Problem: package kf5-kimageformats-5.100.0-1.fc37.x86_64 requires libavif.so.14()(64bit), but none of the providers can be installed - cannot install both libavif-0.10.1-3.fc37.x86_64 and libavif-0.11.1-1.fc37.x86_64 - cannot install the best update candidate for package kf5-kimageformats-5.99.0-4.fc37.x86_64 - problem with installed package libavif-0.11.1-1.fc37.x86_64 ===========================================================================
Well, there are two bugs for this now: https://bugzilla.redhat.com/show_bug.cgi?id=2115733 and https://bugzilla.redhat.com/show_bug.cgi?id=2115927. My conscience is clear, I did what I was supposed to do and raised the red flag. Now we are in a situation where a known serious bug has been pushed to stable. Hopefully 5.18.16 is pushed soon. The damage, unfortunately has been done.
OK, I just know I was running 5.18.13 with no issues, and then got hit with the issue in 5.18.15 (which was also reported in Arch, specifically for 5.18.15 link above). So I'm certainly not the only one reporting this issue specifically for 5.18.15. I'm now running 5.18.16 and the issue is resolved.
Well, AFAIK Fedora didn't build the 5.18.14 kernel - I looked in koji and didn't see it - so for Fedora users, the issue I'm talking about was introduced with the 5.18.15 update. My point was simply that the issue is fixed in 5.18.16 - which you've built in koji and I'm currently running. Since 5.18.16 is already in the pipeline, what's the rush with pushing 5.18.15 to stable when the impact for AMD users is potentially severe. We're talking a few extra days for 5.18.16 to clear. IMHO the risk benefit isn't worth it - but hey, it's your rodeo. ;-)
I don't quite understand what the point is with pushing to stable a well documented hang situation for AMD users when the issue appears to be fixed with the 5.18.16 update that is now in testing. What's the upside? Seems like you're potentially affecting a large number of users for nothing.
Whatever happened to cause the radeon issue: https://bugzilla.redhat.com/show_bug.cgi?id=2108247
has unfortunately been propagated to the 5.19 builds. I've tested with the 5.19 builds and the same error occurs.
Since upgrading to this kernel am receiving numerous radeon lockups - and the only way to recover is to drop the power cycle the system.
Jul 17 21:02:25 charon kernel: radeon 0000:03:00.0: ring 0 stalled for more than 19349msec Jul 17 21:02:25 charon kernel: radeon 0000:03:00.0: GPU lockup (current fence id 0x000000000014756d last fence id 0x0000000000147577 on ring 0) Jul 17 21:02:26 charon kernel: radeon 0000:03:00.0: ring 5 stalled for more than 19594msec Jul 17 21:02:26 charon kernel: radeon 0000:03:00.0: ring 3 stalled for more than 19152msec