On Mon, Apr 07, 2008 at 02:02:55PM +0300, Hasso Tepper wrote: > We don't have float_t and double_t typedefs required by C99 in math.h. As > my knowledge about all consequences are limited, I'd take safest path for > now and would commit this patch (btw, OpenBSD has the same code for now):
It doesn't make sense to define them as float or double on i386. For AMD64 it effectively depends on whether the compiler wants to implement floating point on top of SSE or FPU. Joerg
