Date: Thu, 5 Jun 1997 16:48:40 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: Gyorgy Abonyi <loop AT hu DOT bonus DOT com>
cc: Peter Palotas <blizzar AT hem1 DOT passagen DOT se>, djgpp AT delorie DOT com
Subject: Re: Speed of DJGPP?
In-Reply-To: <199706051312.PAA23996@zeus.hu.bonus.com>
Message-ID: <Pine.SUN.3.91.970605164252.6778A-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Precedence: bulk


On Thu, 5 Jun 1997, Gyorgy Abonyi wrote:

> This's becouse the DOS4GW extender what watcom uses, replaces the most real
> mode DOS functions, with it's own 32-bit versions, so with Watcom the file
> I/O operations don't need 32/16 bit mode switching, what is really slow... 

This is all incorrect.  First, DOS4GW is just another DOS extender, so it
eventually calls DOS.  Therefore, the switch from real to protected mode 
and back is still required.  Second, the time wasted by the compiler in 
mode switches is relatively small, especially if you compile with 
optimizations, so I/O is not the major factor in compilation speed.  Try 
compiling with and without optimizations and see the big difference (in 
GCC, it's around factor of 2); since the amount of I/O is the same no 
matter what level of optimization is involved, you will see that most of 
the time the compiler needs raw CPU power (to parse the code and optimize 
it).