[libc-commits] [libc] [libc][stdfix] Implement divifx functions in LLVM libc (PR #206115)

[via libc-commits]

sohail103 wrote:

> > Also quick question about the organization of the implementation. I followed the existing pattern used by idiv and rdivi by placing the divifx template in libc/src/__support/fixed_point/fx_bits.h but I was wondering if thats where you would prefer these helper templates to live long term. Would it make more sense for algorithms like divifx (and maybe even idiv/rdivi as well) to live in separate headers under libc/src/__support/fixed_point (like sqrt.h does) and let fx_bits.h mainly contain the core utilities.
> 
> Long term I think what we want is following suite with how the floating point functions are organized, which I believe is each of the unique fx functions are under `libc/src/__support/math/` and the helpers in their own equivalent to `libc/src/__support/FPUtil/` (which I think is equivalent to what we have with `libc/src/__support/fixed_point/`). We'll probably want to discuss the layout with other libc folks but I think organizing the file structure could be handled later and in a separate PR.

Thanks for the clarification. I'll continue following the current organization for the remaining functions for now then.

https://github.com/llvm/llvm-project/pull/206115