[all-commits] [llvm/llvm-project] 944f42: [AMDGPU] Improve the lowering of raw_buffer_load_{...

[Konstantina Mitropoulou via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 944f429b210d0d94171dc5df3068d4f1e8fb5d80
      https://github.com/llvm/llvm-project/commit/944f429b210d0d94171dc5df3068d4f1e8fb5d80
  Author: Konstantina Mitropoulou <Konstantina.Mitropoulou at amd.com>
  Date:   2023-02-22 (Wed, 22 Feb 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUCombine.td
    M llvm/lib/Target/AMDGPU/AMDGPUPostLegalizerCombiner.cpp
    M llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.raw.buffer.load.ll
    M llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.ll

  Log Message:
  -----------
  [AMDGPU] Improve the lowering of raw_buffer_load_{i8,i16} and struct_buffer_load_{i8,i16} intrinsics

Currently, raw_buffer_load_{i8,i16} and struct_buffer_load_{i8,i16}
intrinsics are lowered as buffer_load_{u8,u16}. This patch combines
buffer_load_{u8,u16} and sign extension instructions in order to
generate buffer_load_{i8,i16} instructions.

Reviewed By: foad

Differential Revision: https://reviews.llvm.org/D144313