[llvm] Add an all-in-one histogram intrinsic, along with lowering for AArch64 (PR #88106)

[Sam Tebbs via llvm-commits]

https://github.com/SamTebbs33 approved this pull request.

This looks good to me as is and I support the all-in-one approach. I think it would be worth addressing the FIXMEs that are simple enough to do in this patch (moving out code to their own functions and the operand assertion in `getMaskedHistogram`).

https://github.com/llvm/llvm-project/pull/88106