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
From: "Dave Korn" <dk AT artimi DOT com>
To: <cygwin AT cygwin DOT com>
Subject: RE: g77 executable with largish array fails (solved?)
Date: Tue, 26 Oct 2004 14:14:42 +0100
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
In-Reply-To: <417DFEB1.1080604@familiehaase.de>
Message-ID: <NUTMEGk3Y8pEBtKarvz00000242@NUTMEG.CAM.ARTIMI.COM>
X-OriginalArrivalTime: 26 Oct 2004 13:14:43.0734 (UTC) FILETIME=[C22DA360:01C4BB5D]

> -----Original Message-----
> From: cygwin-owner On Behalf Of Gerrit P. Haase
> Sent: 26 October 2004 08:37

> Billinghurst, David (CALCRTS) wrote:
> 
> >>From: David Billinghurst
> >>
> >>The following test case is distilled from the LAPACK timing 
> test cases.  
> >>These used to work on some previous versions of cygwin/g77, 
> but now fail.
> > 
> > 
> > I translated the problem into C using f2c.  It becomes:
> > 
> > $ cat labugc.c
> > int main() {
> >   static float a[39*1024*1024];
> > }
> > 
> > $ gcc -O0 -o labugc.exe labugc.c
> > $ ./labugc
> > C:\cygwin\usr\people\billingd\labugc.exe (3372): *** 
> MapViewOfFileEx(0x728, in_h 0x728) failed, Win32 error 6
> > 
> > 
> > Aha, I says.  Looks like stack overflow.  Google pointed me 
> to an work around
> > gcc -Wl,--stack,200000000 -O0 -o labugc.exe labugc.c
> > 
> > 	David
> 
> Interesting, it works without special flags when using gcc-3.4.1.
> 
> Gerrit

  I can explain that: here's why a static variable might cause a MapViewOfFileEx,
and also why changing the stack size might fix it:

http://www.cygwin.com/ml/cygwin/2004-07/msg00646.html

and so I reckon the difference between 3.3.3 and 3.4.1 in this regard will be
cause by the change in placement of statically zero initialised data.  Gerrit,
does adding -fzero-initialized-in-bss or -fno-zero-initialized-in-bss to the
compile flags cause the problem to return for 3.4.1, and then adding the
-Wl,--stack option fix it again?


    cheers, 
      DaveK
-- 
Can't think of a witty .sigline today....


--
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/