[llvm] [VPlan] Move addExplicitVectorLength to tryToBuildVPlanWithVPRecipes (PR #166164)

[Florian Hahn via llvm-commits]

================
@@ -42,10 +42,8 @@ define i32 @reduction(ptr %a, i64 %n, i32 %start) {
 ; IF-EVL-OUTLOOP-NEXT:    WIDEN-REDUCTION-PHI ir<[[RDX_PHI:%.+]]> = phi vp<[[RDX_START]]>, vp<[[RDX_SELECT:%.+]]>
 ; IF-EVL-OUTLOOP-NEXT:    EMIT-SCALAR vp<[[AVL:%.+]]> = phi [ ir<%n>, vector.ph ], [ vp<[[AVL_NEXT:%.+]]>, vector.body ] 
 ; IF-EVL-OUTLOOP-NEXT:    EMIT-SCALAR vp<[[EVL:%.+]]> = EXPLICIT-VECTOR-LENGTH vp<[[AVL]]>
-; IF-EVL-OUTLOOP-NEXT:    vp<[[ST:%[0-9]+]]> = SCALAR-STEPS vp<[[EVL_PHI]]>, ir<1>, vp<[[EVL]]>
-; IF-EVL-OUTLOOP-NEXT:    CLONE ir<[[GEP1:%.+]]> = getelementptr inbounds ir<%a>, vp<[[ST]]>
-; IF-EVL-OUTLOOP-NEXT:    vp<[[PTR1:%[0-9]+]]> = vector-pointer ir<[[GEP1]]>
----------------
fhahn wrote:

Do you know why did we not remove the redundant vector-pointers before? It is not immediately clear how this is a consequence of moving the transform

https://github.com/llvm/llvm-project/pull/166164