| Summary: | [amdgpu] [powerplay] [runpm] Card doesn't re-init when using powerplay and runpm | ||||||
|---|---|---|---|---|---|---|---|
| Product: | DRI | Reporter: | Mike Lothian <mike> | ||||
| Component: | DRM/AMDgpu | Assignee: | Default DRI bug account <dri-devel> | ||||
| Status: | RESOLVED FIXED | QA Contact: | |||||
| Severity: | normal | ||||||
| Priority: | medium | CC: | mike | ||||
| Version: | DRI git | ||||||
| Hardware: | Other | ||||||
| OS: | All | ||||||
| Whiteboard: | |||||||
| i915 platform: | i915 features: | ||||||
| Attachments: |
|
||||||
The card now re-inits with However when power_dpm_force_performance_level is at auto the speeds don't ramp up When I was in bioshock everything was at it's lowest levels and I was getting about 3fps, when I echo'd high into power_dpm_force_performance_level I went back to 17fps, when I echo'd auto back into power_dpm_force_performance_level the fps stayed high When I leave the game and relaunch (so the card is re-initialised with runpm) it goes back to 3fps This is with the latest code in drm-next-4.6-wip - pcie dpm is still disabled I've done another test I ran DRI_PRIME=1 glxgears, just to keep the card running I then launched BioShock and the sclk & mclk again stayed at their lowest settings, I echo'd high into power_dpm_force_performance_level which changed the sclk & mclk values. When I echo'd auto back into power_dpm_force_performance_level and the clocks stayed high. When I exited the game (with glxgears still running) the clocks went back to their lowest settings, which is what I'd expect So it looks like then the card is re-initialized the auto performance level doesn't work correctly - or is being set to low Hope this helps I've just noticed that I don't need to echo "high" then "auto" to get dpm working, simply echoing "auto" even though that's what it's set to is enough to get the FPS to jump back up to a playable rate I'm not sure which commit it was - but this is now fixed for me Cheers :D |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Created attachment 120567 [details] Dmesg with reinit problems I think the key messages are: [ 61.153149] SMC address is beyond the SMC RAM area. [ 61.153161] SMC address is beyond the SMC RAM area. [ 61.153511] SMC address is beyond the SMC RAM area. [ 61.153523] SMC address is beyond the SMC RAM area. [ 61.153524] SMC address is beyond the SMC RAM area. [ 61.153524] Failed to initialize ARB table index! [ 61.153526] SMC address is beyond the SMC RAM area. [ 61.153528] Failed to notify no display! [ 61.153536] Failed to enable voltage DPM during DPM Start Function! [ 61.153537] Failed to start DPM! [ 61.313104] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: ring 0 test failed (scratch(0xC040)=0xCAFEDEAD) [ 61.313108] [drm:amdgpu_resume [amdgpu]] *ERROR* resume 5 failed -22