[PATCH] D137542: [AMDGPU] Use aperture registers instead of S_GETREG

[Pierre van Houtryve via Phabricator via llvm-commits]

Pierre-vh updated this revision to Diff 474243.
Pierre-vh added a comment.

So I indeed checked ocltst and the previous version crashed.
Now it looks fine, but the verifier crashes in a lot of tests, especially GISel ones.
The issue is that we need to use this register as a 64 bit operand but it's a 32 bit register, so the verifier complains on S_MOV_B64.

How can I solve this? Is there another instruction I could use or do we have to change the RC of the src_shared/private_base register to 64 bit? (ideally it should be available for both, no?)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D137542/new/

https://reviews.llvm.org/D137542

Files:
  llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
  llvm/lib/Target/AMDGPU/SIDefines.h
  llvm/lib/Target/AMDGPU/SIISelLowering.cpp
  llvm/test/CodeGen/AMDGPU/GlobalISel/flat-scratch-init.gfx.ll
  llvm/test/CodeGen/AMDGPU/GlobalISel/implicit-kernarg-backend-usage-global-isel.ll
  llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-addrspacecast.mir
  llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.is.private.ll
  llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.is.shared.ll
  llvm/test/CodeGen/AMDGPU/addrspacecast-known-non-null.ll
  llvm/test/CodeGen/AMDGPU/addrspacecast.ll
  llvm/test/CodeGen/AMDGPU/atomicrmw-expand.ll
  llvm/test/CodeGen/AMDGPU/callee-special-input-sgprs-fixed-abi.ll
  llvm/test/CodeGen/AMDGPU/flat-scratch-init.ll
  llvm/test/CodeGen/AMDGPU/implicit-kernarg-backend-usage.ll
  llvm/test/CodeGen/AMDGPU/llvm.amdgcn.is.private.ll
  llvm/test/CodeGen/AMDGPU/llvm.amdgcn.is.shared.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D137542.474243.patch
Type: text/x-patch
Size: 49791 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20221109/0f880e87/attachment.bin>