[libc-commits] [libc] [libc][math] Implement an integer-only version of double precision sin with 1 ULP errors. (PR #184752)

[via libc-commits]

lntue wrote:

@statham-arm : this is my first try of integer-only implementation for targets without double precision FPUs.  Do you mind checking if 64 & 128 bit integers performance with something like this on such targets is better than emulating double precision computations?  Thanks!
(I suspect that our internal 128-bit integer operations for 32-bit arms could be tuned for better performance / code size)

https://github.com/llvm/llvm-project/pull/184752