[llvm] [LV][VPlan] Introduce VPScalarStore recipe to handle scalar stores in exit block. (PR #93148)

[Mel Chen via llvm-commits]

================
@@ -625,10 +625,10 @@ exit:
 define void @reduc_mul_add_store_same_ptr(ptr %dst, ptr readonly %src) {
 ; CHECK-LABEL: define void @reduc_mul_add_store_same_ptr
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[TMP4:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP3:%.*]])
-; CHECK-NEXT:    store i32 [[TMP4]], ptr %dst, align 4
-; CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.vector.reduce.mul.v4i32(<4 x i32> [[TMP1:%.*]])
-; CHECK-NEXT:    store i32 [[TMP2]], ptr %dst, align 4
----------------
Mel-Chen wrote:

...That's a good question. I think this was the incorrect order of stores. 
@fhahn @igogo-x86 Could you help verify it again?

https://github.com/llvm/llvm-project/pull/93148