Date: Tue, 27 Apr 1999 10:18:22 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: David Anderson <dma AT hpesdma DOT fc DOT hp DOT com>
cc: djgpp-workers AT delorie DOT com
Subject: Re: [dma AT hpesdma DOT fc DOT hp DOT com: Performance Observation]
In-Reply-To: <199904261647.MAA06392@envy.delorie.com>
Message-ID: <Pine.SUN.3.91.990427101707.18849E-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


> From: David Anderson <dma AT hpesdma DOT fc DOT hp DOT com>
> Subject: Performance Observation

> 	And then using the latest & greatest:
> 
> C:\Dave\fft\v2>gcc -O3 -ffast-math -o fft.exe fft.c -lm
> fft.c: In function `main':
> fft.c:80: warning: return type of `main' is not `int'
[snip]
> 	More than 20% slower! Almost 30% larger!
> 
> 	So what's up?

The effect of different optimization switches is highly compiler-
version specific.  There is no promise that the same switches will
cause the same speedups in different GCC versions.

I suggest to play with the plethora of optimization options, like the
FAQ says (section 14.2; did you try the advice there?).  First thing
I would try is to get rid of -O3, it is usually evil.