Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:11581
Return-Path: <nucleuz@gmail.com>
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Delivered-To: mailing list internals@lists.php.net
Received: (qmail 37504 invoked by uid 1010); 27 Jul 2004 07:00:25 -0000
Delivered-To: ezmlm-scan-internals@lists.php.net
Delivered-To: ezmlm-internals@lists.php.net
Received: (qmail 37377 invoked from network); 27 Jul 2004 07:00:24 -0000
Received: from unknown (HELO mproxy.gmail.com) (64.233.170.196)
  by pb1.pair.com with SMTP; 27 Jul 2004 07:00:24 -0000
Received: by mproxy.gmail.com with SMTP id 76so127813rnk
        for <internals@lists.php.net>; Tue, 27 Jul 2004 00:00:24 -0700 (PDT)
Received: by 10.38.88.18 with SMTP id l18mr167578rnb;
        Tue, 27 Jul 2004 00:00:24 -0700 (PDT)
Message-ID: <1628f4cb040727000046bf3a7f@mail.gmail.com>
Date: Tue, 27 Jul 2004 09:00:24 +0200
To: internals@lists.php.net
Cc: Rasmus Lerdorf <rasmus@php.net>
In-Reply-To: <Pine.LNX.4.58.0407260822590.6106@t42p>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
References: <1628f4cb04072600137d82bb80@mail.gmail.com> <Pine.LNX.4.58.0407260822590.6106@t42p>
Subject: Re: [PHP-DEV] Segfault problem with PHP5
From: nucleuz@gmail.com (=?ISO-8859-1?Q?h=E5vard_eide?=)

Trying again with a link to the file:

Just ran it through valgrind with the options that Rasmus suggested
and got a output file: http://www.ikomm.no/valgrind/out.pid27848  I
can see what'
happening, but have no idea where to go from here.
Can someone have a look at it?

Cheers

> > If the segfault was the same everytime it should be easy to track
> > down, but I've had a large amount of different segfaults running
> > apache through gdb. The other thing is that one screen can be loaded
> > different times before the segfault comes as well, sometimes the
> > segfault comes straight away, other times after 5-6 reloads of the
> > same screen.
> 
> This tends to be the sort of problem where Valgrind can be very useful.
> It sounds like memory corruption somewhere.
> 
> Run it like this:
> 
> valgrind --leak-check=yes --show-reachable=yes --num-callers=16 --error-limit=no --tool=memcheck --logfile=out /usr/bin/apache -X
> 
> Then hit the server and have a look at the out.<pid> file.
> 
> -Rasmus
> 


-- 
www.phpvolcano.com