[llvm] [X86] Lower mathlib call ldexp into scalef when avx512 is enabled (PR #69710)

[via llvm-commits]

huhu233 wrote:

* Fix coding style issues
* Support vector versions of ldexp, e.g., `@llvm.ldexp.v8f32.v8i32` -- +avx512vl , `@llvm.ldexp.v4f32.v4i32` -- +avx512f or +avx512vl, etc.

https://github.com/llvm/llvm-project/pull/69710