[llvm] [Analysis] isTriviallyVectorizable - add Intrinsic::clmul along with vectorisation tests (PR #180014)

[David Green via llvm-commits]

================
@@ -0,0 +1,97 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6
+; RUN: opt -passes=loop-vectorize -mtriple=aarch64 -mattr=+crypto,+sve -scalable-vectorization=on -force-vector-width=2 -S %s | FileCheck %s
----------------
davemgreen wrote:

Hi - It was a v2i64 that we ran into on AArch64 that didn't work. We are working through the details but it looks like a fair few architectures were crashing on the same types.

These intrinsics are fairly new, it might be worth going through a number of archs to make sure nothing fails with them.

https://github.com/llvm/llvm-project/pull/180014