AMDGPU vs. Radeon DRM On Linux 4.13 For AMD GCN 1.0/1.1 GPUs

[bridgman]

so your solution for a problem with the open-source driver is to use the closed source driver instead?

If you are looking for a solution the same day (which apparently you were) then yes. Longer term no.

is there an installer to combine the openCL-only module with the open source driver?

There will be.

why not just fix the open source driver?

No other vendor had picked up on it despite us working on it for a couple of years, so we went back to our own OpenCL driver. You know this.

[haagch]

But it is my desktop PC i do need X ???.... it is not a mining only server...

You can always just try it. I got GPU hangs when X was running. Maybe you'll have more luck.

I was talking about the OpenCL driver we ship with AMDGPU-PRO, which should work fine with your 8320 and RX470.

I doubt that. With OpenCL from amdgpu-pro 17.10.429170 ethminer is doing this:

  ℹ  15:18:15|ethminer  Found suitable OpenCL device [ Ellesmere ] with 3221225472  bytes of GPU memory
  ℹ  15:18:15|ethminer  Connecting to stratum server eu1.ethermine.org:4444
Could not resolve hosteu1.ethermine.org:4444, Operation canceled  ℹ  15:18:15|ethminer  Reconnecting in 3 seconds...
  ℹ  15:18:18|ethminer  Connecting to stratum server eu1.ethermine.org:4444
  ℹ  15:18:18|stratum   Connected to stratum server eu1.ethermine.org : 4444
  ℹ  15:18:18|stratum   Starting farm
  ℹ  15:18:18|stratum   Subscribed to stratum server
  ℹ  15:18:19|stratum   Authorized worker 0x96f2e7aedf15582c8321ff03ee0057e229ccca18.Miner1
  ℹ  15:18:19|stratum   Received new job #e1f1711e
 cl  15:18:19|cl-0      Set work. Header #e1f1711e… target 0000000112e0
 cl  15:18:19|cl-0      Initialising miner with seed #e7cc04b2…
  m  15:18:19|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
  m  15:18:19|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
  m  15:18:19|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
  m  15:18:19|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
  m  15:18:19|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
 cl  15:18:20|cl-0      Platform: AMD Accelerated Parallel Processing
 cl  15:18:20|cl-0      Device:   Ellesmere  / OpenCL 1.2 AMD-APP (2348.3)
  ✘  15:18:20|cl-0      clCreateCommandQueue ( -6 )
  ✘  15:18:20|cl-0      OpenCL Error: clEnqueueWriteBuffer -36
  m  15:18:20|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
  m  15:18:20|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
  m  15:18:20|ethminer  Mining on #e1f1711e… : 0.00MH/s [A0+0:R0+0:F0]
[...]
  ℹ  15:18:38|stratum   Received new job #d51a7ad6

I cut out some repeated messages but once it ends up receiving a second job after not being able to complete the first one because of clCreateCommandQueue ( -6 ) and OpenCL Error: clEnqueueWriteBuffer -36 it just sits there not doing any work.

Btw they renamed ethash_cl_miner_kernel.cl => CLMiner_kernel.cl so if you want to add "#pragma OPENCL EXTENSION cl_clang_storage_class_specifiers : enable" for clover, add it to that file.