From: elf AT netcom DOT com (Marc Singer)
Message-Id: <199612192249.OAA03416@netcom7.netcom.com>
Subject: Re: [??] declaring 64bit integer
To: salvador AT natacha DOT inti DOT edu DOT ar (Salvador Eduardo Tropea)
Date: Thu, 19 Dec 1996 14:49:40 -0800 (PST)
Cc: djgpp AT delorie DOT com (DJGPP List Alias)
In-Reply-To: <m0vajuL-000S1dC@natacha.inti.edu.ar> from "Salvador Eduardo Tropea" at Dec 19, 96 06:05:29 pm
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 837       

> > >   How should I declare a 64bit integer w/ djgpp?  Is it even possible?
> > >   I tried "long long" and when I looked at the assembler output, it
> > >   looked like it was using eax & edx, but gcc gave me a warning.
> > 
> > I believe that the Intel x86 compiler does NOT have 64 bit integers.  The
> > Alpha compiler does.
> You can use EAX:EDX as a 64 bit register, but in general the DJGPP 64 
> bits integers are managed just like a 32 bits integer in a 16 bits 
> compiler: Using routines.

Kind of a bummer, though.  I believe this is one of the few things
that we cannot do in C that is a) desirable and b) must be coded in
assembly language.  The bummer is that the 64 bit results of DIV and
MUL can be very useful for extended precision math.  The subroutine
implementation is too great of a performance penalty.

-- Marc