FS#68424 - [mesa] [drm] [amdgpu] [Webkit] complete GPU/Gnome crash on epiphany video player start

Attached to Project: Arch Linux
Opened by Jakub (cartesius) - Monday, 26 October 2020, 12:22 GMT
Last edited by Sven-Hendrik Haase (Svenstaro) - Monday, 07 February 2022, 07:24 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To Jan Alexander Steffens (heftig)
Laurent Carlier (lordheavy)
Architecture All
Severity Critical
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 2
Private No

Details

Description:

GPU crash, ssh/kernel still working, but no signal from GPU at all, no console etc.


amdgpu driver for RX 570

dmesg:

[13676.299993] [drm:amdgpu_dm_atomic_commit_tail [amdgpu]] *ERROR* Waiting for fences timed out!
[13680.917949] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx timeout, signaled seq=547054, emitted seq=547057
[13680.918006] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process WebKitWebProces pid 14884 thread WebKitWebP:cs0 pid 14895
[13680.918008] amdgpu 0000:01:00.0: amdgpu: GPU reset begin!
[13681.315827] amdgpu 0000:01:00.0: [drm:amdgpu_ring_test_helper [amdgpu]] *ERROR* ring kiq_2.1.0 test failed (-110)
[13681.315878] [drm:gfx_v8_0_hw_fini [amdgpu]] *ERROR* KCQ disable failed
[13681.500450] amdgpu: cp is busy, skip halt cp
[13681.684907] amdgpu: rlc is busy, skip halt rlc
[13681.685921] amdgpu 0000:01:00.0: amdgpu: GPU BACO reset
[13681.856639] amdgpu 0000:01:00.0: amdgpu: GPU reset succeeded, trying to resume
[13681.857082] [drm] PCIE GART of 256M enabled (table at 0x000000F400300000).
[13681.857090] [drm] VRAM is lost due to GPU reset!
[13683.260569] [drm] enabling link 0 failed: 15
[13683.299052] [drm] UVD and UVD ENC initialized successfully.
[13683.399076] [drm] VCE initialized successfully.
[13683.406181] [drm] recover vram bo from shadow start
[13683.409551] [drm] recover vram bo from shadow done
[13683.409552] [drm] Skip scheduling IBs!
[13683.409553] [drm] Skip scheduling IBs!
[13683.409561] [drm] Skip scheduling IBs!
[13683.409564] [drm] Skip scheduling IBs!
[13683.409566] [drm] Skip scheduling IBs!
[13683.409569] [drm] Skip scheduling IBs!
[13683.409572] [drm] Skip scheduling IBs!
[13683.409574] [drm] Skip scheduling IBs!
[13683.409577] [drm] Skip scheduling IBs!
[13683.409579] [drm] Skip scheduling IBs!
[13683.409581] [drm] Skip scheduling IBs!
[13683.409585] [drm] Skip scheduling IBs!
[13683.409587] [drm] Skip scheduling IBs!
[13683.409588] [drm] Skip scheduling IBs!
[13683.409590] [drm] Skip scheduling IBs!
[13683.409592] [drm] Skip scheduling IBs!
[13683.409593] [drm] Skip scheduling IBs!
[13683.409595] [drm] Skip scheduling IBs!
[13683.409613] amdgpu 0000:01:00.0: amdgpu: GPU reset(2) succeeded!
[13683.409704] [drm] Skip scheduling IBs!
[13683.410279] [drm:amdgpu_cs_ioctl [amdgpu]] *ERROR* Failed to initialize parser -125!
[13698.805288] rfkill: input handler enabled
[13699.358531] audit: type=1100 audit(1603714090.328:319): pid=15113 uid=0 auid=4294967295 ses=4294967295 msg='op=PAM:authentication grantors=pam_succeed_if,pam_permit acct="gdm" exe="/usr/lib/gdm-session-worker" hostname=chrostik addr=? terminal=/dev/tty1 res=success'
[13699.358743] audit: type=1101 audit(1603714090.328:320): pid=15113 uid=0 auid=4294967295 ses=4294967295 msg='op=PAM:accounting grantors=pam_succeed_if,pam_permit acct="gdm" exe="/usr/lib/gdm-session-worker" hostname=chrostik addr=? terminal=/dev/tty1 res=success'
[13699.359440] audit: type=1103 audit(1603714090.329:321): pid=15113 uid=0 auid=4294967295 ses=4294967295 msg='op=PAM:setcred grantors=pam_permit acct="gdm" exe="/usr/lib/gdm-session-worker" hostname=chrostik addr=? terminal=/dev/tty1 res=success'

Additional info:

* Linux foo 5.9.1-zen2-1-zen #1 ZEN SMP PREEMPT Sun, 18 Oct 2020 02:45:20 +0000 x86_64 GNU/Linux
* libva-mesa-driver 20.2.1-1
* mesa 20.2.1-1


Steps to reproduce:

1. Start Gnome on Wayland.
2. Start epiphany.
3. Play online video, VAAPI enabled.
This task depends upon

Closed by  Sven-Hendrik Haase (Svenstaro)
Monday, 07 February 2022, 07:24 GMT
Reason for closing:  Upstream
Additional comments about closing:  2022-01-08: A task closure has been requested. Reason for request: This is an upstream bug(see https://gitlab.freedesktop.org/mesa/mesa /-/issues/3656)
Comment by Jakub (cartesius) - Monday, 09 November 2020, 14:48 GMT
Crashed again, now with the 5.9.6-zen1-1-zen kernel.
Comment by Christian Martinez (alexithymia) - Wednesday, 30 December 2020, 19:06 GMT
Jakub if you're still encountering this issue and you're running a Raven Ridge APU, you can track the mesa issue here as I'm also affected and a couple of others: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3656
Comment by Magnus (DeArchDev) - Saturday, 08 January 2022, 17:41 GMT
This issue is the same as https://bugs.archlinux.org/task/68402?project=1(this needs verification as this is according to me and not fully verified)

Loading...