Hello there, this might be an odd question, but I figured it couldn't hurt to ask. :-)
I was looking through the documentation and noticed that there are builtins for the trigonometric and other functions on the x86 arch, (sinf, cosf, tanf etc) but not on the x64 arch, although technically it has the better/more standard FPU.
I was wondering if this is deliberate for reasons I just haven't found, or if it is forthcoming, or if Pelles needs some contributions. ;-)
Floating-point calculations on x64 are done with SSE registers, so I'm not convinced the FPU is a win. It would at least involve some clumpsy mappings back and forth, which may kill any speed gains from the FPU. AFAIK, other x64 compilers trust library functions rather than inline code - but I havn't checked them all...