[llvm] [VPlan] Add VPInstruction::StepVector and use it in VPWidenIntOrFpInductionRecipe (PR #129508)

[Luke Lau via llvm-commits]

================
@@ -2410,6 +2410,23 @@ void VPlanTransforms::convertToConcreteRecipes(VPlan &Plan,
         continue;
       }
 
+      if (auto *IVR = dyn_cast<VPWidenIntOrFpInductionRecipe>(&R)) {
+        // Infer an up-to-date type since
+        // optimizeVectorInductionWidthForTCAndVFUF may have truncated the start
+        // and step values.
+        Type *Ty = TypeInfo.inferScalarType(IVR->getStartValue());
+        if (TruncInst *Trunc = IVR->getTruncInst())
+          Ty = Trunc->getType();
+        if (Ty->isFloatingPointTy())
+          Ty = IntegerType::get(Ty->getContext(), Ty->getScalarSizeInBits());
+        VPInstruction *StepVector = new VPInstructionWithType(
+            VPInstruction::StepVector, {}, Ty, R.getDebugLoc());
----------------
lukel97 wrote:

Oh I forgot you had added a builder method for VPInstructionWithType, I've used it in 83fafe3e19868c0fd73c6a50208775c4ea511f80. I had to edit it though to only call the FMF constructor if FMFs were present, otherwise it triggers the `assert(isFPMathOp() && "this op can't take fast-math flags");` in VPInstruction because StepVector doesn't take the flags

https://github.com/llvm/llvm-project/pull/129508