Arch Linux

tcn

Member

Registered: 2011-09-30

Posts: 68

[drm] GPU recovery disabled.

Hi!

This is a surprisingly long standing problem, more precisely since 4.15 all the way up to 4.17. After resuming from sleep (echo -n mem > /sys/power/state) amdgpu is dead (always, reliably).
Here's what dmesg has to say about it:

[   42.802559] PM: suspend exit
[   42.824332] amdgpu 0000:41:00.0: GPU fault detected: 147 0x0bd84802
[   42.824338] amdgpu 0000:41:00.0:   VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x0034F97B
[   42.824341] amdgpu 0000:41:00.0:   VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0C048002
[   42.824345] amdgpu 0000:41:00.0: VM fault (0x02, vmid 6) at page 3471739, read from 'TC0' (0x54433000) (72)
[   52.956306] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx timeout, last signaled seq=1287, last emitted seq=1289
[   52.956316] [drm] IP block:gfx_v8_0 is hung!
[   52.956362] [drm] GPU recovery disabled.

I've also seen fault 146 but other than that it mostly looks the same. 4.14-lts (with dc=0) works fine. Also tried the amd staging next kernel from AUR.

Zenith Extreme, 1950x, RX 460 (polaris 11).

Any advice? Report upstream?

thx
tcn

Last edited by tcn (2018-06-25 08:45:53)

loqs

Member

Registered: 2014-03-06

Posts: 18,055

Re: [drm] GPU recovery disabled.

How recently did you try linux-amd-staging-git?  If linux-amd-staging-drm-next-git has the issue would suggest reporting it upstream.
Also could you elaborate on all the combinations of amdgpu.dc and kernel version you have tried and the results.

tcn

Member

Registered: 2011-09-30

Posts: 68

Re: [drm] GPU recovery disabled.

How recently did you try linux-amd-staging-git?

Couple of days ago. Well, however, while this looks like it's 4.18...

==> Making package: linux-amd-staging-drm-next-git 4.18.754874.486e4f30e2a3-1 (Mon 25 Jun 2018 03:29:39 PM CEST)

...what I actually end up with is 4.17-rc5-something. I just deleted the AUR git (branch is master, BTW) and retried: same (yay -S linux-amd-staging-drm-next-git).

Also could you elaborate on all the combinations of amdgpu.dc and kernel version you have tried and the results.

Well, I don't have dmesg for very combination at hand but let's put it this way: so far I haven't found any combination that wouldn't immediately crash.

How can I tell whether DC *is actually* enabled?

loqs

Member

Registered: 2014-03-06

Posts: 18,055

Re: [drm] GPU recovery disabled.

==> Making package: linux-amd-staging-drm-next-git 4.18.754874.486e4f30e2a3-1 (Mon 25 Jun 2018 03:29:39 PM CEST)

...what I actually end up with is 4.17-rc5-something. I just deleted the AUR git (branch is master, BTW) and retried: same (yay -S linux-amd-staging-drm-next-git).

_kernel_rel=4.18
....
  echo ${_kernel_rel}.$(git rev-list --count HEAD).$(git rev-parse --short HEAD)

Yes the PKGBUILD sets the version to 4.18 even though it is based on a 4.17 release candidate according to the git tree and the top Makefile.

Also could you elaborate on all the combinations of amdgpu.dc and kernel version you have tried and the results.

Well, I don't have dmesg for very combination at hand but let's put it this way: so far I haven't found any combination that wouldn't immediately crash.

How can I tell whether DC *is actually* enabled?

parm:           dc:Display Core driver (1 = enable, 0 = disable, -1 = auto (default)) (int)

If you are not specifying the option it goes to autodetection for your GPU I would expect autodetection to enable it on 4.17+