From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261578AbVALX0Y (ORCPT ); Wed, 12 Jan 2005 18:26:24 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261579AbVALXZr (ORCPT ); Wed, 12 Jan 2005 18:25:47 -0500 Received: from mail.kroah.org ([69.55.234.183]:44173 "EHLO perch.kroah.org") by vger.kernel.org with ESMTP id S261578AbVALXYO (ORCPT ); Wed, 12 Jan 2005 18:24:14 -0500 Date: Wed, 12 Jan 2005 15:21:12 -0800 From: Greg KH To: Jon Smirl Cc: lkml Subject: Re: [PATCH] Export symbol from I2C eeprom driver Message-ID: <20050112232112.GD15085@kroah.com> References: <9e47339105010721347fbeb907@mail.gmail.com> <20050108055315.GC8571@kroah.com> <9e473391050107220875baa32b@mail.gmail.com> <20050108222719.GA3226@kroah.com> <9e473391050108161426b36e4d@mail.gmail.com> <20050110234726.GE3286@kroah.com> <9e4733910501101820388563bb@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <9e4733910501101820388563bb@mail.gmail.com> User-Agent: Mutt/1.5.6i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Jan 10, 2005 at 09:20:49PM -0500, Jon Smirl wrote: > On Mon, 10 Jan 2005 15:47:26 -0800, Greg KH wrote: > > > I don't want to load the driver from the script because the radeon > > > driver is creating a sysfs link into the eeprom directory from the > > > radeon one. > > > > How are you getting the kobject to the eeprom directory from the radeon > > driver? > > > > I own the private I2C bus and eeprom is the only chip that will attach > to the bus. I need to do the link in the driver since there are four > busses and upto two monitors. The driver knows how to pair the head up > with the right bus. > > if (dev_priv->primary_head.connector != ddc_none) > list_for_each(item, > &dev_priv->i2c[dev_priv->primary_head.connector].adapter.clients) { > client = list_entry(item, struct i2c_client, list); > sysfs_create_link(&dev->primary.dev_class->kobj, > &client->dev.kobj, "monitor"); > break; > } Ick. Oh well, sure, that's ok. But I really think that a Kconfig rule could be made for this, instead of trying to pull an exported symbol in. And I have a patch in my queue to delete that id, so you will have to come up with another symbol in the driver to do that with :) thanks, greg k-h