Mail Archives: djgpp/1997/10/31/00:31:47
ron newman wrote:
>
> Sorry if this is a duplicate. My server went down and I think the previous post got trashed.
>
> Anyway, I have a problem in my installation that's causing some of the downloaded test programs to fail.
> Compilation of everything seems to work fine.
>
> I'm using CWSDPMI and DOS 5.0 on an 80486DX2/66.
> The test program 'farptr' and 'nearptr' both fail. the math test program works but the last four
> digits of results are off. (weird)
The test programs were designed to test the functionality of the DJGPP
library; they are not intended as examples for the common user. I just
looked at the code for farptr.c and it's horrible, like something out of
a nightmare. :) It reads the command line in some really strange way;
you probably need to actually pass it some kind of parameter to make it
work. If you look at the code for 'near.c', the last part of main() is
commented as intended to produce a GPF.
I'm not sure which math test program you're referring to though. I
found a couple of items in the tests/libc/ansi/math directory, but not
being a math major, they make little sense to me...
--
---------------------------------------------------------------------
| John M. Aldrich | "Be wary of strong drink. It can |
| aka Fighteer I | make you shoot at tax collectors-- |
| mailto:fighteer AT cs DOT com | and miss." |
| http://www.cs.com/fighteer | - Lazarus Long |
---------------------------------------------------------------------
- Raw text -