Date: Wed, 16 Jun 1999 16:49:31 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: "Dr. =?iso-8859-1?Q?S=F3lyom=20Andr=E1s?=" <solyom AT eik DOT bme DOT hu>
cc: djgpp AT delorie DOT com
Subject: Re: Compilers comparisson, some opinions about the generated , assembler
In-Reply-To: <37679E2C.5A281FD3@eik.bme.hu>
Message-ID: <Pine.SUN.3.91.990616164546.10392B-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Wed, 16 Jun 1999, Dr. =?iso-8859-1?Q?S=F3lyom=20Andr=E1s?= wrote:

> I have the results of two actual runs, one  using Djgpp's GCC in a DOS box and
> the other Borland's C++ Builder with the same CPP source. The program which
> processed a 255 Mbyte text file. Djgpp finished in 10 min 31 sec, the Builder
> in 17 min 3 sec. The cause of this difference must be the inefficiency of the
> Windows implementation of the  FAT32 file system. The same program compiled in
> Linux and run on the same machine using the same FAT32 file system mounted
> under /dos/d finished in 1 min 21 sec...

It's probably something specific to FAT32 which indeed slows down Windows
tremendously.  With FAT16, I usually get only 30% slow-down in DOS/Windows
as compared to Linux on the same machine, even if Linux version of the
program runs on non-FAT partitions, provided that the DOS/Windows system
configuration is optimal.