need help with math co emulation

Back in the late 1970s or early 1980s, IIRC, they took out the floating point emulation from the UNIX OS because it was too much of a pain to support it, and machines like the PDP-11 and Interdata machines all had floating point built in. A few of the oldest models did not, but you could buy a board to do it. They just did not want to support all that stuff, so they stopped.

I suppose this ancient history has nothing to do with Linux, but these days it may be cheaper to by a processor that does floating point than to support the fp emulation. Not, perhaps, for any one individual who has some legacy hardware, but from the point of view of support.

One model of the IBM 1620 (or was it the 1401) did not even have integer arithmetic in it, but did addition and subtraction by table lookup, with the table stored in low memory somewhere. This was in the days when hardware was expensive. In any case, it is my understanding that too often the additionsubtraction table got overwritten by runaway program, and the hardware had to be rebooted. No memory management hardware in those days. Pretty soon those machines got upgraded. Like WinModems, it sometimes does not make sense to skimp on hardware.

-- .~. Jean-David Beyer Registered Linux User 85642. V PGP-Key: 9A2FC99A Registered Machine 241939. ^^-^^ 22:25:00 up 85 days, 16:21, 3 users, load average: 4.21, 4.14, 4.14