[llvm] [AArch64] Use 0-cycle reg2reg MOVs for FPR32, FPR16, FPR8 (PR #144152)

[Jon Roelofs via llvm-commits]

================
@@ -0,0 +1,45 @@
+; RUN: llc < %s -march=arm64 | FileCheck %s -check-prefixes=NOTCPU --match-full-lines
----------------
jroelofs wrote:

The first RUN line should really be two:

```
; RUN: llc < %s -mtriple=arm64-linux-gnu | FileCheck %s -check-prefixes=NOTCPU --match-full-lines
; RUN: llc < %s -mtriple=arm64-apple-macosx -mcpu=generic | FileCheck %s -check-prefixes=NOTCPU --match-full-lines
```

and probably should have a different name for the check prefix, once these are split. I can't think of a good one at the moment.

https://github.com/llvm/llvm-project/pull/144152