From dazoe at dazoe.net Mon Aug 2 01:07:13 2010 From: dazoe at dazoe.net (Dave Akers) Date: Sun, 1 Aug 2010 18:07:13 -0500 Subject: [mpm-itk] Chroot Message-ID: I wanted VirtualHosts running as a separate user/group also in a chroot. There were some strange things going on with ChrootDir. the server would start and chroot only one httpd process but none of the others. After digging in the source and playing around a little I came up with a patch that will allow the ChrootDir to work as expected, and as a little bonus to those who want it, each VirtualHost can be chrooted independently via VHostRoot. I have attached my patch. Do with it what you will but i do hope this will make it into mpm_itk. --dazoe -------------- next part -------------- A non-text attachment was scrubbed... Name: 11-chroot.patch Type: text/x-patch Size: 3243 bytes Desc: not available URL: From dazoe at dazoe.net Mon Aug 2 03:44:04 2010 From: dazoe at dazoe.net (Dave Akers) Date: Sun, 1 Aug 2010 20:44:04 -0500 Subject: [mpm-itk] Chroot In-Reply-To: References: Message-ID: I wanted VirtualHosts running as a separate user/group also in a chroot. There were some strange things going on with ChrootDir. the server would start and chroot only one httpd process but none of the others. After digging in the source and playing around a little I came up with a patch that will allow the ChrootDir to work as expected, and as a little bonus to those who want it, each VirtualHost can be chrooted independently via VHostRoot. I have attached my patch. Do with it what you will but i do hope this will make it into mpm_itk. --dazoe -------------- next part -------------- A non-text attachment was scrubbed... Name: 11-chroot.patch Type: text/x-patch Size: 3243 bytes Desc: not available URL: From sgunderson at bigfoot.com Mon Aug 2 16:10:44 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Mon, 2 Aug 2010 16:10:44 +0200 Subject: [mpm-itk] Chroot In-Reply-To: References: Message-ID: <20100802141044.GA7341@uio.no> On Sun, Aug 01, 2010 at 06:07:13PM -0500, Dave Akers wrote: > I wanted VirtualHosts running as a separate user/group also in a > chroot. Why? You're aware that the chroot jail, at least on Linux, is pretty weak, right? > There were some strange things going on with ChrootDir. the > server would start and chroot only one httpd process but none of the > others. After digging in the source and playing around a little I came > up with a patch that will allow the ChrootDir to work as expected, and > as a little bonus to those who want it, each VirtualHost can be > chrooted independently via VHostRoot. I have attached my patch. Do > with it what you will but i do hope this will make it into mpm_itk. How do you intend to deal with subsequent requests on the same HTTP connection? /* Steinar */ -- Homepage: http://www.sesse.net/ From dazoe at dazoe.net Mon Aug 2 16:18:01 2010 From: dazoe at dazoe.net (Dave Akers) Date: Mon, 2 Aug 2010 09:18:01 -0500 Subject: [mpm-itk] Chroot In-Reply-To: <20100802141044.GA7341@uio.no> References: <20100802141044.GA7341@uio.no> Message-ID: Chroot is just another security step, even if it is a week one. And as far as the subsequent requests on the same connection, they have to be for the same host because itk allready dropped to the user for that host and it will not try to chroot if it is allready running as a user other than root. -dazoe On Aug 2, 2010 9:10 AM, "Steinar H. Gunderson" wrote: > On Sun, Aug 01, 2010 at 06:07:13PM -0500, Dave Akers wrote: >> I wanted VirtualHosts running as a separate user/group also in a >> chroot. > > Why? You're aware that the chroot jail, at least on Linux, is pretty weak, > right? > >> There were some strange things going on with ChrootDir. the >> server would start and chroot only one httpd process but none of the >> others. After digging in the source and playing around a little I came >> up with a patch that will allow the ChrootDir to work as expected, and >> as a little bonus to those who want it, each VirtualHost can be >> chrooted independently via VHostRoot. I have attached my patch. Do >> with it what you will but i do hope this will make it into mpm_itk. > > How do you intend to deal with subsequent requests on the same HTTP > connection? > > /* Steinar */ > -- > Homepage: http://www.sesse.net/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From dazoe at dazoe.net Mon Aug 2 16:18:05 2010 From: dazoe at dazoe.net (Dave Akers) Date: Mon, 2 Aug 2010 09:18:05 -0500 Subject: [mpm-itk] Chroot In-Reply-To: <20100802141044.GA7341@uio.no> References: <20100802141044.GA7341@uio.no> Message-ID: Chroot is just another security step, even if it is a week one. And as far as the subsequent requests on the same connection, they have to be for the same host because itk allready dropped to the user for that host and it will not try to chroot if it is allready running as a user other than root. -dazoe On Aug 2, 2010 9:10 AM, "Steinar H. Gunderson" wrote: > On Sun, Aug 01, 2010 at 06:07:13PM -0500, Dave Akers wrote: >> I wanted VirtualHosts running as a separate user/group also in a >> chroot. > > Why? You're aware that the chroot jail, at least on Linux, is pretty weak, > right? > >> There were some strange things going on with ChrootDir. the >> server would start and chroot only one httpd process but none of the >> others. After digging in the source and playing around a little I came >> up with a patch that will allow the ChrootDir to work as expected, and >> as a little bonus to those who want it, each VirtualHost can be >> chrooted independently via VHostRoot. I have attached my patch. Do >> with it what you will but i do hope this will make it into mpm_itk. > > How do you intend to deal with subsequent requests on the same HTTP > connection? > > /* Steinar */ > -- > Homepage: http://www.sesse.net/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From sgunderson at bigfoot.com Mon Aug 2 16:58:23 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Mon, 2 Aug 2010 16:58:23 +0200 Subject: [mpm-itk] Chroot In-Reply-To: References: <20100802141044.GA7341@uio.no> Message-ID: <20100802145823.GB7341@uio.no> On Mon, Aug 02, 2010 at 09:18:05AM -0500, Dave Akers wrote: > And as far as the subsequent requests on the same connection, they have to > be for the same host because itk allready dropped to the user for that host > and it will not try to chroot if it is allready running as a user other > than root. Sure, but how are you going to deal with, say, lookups to .htaccess files that are no longer in the path? I doubt this is going to work very well for arbitrary modules. /* Steinar */ -- Homepage: http://www.sesse.net/ From c.c.dean at durham.ac.uk Tue Aug 3 17:03:52 2010 From: c.c.dean at durham.ac.uk (Colin Dean) Date: Tue, 03 Aug 2010 16:03:52 +0100 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <20100730235544.GA25932@uio.no> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> Message-ID: <4C582FD8.3000109@durham.ac.uk> On Mon, Jul 26, 2010 at 23:59:57PM +0200, Steinar H. Gunderson wrote: > On Mon, Jul 26, 2010 at 10:37:03AM +0200, Marijn Otte wrote: >>> Of course, the _right_ fix is to fix whatever causes these uid switches in >>> the first place. >> What happens is that a proxy server uses the same connection for requests >> on multiple vhosts, because behind the proxy 2 users are accessing two >> different websites on the same server, at the same time. I think there is >> nothing wrong with that? > > It's fully legal as per the HTTP standards, but definitely not a good idea > performance-wise. Ideally the proxy server should understand that it > shouldn't reuse that connection for different vhosts (ie., you should be able > to tell it so). Of course this is impossible to realize if you don't have > control over the proxy server. The proxy server may also be proxying for requests on the same vhost but different directories (e.g. on our University site with lots of users having their own pages on the same vhost). In that case, it's pretty impractical to configure most proxy servers to re-use connections per-vhost but not per-user-directory. And if we prohibit re-use altogether, performance really suffers. Making ?cannot read .htaccess after uid switch? behave identically as ?cannot setuid after uid switch?, doesn't fix this for us, because the problem is that once setuid() to one user, the httpd process cannot even traverse the directory tree of another user, before it gets as far as trying to read .htaccess files. We couldn't require/enforce that every user makes their directory tree world-accessible. So isn't there a need for "cannot traverse directory tree after uid switch" to exit in the same way? Colin From sgunderson at bigfoot.com Tue Aug 3 17:32:49 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Tue, 3 Aug 2010 17:32:49 +0200 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <4C582FD8.3000109@durham.ac.uk> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> <4C582FD8.3000109@durham.ac.uk> Message-ID: <20100803153249.GA12424@uio.no> On Tue, Aug 03, 2010 at 04:03:52PM +0100, Colin Dean wrote: > So isn't there a need for "cannot traverse directory tree after uid > switch" to exit in the same way? I posted a patch for that a few days ago; did you try it? /* Steinar */ -- Homepage: http://www.sesse.net/ From c.c.dean at durham.ac.uk Wed Aug 4 13:48:19 2010 From: c.c.dean at durham.ac.uk (Colin Dean) Date: Wed, 04 Aug 2010 12:48:19 +0100 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <20100803153249.GA12424@uio.no> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> <4C582FD8.3000109@durham.ac.uk> <20100803153249.GA12424@uio.no> Message-ID: <4C595383.3010003@durham.ac.uk> Steinar H. Gunderson wrote: > On Tue, Aug 03, 2010 at 04:03:52PM +0100, Colin Dean wrote: >> > So isn't there a need for "cannot traverse directory tree after uid >> > switch" to exit in the same way? > I posted a patch for that a few days ago; did you try it? > I applied the patch for server/config.c you posted Sat Jul 31 01:55:44, but that doesn't fix for me. When I try to switch in a browser between 2 PHP scripts on the same vhost but in directories with different user setting via AssignUserId, I always get a 403 Forbidden referring to the PHP script for the second directory on the same connection. I don't actually have any htaccess files in this example, but each user does not have any access rights at the filesystem level to the other user's directory tree. From sgunderson at bigfoot.com Wed Aug 4 18:17:17 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Wed, 4 Aug 2010 18:17:17 +0200 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <4C595383.3010003@durham.ac.uk> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> <4C582FD8.3000109@durham.ac.uk> <20100803153249.GA12424@uio.no> <4C595383.3010003@durham.ac.uk> Message-ID: <20100804161717.GB21201@uio.no> On Wed, Aug 04, 2010 at 12:48:19PM +0100, Colin Dean wrote: > I applied the patch for server/config.c you posted Sat Jul 31 01:55:44, > but that doesn't fix for me. When I try to switch in a browser between 2 > PHP scripts on the same vhost but in directories with different user > setting via AssignUserId, I always get a 403 Forbidden referring to the > PHP script for the second directory on the same connection. I don't > actually have any htaccess files in this example, but each user does not > have any access rights at the filesystem level to the other user's > directory tree. What does the log say in this case? /* Steinar */ -- Homepage: http://www.sesse.net/ From c.c.dean at durham.ac.uk Thu Aug 5 13:17:31 2010 From: c.c.dean at durham.ac.uk (Colin Dean) Date: Thu, 05 Aug 2010 12:17:31 +0100 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <20100804161717.GB21201@uio.no> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> <4C582FD8.3000109@durham.ac.uk> <20100803153249.GA12424@uio.no> <4C595383.3010003@durham.ac.uk> <20100804161717.GB21201@uio.no> Message-ID: <4C5A9DCB.2020605@durham.ac.uk> Steinar H. Gunderson wrote: >> I applied the patch for server/config.c you posted Sat Jul 31 01:55:44, >> > but that doesn't fix for me. When I try to switch in a browser between 2 >> > PHP scripts on the same vhost but in directories with different user >> > setting via AssignUserId, I always get a 403 Forbidden referring to the >> > PHP script for the second directory on the same connection. I don't >> > actually have any htaccess files in this example, but each user does not >> > have any access rights at the filesystem level to the other user's >> > directory tree. > What does the log say in this case? Just permission denied errors, like: [error] [client x.x.x.x] (13)Permission denied: access to /php.myadmin/test.php denied From marijn at e-active.nl Thu Aug 5 14:25:41 2010 From: marijn at e-active.nl (Marijn Otte) Date: Thu, 5 Aug 2010 14:25:41 +0200 Subject: [mpm-itk] Switching vhosts during the same connection in Debian References: <20100723162027.GU17040@stud.ntnu.no><20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> Message-ID: > You can try this patch as a first approximation. Thank you very much for the patch! We installed it and it seems to work very well for our problem (switching between vhosts). We will do a lot more testings and will post the results here. Kind regards, Marijn Otte -----Oorspronkelijk bericht----- Van: mpm-itk-bounces at err.no [mailto:mpm-itk-bounces at err.no] Namens Steinar H. Gunderson Verzonden: zaterdag 31 juli 2010 1:56 Aan: mpm-itk at err.no CC: Marijn Otte Onderwerp: Re: [mpm-itk] Switching vhosts during the same connection in Debian On Fri, Jul 23, 2010 at 06:26:11PM +0200, Steinar H. Gunderson wrote: > I reiterate that the best mpm-itk fix would probably be to make ?cannot read > .htaccess after uid switch? behave identically as ?cannot setuid after uid > switch?, ie. close the HTTP connection straight. I have no idea how easy it > is to do, though. You can try this patch as a first approximation. The logging doesn't actually appear to work, it's pretty much untested, and probably needs some more sanity checking, but it just might help with your issue: --- a/server/config.c 2010-07-21 20:11:07.000000000 +0200 +++ b/server/config.c 2010-07-31 01:53:15.000000000 +0200 @@ -1840,6 +1867,15 @@ else { if (!APR_STATUS_IS_ENOENT(status) && !APR_STATUS_IS_ENOTDIR(status)) { +#ifdef ITK_MPM + if (getuid() != 0) { + ap_log_error(APLOG_MARK, APLOG_WARNING, status, r, + "Couldn't read %s, closing connection.", + filename); + ap_lingering_close(r->connection); + exit(0); + } +#endif ap_log_rerror(APLOG_MARK, APLOG_CRIT, status, r, "%s pcfg_openfile: unable to check htaccess file, " "ensure it is readable", /* Steinar */ -- Homepage: http://www.sesse.net/ _______________________________________________ mpm-itk mailing list mpm-itk at err.no http://lists.err.no/mailman/listinfo/mpm-itk From sgunderson at bigfoot.com Thu Aug 5 14:42:33 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Thu, 5 Aug 2010 14:42:33 +0200 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <4C5A9DCB.2020605@durham.ac.uk> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> <4C582FD8.3000109@durham.ac.uk> <20100803153249.GA12424@uio.no> <4C595383.3010003@durham.ac.uk> <20100804161717.GB21201@uio.no> <4C5A9DCB.2020605@durham.ac.uk> Message-ID: <20100805124233.GB24869@uio.no> On Thu, Aug 05, 2010 at 12:17:31PM +0100, Colin Dean wrote: >> What does the log say in this case? > Just permission denied errors, like: > > [error] [client x.x.x.x] (13)Permission denied: access to > /php.myadmin/test.php denied Hm, but that's for the PHP file, not the .htaccess file? This sounds like a different issue somehow to me; I don't see how the per-directory patch should have changed anything for the .php file itself. /* Steinar */ -- Homepage: http://www.sesse.net/ From dazoe at dazoe.net Fri Aug 6 00:26:55 2010 From: dazoe at dazoe.net (Dave Akers) Date: Thu, 5 Aug 2010 17:26:55 -0500 Subject: [mpm-itk] Chroot In-Reply-To: <20100802145823.GB7341@uio.no> References: <20100802141044.GA7341@uio.no> <20100802145823.GB7341@uio.no> Message-ID: If you knowingly set chrootdir then you are aware that .htaccess outside the chroot no longer have any effect. I have seen people asking for chroot with mpm_itk, my patch fixes/adds it, so why not include it. When using chroot with anything there are always extra steps to go through to set up the chroot, eg: copying libs, copying bins, etc. And a side note if ChrootDir has no effect in mpm_itk then the directive should be removed. On Mon, Aug 2, 2010 at 9:58 AM, Steinar H. Gunderson wrote: > On Mon, Aug 02, 2010 at 09:18:05AM -0500, Dave Akers wrote: >> And as far as the subsequent requests on the same connection, they have to >> be for the same host because itk allready dropped to the user for that host >> and it will not try to chroot if it is allready running as a user other >> than root. > > Sure, but how are you going to deal with, say, lookups to .htaccess files > that are no longer in the path? I doubt this is going to work very well for > arbitrary modules. > > /* Steinar */ > -- > Homepage: http://www.sesse.net/ > > _______________________________________________ > mpm-itk mailing list > mpm-itk at err.no > http://lists.err.no/mailman/listinfo/mpm-itk > From sgunderson at bigfoot.com Fri Aug 6 00:35:48 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Fri, 6 Aug 2010 00:35:48 +0200 Subject: [mpm-itk] Chroot In-Reply-To: References: <20100802141044.GA7341@uio.no> <20100802145823.GB7341@uio.no> Message-ID: <20100805223548.GB26311@uio.no> On Thu, Aug 05, 2010 at 05:26:55PM -0500, Dave Akers wrote: > If you knowingly set chrootdir then you are aware that .htaccess > outside the chroot no longer have any effect. This is not a matter of ?no effect? -- it's a matter of unpredictable and very confusing effects (down to the possibility of introducing security holes). For instance, what are you going to do about the .htaccess cache? > I have seen people asking for chroot with mpm_itk, my patch fixes/adds it, > so why not include it. Because it breaks in very common circumstances, and adds complexity for little gain? Patches are included if someone can convince the upstream author they make sense, not just because someone claims someone has requested a given feature at some point. (As an example, you might notice that mpm-itk is not in Apache upstream. :-) ) > And a side note if ChrootDir has no effect in mpm_itk then the > directive should be removed. You could argue that the directive should live inside an #ifndef MPM_ITK or similar, yes. It's not terribly high on the priority list, though. /* Steinar */ -- Homepage: http://www.sesse.net/ From sgunderson at bigfoot.com Fri Aug 6 01:24:23 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Fri, 6 Aug 2010 01:24:23 +0200 Subject: [mpm-itk] Switching vhosts during the same connection in Debian In-Reply-To: <20100730235544.GA25932@uio.no> References: <20100723162027.GU17040@stud.ntnu.no> <20100723162611.GB2008@uio.no> <20100730235544.GA25932@uio.no> Message-ID: <20100805232423.GC26311@uio.no> On Sat, Jul 31, 2010 at 01:55:44AM +0200, Steinar H. Gunderson wrote: > You can try this patch as a first approximation. The logging doesn't actually > appear to work, it's pretty much untested, and probably needs some more > sanity checking, but it just might help with your issue: Here's an updated version, with some small changes: It fixes the formatting, adds an explanatory comment, fixes the logging, and makes sure the logic doesn't trigger in subrequests. I'm still not 100% sure it cannot trigger on the first request in a connection (given some obscure special case); testing welcome. /* Steinar */ -- Homepage: http://www.sesse.net/ -------------- next part -------------- A non-text attachment was scrubbed... Name: 07-fix-htaccess-issues.patch Type: text/x-diff Size: 2331 bytes Desc: not available URL: From garybrooks at cloudaccess.net Fri Aug 6 04:36:55 2010 From: garybrooks at cloudaccess.net (Gary Brooks) Date: Thu, 5 Aug 2010 22:36:55 -0400 Subject: [mpm-itk] Chroot In-Reply-To: References: <20100802141044.GA7341@uio.no> <20100802145823.GB7341@uio.no> Message-ID: Interesting one, though I think chrooting would slow down the server. On Thu, Aug 5, 2010 at 6:26 PM, Dave Akers wrote: > If you knowingly set chrootdir then you are aware that .htaccess > outside the chroot no longer have any effect. I have seen people > asking for chroot with mpm_itk, my patch fixes/adds it, so why not > include it. When using chroot with anything there are always extra > steps to go through to set up the chroot, eg: copying libs, copying > bins, etc. > And a side note if ChrootDir has no effect in mpm_itk then the > directive should be removed. > > On Mon, Aug 2, 2010 at 9:58 AM, Steinar H. Gunderson > wrote: > > On Mon, Aug 02, 2010 at 09:18:05AM -0500, Dave Akers wrote: > >> And as far as the subsequent requests on the same connection, they have > to > >> be for the same host because itk allready dropped to the user for that > host > >> and it will not try to chroot if it is allready running as a user other > >> than root. > > > > Sure, but how are you going to deal with, say, lookups to .htaccess files > > that are no longer in the path? I doubt this is going to work very well > for > > arbitrary modules. > > > > /* Steinar */ > > -- > > Homepage: http://www.sesse.net/ > > > > _______________________________________________ > > mpm-itk mailing list > > mpm-itk at err.no > > http://lists.err.no/mailman/listinfo/mpm-itk > > > > _______________________________________________ > mpm-itk mailing list > mpm-itk at err.no > http://lists.err.no/mailman/listinfo/mpm-itk > -- *Gary Brooks | CEO* garybrooks at cloudaccess.net *Phone: * +1-231-844-4053 Extension 101 *Cell Phone:* +1-231-818-0078 *Skype id: *garyjaybrooks2000 *Fax: * 313-899-7032 *Web: * http://www.cloudaccess.net *Address:* 10850 Traverse Hwy, Suite 1135 | Traverse City, Michigan 49684 -------------- next part -------------- An HTML attachment was scrubbed... URL: From forum at hubbitus.com.ru Tue Aug 31 17:12:09 2010 From: forum at hubbitus.com.ru (Pavel Alexeev (aka Pahan-Hubbitus)) Date: Tue, 31 Aug 2010 19:12:09 +0400 Subject: [mpm-itk] Segfault in /opt/1C/v8.1/i386/core81.so Message-ID: <4C7D1BC9.3050603@ru.bir.ru> I use soap apache module from 1C distribution ( http://www.1c.ru/eng/title.htm ) together MPM-itk. And in logs seen many errors like: [Tue Aug 31 18:48:36 2010] [error] child died with signal 11 [Tue Aug 31 18:48:37 2010] [error] child died with signal 11 [Tue Aug 31 18:48:38 2010] [error] child died with signal 11 [Tue Aug 31 18:48:39 2010] [error] child died with signal 11 [Tue Aug 31 18:48:40 2010] [error] child died with signal 11 [Tue Aug 31 18:48:41 2010] [error] child died with signal 11 [Tue Aug 31 18:48:42 2010] [error] child died with signal 11 [Tue Aug 31 18:48:43 2010] [error] child died with signal 11 [Tue Aug 31 18:48:44 2010] [error] child died with signal 11 [Tue Aug 31 18:48:45 2010] [error] child died with signal 11 [Tue Aug 31 18:48:46 2010] [error] child died with signal 11 pure virtual method called terminate called without an active exception [Tue Aug 31 18:55:03 2010] [error] child died with signal 6 pure virtual method called terminate called without an active exception [Tue Aug 31 18:55:03 2010] [error] child died with signal 6 pure virtual method called terminate called without an active exception [Tue Aug 31 18:55:04 2010] [error] child died with signal 6 pure virtual method called terminate called without an active exception [Tue Aug 31 19:00:02 2010] [error] child died with signal 6 This module is propritary software and I have not sources. I setup core dumps, and see this backtrace from it: Loaded symbols for /opt/1C/v8.1/i386/backbas.so Core was generated by `/usr/sbin/httpd.itk'. Program terminated with signal 11, Segmentation fault. #0 0x012925b9 in core::Thread::join() () from /opt/1C/v8.1/i386/core81.so (gdb) (gdb) bt #0 0x012925b9 in core::Thread::join() () from /opt/1C/v8.1/i386/core81.so #1 0x00bf241c in ?? () from /opt/1C/v8.1/i386/wscntr.so #2 0x00bfa4ab in ?? () from /opt/1C/v8.1/i386/wscntr.so #3 0x00bf83d4 in ?? () from /opt/1C/v8.1/i386/wscntr.so #4 0x00bf8461 in ?? () from /opt/1C/v8.1/i386/wscntr.so #5 0x01af9e49 in core::create_string_value(stlp_std::basic_string, stlp_std::allocator > const&) () from /opt/1C/v8.1/i386/ws.so #6 0x01afa1fe in core::create_string_value(stlp_std::basic_string, stlp_std::allocator > const&) () from /opt/1C/v8.1/i386/ws.so #7 0x01afa4eb in core::create_string_value(stlp_std::basic_string, stlp_std::allocator > const&) () from /opt/1C/v8.1/i386/ws.so #8 0x00433207 in ?? () from /opt/1C/v8.1/i386/wsap22.so #9 0x005980f9 in exit () from /lib/i686/nosegneg/libc.so.6 #10 0x00bca9a9 in ?? () #11 0x00bcae21 in ?? () #12 0x00bcb823 in ap_mpm_run () #13 0x00ba088e in main () Is it 1C problem or MPM-ITK? Can I do with it something? -- With best wishes, Pavel Alexeev aka Pahan-Hubbitus. For fast contact with me you could use Jabber: Hubbitus at jabber.ru -------------- next part -------------- An HTML attachment was scrubbed... URL: From sgunderson at bigfoot.com Tue Aug 31 17:27:23 2010 From: sgunderson at bigfoot.com (Steinar H. Gunderson) Date: Tue, 31 Aug 2010 17:27:23 +0200 Subject: [mpm-itk] Segfault in /opt/1C/v8.1/i386/core81.so In-Reply-To: <4C7D1BC9.3050603@ru.bir.ru> References: <4C7D1BC9.3050603@ru.bir.ru> Message-ID: <20100831152723.GA24506@samfundet.no> On Tue, Aug 31, 2010 at 07:12:09PM +0400, Pavel Alexeev (aka Pahan-Hubbitus) wrote: > #0 0x012925b9 in core::Thread::join() () from /opt/1C/v8.1/i386/core81.so > (gdb) The crash here is definitely in the 1C module. > Is it 1C problem or MPM-ITK? Can I do with it something? You'll need to talk to the vendor who sold you the module. /* Steinar */ -- Homepage: http://www.sesse.net/