[llvm] af0b0ce - [AMDGPU] Fix SIFoldOperandsImpl::tryFoldZeroHighBits when met non-reg src1 operand. (#133761)

[via llvm-commits]

Author: Valery Pykhtin
Date: 2025-04-01T10:27:58+02:00
New Revision: af0b0ce665e2a36dd60c2aa70e78cc54cf20b7e3

URL: https://github.com/llvm/llvm-project/commit/af0b0ce665e2a36dd60c2aa70e78cc54cf20b7e3
DIFF: https://github.com/llvm/llvm-project/commit/af0b0ce665e2a36dd60c2aa70e78cc54cf20b7e3.diff

LOG: [AMDGPU] Fix SIFoldOperandsImpl::tryFoldZeroHighBits when met non-reg src1 operand. (#133761)

This happens when a constant is propagated to a V_AND 0xFFFF, reg
instruction.

Fixes failures like:

```
llc: /github/llvm-project/llvm/include/llvm/CodeGen/MachineOperand.h:366: llvm::Register llvm::MachineOperand::getReg() const: Assertion `isReg() && "This is not a register operand!"' failed.
Stack dump:
0.      Program arguments: /github/llvm-project/build/Debug/bin/llc -mtriple=amdgcn -mcpu=gfx1101 -verify-machineinstrs -run-pass si-fold-operands /github/llvm-project/llvm/test/CodeGen/AMDGPU/fold-zero-high-bits-skips-non-reg.mir -o -
1.      Running pass 'Function Pass Manager' on module '/github/llvm-project/llvm/test/CodeGen/AMDGPU/fold-zero-high-bits-skips-non-reg.mir'.
2.      Running pass 'SI Fold Operands' on function '@test_tryFoldZeroHighBits_skips_nonreg'
...
#12 0x00007f5a55005cfc llvm::MachineOperand::getReg() const /github/llvm-project/llvm/include/llvm/CodeGen/MachineOperand.h:0:5
#13 0x00007f5a555c6bf5 (anonymous namespace)::SIFoldOperandsImpl::tryFoldZeroHighBits(llvm::MachineInstr&) const /github/llvm-project/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp:1459:36
#14 0x00007f5a555c63ad (anonymous namespace)::SIFoldOperandsImpl::run(llvm::MachineFunction&) /github/llvm-project/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp:2455:11
#15 0x00007f5a555c6780 (anonymous namespace)::SIFoldOperandsLegacy::runOnMachineFunction
```

Added: 
    llvm/test/CodeGen/AMDGPU/fold-zero-high-bits-skips-non-reg.mir

Modified: 
    llvm/lib/Target/AMDGPU/SIFoldOperands.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp b/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
index cc15dd7cb495c..46bd5d8044c45 100644
--- a/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
+++ b/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
@@ -1453,7 +1453,7 @@ bool SIFoldOperandsImpl::tryFoldZeroHighBits(MachineInstr &MI) const {
     return false;
 
   std::optional<int64_t> Src0Imm = getImmOrMaterializedImm(MI.getOperand(1));
-  if (!Src0Imm || *Src0Imm != 0xffff)
+  if (!Src0Imm || *Src0Imm != 0xffff || !MI.getOperand(2).isReg())
     return false;
 
   Register Src1 = MI.getOperand(2).getReg();

diff  --git a/llvm/test/CodeGen/AMDGPU/fold-zero-high-bits-skips-non-reg.mir b/llvm/test/CodeGen/AMDGPU/fold-zero-high-bits-skips-non-reg.mir
new file mode 100644
index 0000000000000..b1aa88969c5bb
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/fold-zero-high-bits-skips-non-reg.mir
@@ -0,0 +1,17 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1101 -run-pass si-fold-operands %s -o - | FileCheck %s
+---
+name: test_tryFoldZeroHighBits_skips_nonreg
+tracksRegLiveness: true
+body: |
+  bb.0:
+    ; CHECK-LABEL: name: test_tryFoldZeroHighBits_skips_nonreg
+    ; CHECK: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
+    ; CHECK-NEXT: [[REG_SEQUENCE:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[V_MOV_B32_e32_]], %subreg.sub0, [[V_MOV_B32_e32_]], %subreg.sub1
+    ; CHECK-NEXT: [[V_AND_B32_e64_:%[0-9]+]]:vgpr_32 = V_AND_B32_e64 65535, 0, implicit $exec
+    ; CHECK-NEXT: S_NOP 0, implicit [[V_AND_B32_e64_]]
+  %0:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
+  %1:vreg_64 = REG_SEQUENCE %0, %subreg.sub0, %0, %subreg.sub1
+  %2:vgpr_32 = V_AND_B32_e64 65535, %1.sub0, implicit $exec
+  S_NOP 0, implicit %2
+...