PC Graphics

[dimam]

Please answer me because I don't know if I should wait for next driver version or change the card by warranty.

dmesg | grep amdgpu output shows this:

[ 2.017729] [drm] amdgpu kernel modesetting enabled.
[ 2.017730] [drm] amdgpu version: 18.50.1.418
[ 2.019784] fb: switching to amdgpudrmfb from VESA VGA
[ 2.020436] amdgpu 0000:01:00.0: Invalid PCI ROM header signature: expecting 0xaa55, got 0xffff
[ 2.020733] amdgpu 0000:01:00.0: VRAM: 2048M 0x000000F400000000 - 0x000000F47FFFFFFF (2048M used)
[ 2.020734] amdgpu 0000:01:00.0: GART: 256M 0x000000FF00000000 - 0x000000FF0FFFFFFF
[ 2.020829] [drm] amdgpu: 2048M of VRAM memory ready
[ 2.020830] [drm] amdgpu: 7974M of GTT memory ready.
[ 2.290959] fbcon: amdgpudrmfb (fb0) is primary device
[ 2.353733] amdgpu 0000:01:00.0: fb0: amdgpudrmfb frame buffer device
[ 2.368822] [drm] Initialized amdgpu 3.27.0 20150101 for 0000:01:00.0 on minor 0
[ 180.678899] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0ff8770c for process chromium-browse pid 5490 thread chromium-browse pid 5490
[ 180.678904] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001203FF
[ 180.678906] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0A07700C
[ 180.678908] amdgpu 0000:01:00.0: VM fault (0x0c, vmid 5, pasid 32770) at page 1180671, read from 'SDM0' (0x53444d30) (119)
[ 228.238282] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0ff8770c for process chromium-browse pid 5490 thread chromium-browse pid 5490
[ 228.238288] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001203FF
[ 228.238289] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0207700C
[ 228.238292] amdgpu 0000:01:00.0: VM fault (0x0c, vmid 1, pasid 32770) at page 1180671, read from 'SDM0' (0x53444d30) (119)
[ 231.808255] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0ff8770c for process chromium-browse pid 5490 thread chromium-browse pid 5490
[ 231.808261] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001203FF
[ 231.808263] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0607700C
[ 231.808266] amdgpu 0000:01:00.0: VM fault (0x0c, vmid 3, pasid 32770) at page 1180671, read from 'SDM0' (0x53444d30) (119)
[ 669.251439] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0ff8770c for process chromium-browse pid 5490 thread chromium-browse pid 5490
[ 669.251446] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001203FF
[ 669.251449] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0E07700C
[ 669.251453] amdgpu 0000:01:00.0: VM fault (0x0c, vmid 7, pasid 32770) at page 1180671, read from 'SDM0' (0x53444d30) (119)
[ 1300.663426] amdgpu 0000:01:00.0: GPU fault detected: 146 0x0ff8770c for process chromium-browse pid 5490 thread chromium-browse pid 5490
[ 1300.663430] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x001485FF
[ 1300.663433] amdgpu 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0807700C
[ 1300.663436] amdgpu 0000:01:00.0: VM fault (0x0c, vmid 4, pasid 32770) at page 1345023, read from 'SDM0' (0x53444d30) (119)

[elstaci]

From a Linux website giving the correct steps in installing AMDGPU-PRO drivers in Ubuntu 18.04: https://linuxconfig.org/how-to-install-the-latest-amd-radeon-drivers-on-ubuntu-18-04-bionic-beaver-l...

Just in case you incorrectly installed the AMD Drivers.

To test to see if your GPU card is defective, Run it in Windows or another computer. If it runs in Windows or in another computer than it probably is a driver issue or it is installed wrong or it is not compatible with the kernel you have installed.

[dimam]

I don't have Windows x64. Is there any tool for Linux to check the hardware? I mean video memory and so on.

I did try memtestCL for video memory, but I'm not sure it's correct. Seems that tool is able to say "all is ok" for 3Gb of video memory of my 2Gb rx460.