[libc-commits] [libc] [libc] Extend fputil::sqrt to use floating point instructions for arm32. (PR #134499)

[Mikhail R. Gadelha via libc-commits]

mikhailramalho wrote:

> > Does it have to be element wise? Isn't this only required if the argument is a vector?
> 
> Yes, `__builtin_sqrt` is whatever GNU decided it to do, which implies some standard handling. `__builtin_elementwise_sqrt` is more straightforward and type generic.

I see.

Can we move the implementation out of the platform-specific dirs? Something like https://github.com/llvm/llvm-project/blob/main/libc/src/__support/FPUtil/FMA.h but for sqrt?

https://github.com/llvm/llvm-project/pull/134499