home *** CD-ROM | disk | FTP | other *** search
- .OP
- Neurology Dept., SUNY
- HSC T12 Rm020
- Stony Brook, N.Y. 11794
- Dec.31, 1985
- 516-444-1475
- C Users' Group
- Box 97
- McPherson, Kansas 67460
- 316-241-1065
-
- To whom it may concern:
-
- These files show how to revise the Software Toolworks' C/80
- floating point library (for 8080/Z80) so that it uses the AMD
- 9511A (Intel 8231A) hardware math chip. I planned to submit this
- disk to the CUG after my article appeared in Micro/Systems, but
- lo! the MCHIP80 disk (CUG vol.162) was released first.
- However, I don't think either MCHIP80 or my patch is
- redundant for two reasons. First, my patch is easier to use.
- Unlike MCHIP80, which forbids initialization of floats and
- requires replacement of atof() and ftoa(), my patch doesn't
- require any program redesign--anything that is legal with the
- original floating point library is OK with my patch. Second,
- MCHIP80 illustrates a generally applicable stratagem for
- modifying the performance of a high level language: intercepting
- calls to library routines by cleverly sequencing the files
- submitted to a linking loader.
- Anyway, Walt Bilofsky suggested that I send these to you
- when I discussed it with him.
-
- Yours truly,
-
-
-
- Ted Carnevale