@futurebird Look at how it's done in a real libm that's just aiming for low ulp error not perfect rounding. On |x|≤π/2 it's just low order polynomial in x², times x.
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
replies: