Greetings,
I tried to solve one of the very few remaining issues related to
Gcc 10 migration in mapsembler2 2.2.4+dfsg1-1 (#957519). The
updated version is available on Salsa for review:
https://salsa.debian.org/med-team/mapsembler2
The autopkgtests suite revealed issues with 32bits architectures
support, so after detecting a missing memory check, and trying
tuning program parameters, I ended up having to request a
removal (#972436). The autopkgtest result in Salsa CI shown the
same symptom as on 32 bits curiously: a failure to allocate 4G.
The reprotest shows timestamps in the build results but I
haven't investigated further.
The changelog overall is:
* Team upload.
* added no_std_hash_redefinition.patch and gcc10.patch to
address compilation failures with Gcc 10. (Closes: #957519)
* package removed from 32 bits architecture repostories due to
test failure and end of maintenance upstream.
* added check_memory.patch to clarify memory allocation errors
in autopkgtests, but also in general use.
* excluded cmake build directory in gatb-core examples, as
they are unusable, and the CMakeCache.txt triggers a lintian
error.
* made the build more verbose to check for build flags.
Thanks Pranav for having provided the test suite, issues could
have remained under the radar otherwise.
Kind Regards,
--
Étienne Mollier <etienne.mollier@mailoo.org>
Old rsa/3072: 5ab1 4edf 63bb ccff 8b54 2fa9 59da 56fe fff3 882d
New rsa/4096: 8f91 b227 c7d6 f2b1 948c 8236 793c f67e 8f0d 11da
Sent from /dev/pts/4, please excuse my verbosity.
Attachment:
signature.asc
Description: PGP signature