[all-commits] [llvm/llvm-project] b9ae01: AMDGPU: Add NextUseAnalysis Pass (#178873)

[macurtis-amd via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: b9ae01500d8b8c6c5b6f35d09bacd11325e67c8f
      https://github.com/llvm/llvm-project/commit/b9ae01500d8b8c6c5b6f35d09bacd11325e67c8f
  Author: macurtis-amd <macurtis at amd.com>
  Date:   2026-04-16 (Thu, 16 Apr 2026)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPU.h
    A llvm/lib/Target/AMDGPU/AMDGPUNextUseAnalysis.cpp
    A llvm/lib/Target/AMDGPU/AMDGPUNextUseAnalysis.h
    M llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
    M llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
    M llvm/lib/Target/AMDGPU/CMakeLists.txt
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/acyclic-014bb.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/acyclic-770bb.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/acyclic-cfg-with-self-loop.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/acyclic-phi-merge-distances.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/complex-acyclic-cfg-with-4-self-loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/complex-control-flow-11blocks.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/complex-control-flow-15blocks.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/complex-single-loop-a.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/complex-single-loop.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/double-nested-loops-complex-cfg.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/if_else_with_loops_nested_in_2_outer_loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/inner_cfg_in_2_nested_loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/loop_nested_in_3_outer_loops_complex_cfg.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/nested-loops-with-side-exits-a.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/sequence_2_loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/simple-loop-3blocks.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/spill-vreg-many-lanes.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_basic_case.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_do_not_spill_restore_inside_loop.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_emit_restore_in_common_dominator.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_emit_restore_in_loop_preheader1.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_emit_restore_in_loop_preheader2.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_emit_restore_in_loop_preheader3.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_emit_restore_in_loop_preheader4.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_keep_spilled_reg_live.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_multiple_spills1.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_multiple_spills2.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_multiple_spills3.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_nested_loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_spill_in_common_dominator_and_optimize_restores.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_spill_loop_livethrough_reg.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/test_ers_spill_loop_value_in_exit_block.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/three-tier-ranking-nested-loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/triple-nested-loops.mir
    A llvm/test/CodeGen/AMDGPU/NextUseAnalysis/two-sequential-loops.mir

  Log Message:
  -----------
  AMDGPU: Add NextUseAnalysis Pass (#178873)

Based on
- https://github.com/llvm/llvm-project/pull/156079 and
- https://github.com/llvm/llvm-project/pull/171520

See those PRs for background.

Provides a compatibility mode option
`--amdgpu-next-use-analysis-compatibility-mode` that produces results
that match either PR #156079 (`compute`) or PR #171520 (`graphics`).

Co-authored-by: alex-t <atimofee at amd.com>
Co-authored-by: Konstantina Mitropoulou <KonstantinaMitropoulou at amd.com>

---------

Co-authored-by: Konstantina Mitropoulou <KonstantinaMitropoulou at amd.com>



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