Directed rounding floating point on DS3100

[Thomas Epperly]

I will be writing some mathematical software on a DECstation 3100 that
will need direct rounding control(ie. round operations towards positive
infinity, negative infinity or zero).  I have found what appears to be the
appropriate include file, but no man pages or other documentation.
/usr/include/machine/fpu.h defines some structures and declares some
external routines, but it doesn't give the argument lists for the
routines.  If this information is available, please send it or tell me
where to find it.

Thanks,
Tom Epperly
Graduate student at Univ. of Wisconsin
--