Mailing-List: contact cygwin-apps-help AT cygwin DOT com; run by ezmlm
Sender: cygwin-apps-owner AT cygwin DOT com
List-Subscribe: <mailto:cygwin-apps-subscribe AT cygwin DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-apps/>
List-Post: <mailto:cygwin-apps AT cygwin DOT com>
List-Help: <mailto:cygwin-apps-help AT cygwin DOT com>, <http://sources.redhat.com/lists.html#faqs>
Delivered-To: mailing list cygwin-apps AT cygwin DOT com
Message-ID: <011c01c1a6fe$cca1e2f0$0d00a8c0@mchasecompaq>
From: "Michael A Chase" <mchase AT ix DOT netcom DOT com>
To: "Robert Collins" <robert DOT collins AT itdomain DOT com DOT au>,
   <cygwin-apps AT sources DOT redhat DOT com>
References: <026501c1a6fd$95bc5910$0200a8c0 AT lifelesswks>
Subject: Re: setup crashing - fixed?
Date: Sat, 26 Jan 2002 22:49:09 -0800
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 6.00.2600.0000
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000

----- Original Message ----- 
From: "Robert Collins" <robert DOT collins AT itdomain DOT com DOT au>
To: <cygwin-apps AT sources DOT redhat DOT com>
Sent: Saturday, January 26, 2002 22:41
Subject: setup crashing - fixed?


> Right, I think I found the cause. At least I can successfully download
> from *every* available site, and before I had repeatable problems doing
> that.
> 
> Corinna, can you CVS update and try a fresh build?
> 
> I found a bug that could have caused the issue:
> 
> site (char const *newkey)
> {
>   key = new char[strlen (newkey + 1)];
>                                ^^^^^
>     strcpy (key, newkey);
> };
> 
> However, fixing that DID NOT fix the problem (although I had isolated
> the fault to this constructor by means of replacement new and new[]
> calls using HeapAlloc, combined with HeapValidate().)
> 
> Moving the constructor body out of the header fixed the problem. I don't
> have time to generate a testcase, but if anyone does have time, I'm sure
> the gcc team would appreciate knowing about this. The fault occured when
> that constructor was called from a template instantiation of list.h.

I hope you moved the '+ 1' outside the strlen() as well.
-- 
Mac :})
** I normally forward private questions to the appropriate mail list. **
Give a hobbit a fish and he eats fish for a day.
Give a hobbit a ring and he eats fish for an age.