FP format, and test suite (was: [cc65] Floating point support?)

From: Spiro Trikaliotis <ml-cc651trikaliotis.net>
Date: 2004-09-28 10:31:40
Hello,

* On Tue, Sep 28, 2004 at 09:33:57AM +0200 Marc Rintsch wrote:

> > AFAIR, IEEE 754 does *not* tell use the format of the FP values! 

> I always thought one benefit of the IEEE format is, that it's designed to 
> implement comparisons (<, >, =) of n-bit floats as comparisons of n-bit 
> signed integers?  That's only possible if the MSB is the sign bit.

Yes - in fact, you're right. :-(

This is another example of "never trust your teacher/Professor". ;-) He
once told us that the order of the entries were not given by 754. I
never checked for myself, but I trusted him. He even told us that this
way is the most common, and that many sources would state it was the
only one, but that this were not given by 754.

While I do not have access to 754 (do not want to pay for it just to
find out this thing), the official page
http://grouper.ieee.org/groups/754/ links to pages which describe the FP
format with sign = MSB, exponent, fractional part - order (from MSB).


BTW: If anyone wants to implemente IEEE 754 FP, he might want to check
the test suite of http://www.math.utah.edu/~beebe/software/ieee/. ;-)

Some more resources on 754-1985 can be found here:
http://cch.loria.fr/documentation/IEEE754/

BTW: 754 seems to be going through revisions, so it might change in the
future.

Regards,
   Spiro.

-- 
Spiro R. Trikaliotis
http://www.trikaliotis.net/
----------------------------------------------------------------------
To unsubscribe from the list send mail to majordomo@musoftware.de with
the string "unsubscribe cc65" in the body(!) of the mail.
Received on Tue Sep 28 10:31:49 2004

This archive was generated by hypermail 2.1.8 : 2004-09-28 10:31:58 CEST