From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:54667) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Sdfk3-0006uT-Er for qemu-devel@nongnu.org; Sun, 10 Jun 2012 06:51:32 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Sdfk1-000520-GT for qemu-devel@nongnu.org; Sun, 10 Jun 2012 06:51:30 -0400 Received: from moutng.kundenserver.de ([212.227.126.171]:52366) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Sdfk1-00050u-76 for qemu-devel@nongnu.org; Sun, 10 Jun 2012 06:51:29 -0400 Message-ID: <4FD47C31.8080701@rdsoftware.de> Date: Sun, 10 Jun 2012 12:51:29 +0200 From: Erik Rull MIME-Version: 1.0 References: <4FD22966.7020307@rdsoftware.de> <4FD2426D.3080607@redhat.com> <4FD266EF.60502@rdsoftware.de> <4FD26F71.2030803@redhat.com> In-Reply-To: <4FD26F71.2030803@redhat.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] USB Hostport Differences List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Hans de Goede Cc: "qemu-devel@nongnu.org" , Gerd Hoffmann Hans de Goede wrote: > Hi, > > On 06/08/2012 10:56 PM, Erik Rull wrote: >> Hans de Goede wrote: >>> Hi, >>> >>> On 06/08/2012 06:33 PM, Erik Rull wrote: >>>> Hi all, >>>> >>>> when assigning USB host devices to a guest using the hostport option, >>>> there seem to be different formats, when calling info usbhost: >>>> >>>> - On my vanilla kernel linux there is a hostport format e.g. "1.5" or >>>> "1.2" >>>> - On my Debian 6.0 full blown linux there is a hostport format "2" or >>>> "4", that means, there are no dots and only one number >>>> >>> >>> Hmm, is this on the same machine, with the usb devices hooked up the same >>> way? Normally these differences come >>> from there being hubs in the chain, ie "2" or "4" indicate devices plugged >>> directly into a root hub, >>> "1.5" and "1.2" mean devices plugged into a hub, which itself is plugged >>> into root port "1" >>> >>> Regards, >>> >>> Hans >> >> No it's on different machines - first with Intel Hardware, second with >> AMD Hardware. I have no hubs attached, I use the ports that are offered >> directly on the mainboard (EPIC Nano Connectors / ATX rear side panel). >> On all other Intel systems I see the same effect with the "1.x" ports. >> Only the AMD system shows the single numbers. >> > > Those would be sandy bridge or newer Intel machines then, these no longer have > an EHCI usb controller + UHCI companion controllers, but only an EHCI > controller, > so the root ports are USB-2 only. In order for USB-1 devices to still work the > chipset has an integrated USB-2 hub (which can handle USB-1 ports), of you do > lsusb you should see something like this in there: > > Bus 001 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub > > Which is why you get the 1.x for the motherboard ports, because there is > actually > a hub between the root hub and the ports. > > Regards, > > Hans > Hi Hans, thanks for the explanation - so I will try to find out which way of USB handling is given on the current system and then add the correct hostport filters. Is there an easier way of finding this kind of architecture automatically beside grep'ing for such a Hub? Best regards, Erik