Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755386AbcCTLhp (ORCPT ); Sun, 20 Mar 2016 07:37:45 -0400 Received: from mail-pf0-f194.google.com ([209.85.192.194]:35575 "EHLO mail-pf0-f194.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754117AbcCTLhg (ORCPT ); Sun, 20 Mar 2016 07:37:36 -0400 Date: Sun, 20 Mar 2016 19:31:47 +0800 From: Peter Chen To: Linus Torvalds Cc: Greg KH , Peter Chen , Philipp Zabel , Alan Stern , Rob Herring , Arnd Bergmann , Andrew Morton , Linux Kernel Mailing List , USB list Subject: Re: [GIT PULL] USB driver patches for 4.6-rc1 Message-ID: <20160320113147.GB30461@shlinux2.ap.freescale.net> References: <20160317000918.GA12818@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1248 Lines: 32 On Fri, Mar 18, 2016 at 04:01:51PM -0700, Linus Torvalds wrote: > On Fri, Mar 18, 2016 at 3:51 PM, Linus Torvalds > wrote: > > > > The commit that ends up being marked bad is odd, but there it is: > > 69bec7259853 "USB: core: let USB device know device node". > > Confirmed. Not only did it bisect to that, reverting it on top of the > current kernel fixes my machine. > > So that commit is somehow buggy. I don't see what it does that would > break even with OF disabled, but something does. > > I'll just revert it. The way it is done seems bogus anyway. It looks > at of_node when OF is disabled, but generally that isn't even > initialized as far as I can tell, and we have things like > dev_of_node() helpers to make sure you don't do that. > I am sorry to make things break, Nicolai Stange's found the root cause for this problem, and his patch fixed it. USB device structure (both struct usb_hcd and struct usb_device) is initialized by kzalloc, so the struct device in it is initialized by zero, and will not cause non-initialized for USB device, but you are right, a good practice is using dev_of_node for all devices in case the struct device is not zero-initialized. -- Best Regards, Peter Chen