compute-runtime VS opencl-intercept-layer

Well... Fuck! https://github.com/intel/compute-runtime/issues/643

For what it's worth, I was able to get IOMMU enabled and iGPU passthrough working for Plex on an Ubuntu 22.04 LXC container with a fresh install of Proxmox 8.0.3 (kenrel 6.2.16-3-pve) on an Intel i5-13400, ASRock B760M Pro RS/D4, also using Intel drivers released early today. I largely followed this guide.

For this little project unless someone chimes in with experience or can point to regarding the RX 6600 XT / 6650 XT / perhaps 7600 I might go with https://github.com/intel/compute-runtime . (They don't have perfect documentation re supported gpus either, the readme table doesn't list DG2, you need to go to releases to see that) . Phoronix reported it as pretty stable when they tested it with kernel 6.2 in March.

Update: I was able to fix my issue. I'm using Ubuntu 22.04.2 LTS and have the newest available kernel, 6.3.1. Installing the drivers via apt does not work, instead I needed to use https://github.com/intel/compute-runtime/releases/

Drivers: https://github.com/intel/compute-runtime/releases

wget https://github.com/intel/compute-runtime/releases/download/22.43.24595.30/intel-level-zero-gpu-dbgsym_1.3.24595.30_amd64.ddeb

For OpenCL HDR/DV tone-mapping, install the extra compute-runtime if running on host.

Somewhat selfishly, one of the tools I see missing from your guide is the OpenCL Intercept Layer. This is an open-source tool we've developed and we use it every day to debug and analyze OpenCL programs.

I'd also suggest taking a look at the OpenCL Intercept Layer too, if you haven't already. It can do things like logging OpenCL errors, calling clFlush or clFinish after each enqueue, and more, all without modifying source or rebuilding.

The OpenCL Intercept Layer can get it for you (documentation link).

Are you sure your program is executing completely and without errors? I'd strongly recommend using a tool like the OpenCL Intercept Layer to verify that everything is running correctly. You can even it to dump your input and result matrices to verify they are correct, if desired.

We use the OpenCL Intercept Layer extensively. It's open-source, vendor-independent, and cross-platform. I wrote a tutorial to demonstrate common usages, if you want to see what it can do.

At some risk of redundancy, the one tool I would definitely recommend is the OpenCL Intercept Layer, since it will make it a lot easier to debug problems when they inevitably occur, or to optimize your program once it's working. Good luck and have fun!