[Libre-soc-dev] LLVM Adds A SPIR-V CPU Runner For Handling GPU Kernels On The CPU

[vivek pandya]

On Tue, Oct 27, 2020 at 4:45 AM Cole Poirier <colepoirier at gmail.com> wrote:

> On Monday, October 26, 2020, Luke Kenneth Casson Leighton <lkcl at lkcl.net>
> wrote:
>
> > On 10/26/20, Cole Poirier <colepoirier at gmail.com> wrote:
> > > Obviously not comparable to Kazan, but interesting for debugging
> purposes
> > > at least.
> >
> > no, it's exactly what we need.  it was one of the missing pieces, once
> > the Shader binary is compiled, how do you execute it (in the
> > CPU-which-is-also-a-GPU) and as best i can tell that's precisely what
> > this is.
>
>
>
> Oh that’s fantastic! Maybe I should reach out to its developers and see if
> they would be interested in doing similar work but paid by NLNET for our
> project? i.e. working on Kazan or its Mesa cpp equivalent?
>

Just be transparent, my work on this side has been stalled for a while, I
have tried to copy AMDGPU's NIR-to-LLVM converting code (either converter
is so big code or my engineering skills are so bad), I started with basic
arithmetic instructions and few intrinsics
however when trying a simple shader I found there are still lots of basic
intrinsics missing and I have to add them before I could see LLVM IR
generated.
if I could find some time and get something ready I will commit it.

Also there is bit difference in the approach taken by me and AMDGPU's LLVM
codegen . I am using a ORCJit instance and just trying to get LLVM IR for
now.
AMDGPU team has created a complete LLVM codegen pipeline where they can
even choose to schedule some optimization and generate AMD assembly.


-Vivek

>
> Cole
> _______________________________________________
> Libre-soc-dev mailing list
> Libre-soc-dev at lists.libre-soc.org
> http://lists.libre-soc.org/mailman/listinfo/libre-soc-dev
>