[llvm] [LV] codegen for tail-folded epilogue loop (PR #208764)

[David Sherwood via llvm-commits]

================
@@ -0,0 +1,104 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6
+; RUN: opt -p loop-vectorize -epilogue-tail-folding-policy=prefer-fold-tail \
+; RUN: -force-vector-width=16 -epilogue-vectorization-force-VF=8 -debug -mcpu=neoverse-v1 -S %s | FileCheck %s
+
+; RUN: opt -S -p loop-vectorize -debug --disable-output -epilogue-tail-folding-policy=prefer-fold-tail \
----------------
david-arm wrote:

You need to add `REQUIRES: asserts` to the top of the file when testing debug output.

Also, can you restrict the debug output to just the loop vectoriser output to make the test faster? For example, `-debug-only=loop-vectorize`

https://github.com/llvm/llvm-project/pull/208764