[clang] [CIR][AArch64] Upstream vrshrd_n_s64/u64 and vrshr_n_v vector rounding shift right (PR #185992)

[Md Mouzam Arfi Hussain via cfe-commits]

ArfiH wrote:

@banach-space  @andykaylor 
Thank you for the feedback. I have addressed all the comments:

- `vrshr_n_v`/`vrshrq_n_v` and `vrshrd_n_u64` case labels are no longer moved.
- `vrshrd_n_s64/u64` is now implemented via `emitNeonCall`, mirroring Original Codegen.
- `emitNeonCallToOp`/`emitNeonCall` are no longer moved — forward declarations
  are used instead so they can be called from `emitCommonNeonBuiltinExpr`.
- Unrelated changes in both test files have been reverted.
- Duplicate `LLVM-LABEL`/`CIR-LABEL` pairs replaced with `ALL-LABEL`.
Please let me know if you have any feedback.

https://github.com/llvm/llvm-project/pull/185992