bill@wrangler.WLK.COM (Bill Kennedy) (03/22/91)
Last week I started caterwauling about getting a royal runaround from
Wyle Labs regarding the behavior of the 68881 after the upgrade to
release Vr3. In a nutshell, in Vr2 the FPP worked and in Vr3 sysdef
claimed that it wasn't there even though sus firmware said it was. In
particular I was disturbed when Wyle told me that I would need a $2,000
option card in order to get my FPP back, they were wrong.
I have gotten a lot of help and encouragement from the folks in Columbia
and a lot more misinformation from Wyle Labs. Despite the latter, the
sales rep continued to try to get to the truth, but the answer came from
Sea Change Corporation in Canada. I bought their TCP+ product (with
which I'm well pleased) and although they had no involvement whatsoever
with the FPP problem they found the answer for me. The code fragment
(it's small so I'm enclosing it) below, compiled with the -fH flag verifies
that the FPP is present and working.
So the bottom line is that sysdef lies about the FPP, the technical gurus
at Wyle Labs all have cellophane navels, and Sea Change once again comes
the the rescue (they really do wear white hats). Many thanks to all who
helped and tried to help. If you're in doubt about whether or not your
sysdef lies about your FPP compile this with -fH and the a.out should
either drop a core (no FPP) or give you roughly pi and an indication that
the FPP is alive and well (sorry to omit the man page :-)
#include <stdio.h>
main(argc,argv)
int argc;
char **argv;
{
int retcode;
double test, ans;
test = 355.0/113.0;
ans = test * 1.0;
printf("test is %7.5f ans is %7.5f\n",test,ans);
if ( -1 == _sysm68k(3,retcode) )
printf("The FPP is true\n");
else
printf("The FPP is false\n");
}
--
Bill Kennedy usenet {att,cs.utexas.edu,pyramid!daver}!ssbn.wlk.com!bill
internet bill@ssbn.WLK.COM or attmail!ssbn!bill