On 03/01/2024 20.13, Johannes Schauer Marin Rodrigues wrote:
now that I know that BUILT_MODULE_NAME[0] and DEST_MODULE_LOCATION[0] break it, I found other packages via codesearch.d.n that use that syntax and I was wondering whether they break as well.
I did the same test after seeing your bug report. Arghhhh.... I'm pretty sure I ran install tests with all *-dkms packages from sid and dkms from experimental (-2 at that time) without seeing any issues ... but probably I messed up the test setup (or the chroot where I tested).
I also tried installing all of the above packages with dkms from unstable patched with my patch from my last email and then they succeed to install again, thus indicating that this is all the same problem class.
Thanks for your analysis which isolated the broken change.
Since this problem is making several dozen packages in the archive uninstallable, I'm raising the severity of this bug to serious. Apologies if I'm wrong but better safe then sorry, right? Maybe this prevents users of apt-listbugs from installing dkms from unstable and have their dkms modules fail. Please lower the severity if I raised it in error. Apologies in that case.
Perfectly fine. I decided working on the fixes (there was another regression reported, too) before playing with the bts :-)
I'd also like to suggest the following patch to be included into the dkms package which would've spotted this problem early on:
That was the first change I did today - making sure the problem is reproducible with the test packages. :-) I was really confused when you reported dkms-test-dkms working fine - which had a similarily trivial dkms.conf than your package (except for some supposedly irrelevant '[0]').
BTW, even if your package isn't in Debian, I'd suggest adding Testsuite: autopkgtest-pkg-dkms Andreas
