[llvm] [VPlan] Impl VPlan-based pattern match for ExtendedRed and MulAccRed (NFCI) (PR #113903)

[Alexey Bataev via llvm-commits]

================
@@ -2196,30 +2197,49 @@ InstructionCost VPReductionRecipe::computeCost(ElementCount VF,
   TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput;
   unsigned Opcode = RdxDesc.getOpcode();
 
-  // TODO: Support any-of and in-loop reductions.
+  // TODO: Support any-of reductions.
   assert(
       (!RecurrenceDescriptor::isAnyOfRecurrenceKind(RdxKind) ||
        ForceTargetInstructionCost.getNumOccurrences() > 0) &&
       "Any-of reduction not implemented in VPlan-based cost model currently.");
-  assert(
-      (!cast<VPReductionPHIRecipe>(getOperand(0))->isInLoop() ||
-       ForceTargetInstructionCost.getNumOccurrences() > 0) &&
-      "In-loop reduction not implemented in VPlan-based cost model currently.");
 
   assert(ElementTy->getTypeID() == RdxDesc.getRecurrenceType()->getTypeID() &&
          "Inferred type and recurrence type mismatch.");
 
-  // Cost = Reduction cost + BinOp cost
-  InstructionCost Cost =
-      Ctx.TTI.getArithmeticInstrCost(Opcode, ElementTy, CostKind);
+  // Note that TTI should model the cost of moving result to the scalar register
+  // and the binOp cost in the getReductionCost().
   if (RecurrenceDescriptor::isMinMaxRecurrenceKind(RdxKind)) {
     Intrinsic::ID Id = getMinMaxReductionIntrinsicOp(RdxKind);
-    return Cost + Ctx.TTI.getMinMaxReductionCost(
-                      Id, VectorTy, RdxDesc.getFastMathFlags(), CostKind);
+    return Ctx.TTI.getMinMaxReductionCost(Id, VectorTy,
+                                          RdxDesc.getFastMathFlags(), CostKind);
   }
 
-  return Cost + Ctx.TTI.getArithmeticReductionCost(
-                    Opcode, VectorTy, RdxDesc.getFastMathFlags(), CostKind);
+  return Ctx.TTI.getArithmeticReductionCost(
+      Opcode, VectorTy, RdxDesc.getFastMathFlags(), CostKind);
+}
+
+InstructionCost
+VPExtendedReductionRecipe::computeCost(ElementCount VF,
+                                       VPCostContext &Ctx) const {
+  const RecurrenceDescriptor &RdxDesc = getRecurrenceDescriptor();
+  unsigned Opcode = RdxDesc.getOpcode();
+  Type *RedTy = Ctx.Types.inferScalarType(this);
+  auto *SrcVecTy =
+      cast<VectorType>(ToVectorTy(Ctx.Types.inferScalarType(getVecOp()), VF));
+  TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput;
+
+  return Ctx.TTI.getExtendedReductionCost(Opcode, isZExt(), RedTy, SrcVecTy,
+                                          RdxDesc.getFastMathFlags(), CostKind);
+}
+
+InstructionCost VPMulAccRecipe::computeCost(ElementCount VF,
+                                            VPCostContext &Ctx) const {
+  Type *RedTy = Ctx.Types.inferScalarType(this);
+  VectorType *SrcVecTy =
+      cast<VectorType>(ToVectorTy(Ctx.Types.inferScalarType(getVecOp0()), VF));
----------------
alexey-bataev wrote:

```suggestion
  auto *SrcVecTy =
      cast<VectorType>(ToVectorTy(Ctx.Types.inferScalarType(getVecOp0()), VF));
```

https://github.com/llvm/llvm-project/pull/113903