[all-commits] [llvm/llvm-project] d89252: [AMDGPU] Break-up large PHIs for DAGISel

[Pierre van Houtryve via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: d892521076fec823687b0608f9897256cecf5535
      https://github.com/llvm/llvm-project/commit/d892521076fec823687b0608f9897256cecf5535
  Author: pvanhout <pierre.vanhoutryve at amd.com>
  Date:   2023-03-28 (Tue, 28 Mar 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
    M llvm/test/CodeGen/AMDGPU/agpr-copy-no-free-registers.ll
    A llvm/test/CodeGen/AMDGPU/amdgpu-codegenprepare-break-large-phis.ll
    M llvm/test/CodeGen/AMDGPU/collapse-endcf.ll
    M llvm/test/CodeGen/AMDGPU/cross-block-use-is-not-abi-copy.ll
    M llvm/test/CodeGen/AMDGPU/debug-value.ll
    M llvm/test/CodeGen/AMDGPU/early-if-convert.ll
    M llvm/test/CodeGen/AMDGPU/extract-subvector-16bit.ll
    M llvm/test/CodeGen/AMDGPU/insert_vector_elt.ll
    M llvm/test/CodeGen/AMDGPU/loop-live-out-copy-undef-subrange.ll
    M llvm/test/CodeGen/AMDGPU/mfma-loop.ll
    M llvm/test/CodeGen/AMDGPU/subreg-coalescer-undef-use.ll
    M llvm/test/CodeGen/AMDGPU/tuple-allocation-failure.ll
    M llvm/test/CodeGen/AMDGPU/v1024.ll
    M llvm/test/CodeGen/AMDGPU/wave32.ll
    M llvm/test/CodeGen/AMDGPU/wqm.ll

  Log Message:
  -----------
  [AMDGPU] Break-up large PHIs for DAGISel

DAGISel uses CopyToReg/CopyFromReg to lower PHI nodes. With large PHIs, this can result in poor codegen.
This is because it introduces a need to have a build_vector before copying the PHI value, and that build_vector may have many undef elements. This can cause very high register pressure and abnormal stack usage in some cases.

This scalarization/phi "break-up" can be easily tuned/disabled through CL options in case it's not beneficial for some users.
It's also only enabled for DAGIsel and GlobalISel handles PHIs much better (as it works on the whole function).

This can both scalarize (break a vector into its elements) and simplify (break a vector into smaller, more manageable subvectors) PHIs.

Fixes SWDEV-321581

Reviewed By: kzhuravl

Differential Revision: https://reviews.llvm.org/D143731