[llvm] [LV] Support multiplies by constants when forming scaled reductions. (PR #161092)

[David Sherwood via llvm-commits]

================
@@ -1741,6 +1741,16 @@ void LoopVectorizationPlanner::printPlans(raw_ostream &O) {
 }
 #endif
 
+bool llvm::canConstantBeExtended(const ConstantInt *CI, Type *NarrowType,
+                                 TTI::PartialReductionExtendKind ExtKind) {
+  APInt TruncatedVal = CI->getValue().trunc(NarrowType->getScalarSizeInBits());
----------------
david-arm wrote:

Do we have a negative partial reduction test for `sext i8 128 to i32`?

https://github.com/llvm/llvm-project/pull/161092