Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com
Message-Id: <6.0.1.1.0.20040718091528.01f24730@imap.myrealbox.com>
X-Sender: tprince AT imap DOT myrealbox DOT com
Date: Sun, 18 Jul 2004 09:24:35 -0700
To: "Alex Vinokur" <alexvn AT connect DOT to>, cygwin AT cygwin DOT com
From: Tim Prince <tprince AT computer DOT org>
Subject: Re: Performance: g++ Cygwin vs. other compilers (copying char[] to vector)
In-Reply-To: <cde649$qb4$1@sea.gmane.org>
References: <cddooq$ssi$1 AT sea DOT gmane DOT org> <6 DOT 0 DOT 1 DOT 1 DOT 0 DOT 20040718063702 DOT 01f796e8 AT imap DOT myrealbox DOT com> <cde649$qb4$1 AT sea DOT gmane DOT org>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
X-IsSubscribed: yes

At 08:48 AM 7/18/2004, Alex Vinokur wrote:


>"Tim Prince" <tprince AT computer DOT org> wrote in message 
>news:6 DOT 0 DOT 1 DOT 1 DOT 0 DOT 20040718063702 DOT 01f796e8 AT imap DOT myrealbox DOT com...
> > At 05:00 AM 7/18/2004, Alex Vinokur wrote:
> >
> > >Hi,
> > >
> > >How to explain so considerable difference in performance: g++ Cygwin vs.
> > >other compilers in tests below?
>[snip]
>
> > I don't find your compile options, or whether you have profiled.  For g++
> > under cygwin,
>[snip]
>
>g++ *.cpp -o cps_cyg.exe   // g++ Cygwin
>g++ -mno-cygwin *.cpp -o cps_mgw.exe  // g++ Mingw
>gpp *.cpp -o cps_dj.exe  // g++ Djgpp
>cl /EHsc *.cpp -o cps_ms.exe   // C++ Microsoft
>dmc -I. -IC:/dm/stlport/stlport -Ae *.cpp -o cps_dm.exe  // C++ Digital Mars

Microsoft C default is a good compromise between compilation speed and 
performance.  g++ Cygwin aims for compilation speed and no transformations 
which inhibit debugging.  Performance simply is not comparable without 
normal optimization:
g++ -O3 -Drestrict=__restrict__ -funroll-loops -march=pentium4 -mfpmath=sse 
*.cpp
CL /EHsc /Ox /arch:SSE2 *.cpp
I have no idea about Digital Mars, but STLport does have more optimization 
than MS.
Nor do I know if any of your versions of g++ tinker with default optimization.
I believe clock() is implemented differently between cygwin and msvcrt, and 
you may have additional variations represented here.

Tim Prince 


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/