[llvm] [VPlan] Fold NOT into predicate of wide compares. (PR #129430)
David Sherwood via llvm-commits
llvm-commits at lists.llvm.org
Wed Mar 5 08:37:55 PST 2025
================
@@ -1149,13 +1147,12 @@ define void @ptr_induction_remove_dead_recipe(ptr %start, ptr %end) {
; CHECK-NEXT: CLONE ir<%ptr.iv.next> = getelementptr inbounds vp<[[PTR_IV]]>, ir<-1>
; CHECK-NEXT: vp<[[VEC_PTR:%.+]]> = reverse-vector-pointer inbounds ir<%ptr.iv.next>, vp<[[VF]]>
; CHECK-NEXT: WIDEN ir<%l> = load vp<[[VEC_PTR]]>
-; CHECK-NEXT: WIDEN ir<%c.1> = icmp eq ir<%l>, ir<0>
-; CHECK-NEXT: EMIT vp<[[NEG:%.+]]> = not ir<%c.1>
+; CHECK-NEXT: WIDEN ir<%c.1> = icmp ne ir<%l>, ir<0>
; CHECK-NEXT: Successor(s): pred.store
; CHECK-EMPTY:
; CHECK-NEXT: <xVFxUF> pred.store: {
; CHECK-NEXT: pred.store.entry:
-; CHECK-NEXT: BRANCH-ON-MASK vp<[[NEG]]>
+; CHECK-NEXT: BRANCH-ON-MASK ir<%c.1>
----------------
david-arm wrote:
nit: In the changes to llvm/test/Transforms/LoopVectorize/AArch64/masked-op-cost.ll you removed any reference to IR names. Not sure if you also want to do that here?
https://github.com/llvm/llvm-project/pull/129430
More information about the llvm-commits
mailing list