Bug#997884: libgclib breaks cdbfasta autopkgtest: ABI breakage

2021-10-27 Thread Andreas Tille
Hi Étienne,

thanks a lot for cleaning up the hassle I've created by naive upload.
I'm busy with real life this week - so feel free to upload.  Hope it
does not reside in new as long as other packages there.

Kind regards
Andreas.

Am Tue, Oct 26, 2021 at 10:46:26PM +0200 schrieb Étienne Mollier:
> Control: tags -1 confirmed
> 
> So, there is indeed an ABI breakage in libgclib.  I pushed some
> changes on Salsa to bump the SONAME to version 3 [1].  Since
> this would require a round trip through NEW, I also made some
> welcome updates to the copyright file.  I haven't uploaded yet,
> in case someone wishes to have a second look to make sure
> nothing is missing, notably in the copyright file, and I will
> try to recall to dput this tomorrow evening (UTC+2) if alright
> and not sponsored already.
> 
> [1]: https://salsa.debian.org/med-team/libgclib/
> 
> In hope this helps,
> Cheers,  :)
> -- 
> Étienne Mollier 
> Fingerprint:  8f91 b227 c7d6 f2b1 948c  8236 793c f67e 8f0d 11da
> Sent from /dev/pts/2, please excuse my verbosity.



> ___
> Debian-med-packaging mailing list
> debian-med-packag...@alioth-lists.debian.net
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-med-packaging


-- 
http://fam-tille.de



Bug#997884: libgclib breaks cdbfasta autopkgtest: ABI breakage

2021-10-26 Thread Étienne Mollier
Control: tags -1 confirmed

So, there is indeed an ABI breakage in libgclib.  I pushed some
changes on Salsa to bump the SONAME to version 3 [1].  Since
this would require a round trip through NEW, I also made some
welcome updates to the copyright file.  I haven't uploaded yet,
in case someone wishes to have a second look to make sure
nothing is missing, notably in the copyright file, and I will
try to recall to dput this tomorrow evening (UTC+2) if alright
and not sponsored already.

[1]: https://salsa.debian.org/med-team/libgclib/

In hope this helps,
Cheers,  :)
-- 
Étienne Mollier 
Fingerprint:  8f91 b227 c7d6 f2b1 948c  8236 793c f67e 8f0d 11da
Sent from /dev/pts/2, please excuse my verbosity.


signature.asc
Description: PGP signature


Bug#997884: libgclib breaks cdbfasta autopkgtest: ABI breakage?

2021-10-26 Thread Étienne Mollier
Control: affects -1 src:gffread 0.12.1-4

Hi Paul,

On Tue, 26 Oct 2021 16:58:28 +0200 Paul Gevers  wrote:
> [1] https://qa.debian.org/excuses.php?package=libgclib
> 
> https://ci.debian.net/data/autopkgtest/testing/amd64/c/cdbfasta/16212088/log.gz
> 
> + cdbfasta ebola.fasta
> cdbfasta: symbol lookup error: cdbfasta: undefined symbol: _Z8GReallocPPvm
> autopkgtest [22:12:39]: test run-unit-tests

Thank you for the heads up!  This looks to also affect gffread
0.12.1-4 [2].  Looking that up…

[2]: 
https://ci.debian.net/data/autopkgtest/testing/amd64/g/gffread/16212089/log.gz

Have a good evening,  :)
-- 
Étienne Mollier 
Fingerprint:  8f91 b227 c7d6 f2b1 948c  8236 793c f67e 8f0d 11da
Sent from /dev/pts/6, please excuse my verbosity.


signature.asc
Description: PGP signature


Bug#997884: libgclib breaks cdbfasta autopkgtest: ABI breakage?

2021-10-26 Thread Paul Gevers
Source: libgclib
Control: found -1 libgclib/0.12.7+ds-2
Severity: serious
Tags: sid bookworm
X-Debbugs-CC: debian...@lists.debian.org
User: debian...@lists.debian.org
Usertags: breaks
Control: affects -1 src:cdbfasta

Dear maintainer(s),

With a recent upload of libgclib the autopkgtest of cdbfasta fails in
testing when that autopkgtest is run with the binary packages of
libgclib from unstable. It passes when run with only packages from
testing. In tabular form:

   passfail
libgclib   from testing0.12.7+ds-2
cdbfasta   from testing1.00+git20181005.014498c+dfsg-2
all others from testingfrom testing

I copied some of the output at the bottom of this report. Looking at the
message and the changelog, it seems that symbols were changed, but this
one was apparently used by cdbfasta. I think this points at an ABI
breakage and the SONAME of libgclib must be bumped and a transition must
be followed through. (I could be wrong).

Currently this regression is blocking the migration of libgclib to
testing [1]. Can you please investigate the situation?

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=libgclib

https://ci.debian.net/data/autopkgtest/testing/amd64/c/cdbfasta/16212088/log.gz

+ cdbfasta ebola.fasta
cdbfasta: symbol lookup error: cdbfasta: undefined symbol: _Z8GReallocPPvm
autopkgtest [22:12:39]: test run-unit-tests




OpenPGP_signature
Description: OpenPGP digital signature