usertags: regression

Dear maintainers,

With a recent upload of t-coffee the autopkgtest of t-coffee fails in
testing when that autopkgtest is run with the binary packages of
t-coffee from unstable. It passes when run with only packages from
testing. In tabular form:
   passfail
t-coffee   from testing11.00.8cbe486-7
all others from testingfrom testing

I copied some of the output at the bottom of this report.

Currently this regression is contributing to the delay of the migration
to testing [1]. Can you please investigate the situation and fix it? If
needed, please change the bug's severity.

More information about this bug and the reason for filing it can be found on
https://wiki.debian.org/ContinuousIntegration/RegressionEmailInformation

Paul

[1] https://qa.debian.org/excuses.php?package=t-coffee

https://ci.debian.net/data/autopkgtest/testing/amd64/t/t-coffee/1045746/log.gz

# Command Line: t_coffee sample_seq1.fasta  [PROGRAM:T-COFFEE]

# TERMINATION STATUS: FAILURE [PROGRAM: T-COFFEE pid 1242 ppid 1131
#CL: t_coffee sample_seq1.fasta



signature.asc
Description: OpenPGP digital signature