On Friday, 31 March 2023 12:57:44 CEST Stefan K wrote: > when I searching my error message I got a lot of hits.. > In arch it's solved: > https://bugs.archlinux.org/task/76934 > > the patch: > https://gitlab.freedesktop.org/superm1/linux/-/commit/2145b4de3fea9908cda6be > f0693a797cc7f4ddfc.patch
Oh boy, that brought back some memories ... That patch would likely indeed have solved your problem, but I'm going to take a guess and assume that you didn't actually try to apply that patch and verified that it then worked. (because it would be very difficult to apply) That patch was HUGE and therefor unacceptable to be backported to 6.1 and this issue is AFAIUI one of the reasons that it took longer then normal to declare 6.1 a LTS release. The memories refer to https://bugs.debian.org/1028451 and that got solved by a different set of patches which were accepted (and the result of *further* discussions around that issue). https://gitlab.freedesktop.org/drm/amd/-/issues/2171#note_1721571 which is the "Forwarded" of that bug, also refers to "6.1.5.arch2-1". Further discussion on that issue resulted in the aforementioned patches which were accepted. But I believe there were some corner cases expected that would not be solved and I'm guessing your issue is one of those. On Wednesday, 29 March 2023 09:59:22 CEST Stefan K wrote: > and syslog said: > [drm:dc_link_allocate_mst_payload [amdgpu]] *ERROR* Failure: pbn_per_slot==0 > not allowed. Cannot continue, returning DC_UNSUPPORTED_VALUE. That's from your initial bug report and that may be key to finding out if/why your case is indeed such a corner case and from there find. https://bugs.debian.org/1033050 may be the same or a similar issue. @Bernhard: Do you think it's the same issue? If so, that may be useful as your report contains a kernel crash with a stack trace. I don't know what to do with that, but others can. Certainly the upstream devs which this issue likely needs to be reported to.
signature.asc
Description: This is a digitally signed message part.
