From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kay Sievers Subject: Re: [PATCH] atm: correct sysfs 'device' link creation and parent relationships Date: Tue, 7 Dec 2010 08:35:53 +0100 Message-ID: References: <1291587480.17538.12.camel@dcbw.foobar.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org, Duncan Sands , linux-usb@vger.kernel.org, Chas Williams To: Dan Williams Return-path: Received: from mail-gx0-f180.google.com ([209.85.161.180]:44785 "EHLO mail-gx0-f180.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752425Ab0LGHgJ convert rfc822-to-8bit (ORCPT ); Tue, 7 Dec 2010 02:36:09 -0500 In-Reply-To: <1291587480.17538.12.camel@dcbw.foobar.com> Sender: netdev-owner@vger.kernel.org List-ID: On Sun, Dec 5, 2010 at 23:17, Dan Williams wrote: > The ATM subsystem was incorrectly creating the 'device' link for ATM > nodes in sysfs. =C2=A0This led to incorrect device/parent relationshi= ps > exposed by sysfs and udev. =C2=A0Instead of rolling the 'device' link= by hand > in the generic ATM code, pass each ATM driver's bus device down to th= e > sysfs code and let sysfs do this stuff correctly. Looks good to me. It's wrong for any subsystem to ever create a file or link named "device". It's a built-in driver-core managed property which appears automatically as soon as the parent-device relations are correctly established. Udev/libudev intentionally refuses to follow any devices pointed to by "broken" device links. So this patch is needed to make ATM devices working properly with usual userspace tools. Thanks, Kay