[libc-commits] [libc] [llvm] [libc][math] Add float-only implementation for sinf / cosf. (PR #161680)

[Joseph Huber via libc-commits]

jhuber6 wrote:

Well the `sinf` test shows that the expected result is `-0.0`. I'm not sure why this change results in that unfortunately, but it seems pretty straightforward. This is a major improvement for the GPU so it would be nice to figure out why it doesn't work for this case.

https://github.com/llvm/llvm-project/pull/161680