[all-commits] [llvm/llvm-project] ad435b: [clang][CodeGen][SPIR-V][AMDGPU] Tweak AMDGCNSPIRV...

[Alex Voicu via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: ad435bcc14f42dc97286c717cd12446a0facb2ee
      https://github.com/llvm/llvm-project/commit/ad435bcc14f42dc97286c717cd12446a0facb2ee
  Author: Alex Voicu <alexandru.voicu at amd.com>
  Date:   2024-08-21 (Wed, 21 Aug 2024)

  Changed paths:
    M clang/lib/CodeGen/Targets/SPIR.cpp
    M clang/test/CodeGenCUDA/amdgpu-kernel-arg-pointer-type.cu
    M clang/test/CodeGenCUDA/kernel-args.cu

  Log Message:
  -----------
  [clang][CodeGen][SPIR-V][AMDGPU] Tweak AMDGCNSPIRV ABI to allow for the correct handling of aggregates passed to kernels / functions. (#102776)

The AMDGPU kernel ABI is not directly representable in SPIR-V, since it
relies on passing aggregates `byref`, and SPIR-V only encodes `byval`
(which the AMDGPU BE disallows for kernel arguments). As a temporary
solution to this mismatch, we add special handling for AMDGCN flavoured
SPIR-V, whereby aggregates are passed as direct, both to kernels and to
normal functions. This is not ideal (there are pathological cases where
performance is heavily impacted), but empirically robust and guaranteed
to work as the AMDGPU BE retains handling of `direct` passing for legacy
reasons.

We will revisit this in the future, but as it stands it is enough to
pass a wide array of integration tests and generates correct SPIR-V and
correct reverse translation into LLVM IR. The
amdgpu-kernel-arg-pointer-type test is updated via the automated script,
and thus becomes quite noisy.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications