info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Fixed Pnt. Divide
27 views
Skip to first unread message
Bill Kratzer
Jul 1, 1992, 4:25:05 PM7/1/92
to
I am in the process of converting a program from floating point arithmetic
to fixed point. I am using the 16.16 fixed point format (16 bits in each of
the whole and fractional parts of the number). The problem that I run into is
that I want the program to be backwards compatible with non 386/486 computers,
so I can not use the 32 bit IDIV function. I need to use 16 bit instructions
only. Multiplication was rather easy, with only a few 16 bit multiplies and
a few adds, but I can not figure out how to do division. Any suggestions?
to fixed point. I am using the 16.16 fixed point format (16 bits in each of
the whole and fractional parts of the number). The problem that I run into is
that I want the program to be backwards compatible with non 386/486 computers,
so I can not use the 32 bit IDIV function. I need to use 16 bit instructions
only. Multiplication was rather easy, with only a few 16 bit multiplies and
a few adds, but I can not figure out how to do division. Any suggestions?
Thanx in advance,
Bill K.
-------------------------------------------------------------------------
Bill Kratzer
wkra...@cs.psu.edu WMK101@PSUVM
kra...@gauss.math.psu.edu WMK...@psuvm.psu.edu
0 new messages