https://github.com/arsenm updated https://github.com/llvm/llvm-project/pull/153023
>From 37c81647c1b025d4c33ebc02cd83e1e084131b6e Mon Sep 17 00:00:00 2001 From: Matt Arsenault <[email protected]> Date: Mon, 11 Aug 2025 10:47:44 +0900 Subject: [PATCH 1/2] AMDGPU: Handle rewriting VGPR MFMA fed from AGPR copy Previously we handled the inverse situation only. >From be46142733bec5b9fe7bc13c15542088740443be Mon Sep 17 00:00:00 2001 From: Matt Arsenault <[email protected]> Date: Mon, 11 Aug 2025 18:22:09 +0900 Subject: [PATCH 2/2] AMDGPU: Handle V->A MFMA copy from case with immediate src2 Handle a special case for copies from AGPR VGPR on the MFMA inputs. If the "input" is really a subregister def, we will not see the usual copy to VGPR for src2, only the read of the subregister def. Not sure if this pattern appears in practice. --- llvm/lib/Target/AMDGPU/AMDGPURewriteAGPRCopyMFMA.cpp | 5 ++++- .../CodeGen/AMDGPU/rewrite-vgpr-mfma-to-agpr-copy-from.mir | 4 ++-- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/llvm/lib/Target/AMDGPU/AMDGPURewriteAGPRCopyMFMA.cpp b/llvm/lib/Target/AMDGPU/AMDGPURewriteAGPRCopyMFMA.cpp index aaecfa1e33f06..5bda81241722b 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPURewriteAGPRCopyMFMA.cpp +++ b/llvm/lib/Target/AMDGPU/AMDGPURewriteAGPRCopyMFMA.cpp @@ -370,8 +370,11 @@ bool AMDGPURewriteAGPRCopyMFMAImpl::tryFoldCopiesFromAGPR( Register CopyDstReg = UseMI.getOperand(0).getReg(); if (!CopyDstReg.isVirtual()) continue; + for (MachineOperand &CopyUseMO : MRI.reg_nodbg_operands(CopyDstReg)) { + if (!CopyUseMO.readsReg()) + continue; - for (MachineInstr &CopyUseMI : MRI.use_instructions(CopyDstReg)) { + MachineInstr &CopyUseMI = *CopyUseMO.getParent(); if (isRewriteCandidate(CopyUseMI)) { if (tryReassigningMFMAChain(CopyUseMI, CopyDstReg, VRM.getPhys(CopyDstReg))) diff --git a/llvm/test/CodeGen/AMDGPU/rewrite-vgpr-mfma-to-agpr-copy-from.mir b/llvm/test/CodeGen/AMDGPU/rewrite-vgpr-mfma-to-agpr-copy-from.mir index 1c5e0e362e359..bcd0e027b209e 100644 --- a/llvm/test/CodeGen/AMDGPU/rewrite-vgpr-mfma-to-agpr-copy-from.mir +++ b/llvm/test/CodeGen/AMDGPU/rewrite-vgpr-mfma-to-agpr-copy-from.mir @@ -187,8 +187,8 @@ body: | ; CHECK-NEXT: [[COPY1:%[0-9]+]]:av_64_align2 = COPY $vgpr0_vgpr1 ; CHECK-NEXT: [[COPY2:%[0-9]+]]:av_64_align2 = COPY $vgpr2_vgpr3 ; CHECK-NEXT: [[GLOBAL_LOAD_DWORDX4_:%[0-9]+]]:areg_128_align2 = GLOBAL_LOAD_DWORDX4 [[COPY]], 0, 0, implicit $exec :: (load (s128), addrspace 1) - ; CHECK-NEXT: [[COPY3:%[0-9]+]]:vreg_128_align2 = COPY [[GLOBAL_LOAD_DWORDX4_]] - ; CHECK-NEXT: [[COPY3:%[0-9]+]].sub0_sub1:vreg_128_align2 = V_MFMA_F64_4X4X4F64_vgprcd_e64 [[COPY1]], [[COPY2]], 0, 0, 0, 0, implicit $mode, implicit $exec + ; CHECK-NEXT: [[COPY3:%[0-9]+]]:areg_128_align2 = COPY [[GLOBAL_LOAD_DWORDX4_]] + ; CHECK-NEXT: [[COPY3:%[0-9]+]].sub0_sub1:areg_128_align2 = V_MFMA_F64_4X4X4F64_e64 [[COPY1]], [[COPY2]], 0, 0, 0, 0, implicit $mode, implicit $exec ; CHECK-NEXT: GLOBAL_STORE_DWORDX4 [[COPY]], [[COPY3]], 0, 0, implicit $exec :: (store (s128), addrspace 1) ; CHECK-NEXT: SI_RETURN %0:vreg_64_align2 = COPY $vgpr4_vgpr5 _______________________________________________ llvm-branch-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
