[llvm] [APFloat] Add exp functions for single and double using exp/expf implementations from LLVM libc. (PR #190667)

[Mehdi Amini via llvm-commits]

joker-eph wrote:

It's not fixed: 
```

/vol/worker/mlir-nvidia/mlir-nvidia-gcc7/llvm.src/llvm/../libc/src/__support/FPUtil/rounding_mode.h: In function ‘constexpr bool __llvm_libc_apfloat::fputil::fenv_is_round_down()’:
/vol/worker/mlir-nvidia/mlir-nvidia-gcc7/llvm.src/llvm/../libc/src/__support/FPUtil/rounding_mode.h:95:39: error: call to non-constexpr function ‘bool __llvm_libc_apfloat::fputil::generic::fenv_is_round_down()’
     return generic::fenv_is_round_down();
            ~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
```

https://lab.llvm.org/buildbot/#/builders/116/builds/28500

https://github.com/llvm/llvm-project/pull/190667