Mailing-List: contact cygwin-apps-help AT sourceware DOT cygnus DOT com; run by ezmlm
Sender: cygwin-apps-owner AT sourceware DOT cygnus DOT com
List-Subscribe: <mailto:cygwin-apps-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-apps/>
List-Post: <mailto:cygwin-apps AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-apps-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/lists.html#faqs>
Delivered-To: mailing list cygwin-apps AT sources DOT redhat DOT com
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Subject: RE: accept() failed message on kde 1.1.2 ??
content-class: urn:content-classes:message
X-MimeOLE: Produced By Microsoft Exchange V6.0.4417.0
Date: Mon, 18 Jun 2001 16:17:39 +1000
Message-ID: <EA18B9FA0FE4194AA2B4CDB91F73C0EF7A06@itdomain002.itdomain.net.au>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: accept() failed message on kde 1.1.2 ??
Thread-Index: AcD3u/Sh4kc8BMurRyy+fRKGo/MSUQAAPRjw
From: "Robert Collins" <robert DOT collins AT itdomain DOT com DOT au>
To: <cygwin-apps AT sources DOT redhat DOT com>
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id CAA28805

> -----Original Message-----
> From: egor duda [mailto:deo AT logos-m DOT ru]
> Sent: Monday, June 18, 2001 4:07 PM
> To: Christopher Faylor
> Subject: Re: accept() failed message on kde 1.1.2 ??
> 
> 
> Hi!
> 
> Sunday, 17 June, 2001 Christopher Faylor cgf AT redhat DOT com wrote:
> 
> CF> Is there any traction on this problem?  I see that Suhaib 
> has released
> CF> a new version of XFree86 which turns off UNIX-domain 
> sockets, which
> CF> is unfortunate.
> 
> CF> Do we know what is causing the problems that we're seeing?
> 
> i still can't reproduce it :(, so only one thing i can suggest is to
> disable security checks for now. I hope, however, that somebody can
> either debug this problem himself or provide a reproducible testcase,
> so i can debug it myself.
> 
> BTW, Ralf, are those sockets that fail in you test work in blocking or
> non-blocking mode?
> 

Egor, I've been trying to grok the code. Can you tell me if I've got
this right?

on bind() you create a cookie and write it to the file. You also create
a event with a name related to the cookie.
on connect() you read the file and attempt to open the previously
created event. A failure to open is used to indicate failure. 

I ask because I've been playing around with this on win2k sp1, ntsec on.
With XFree86 I get a "inode has changed" error on the /tmp/.X11-unix/foo
socket. Removing the .X11-unix directory fixes that. I suspect it's
related to problem Ralf is describing.

Anyway back to the AF_UNIX sockets - two questions:
1) what stops someone using the NT namespace to see the available events
with a name format like the secure socket event names?
2) Why not allow the accept() to happen at the win32 level, but read and
write the cookie between cygwin, before letting the user program know
that cygwin is involved? It seems to me that that would be a somewhat
easier solution. 

Rob