Mailing List Archive


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[tlug] What does 80387 have to do with IEEE 754 issues? It's moot. (was Re: issues with format of double (or IEEE754))



Botond Botyanszki wrote:

> On Wed, 26 Jul 2006 10:33:16 -0400
> Jim <jep200404@example.com> wrote:
> 
> > > on an i386 mashine...
> > 
> > ISA bus? 
> > ...
> > Do you have the 80387 coprocessor? 
> 
> How is this related to the problem he described?!

The ISA bus is not related to his problem. 

The use (or not) of a '387 _could_ have been related to his problem. 
Back then there had been small (but incompatible) differences 
in how floating point numbers are stored. If one used a '387, 
the numbers would likely be stored in the '387's native format. 
If one did not use a '387, one could use any format 
although one would likely use a emulation library that would 
likely use the same format as the '387. 

I now see at http://en.wikipedia.org/wiki/Intel_80387, 
that the '387 was Intel's first FPU to fully implement IEEE 754, 
hinting that there had been incompatibilities just before 
(in the '287 which predates Linux). 

Of course, this is all mooted by his Hajek's revelation that he 
is using a different CPU (and FPU). 



Home | Main Index | Thread Index

Home Page Mailing List Linux and Japan TLUG Members Links