Tracking down cause of crash ?graphics driver

arubin · 05-14-2024, 09:47 AM

I am trying to track down the cause of crashes on a fairly new PC. Every couple of weeks there is an episode where the mouse cursor freezes, the computer becomes unresponsive and after a few seconds the screen goes black. I suspect, though I do not know how to prove it that this is a graphics driver issue. The graphics card is a Radeon RX 6600. I am running Slackware 15 with multilib up to date with slackpkg. I use kde. My impression is that switching to Wayland reduced the frequency of crashes but not sure. I can be doing anything when the crash occurs. Memtest does not show any problems.

I haven't found anything that I found helpful in any logs but perhaps I don't know what I'm looking for.

Is there any particular thing I should be looking for?

Are there any known issues with Radeon and Linux? I had thought this kernel 5.15.145 worked well with Radeon.

business_kid · 05-14-2024, 10:15 AM

I'm on similar kit with the RX 6600XT. I ran Slackware64 & Multilib for two years on 5.15.63. No issues at all.

Keep an eye on /var/log/syslog, and htop while running. When it freezes, switch to the htop window. Another trick is to set yourself on runlevel 3, & use startx. Then, when it hangs, hit Ctrl_Alt_F1 and look for output.

Thom1b · 05-14-2024, 10:16 AM

If your PC is new, I suggest you to use a recent kernel. When I bought my current PC, I had several hardware bugs even with latest kernel (5.12 at the time). I had to wait for 5.14 to have stable drivers.

jayjwa · 05-14-2024, 12:21 PM

The two things that come to mind is if you see something of the sort

Code:

kernel: [466414.859438] amdgpu 0000:09:00.0: amdgpu: PCIE GART of 1024M enabled (table at 0x000000F400800000).
kernel: [466415.083620] [drm:si_dpm_set_power_state [amdgpu]] *ERROR* si_enable_smc_cac failed
kernel: [466415.473694] [drm] UVD initialized successfully.
kernel: [466415.946024] amdgpu 0000:09:00.0: amdgpu: recover vram bo from shadow start
kernel: [466415.947886] amdgpu 0000:09:00.0: amdgpu: recover vram bo from shadow done
kernel: [466415.947926] [drm] Skip scheduling IBs!
kernel: [466415.948001] [drm] Skip scheduling IBs!
kernel: [466415.994745] amdgpu 0000:09:00.0: amdgpu: GPU reset(1) succeeded!

. I had these with crashes (Ryzen/AMD/Radeon HD 8570) as you describe ~ 2 days - 1 or 2 weeks without warning. Moving to Wayland, I've only had one of those since Feb. The system stays up until I shut it down for a new kernel. This was across multiple Mesa versions, kernel version, window managers, etc. The only constant was amdgpu. The other is the well-known RCU no callbacks issue, which is already documented online in various places.

arubin · 05-14-2024, 01:12 PM

I have found the following. Could this be it?

Quote:

May 12 15:58:08 Lavankossot kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
May 12 15:58:09 Lavankossot last message buffered 1 times
May 12 15:58:09 Lavankossot kernel: sched: RT throttling activated
May 12 15:58:09 Lavankossot kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
May 12 15:58:12 Lavankossot last message buffered 10 times
May 12 15:58:12 Lavankossot kernel: INPUT packet died: IN=eth0 OUT= MAC=08:bf:b8:36:1a:58:e8:6f:38:31:21:95:08:00 SRC=192.168.1.166 DST=192.168.1.51 LEN=524 TOS=0x00 PREC=0x00 TTL=64 ID=8451 DF PROTO=UDP SPT=51613 DPT=54040 LEN=504
May 12 15:58:12 Lavankossot kernel: INPUT packet died: IN=eth0 OUT= MAC=08:bf:b8:36:1a:58:e8:6f:38:31:21:95:08:00 SRC=192.168.1.166 DST=192.168.1.51 LEN=306 TOS=0x00 PREC=0x00 TTL=64 ID=8524 DF PROTO=UDP SPT=40515 DPT=54040 LEN=286
May 12 15:58:13 Lavankossot kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
May 12 15:58:13 Lavankossot last message buffered 1 times
May 12 15:58:13 Lavankossot kernel: [drm:amdgpu_dm_atomic_commit_tail [amdgpu]] *ERROR* Waiting for fences timed out!
May 12 15:58:13 Lavankossot kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
May 12 15:58:13 Lavankossot kernel: INPUT packet died: IN=eth0 OUT= MAC=08:bf:b8:36:1a:58:7a:9f:29:c6:f7:8a:08:00 SRC=192.168.1.227 DST=192.168.1.51 LEN=344 TOS=0x00 PREC=0x00 TTL=64 ID=62732 PROTO=UDP SPT=1900 DPT=54040 LEN=324
May 12 15:58:13 Lavankossot kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
May 12 15:58:14 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: Failed to disable gfxoff!
May 12 15:58:19 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: SMU: I'm not done with your previous command!
May 12 15:58:19 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: Failed to disable gfxoff!
May 12 15:58:24 Lavankossot kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, signaled seq=169044, emitted seq=169046
May 12 15:58:24 Lavankossot kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process kwin_wayland pid 1882 thread kwin_wayla:cs0 pid 1884
May 12 15:58:29 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: SMU: I'm not done with your previous command!
May 12 15:58:29 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: Failed to disable gfxoff!
May 12 15:58:30 Lavankossot kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
May 12 15:58:31 Lavankossot last message buffered 2 times
May 12 15:58:33 Lavankossot kernel: amdgpu 0000:0c:00.0: [drm:amdgpu_ring_test_helper [amdgpu]] *ERROR* ring kiq_2.1.0 test failed (-110)
May 12 15:58:33 Lavankossot kernel: [drm:gfx_v10_0_hw_fini [amdgpu]] *ERROR* KGQ disable failed
May 12 15:58:34 Lavankossot kernel: amdgpu 0000:0c:00.0: [drm:amdgpu_ring_test_helper [amdgpu]] *ERROR* ring kiq_2.1.0 test failed (-110)
May 12 15:58:34 Lavankossot kernel: [drm:gfx_v10_0_hw_fini [amdgpu]] *ERROR* KCQ disable failed
May 12 15:58:34 Lavankossot kernel: [drm:gfx_v10_0_hw_fini [amdgpu]] *ERROR* failed to halt cp gfx
May 12 15:58:36 Lavankossot kernel: INPUT packet died: IN=eth0 OUT= MAC=01:00:5e:00:00:fb:e8:6f:38:31:21:95:08:00 SRC=192.168.1.166 DST=224.0.0.251 LEN=68 TOS=0x00 PREC=0x00 TTL=255 ID=31962 DF PROTO=UDP SPT=5353 DPT=5353 LEN=48
May 12 15:58:40 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: SMU: I'm not done with your previous command!
May 12 15:58:40 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: Failed to disable smu features.
May 12 15:58:40 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: Fail to disable dpm features!
May 12 15:58:40 Lavankossot kernel: [drm:amdgpu_device_ip_suspend_phase2 [amdgpu]] *ERROR* suspend of IP block <smu> failed -62
May 12 15:58:41 Lavankossot kernel: [drm] psp gfx command DESTROY_TMR(0x7) failed and response status is (0x80000306)
May 12 15:58:46 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: SMU: I'm not done with your previous command!
May 12 15:58:46 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: GPU mode1 reset failed
May 12 15:58:46 Lavankossot kernel: amdgpu 0000:0c:00.0: amdgpu: ASIC reset failed with error, -62 for drm dev, 0000:0c:00.0
May 12 15:58:59 Lavankossot kernel: [drm] failed to load ucode SMC(0x18)
May 12 15:58:59 Lavankossot kernel: [drm] psp gfx command LOAD_IP_FW(0x6) failed and response status is (0x80000306)
May 12 15:59:01 Lavankossot kernel: [drm] psp gfx command AUTOLOAD_RLC(0x21) failed and response status is (0x0)
May 12 15:59:01 Lavankossot kernel: [drm

sp_load_non_psp_fw [amdgpu]] *ERROR* Failed to start rlc autoload
May 12 15:59:01 Lavankossot kernel: [drm

sp_resume [amdgpu]] *ERROR* PSP resume failed
May 12 15:59:01 Lavankossot kernel: [drm:amdgpu_device_fw_loading [amdgpu]] *ERROR* resume of IP block <psp> failed -22
May 12 15:59:01 Lavankossot kernel: [drm:amdgpu_cs_ioctl [amdgpu]] *ERROR* Failed to initialize parser -125!
May 12 15:59:01 Lavankossot last message buffered 9 times
May 12 15:59:01 Lavankossot kernel: snd_hda_intel 0000:0c:00.1: CORB reset timeout#2, CORBRP = 65535
May 12 15:59:01 Lavankossot ModemManager[2180]: <warn> could not acquire the 'org.freedesktop.ModemManager1' service name
May 12 15:59:01 Lavankossot pulseaudio[2138]: [pulseaudio] stdin-util.c: Lost I/O connection in module "module-gsettings"
May 12 15:59:11 Lavankossot kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring sdma0 timeout, signaled seq=12351, emitted seq=12353
May 12 15:59:11 Lavankossot kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process pid 0 thread pid 0
May 12 15:59:11 Lavankossot pulseaudio[2138]: [pulseaudio] x11wrap.c: X11 I/O error handler called
May 12 15:59:11 Lavankossot pulseaudio[2138]: [pulseaudio] x11wrap.c: X11 I/O error exit handler called, preparing to tear down X11 modules