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

[Florian Hahn via llvm-commits]

================
@@ -2376,30 +2382,59 @@ InstructionCost VPReductionRecipe::computeCost(ElementCount VF,
   unsigned Opcode = RdxDesc.getOpcode();
   FastMathFlags FMFs = getFastMathFlags();
 
-  // 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, Ctx.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, FMFs, Ctx.CostKind);
+    return Ctx.TTI.getMinMaxReductionCost(Id, VectorTy, FMFs, Ctx.CostKind);
   }
 
-  return Cost + Ctx.TTI.getArithmeticReductionCost(Opcode, VectorTy, FMFs,
-                                                   Ctx.CostKind);
+  if (ElementTy->isFloatingPointTy())
+    return Ctx.TTI.getArithmeticReductionCost(Opcode, VectorTy, FMFs,
+                                              Ctx.CostKind);
+  // Cannot get correct cost when quering TTI with FMFs not contains `reassoc`
+  // for non-FP reductions.
----------------
fhahn wrote:

I am not following the comment/code entirely. IIUC not flags should be set in FMFs, but there is different costs for integer inductions depending on whether empty FMFs or `std::nullopt` is passed?

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