[llvm] [AMDGPU] [GlobalIsel] Combine Fmul with Select into ldexp instruction. (PR #120104)

[Jay Foad via llvm-commits]

================
@@ -445,3 +452,69 @@ void AMDGPUCombinerHelper::applyExpandPromotedF16FMed3(MachineInstr &MI,
   Builder.buildFMinNumIEEE(MI.getOperand(0), B1, C1);
   MI.eraseFromParent();
 }
+
+bool AMDGPUCombinerHelper::matchCombineFmulWithSelectToFldexp(
+    MachineInstr &MI, MachineInstr &Sel,
+    std::function<void(MachineIRBuilder &)> &MatchInfo) {
+  assert(MI.getOpcode() == TargetOpcode::G_FMUL);
+  assert(Sel.getOpcode() == TargetOpcode::G_SELECT);
+  assert(MI.getOperand(2).getReg() == Sel.getOperand(0).getReg());
+
+  Register Dst = MI.getOperand(0).getReg();
+  LLT DestTy = MRI.getType(Dst);
+  LLT ScalarDestTy = DestTy.getScalarType();
+
+  if ((ScalarDestTy == LLT::float64() || ScalarDestTy == LLT::float32() ||
----------------
jayfoad wrote:

Invert this condition and return false early if it is not satisfied.

https://github.com/llvm/llvm-project/pull/120104