Bug#1070919: compyle: FTBFS in bullseye

Antonio Valentino antonio.valentino at tiscali.it
Tue May 14 22:04:40 BST 2024


Dear Santiago,


Il 14/05/24 22:55, Santiago Vila ha scritto:
> El 14/5/24 a las 7:15, Antonio Valentino escribió:
>> On Sat, 11 May 2024 21:46:59 +0200 Santiago Vila <sanvila at debian.org> 
>> wrote:
>>> E       error: unknown target CPU 'generic'
>>
>> I'm sorry but I have no clue about this issue.
>> Looking at the log It seems that the CPU is not recognized.
> 
> Yes, that's what it seems.
> 
>> Not sure, however if the issue is in compyle or in pyopencl.
>>
>> Do you have an idea if the updated versions of the package (e.g. the 
>> ones currently in stable or in unstable) build properly on the same 
>> platform?
> 
> I tried building compyle_0.8.1-4 (bookworm version)
> in a bullseye chroot, and it failed with a similar error.
> 
> On the other hand, both packages (compyle and pyopencl) build fine
> in bookworm (that's why the bugs are marked as closed in the bookworm 
> version)
> in the same platform.
> 
> So yes, this seems to be a problem in pyopencl.
> 
> Because this is oldstable and packages are not expected to change,
> I think that whoever wants to build this from source will be able
> to use nocheck. Therefore, I think it would be ok to forget
> about this one.
> 
> If you are curious, I'm filing these bugs because the last point
> release of bullseye will be the last one, so this is the last
> opportunity to have a distribution which builds from source.
> 
> Release Policy does not apply anymore, so it is up to the
> individual maintainers to decide if they want to fix the bugs
> or not.
> 
> Thanks.

Thanks for dedicating time on this.
Unfortunately I do not have better ideas so I think that we could leave 
things as they are now for the time being.

kind regards
-- 
Antonio Valentino



More information about the debian-science-maintainers mailing list