AMDGPU: Fix test relying on incompatible attributes

[arsenm]

Summary

This combination of amdgpu-waves-per-eu and
amdgpu-flat-work-group-size cannot be satisfied at the same time, so
this was using the default.

[arsenm]

Add test, although it's hard to directly observe this. Also add some documentation

[rampitec]

llvm/docs/AMDGPUUsage.rst

872 ↗

(On Diff #373223)

This needs reformat to keep column width.

[arsenm]

Correct patch

[rampitec]

A test only NFC change? LGTM anyway.

[arsenm]

4c2ee571482e066c0a524bd393a4c61aeaecff8a

llvm/test/CodeGen/AMDGPU/large-work-group-promote-alloca.ll

entry:
   %tmp3 = load i8, i8* %arrayidx12, align 1
   %arrayidx13 = getelementptr inbounds i8, i8 addrspace(1)* %out, i64 1
   store i8 %tmp3, i8 addrspace(1)* %arrayidx13, align 1
   ret void
 }
 
 attributes #0 = { nounwind "amdgpu-flat-work-group-size"="63,63" }
 attributes #1 = { nounwind "amdgpu-waves-per-eu"="1,3" "amdgpu-flat-work-group-size"="256,256" }
-attributes #2 = { nounwind "amdgpu-waves-per-eu"="1,1" "amdgpu-flat-work-group-size"="1024,1024" }
+attributes #2 = { nounwind "amdgpu-waves-per-eu"="1,9" "amdgpu-flat-work-group-size"="1024,1024" }
 attributes #3 = { nounwind "amdgpu-waves-per-eu"="1,10" }
 attributes #4 = { nounwind "amdgpu-waves-per-eu"="1,10" }
 attributes #5 = { nounwind "amdgpu-waves-per-eu"="1,6" "amdgpu-flat-work-group-size"="64,64" }
 attributes #6 = { nounwind "amdgpu-waves-per-eu"="1,8" "amdgpu-flat-work-group-size"="64,64" }
 attributes #7 = { nounwind "amdgpu-waves-per-eu"="1,9" "amdgpu-flat-work-group-size"="64,64" }