From f4c952a9c32eb12080c02bc29c027686dfb7f940 Mon Sep 17 00:00:00 2001
From: Paul Trojahn <paul.trojahn@amd.com>
Date: Mon, 14 Jul 2025 11:21:00 +0200
Subject: [PATCH] [AMDGPU] Check legality of both operands before commute

When trying to fold an SGPR into a DPP add, si-fold-operands correctly
realizes that this is not possible and then tries to commute which
mistakenly succeeds, creating a dpp add with two SGPRs. We need to
check both operands if they are legal in their new position.

This crashes a test in triton on gfx12:
ttps://github.com/triton-lang/triton/blob/345c633787e90a7f94864de3035346eb5de1781f/python/test/unit/language/test_core.py#L2718
---
 llvm/lib/Target/AMDGPU/SIInstrInfo.cpp        |  6 +++--
 .../test/CodeGen/AMDGPU/fold-commute-sgpr.mir | 24 +++++++++++++++++++
 2 files changed, 28 insertions(+), 2 deletions(-)
 create mode 100644 llvm/test/CodeGen/AMDGPU/fold-commute-sgpr.mir

diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
index 4c5f938831243..8e985ea6af53f 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -2807,12 +2807,14 @@ bool SIInstrInfo::isLegalToSwap(const MachineInstr &MI, unsigned OpIdx0,
   if ((int)OpIdx1 != Src0Idx && MO0->isReg()) {
     if (!DefinedRC1)
       return OpInfo1.OperandType == MCOI::OPERAND_UNKNOWN;
-    return isLegalRegOperand(MI, OpIdx1, *MO0);
+    return isLegalRegOperand(MI, OpIdx1, *MO0) &&
+           (!MO1->isReg() || isLegalRegOperand(MI, OpIdx0, *MO1));
   }
   if ((int)OpIdx0 != Src0Idx && MO1->isReg()) {
     if (!DefinedRC0)
       return OpInfo0.OperandType == MCOI::OPERAND_UNKNOWN;
-    return isLegalRegOperand(MI, OpIdx0, *MO1);
+    return (!MO0->isReg() || isLegalRegOperand(MI, OpIdx1, *MO0)) &&
+           isLegalRegOperand(MI, OpIdx0, *MO1);
   }
 
   // No need to check 64-bit literals since swapping does not bring new
diff --git a/llvm/test/CodeGen/AMDGPU/fold-commute-sgpr.mir b/llvm/test/CodeGen/AMDGPU/fold-commute-sgpr.mir
new file mode 100644
index 0000000000000..c6bc248f13388
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/fold-commute-sgpr.mir
@@ -0,0 +1,24 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5
+# RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 -run-pass=si-fold-operands -verify-machineinstrs -o - %s | FileCheck %s
+
+---
+name: fold_commute_sgprs
+body:             |
+  bb.0:
+    liveins: $sgpr0, $sgpr1
+    ; CHECK-LABEL: name: fold_commute_sgprs
+    ; CHECK: liveins: $sgpr0, $sgpr1
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr0
+    ; CHECK-NEXT: [[DEF:%[0-9]+]]:sreg_32 = IMPLICIT_DEF
+    ; CHECK-NEXT: [[COPY1:%[0-9]+]]:vgpr_32 = COPY [[DEF]]
+    ; CHECK-NEXT: [[COPY2:%[0-9]+]]:vgpr_32 = COPY [[COPY]]
+    ; CHECK-NEXT: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr1
+    ; CHECK-NEXT: [[V_ADD_NC_U16_fake16_e64_dpp:%[0-9]+]]:vgpr_32 = V_ADD_NC_U16_fake16_e64_dpp [[COPY1]], 0, [[COPY2]], 0, [[COPY3]], 0, 0, 280, 15, 15, 1, implicit $exec
+    %0:sreg_32 = COPY $sgpr0
+    %1:sreg_32 = IMPLICIT_DEF
+    %2:vgpr_32 = COPY %1:sreg_32
+    %3:vgpr_32 = COPY %0:sreg_32
+    %4:sreg_32 = COPY $sgpr1
+    %5:vgpr_32 = V_ADD_NC_U16_fake16_e64_dpp %2:vgpr_32, 0, %3:vgpr_32, 0, %4:sreg_32, 0, 0, 280, 15, 15, 1, implicit $exec
+...