I am using 4.6-staging because I need the audio driver it provides. I have been trying to lower the voltages of the card by bios editing and there seems to be no issues with windows, but I dont seem to get higher DPM states in use at linux. Something is broken. dmesg is spammed with these amdgpu-pro, mint 18, rx480. Latest 4.6 kernel from git (agd5f/linux). [ 1214.592256] VM fault (0x14, vmid 6) at page 1071080, write from 'CB0' (0x43423000) (96) [ 1214.592262] amdgpu 0000:01:00.0: GPU fault detected: 146 0x059aa504 [ 1214.592262] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057D2 [ 1214.592263] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0D0D0014 [ 1214.592264] VM fault (0x14, vmid 6) at page 1071058, write from 'CB7' (0x43423700) (208) [ 1214.592269] amdgpu 0000:01:00.0: GPU fault detected: 146 0x059a5504 [ 1214.592270] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057D8 [ 1214.592271] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0D020014 [ 1214.592271] VM fault (0x14, vmid 6) at page 1071064, write from 'CB2' (0x43423200) (32) [ 1214.592642] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0f2ad014 [ 1214.592643] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057E9 [ 1214.592643] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0C0D000C [ 1214.592644] VM fault (0x0c, vmid 6) at page 1071081, read from 'CB7' (0x43423700) (208) [ 1214.592650] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0f2aa014 [ 1214.592650] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057D8 [ 1214.592651] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0C02000C [ 1214.592652] VM fault (0x0c, vmid 6) at page 1071064, read from 'CB2' (0x43423200) (32) [ 1214.592657] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0f60e00c [ 1214.592658] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057EC [ 1214.592659] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0D0E0014 [ 1214.592659] VM fault (0x14, vmid 6) at page 1071084, write from 'CB6' (0x43423600) (224) [ 1214.592665] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0f28a00c [ 1214.592666] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057D0 [ 1214.592666] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0D0E0014 [ 1214.592667] VM fault (0x14, vmid 6) at page 1071056, write from 'CB6' (0x43423600) (224) [ 1214.592673] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0ec25014 [ 1214.592673] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001057D2 [ 1214.592674] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0D060014 [ 1214.592675] VM fault (0x14, vmid 6) at page 1071058, write from 'CB0' (0x43423000) (96) [ 1214.661835] steam[6544]: segfault at 4184ddd7 ip 00000000f72ea771 sp 00000000ff9ce690 error 4 in libc-2.23.so[f727d000+1af000] [ 1353.103633] failed to send pre message 200 ret is 0 [ 1353.317672] failed to send message 200 ret is 0 [ 1353.536394] failed to send pre message 201 ret is 0 [ 1353.754228] failed to send message 201 ret is 0 [ 1961.904046] failed to send pre message 200 ret is 0 [ 1962.120865] failed to send message 200 ret is 0 [ 1962.342261] failed to send pre message 201 ret is 0 [ 1962.560233] failed to send message 201 ret is 0
Created attachment 126129 [details] bug in freesync code too?
I was able to get DPM states working again. I don't know what actually broke them. but the freesync log crash is still valid.
Is this still a thing?
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/11.
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.