In a recent note, "Shmuel Metz (Seymour J.)" said: > Date: Mon, 26 Sep 2005 08:18:47 -0300 > > >Ah! The JFFO instruction which appeared on the decsystem KA-10 in > >1968 (according to an unscientific Google search). I suppose it's OK > >for IBM to use it now, since any patent has presumably expired. > > I'm sure that IBM could trot out prior art; CDC had such an > instruction before DEC. I suspect that IBM did as well, on the 7030, > but don't have a manual. > Are you not thinking, rather, of the Cxi instruction, which counted '1'b bits? Of course, this be incorporated in a sequence of decrement; xor; count.
How about using floating point instructions? Normalize, then extract the shift count from the exponent. We'll need to see whether John G. discloses his technique. -- gil -- StorageTek INFORMATION made POWERFUL ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
