Re: [cc65] floating point support redux

On Donnerstag 08 April 2010, Ullrich von Bassewitz wrote:
> On Thu, Apr 08, 2010 at 06:29:47PM +0200, Groepaz wrote:
> > - the compiler uses ieee754 floats. same calling conventions as for 32bit
> > longs can be used. (optionally, doubles can be made the same as floats)
> 
> I've fiddled with that idea for quite some time and I still do like it. But
>  it has one crucial problem: The C standard demands a higher precision for
>  doubles than what 32 bit IEEE floats deliver. So while it has several
>  advantages (one being that the compiler and runtime already know how to
>  handle 32 bit values), not being standards compliant is really something
>  to ponder.
> 
> A standards compliant double needs 34 bits in the mantissa (10 digits of
> precision), so it needs 64 bit IEEE floats or some non standard format.

yes ofcourse ... in case of cc65, i would still mostly ignore actual doubles, 
and threat them the same as floats, because well... it doesnt matter :) it 
could just aswell not support doubles at all, simply because they are way to 
slow, and having a compiler option like above would be a simple convenience 
thing, saving a typedef or so =P

-- 

http://www.hitmen-console.org    http://magicdisk.untergrund.net
http://www.pokefinder.org        http://ftp.pokefinder.org

The general rule about people on the internet seems to be 'attractive, single, 
mentally stable'... now,choose two

----------------------------------------------------------------------
To unsubscribe from the list send mail to majordomo@musoftware.de with
the string "unsubscribe cc65" in the body(!) of the mail.