[all-commits] [llvm/llvm-project] c8d8eb: [AMDGPU] Allow hoising of V_READFIRSTLANE_B32 for ...

[Stanislav Mekhanoshin via All-commits]

  Branch: refs/heads/users/rampitec/readanylane
  Home:   https://github.com/llvm/llvm-project
  Commit: c8d8ebbb5cd9c195c94879b281c28cdbfb5a77a1
      https://github.com/llvm/llvm-project/commit/c8d8ebbb5cd9c195c94879b281c28cdbfb5a77a1
  Author: Stanislav Mekhanoshin <Stanislav.Mekhanoshin at amd.com>
  Date:   2026-02-10 (Tue, 10 Feb 2026)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIISelLowering.cpp
    M llvm/lib/Target/AMDGPU/VOP1Instructions.td
    M llvm/test/CodeGen/AMDGPU/dag-preserve-disjoint-flag.ll
    A llvm/test/CodeGen/AMDGPU/hoist-readfirstlane.ll
    M llvm/test/CodeGen/AMDGPU/isel-amdgcn-cs-chain-intrinsic-w32.ll
    M llvm/test/CodeGen/AMDGPU/isel-amdgcn-cs-chain-intrinsic-w64.ll
    M llvm/test/CodeGen/AMDGPU/isel-amdgpu-cs-chain-intrinsic-dyn-vgpr-w32.ll
    M llvm/test/CodeGen/AMDGPU/llvm.amdgcn.make.buffer.rsrc.ll

  Log Message:
  -----------
  [AMDGPU] Allow hoising of V_READFIRSTLANE_B32 for uniform operand

readfirstlane can be moved across control flow for uniform inputs.
The MachineInstr::NoConvergent attribute allows hoisting
which is otherwise prohibited for a convergent instruction.



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