[llvm] [NVPTX] Support llvm.{exp2, log2} for f32/f16/bf16 and vectors (PR #120519)

[Princeton Ferro via llvm-commits]

Prince781 wrote:

Update:

- Lowering to `ex2.approx` and `lg2.approx` now **on** default. Use `-nvptx-approx-exp2f32=false` and `-nvptx-approx-log2f32=false` to disable.

https://github.com/llvm/llvm-project/pull/120519