[Bug 278212] amdgpu / AMD Radeon XT 6900 - kernel panic (page fault) on FreeBSD 14.0-RELEASE-p6

From: <bugzilla-noreply_at_freebsd.org>
Date: Sun, 07 Apr 2024 01:49:46 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=278212

            Bug ID: 278212
           Summary: amdgpu / AMD Radeon XT 6900 - kernel panic (page
                    fault) on FreeBSD 14.0-RELEASE-p6
           Product: Base System
           Version: 14.0-RELEASE
          Hardware: Any
                OS: Any
            Status: New
          Severity: Affects Only Me
          Priority: ---
         Component: kern
          Assignee: bugs@FreeBSD.org
          Reporter: jon@xyinn.org

Created attachment 249782
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=249782&action=edit
core dump

Hello all,

I few weeks ago I started experiencing a kernel panic with my AMD Radeon XT
6900 XT (although I don't necessarily think it's a regression, but maybe
something related to my workflow with my KVM). I'm on FreeBSD 14.0-RELEASE-p6,
and was compiled the sources in order for me to eventually successfully capture
a kernel dump (attached).

root@leslie:/usr/src # uname -a
FreeBSD leslie 14.0-RELEASE-p6 FreeBSD 14.0-RELEASE-p6 #0
releng/14.0-n265417-d338712beb16: Tue Apr  2 19:26:31 EDT 2024    
root@leslie:/usr/obj/usr/src/amd64.amd64/sys/GENERIC amd64

drm-kmod-20220907_1
gpu-firmware-amd-kmod-sienna-cichlid-20230625

I usually notice this happening when I start up my FreeBSD desktop while the
KVM is currently active on my other laptop. I let FreeBSD (cold start since
resuming from sleep doesn't work on this desktop) start up in the background
and continue using my laptop in the meantime. Once ready, I hit the KVM button,
and everything starts to switch over to the desktop, after a few seconds of
waiting, instead of me seeing the FreeBSD login prompt, I sometimes see the
initial system start up screen, or some other early stage FreeBSD process.
Essentially the system had a kernel panic and restarted. Once I wait for the
system to start up again, I can see that it mentions a page fault, and the
system works normally. I do believe this may have happened to me while I was
using the system normally, but it seems to have a higher probability of
happening with that KVM switch workflow.

Please let me know if there is anything else I can provide. Thank you!

-- 
You are receiving this mail because:
You are the assignee for the bug.