X-Recipient: archive-cygwin AT delorie DOT com
X-SWARE-Spam-Status: No, hits=-1.6 required=5.0	tests=AWL,BAYES_00,TW_CG
X-Spam-Check-By: sourceware.org
Message-ID: <4D3C8109.4030808@bopp.net>
Date: Sun, 23 Jan 2011 13:27:05 -0600
From: Jeremy Bopp <jeremy AT bopp DOT net>
User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.13) Gecko/20101208 Thunderbird/3.1.7
MIME-Version: 1.0
To: cygwin AT cygwin DOT com
Subject: Re: failing to clone a git repo via ssh
References: <4D3BFC69 DOT 7080405 AT cpan DOT org> <4D3C1D54 DOT 7050909 AT cpan DOT org>
In-Reply-To: <4D3C1D54.7050909@cpan.org>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-IsSubscribed: yes
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Id: <cygwin.cygwin.com>
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

On 01/23/2011 06:21 AM, Rafael Kitover wrote:
> I just realized this bug is replicatable without having ssh access to
> our repo, you just need the cygwin sshd enabled, and the guy with access
> to the gitosis went off somewhere anyway...
> 
> Here are the steps:
> 
> cd ~
> mkdir tmp
> cd tmp
> git clone git://git.shadowcat.co.uk/dbsrgits/DBIx-Class.git dbic_git_url
> git clone `whoami`@localhost:tmp/dbic_git_url dbic_ssh

Sadly, the early EOFs problem is an identified but unfixed issue:

http://cygwin.com/ml/cygwin/2010-07/msg00413.html
http://cygwin.com/ml/cygwin/2010-10/msg00044.html

Since my last report in that thread, I did try a few other configurations:

* msysGit with msysGit's ssh
* msysGit with Cygwin's ssh
* Cygwin's git with msysGit's ssh

All of these combinations avoided the early EOFs problem no matter how
many times I repeated my testing.  As cgf said, this does appear to be a
problem in Cygwin's pipe code, but it's very strange that it only seems
to be triggered with Cygwin's git + Cygwin's ssh.  My guess is that
there is some kind of race condition in the pipe setup code when both
ends of the pipe are Cygwin processes, but I'm admittedly unfamiliar
with Cygwin's pipe code.

-Jeremy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple