🚀 go-pugleaf

In a previous article, bai@msiadmin.cit.cornell.edu (Dov Bai-MSI Visitor) writes:
>In article <1992Sep08.134945.5154@balr.com> andy@balr.com (Andrew Little) writes:
>>In article <92Sep04.024410.14613@acs.ucalgary.ca> bauwens@acs.ucalgary.ca (Luc Bauwens) writes:
>>>In article <1992Sep3.131306.13751@tc.cornell.edu> bai@msiadmin.cit.cornell.edu (Dov Bai-MSI Visitor) writes:
>>>
>>>>>>>The following short program compiled with MS-FORTRAN 5.0 with 
>>>>>>>coprocessor crashes OS2 Ver. 2 on my machine:
>>>
>>>I am inclined to believe the other poster (sorry, forgot his name)
>>>who says it may be hardware-related.
>>>
>>>But in my case, it's not an issue of the known bug in some 486
>>>motherboards: I am running this on a 386-25 (an old Mylex motherboard
>>>that I have normally found quite reliable) and (*here it gets
>>>interesting?*) an IIT FPU *not an Intel*.
>>                   ^^^^^^^
>>FYI, we have been able to crash our 386 Unix box (Interactive 3.2)
>>with divide by zeros when we used this FPU, and don't have the same
>>problem with Intel.
>
>The Coprocessor that crashes OS/2 2.0 on my machine is Cyrix.

I am interested in floating-point arithmetic in general.  I am wondering
if this behavior can occur with Intel math coprocessors.  (Someone has
already mentioned that it can occur on poorly-designed 486 motherboards.)
If not, is it due to a bug in IIT and Cyrix chips, and if so, has/will it
be corrected?  What other brands of coprocessors exhibit this behavior?
------------
Sam Figueroa (figueroa@cs.nyu.edu)
This is not a .signature virus.
Do NOT copy me into your .signature file!