[clang] [llvm] [AArch64] Add intrinsics support for SVE2p2 instructions (PR #163575)

[via llvm-commits]

================
@@ -87,6 +87,179 @@ define <vscale x 8 x bfloat> @compact_bf16(<vscale x 8 x i1> %pg, <vscale x 8 x
   ret <vscale x 8 x bfloat> %out
 }
 
----------------
CarolineConcatto wrote:

I believe we've missed the tests for COMPACT  byte and half word

https://github.com/llvm/llvm-project/pull/163575