>Not sure. I suspect it's a glibc bug, as I found that modfl wasn't prototyped >on arm.
Right. Well, I'm not sure I'd class it as a bug: glibc just doesn't provide long double math functions for ARM at the moment. And, in turn, the reason for this is that GCC doesn't actually emit extended precision operations: "long double" is equivalent to "double". NWFPE does actually support extended precision if configured appropriately, so in theory one could turn on this support in GCC. But I suspect it would land us in a nasty compatibility quagmire, and in fact Debian kernels tend to ship with this NWFPE option disabled. p.
