Hi Mathieu, Mathieu Malaterre, on 2026-02-17: > Le lun. 16 févr. 2026 à 22:30, Étienne Mollier <[email protected]> a écrit : > > Greetings, > > > > In #1127756, Adrian Bunk spotted a build failure of dcmtk on > > s390x. I have spent a bit of time trying to check whether > > addressing the issue was manageable. After determining I would > > get nowhere, I have contacted upstream via their contact mail > > address dedicated to bug report ; see their CONTRIBUTING.md > > file. If the calendar plays well, dcmtk 3.7.0 may be sprint > > material for the upcoming weekend. If not, well, at least I > > woule have tried. > > I am pretty sure this is just a false positive. The code base of DCMTK > is compatible with big-endian. I suspect you'll get an answer from the > team real time soon. In case they are busy I suggest you go ahead with > the experimental upload with the test disabled.
Thanks for your thoughts, although the feedback I got sounded like there was indeed a problem in produced data (as opposed to expected data I believe). From DCMTK Team, on 2026-02-17: > > > In this case, the pixel data of the dataset produced in the failing unit > > > test indeed contains incorrect pixel data with incorrect byte order. But I agree it is time to move forwards with the migration to version 3.7.0 and this issue should probably not be a blocker for a Debian experimental upload at least. I triggered a last s390x test build to make sure that my skip implementation goes through and will upload at some point tomorrow evening. Have a nice day, :) -- .''`. Étienne Mollier <[email protected]> : :' : pgp: 8f91 b227 c7d6 f2b1 948c 8236 793c f67e 8f0d 11da `. `' sent from /dev/pts/2, please excuse my verbosity `- on air: Beethoven - Sonate pour piano n°8
signature.asc
Description: PGP signature
