Re: Infinity (to the nearest integer)
"Karp, Alan H" <[email protected]>
| Newsgroups | gmane.comp.lang.e.general |
|---|---|
| Message-ID | <A0BE9926384D0940BBAD4B60190CB29A7181FB312C@GVW0433EXB.americas.hpqcorp.net> |
Kevin Reid wrote: > > Suppose we do want to produce NaN. Why do we need integer NaNs > distinct from float NaNs? > Because every bit pattern in an integer register is a valid fixed point number, which leaves no way to denote NaN. We proposed that Itanium have an extra bit on each integer register, but that idea got shot down because of the performance impact on register spill. (I proposed calling such things gNaT for generally Not a Thing, which got shot down even faster.) Besides, that wouldn't be portable across architectures, which is important, as Jonathan points out. ________________________ Alan Karp Principal Scientist Virus Safe Computing Initiative Hewlett-Packard Laboratories 1501 Page Mill Road Palo Alto, CA 94304 (650) 857-3967, fax (650) 857-7029 http://www.hpl.hp.com/personal/Alan_Karp