Mailing-List: contact dev-help@httpd.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@httpd.apache.org
Message-ID: <3ED8E72D.2080703@algroup.co.uk>
Date: Sat, 31 May 2003 18:32:29 +0100
From: Ben Laurie <ben@algroup.co.uk>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-GB;
 rv:1.4b) Gecko/20030507
MIME-Version: 1.0
To: dev@httpd.apache.org
Subject: Re: FW: Apache socket question
References: <PIEHKKDELELLBCEHLPLFOEAJCKAA.atli@gangverk.is>
 <2147483647.1054284774@[10.0.1.185]> <3ED78D1D.6090606@attglobal.net>
In-Reply-To: <3ED78D1D.6090606@attglobal.net>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Jeff Trawick wrote:

>> While this is probably a bad idea, I bet that if you could trick
>> ap_lingering_close() not to close your socket when the connection is
>> closed, you might be able to get away with this.
>>
>> The relevent line in server/connection.c:214 is:
>>
>> apr_socket_t *csd = ap_get_module_config(c->conn_config, &core_module);
>>
>> So, something like:
>>
>> ap_set_module_config(c->conn_config, &core_module, NULL);
>>
>> in your module might do the trick.  But, don't expect us to support
>> this.  =) -- justin
> 
> 
> in fact, we might intentionally do something to make that blow up :)
> 
> my suggestion on apache-modules was to set c->aborted in the
> pre-connection hook, presumably after passing the descriptor
> successfully to another process
> 
> ap_lingering_close() doesn't perform the shutdown() if c->aborted

Icky! Presumably the nice thing to do would be to invent a flag that
inhibits the shutdown?

Cheers,

Ben.

-- 
http://www.apache-ssl.org/ben.html       http://www.thebunker.net/

"There is no limit to what a man can do or how far he can go if he
doesn't mind who gets the credit." - Robert Woodruff