[llvm] [LV] Prefer fixed VFs for epilogue vectorization (PR #211814)

[David Sherwood via llvm-commits]

================
@@ -16,7 +16,7 @@ define void @iv_casts(ptr %dst, ptr %src, i32 %x, i64 %N) #0 {
 ; DEFAULT-NEXT:    [[TMP2:%.*]] = shl nuw i64 [[TMP1]], 3
 ; DEFAULT-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
 ; DEFAULT-NEXT:    [[TMP7:%.*]] = call i64 @llvm.vscale.i64()
-; DEFAULT-NEXT:    [[TMP31:%.*]] = shl nuw i64 [[TMP7]], 3
+; DEFAULT-NEXT:    [[TMP9:%.*]] = shl nuw i64 [[TMP7]], 3
----------------
david-arm wrote:

Changes in this file can be reverted.

https://github.com/llvm/llvm-project/pull/211814