[all-commits] [llvm/llvm-project] 89a94b: [AMDGPU] Change FLAT SADDR to VADDR form in moveTo...

[Stanislav Mekhanoshin via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 89a94be16b01154e421d86d2e15e48d87b8d07db
      https://github.com/llvm/llvm-project/commit/89a94be16b01154e421d86d2e15e48d87b8d07db
  Author: Stanislav Mekhanoshin <Stanislav.Mekhanoshin at amd.com>
  Date:   2021-05-03 (Mon, 03 May 2021)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
    M llvm/lib/Target/AMDGPU/SIInstrInfo.h
    M llvm/lib/Target/AMDGPU/SIInstrInfo.td
    A llvm/test/CodeGen/AMDGPU/global-load-saddr-to-vaddr.ll
    A llvm/test/CodeGen/AMDGPU/move-load-addr-to-valu.mir

  Log Message:
  -----------
  [AMDGPU] Change FLAT SADDR to VADDR form in moveToVALU

Instead of legalizing saddr operand with a readfirstlane
when address is moved from SGPR to VGPR we can just
change the opcode.

Differential Revision: https://reviews.llvm.org/D101405