The -ffast-math option inlines for example sqrt() to a single FPU instruction, without any library call. The drawback is that errno is not set, so it is a violation of the C standard. However most programs are OK with that, including POV-Ray.
How GCC knows what instructions to inline? I guess there's support for m68k target for this option? And has -m68881 / -m68060 / -msoft-float any effect on --fast-math?
NB: I'm going to send my math-68881.h patch for C99 to GCC people, and I will also ask them if that file can really be useful for someone nowadays.
I'm really curious about the answer.
--