From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750734AbVJRPgt (ORCPT ); Tue, 18 Oct 2005 11:36:49 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750800AbVJRPgt (ORCPT ); Tue, 18 Oct 2005 11:36:49 -0400 Received: from cpu1185.adsl.bellglobal.com ([207.236.110.166]:53930 "EHLO mail.rtr.ca") by vger.kernel.org with ESMTP id S1750734AbVJRPgt (ORCPT ); Tue, 18 Oct 2005 11:36:49 -0400 Message-ID: <4355168A.8010507@rtr.ca> Date: Tue, 18 Oct 2005 11:36:42 -0400 From: Mark Lord User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.12) Gecko/20050923 Debian/1.7.12-0ubuntu05.04 X-Accept-Language: en, en-us MIME-Version: 1.0 To: Linux Kernel Subject: 2.6.14-RC2 Bug? HiSpeed USB devices demoted to low-speed after resume Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org For both suspend-to-disk ("software suspend") and suspend-to-ram, the 2.6.14-RC2 kernel appears to disable ehci_hcd on resume. That is, plugging in a hi-speed Mass Storage device *after* resume results in these syslog messages: >kernel: usb 3-1: new full speed USB device using uhci_hcd and address 2 >kernel: usb 3-1: not running at top speed; connect to a high speed hub Very odd. The 2.6.13 kernel behaves much better, as does 2.6.14-RC2 *before* suspending: >kernel: usb 5-5: new high speed USB device using ehci_hcd and address 3 Unloading and reloading ehci_hcd restores high-speed USB operation, until the next suspend/resume. WTF?