From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Seidel Subject: Re: Bug: "git-svn fetch" segfaults on unknown user Date: Thu, 6 Nov 2008 16:48:09 +0100 Message-ID: <20081106154808.GA27102@merkur.sol.de> References: <20081106143826.GA26900@merkur.sol.de> <237967ef0811060643g1037b8dbuf3660dfa8e3ff625@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii To: git@vger.kernel.org X-From: git-owner@vger.kernel.org Thu Nov 06 16:50:37 2008 connect(): Connection refused Return-path: Envelope-to: gcvg-git-2@gmane.org Received: from vger.kernel.org ([209.132.176.167]) by lo.gmane.org with esmtp (Exim 4.50) id 1Ky77u-0005I2-16 for gcvg-git-2@gmane.org; Thu, 06 Nov 2008 16:50:30 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750972AbYKFPtK (ORCPT ); Thu, 6 Nov 2008 10:49:10 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751158AbYKFPtI (ORCPT ); Thu, 6 Nov 2008 10:49:08 -0500 Received: from fmmailgate03.web.de ([217.72.192.234]:55209 "EHLO fmmailgate03.web.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750892AbYKFPtH (ORCPT ); Thu, 6 Nov 2008 10:49:07 -0500 Received: from smtp06.web.de (fmsmtp06.dlan.cinetic.de [172.20.5.172]) by fmmailgate03.web.de (Postfix) with ESMTP id E658DF3F5818 for ; Thu, 6 Nov 2008 16:49:05 +0100 (CET) Received: from [77.179.124.209] (helo=merkur.sol.de) by smtp06.web.de with asmtp (TLSv1:AES256-SHA:256) (WEB.DE 4.109 #226) id 1Ky76X-0004cI-00 for git@vger.kernel.org; Thu, 06 Nov 2008 16:49:05 +0100 Received: from jens by merkur.sol.de with local (Exim 4.69) (envelope-from ) id 1Ky75d-00073P-Os for git@vger.kernel.org; Thu, 06 Nov 2008 16:48:10 +0100 Mail-Followup-To: git@vger.kernel.org Content-Disposition: inline In-Reply-To: <237967ef0811060643g1037b8dbuf3660dfa8e3ff625@mail.gmail.com> User-Agent: Mutt/1.5.18 (2008-05-17) X-Sender: tux-master@web.de X-Provags-ID: V01U2FsdGVkX1/OxUoe3TlW2aohS9Gqpkyx2+g6GR+J5ZL19WlG 6YCZOpCl5byk95RcDrCz71gZbJfaGxMzjk9eUR0Dt/JAnYV1eN jv/2IYiUo= Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Thu, Nov 06, 2008 at 03:43:18PM +0100, Mikael Magnusson wrote: > 2008/11/6 Jens Seidel : > > I noticed the following bug in git-svn version 1.6.0.3 (svn 1.5.5): > > > > "git-svn fetch" segfaults if I provide an incomplete svn.authorsfile config > > file setting with account<-->email mappings. This happens if an unknown user > > account is found in the Subversion log. I should really provide more information: I imported my Subversion repository in the past succesfully from my server (svn+ssh protocol). Since I'm new to git (and still confused by the large amount of git tools) I decided to start on a svnsync'ed copy of my repository (file protocol) to avoid any harm. I added also some new config settings including svn.authorsfile and fetching revisions started to fail. The last output was: r38 = 4dca32e3d82a32206fd11545d4f270e118f51f12 (trunk@166) M myproject/file Revision 39 contained the unknown user account so I suspected the error is related to an incomplete authors file. I used following versions of Subversion: Client: 1.5.5 Server: 1.5.3 After adding the missing account to the authors file it fetched all revisions, called the garbage collector and stilled crashed with a core dump. Last output: Checked out HEAD: file:///home/jens/Subversion-Repository.synced/myproject/trunk r1700 Abgebrochen (core dumped) This happened also after I removed the svn.authorsfile config setting. My current settings: [user] name = Jens Seidel email = jensseidel@users.sf.net [color] diff = auto branch = auto interactive = auto [svn] findcopiesharder = true rmdir = true "gdb -c core" is useless, it just mentions the error occured in "perl git-svn fetch" and there is no stacktrace (don't know how to debug perl). I removed now both [svn] options and it failed again. So it seems a problem with file:// access while svn+ssh:// works well. > Are you sure this isn't the regular svn 1.5.x crash? Ie, does > compiling subversion with --disable-runtime-module-search fix it? Huch? I follow the Subversion developer list and I'm not not aware of such a problem. Will try it now ... PS: Compared to previous releases of git the import is at least 10 times faster. Good job! Jens