[PATCH] D92483: AMDGPU - Use MUBUF instructions for global address space access

Mon Dec 21 15:52:41 PST 2020

t-tye added a comment.

In D92483#2466989 <https://reviews.llvm.org/D92483#2466989>, @rampitec wrote:

> In D92483#2466967 <https://reviews.llvm.org/D92483#2466967>, @t-tye wrote:
>
>> In D92483#2466961 <https://reviews.llvm.org/D92483#2466961>, @rampitec wrote:
>>
>>> In D92483#2466959 <https://reviews.llvm.org/D92483#2466959>, @t-tye wrote:
>>>
>>>> We did discuss this further and decided that gfx6* should report it supports amdhsa with the restriction that it does not support generic addresses. AMDGPUUsage already notes the restriction in all relevant sections. The LIT tests would also need updating accordingly. The COMgr tests would also need updating.
>>>
>>> AMDGPUUsage does not list amdhsa for all 3 SI targets, yet it is going to be accepted with this patch. If that is so then amdhsa has to be listed in the table for these targets.
>>
>> The table is indicating what runtimes support which ABIs. Since ROCm does not support gfx6* it would still be correct to not list it in the table. It may be useful to have another column to indicate which ABIs are supported by which processors.
>
> Pfff... I do not like it, but OK.

I am not completely happy with it either, but wanted to address questions of which runtimes support which processors. So I looked at what ROCr, VDI and PAL runtimes currently support and added that to the column that already existed but was woefully out of date and did not include PAL information. MESA information is also still missing.

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D92483/new/

https://reviews.llvm.org/D92483