Bug#970875: [Debian-med-packaging] Bug#970875: mash: inconsistent results on 32 vs 64 bits architectures
Hi Sascha, Sascha Steinbiss, on 2021-02-20 12:57:25 +0100: > It looks like, in general, results are known to be different between > 64bit and 32bit archs. I've looked at reported issues upstream and > found: https://github.com/marbl/Mash/issues/109 > > So I don't think this is necessarily a bug in Debian, we even address it > in https://salsa.debian.org/med-team/mash/-/blob/master/debian/ > > Should we tag this 'upstream'? Ah, good finding! Yes, I believe it would make sense to tag it upstream. My main concern was that the inconsistency in results was seemingly visible on at least one other package. The difference was enough to zero all out in the mash related kleborate test, but I probably should double check it's test suite to make sure the bug is attributed adequately. Kind Regards, -- 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#970875: [Debian-med-packaging] Bug#970875: mash: inconsistent results on 32 vs 64 bits architectures
Hi, > While investigating FTBFS of `kleborate' on i386 and armhf, I > found out that `mash' seemed to output inconsistent results > depending on the underlying CPU architecture it is running on. [...] > Since the output obtained on amd64 is considered appropriate by > `kleborate' test suite, I heavily suspect that the result on > i386 should correspond to the results on amd64. The issue may > be repeatable on armhf, but haven't checked yet. It looks like, in general, results are known to be different between 64bit and 32bit archs. I've looked at reported issues upstream and found: https://github.com/marbl/Mash/issues/109 So I don't think this is necessarily a bug in Debian, we even address it in https://salsa.debian.org/med-team/mash/-/blob/master/debian/ Should we tag this 'upstream'? Cheers Sascha
