From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mx1.redhat.com ([209.132.183.28]:62450 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750849AbZKHHgQ (ORCPT ); Sun, 8 Nov 2009 02:36:16 -0500 Message-ID: <4AF6762F.6040208@redhat.com> Date: Sun, 08 Nov 2009 08:41:35 +0100 From: Hans de Goede MIME-Version: 1.0 To: =?ISO-8859-2?Q?N=E9meth_M=E1rton?= CC: Jean-Francois Moine , V4L Mailing List , Thomas Kaiser , Theodore Kilgore , Kyle Guinn Subject: Re: [PATCH 3/3] gspca pac7302/pac7311: separate the two subdrivers References: <4AEE04DE.2060300@freemail.hu> <4AF11D25.1080607@freemail.hu> In-Reply-To: <4AF11D25.1080607@freemail.hu> Content-Type: text/plain; charset=ISO-8859-2; format=flowed Content-Transfer-Encoding: 8bit Sender: linux-media-owner@vger.kernel.org List-ID: Hi Németh, I'm getting similar reports for other (non pac73xx) model webcams with 2.6.32 on eeepc's and other computers using the uhci usb driver. Can you try installing an older kernel (so 2.6.30) and then building and installing the latest v4l-dvb tree (with your changes in) over that ? I think what you are seeing here is a bug outside of the v4l subsystem, this is just a hunch though. Regards, Hans On 11/04/2009 07:20 AM, Németh Márton wrote: > Dear Jef, > > although I tested my patch on my development computer together with Labtec > Webcam 2200 (gspca_pac7302 driver) it seems that the patch may cause regression > on some computers. For example I tested the gspca_pac7302 driver from > http://linuxtv.org/hg/~jfrancois/gspca/ on top of Linux kernel 2.6.32-rc5 on > an EeePC 901. I get the following error message in dmesg: > > [ 4476.992201] usb 3-2: new full speed USB device using uhci_hcd and address 11 > [ 4477.230485] usb 3-2: New USB device found, idVendor=093a, idProduct=2626 > [ 4477.230507] usb 3-2: New USB device strings: Mfr=0, Product=0, SerialNumber=0 > [ 4477.231139] usb 3-2: configuration #1 chosen from 1 choice > [ 4477.417456] Linux video capture interface: v2.00 > [ 4477.437131] gspca: main v2.7.0 registered > [ 4477.443214] gspca: probing 093a:2626 > [ 4477.453491] gspca: /dev/video0 created > [ 4477.453541] gspca: probing 093a:2626 > [ 4477.453549] gspca: intf != 0 > [ 4477.453598] gspca: probing 093a:2626 > [ 4477.453605] gspca: intf != 0 > [ 4477.453755] usbcore: registered new interface driver pac7302 > [ 4477.453771] pac7302: registered > [ 4489.552153] gspca: set alt 8 err -71 > > I bisected the problem on EeePC 901 and the changeset 13373:99c23949b411 > (gspca - pac7302/pac7311: Separate the two subdrivers.) was marked as the first bad > commit. > > On my development computer the same configuration works correctly: > > [ 7872.020222] usb 3-1: new full speed USB device using uhci_hcd and address 4 > [ 7872.251240] usb 3-1: configuration #1 chosen from 1 choice > [ 7872.744755] Linux video capture interface: v2.00 > [ 7872.785032] gspca: main v2.7.0 registered > [ 7872.797061] gspca: probing 093a:2626 > [ 7872.807577] gspca: /dev/video0 created > [ 7872.809747] usbcore: registered new interface driver pac7302 > [ 7872.809798] pac7302: registered > > Is the separated driver working for you? > Do you have any idea what could went wrong? Maybe some timing problem? > > Regards, > > Márton Németh > > Németh Márton wrote: >> From: Márton Németh >> >> All PAC7311 specific functions remain in pac7311.c. All PAC7302 specific >> functions are moved to pac7302.c. The USB device table is also divided into >> two parts. This makes it possible to remove the sensor specific decisions >> from different functions and also remove sensor infromation from the USB >> device table. >> >> The common functions are just copied to both subdrivers. These common >> functions can be separated later to a common file or helper module. >> >> Signed-off-by: Márton Németh >> Cc: Thomas Kaiser >> Cc: Theodore Kilgore >> Cc: Kyle Guinn