From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Ahern Subject: Re: USB EHCI patch for 0.14.0? Date: Fri, 11 Mar 2011 09:34:13 -0700 Message-ID: <4D7A4F05.2020801@cisco.com> References: <0NZZCL-1Py46m2UCX-0000eH@icpu525.kundenserver.de> <4D7A474D.5050807@cisco.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: erik.rull@rdsoftware.de, kvm@vger.kernel.org To: Markus Armbruster Return-path: Received: from sj-iport-1.cisco.com ([171.71.176.70]:31382 "EHLO sj-iport-1.cisco.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752546Ab1CKQdc (ORCPT ); Fri, 11 Mar 2011 11:33:32 -0500 In-Reply-To: Sender: kvm-owner@vger.kernel.org List-ID: On 03/11/11 09:17, Markus Armbruster wrote: > David Ahern writes: > >> On 03/11/11 08:39, Markus Armbruster wrote: >>> erik.rull@rdsoftware.de writes: >>> >>>> Hi David, >>>> >>>> I did a second iteration and it looked way better, maybe my first attempts >>>> were somehow buggy. >>>> >>>> First - please review your DPRINTF in the usb-ehci.c, there is a variable >>>> "dev" undefined in line 504/505 when enabling the debugging defines at the >>>> top of the file, the compiler complains there. >>>> >>>> I tested again with your hints, here my results: >>>> >>>> - using -device usb-host causes windows not to boot completely, using >>>> -usbdevice host:auto:*.* is fine! >>>> >>>> - using -usbdevice tablet is better than using -device usb-tablet >>> >>> Better inhowfar? >> >> As I recall the USB code is the ignored step-child in qemu; it was never >> properly/fully integrated into the qdev stuff. The paths appear to >> differ in how usb devices are handled. I for one have always stuck with >> the -usbdevice route. > > Please report any deficiencies so we can fix this. Erik reported a few above. David > > -usbdevice is around for backward compatibility only. There are already > cases where -device provides you more control than -usbdevice. This > feature gap will only grow.