Messages

1

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: March 27, 2023, 02:00:19 pm »

By accident, I found this:
https://quickbuild.io/~raptor-engineering-public/+archive/ubuntu/rocm-power

I had a look at the packages included in the PPA. It seems that they have build the rocminfo tool but not any of the functionality to get ROCm OpenCL actually working. (Maybe I'm wrong there - I didn't look inside the deb packages, only at the package names.) So I tried it out on Gentoo and rocminfo now builds and runs with no modifications using the existing ebuild file.

rocm-opencl-runtime and rocclr still have arch-specific code for assembly and SSE intrinsics but maybe it is small enough to fix the remainder...

2

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: March 26, 2023, 05:47:47 am »

Still no progress on my own attempt, unfortunately.

On the plus side, at least I have managed to get POCL (dummy driver to run OpenCL code on the CPU), the hardware-independent parts of OpenCL (e.g. headers, ICD etc.), and PyOpenCL all keyworded for ppc64 on Gentoo.

3

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: March 26, 2023, 05:45:05 am »

Still no progress on my own attempt, unfortunately. AMD are releasing new versions faster than I can merge the little work I have already done.

I found this other page detailing another person's attempts to build ROCm on POWER9, which might be worth looking at.

 https://systems.nic.uoregon.edu/internal-wiki/index.php?title=Rocm_on_power9

4

Talos II / Re: mixing memory sizes on the same CPU?

« on: October 27, 2022, 09:47:39 am »

I personally have 2x 32 GB and 2x 64 GB connected to a single CPU (inserted in the recommended slots as per the Talos II motherboard user guide) and have had no problems using all the memory capacity. I haven't tried any memory benchmarks to detect if there is any slowdown.

5

GPU Compute / Accelerators / Re: GPU options as of December 2021

« on: May 08, 2022, 08:42:52 am »

Linux kernel 5.15 does not work but 5.10 (compiled with 4k pages) does.

That's the solution

Thanks @MPC7500, adding that boot option fixed the problem for me.

6

GPU Compute / Accelerators / Re: GPU options as of December 2021

« on: January 19, 2022, 12:35:11 am »

I have a W5500 which will probably have similar compatibility issues with a W5700.

Firmware is "amdgpu/navi14*.bin". When I started out I had some issues with some firmware releases being incompatible but every firmware revision since ~~ October 2021 has worked.

Linux kernel 5.15 does not work but 5.10 (compiled with 4k pages) does.

The Fedora 34 installer live CD (and a system installed to disk from that live CD without enabling network access so it used the same kernel and not an upgraded one) was able to use the GPU despite being 64k pages, but I haven't found any other 64k setup that gets the GPU working.

7

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: January 19, 2022, 12:23:02 am »

Any update on this? OpenCL is on my list of things I'd like to have working.

It took me a bit longer than I planned to get my development environment set up.

Some of the assembly sections (e.g. saving and restoring the stack pointer) were simple enough. Right now I am trying to work out how to port the assembly sections that handle the floating point exception status. (I haven't yet found any PowerPC assembly documentation that deals with this.)

If I get completely stuck then I will just post a version with broken floating point exception handling that crashes if there is a divide by zero etc.

8

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: October 29, 2021, 12:04:27 am »

I assume this is already outdated?
https://wiki.nikhef.nl/grid/AMD_GPU_on_IBM_POWER

Also I found this:
https://www.phoronix.com/scan.php?page=news_item&px=AMD-AOMP-On-POWER

Both those links refer to the AOMP compiler. I don't remember if I specifically tried AOMP, but my experience was that compilers like "llvm-roc" would build without any problems - the issues with x86 assembly or intrinsics are all in the "runtime" parts.

Once I have a full development environment set up to my liking then I will look into this in more detail.

9

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: October 24, 2021, 11:25:33 pm »

Thanks for the tips. While following the link you gave about memcpy I found this comment: https://stackoverflow.com/questions/8823267/linking-against-older-symbol-version-in-a-so-file?rq=1#comment100378482_8862631

This doesn't work if you're compiling on an architecture that wasn't built back in 2002 when x86-64 was first added - you'll get an error that the requested versioned symbols are not available.

I guess ppc64le didn't exist in 2002 so the 2.2.5 symbol version does not exist.

I will try omitting the .symver assembly and see how it behaves - if everything is being built from source then maybe it will be OK. If that doesn't work then I guess I can try adding a source file with an implementation of a simple non-IFUNC memcpy so that the ROCm library will use that instead of the glibc one.

10

GPU Compute / Accelerators / Re: AMD OpenCL / ROCm

« on: October 24, 2021, 09:49:19 am »

There are several short assembly snippets in "os_posix.cpp", one example is https://github.com/ROCm-Developer-Tools/ROCclr/blob/df870b565cf7f7d6d5fc8dd66aa07cd868874f9b/os/os_posix.cpp#L694

Another assembly snippet is https://github.com/RadeonOpenCompute/ROCm-OpenCL-Runtime/blob/12d926d06d36fe74876a82f8f8e1ce8ce7902728/amdocl/glibc_functions.cpp#L30 which seems to be requiring the memcpy ABI from a particular version of Glibc, with no explanation as to why that particular ABI is needed.

11

General CPU Discussion / Re: max memory speeds

« on: October 22, 2021, 01:51:10 pm »

I have a related question: I have noticed that sometimes 3200 MHz RAM is cheaper than the corresponding 2666 MHz DIMM. What happens if you use a 3200 MHz DIMM with a Talos / POWER9 board? Will every 3200 MHz chip support 2666 MHz operation or will it end up running at some slower speed like 2400 MHz or even 1600 MHz?

12

Talos II / Re: Order processing time

« on: October 22, 2021, 01:41:10 pm »

I ordered a motherboard + CPU set just over a month ago. It shipped within a week of payment clearing, but got stuck in customs for two weeks. If you are inside the US then you won't have to deal with customs of course.

13

GPU Compute / Accelerators / AMD OpenCL / ROCm

« on: October 22, 2021, 01:35:41 pm »

I'm trying to get OpenCL working on a recent AMD graphics card (Navi 14 chipset). I have tried downloading and compiling the Radeon Open Compute (ROCm) framework from https://github.com/RadeonOpenCompute .

There are a number of hardcoded -D__x86_64__ declarations in the CMakeLists.txt. After clearing those out there are maybe five or six places where functions include a couple of lines of x86 assembly, doing things that ought to be equally possible on POWER9.

Does anyone know of any existing efforts to port ROCm or another OpenCL runtime to POWER9?

Failing that, can someone suggest a good place to learn the basics of POWER9 assembly (whatever dialect is used by GCC)?