AMDGPU fatal error during GPU init (SOLVED)

After the last round of manjaro updates I can only see a black screen after grub if I try to start with the AMDGPU driver loaded. Happens with both kernel 4.19 and 4.20, Was working fine on both before the update. I’m pretty sure the kernels themselves did not update but mesa did.

I have tried re installing manjaro completely.

A similar thing happened to me before only the solution was to disable amdgpu.dpm (same problem with all distros I tried, fedora-silverblue etc…).

I am running a MSI 390x graphics card.

I have tried disableing amdgpu.dc and amdgpu.dpm in various combonations.

I have unstable performance with the radeon driver so any help would be greatly appreciated.

This is the output from journalctl | grep error (Where the AMDGPU fatal error during GPU init appears)

`Jan 20 18:18:16 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 20 18:18:17 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 20 18:18:17 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 20 18:18:17 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 20 18:18:17 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan 20 22:52:07 jayden-Manjaro dunst[896]: XIO: fatal IO error 4 (Interrupted system call) on X server ":0"`

`Jan 21 08:24:18 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 08:24:18 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 08:24:18 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 08:24:18 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 08:24:18 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan 21 14:41:39 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:41:39 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:41:39 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 14:41:39 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:41:39 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan 21 14:41:40 jayden-Manjaro kernel: [drm:dm_pp_get_static_clocks [amdgpu]] *ERROR* DM_PPLIB: invalid powerlevel state: 0!`

`Jan
 21 14:41:40 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:41:40 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:41:40 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan 21 14:41:40 jayden-Manjaro kernel: [drm:amdgpu_vce_ring_test_ring [amdgpu]] *ERROR* amdgpu: ring 12 test failed`

`Jan
 21 14:41:40 jayden-Manjaro kernel: [drm:amdgpu_device_init.cold.14 
[amdgpu]] *ERROR* hw_init of IP block <vce_v2_0> failed -110`

`Jan 21 14:41:40 jayden-Manjaro kernel: amdgpu 0000:01:00.0: Fatal error during GPU init`

`Jan 21 14:41:40 jayden-Manjaro kernel: amdgpu: probe of 0000:01:00.0 failed with error -110`

`Jan 21 14:42:57 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:42:57 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:42:57 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 14:42:57 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:42:57 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan 21 14:42:58 jayden-Manjaro kernel: [drm:dm_pp_get_static_clocks [amdgpu]] *ERROR* DM_PPLIB: invalid powerlevel state: 0!`

`Jan
 21 14:42:58 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:42:58 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:42:58 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan 21 14:44:30 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:44:30 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:44:30 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 14:44:30 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:44:30 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 14:44:31 jayden-Manjaro systemd-vconsole-setup[434]: KD_FONT_OP_GET 
failed while trying to get the font metadata: Input/output error`

`Jan 21 14:46:16 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:46:16 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:46:16 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 14:46:16 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:46:16 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan 21 14:48:12 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:48:12 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:48:12 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan
 21 14:48:12 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20180810/psargs-330)`

`Jan
 21 14:48:12 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20180810/psparse-516)`

`Jan 21 14:48:15 jayden-Manjaro kernel: [drm:radeon_dp_link_train [radeon]] *ERROR* channel eq failed: 5 tries`

`Jan 21 14:48:15 jayden-Manjaro kernel: [drm:radeon_dp_link_train [radeon]] *ERROR* channel eq failed`

`Jan 21 14:54:14 jayden-Manjaro kernel: ata7.00: irq_stat 0x08000040, interface fatal error, connection status changed`

`Jan 21 14:54:14 jayden-Manjaro kernel: ata7: SError: { RecovComm PHYRdyChg CommWake DevExch }`

`res 40/00:d8:00:48:65/00:00:16:00:00/40 Emask 0x10 (ATA bus error)`

`Jan 21 14:54:15 jayden-Manjaro kernel: print_req_error: I/O error, dev sde, sector 398444544`

`Jan 21 14:54:15 jayden-Manjaro kernel: ata7.00: irq_stat 0x08000040, interface fatal error, connection status changed`

`Jan 21 14:54:15 jayden-Manjaro kernel: ata7: SError: { RecovComm PHYRdyChg CommWake DevExch }`

`res 40/00:98:00:88:c9/00:00:2d:00:00/40 Emask 0x10 (ATA bus error)`

`Jan 21 14:54:16 jayden-Manjaro kernel: print_req_error: I/O error, dev sde, sector 804997120`

`Jan 21 14:55:33 jayden-Manjaro kernel: ata7.00: irq_stat 0x08000040, interface fatal error, connection status changed`

`Jan 21 14:55:33 jayden-Manjaro kernel: ata7: SError: { RecovComm PHYRdyChg CommWake DevExch }`

`res 40/00:70:00:c8:59/00:00:63:00:00/40 Emask 0x10 (ATA bus error)`

`Jan 21 14:55:34 jayden-Manjaro kernel: print_req_error: I/O error, dev sde, sector 110594048`

`Jan 21 14:57:28 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:57:28 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 14:57:28 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan
 21 14:57:28 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 14:57:28 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan 21 14:57:28 jayden-Manjaro kernel: [drm:dm_pp_get_static_clocks [amdgpu]] *ERROR* DM_PPLIB: invalid powerlevel state: 0!`

`Jan
 21 14:57:28 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:57:28 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:57:28 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan 21 14:57:29 jayden-Manjaro kernel: [drm:amdgpu_vce_ring_test_ring [amdgpu]] *ERROR* amdgpu: ring 12 test failed`

`Jan
 21 14:57:29 jayden-Manjaro kernel: [drm:amdgpu_device_init.cold.17 
[amdgpu]] *ERROR* hw_init of IP block <vce_v2_0> failed -110`

`Jan 21 14:57:29 jayden-Manjaro kernel: amdgpu 0000:01:00.0: Fatal error during GPU init`

`Jan 21 14:57:29 jayden-Manjaro kernel: amdgpu: probe of 0000:01:00.0 failed with error -110`

`Jan 21 14:58:09 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:58:09 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 14:58:09 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan
 21 14:58:09 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 14:58:09 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan 21 14:58:09 jayden-Manjaro kernel: [drm:dm_pp_get_static_clocks [amdgpu]] *ERROR* DM_PPLIB: invalid powerlevel state: 0!`

`Jan
 21 14:58:09 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:58:09 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:58:09 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan 21 14:58:10 jayden-Manjaro kernel: [drm:amdgpu_vce_ring_test_ring [amdgpu]] *ERROR* amdgpu: ring 12 test failed`

`Jan
 21 14:58:10 jayden-Manjaro kernel: [drm:amdgpu_device_init.cold.17 
[amdgpu]] *ERROR* hw_init of IP block <vce_v2_0> failed -110`

`Jan 21 14:58:10 jayden-Manjaro kernel: amdgpu 0000:01:00.0: Fatal error during GPU init`

`Jan 21 14:58:10 jayden-Manjaro kernel: amdgpu: probe of 0000:01:00.0 failed with error -110`

`Jan 21 14:59:01 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 14:59:01 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 14:59:01 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan
 21 14:59:01 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 14:59:01 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan 21 14:59:01 jayden-Manjaro kernel: [drm:dm_pp_get_static_clocks [amdgpu]] *ERROR* DM_PPLIB: invalid powerlevel state: 0!`

`Jan
 21 14:59:01 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:59:01 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan
 21 14:59:01 jayden-Manjaro kernel: [drm] dce110_link_encoder_construct:
 Failed to get encoder_cap_info from VBIOS with error code 4!`

`Jan 21 15:00:02 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 15:00:02 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 15:00:02 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan
 21 15:00:02 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 15:00:02 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan 21 15:01:57 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 15:01:57 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 15:01:57 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan
 21 15:01:57 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 15:01:57 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan 21 15:04:10 jayden-Manjaro kernel: RAS: Correctable Errors collector initialized.`

`Jan
 21 15:04:10 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 15:04:10 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan
 21 15:04:10 jayden-Manjaro kernel: ACPI BIOS Error (bug): Could not 
resolve [\_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND 
(20181003/psargs-330)`

`Jan
 21 15:04:10 jayden-Manjaro kernel: ACPI Error: Method parse/execution 
failed \_SB.PCI0.SAT0.SPT4._GTF, AE_NOT_FOUND (20181003/psparse-516)`

`Jan 21 15:04:11 jayden-Manjaro kernel: [drm:radeon_dp_link_train [radeon]] *ERROR* channel eq failed: 5 tries`

`Jan 21 15:04:11 jayden-Manjaro kernel: [drm:radeon_dp_link_train [radeon]] *ERROR* channel eq failed`

Could these be related? though it probably does not help much seeing as one of the people asking still has no real answer. Worth a look and a shot though.

1 Like

Sounds similar to what I experienced, It was around 4.17 to 4.18 I had to disable dpm or I couldn’t see anything past grub

I found some people having the same issue here https://bugs.freedesktop.org/show_bug.cgi?id=108781

There seems to be some sort of fix at the bottom of the comment chain. I know nothing so just pointing it out, I can’t actually help.

1 Like

You mentioned you tried various combinations of the amdgpu.dc and amdgpu.dpm kernel parameters, but have you tried removing the amdgpu.dpm parameter completely?

I’ve been using the following kernel parameters since around kernel 4.19.x and have not had any problems.

amdgpu.dc=1 radeon.cik_support=0 amdgpu.cik_support=1

All kernel updates in between have worked with those kernel parameters for me and I a now running kernel 4.20.3

It’s just a guess but I think there are 3 settings/code paths :

amdgpu.dpm=1 forces Dynamic Power Management on
amdgpu.dpm=0 forces Dynamic Power Management off

And not specifying the parameter at all leaves it up to whatever the default is for the driver/firmware which may be a different code path that avoids the bug.

2 Likes

Thanks, cant believed I missed that.
The first step 1) Removed amdgpu.dpm=x completely from linux commandline and updated grub.
Fixed everything, I had amdgpu.dpm=0 in there and that broke it
Now running AMDGPU driver and successfully booting.

2 Likes

I just tried that as your post came through, It worked, Thanks anyway =)

1 Like

It has been noted, again i know nothing, that this can cause some sort of long standing stability bug to reappear. So maybe look out for that, whatever it is, it is mentioned in the chain somewhere.

1 Like

Please be mindful of dates and don’t Necro threads