Has anyone ever thought of working on the math coproc emulation code in
NetBSD to try to make it emulate the floating processor more closely, so
libm can be compiled be default to use the fpu instructions?

I at least am getting tired of having to edit that Makefile every time I
sup to get some form of speed out of math.

If not, I may attack this eventually, but it will most certainly be a while
before I have the chance to do much with it.

Thanks,
--Michael