Web lists-archives.com

Re: [Spca50x-devs] VX 1000; USB controller crash




On Thu, 22 May 2008 17:30:20 +0200, Felix Obenhuber <lists@xxxxxxxxxxxx> 
wrote:
>You're right. I'm using the v4l-dvb patch for a em2880 device. So I compiled a
>vanilla 2.6.23 to gain a environment closer to your's. Unfortunately the
>behaviour of the modules is almost equal:
	[snip]
>usb 3-2: new full speed USB device using uhci_hcd and address 2
>usb 3-2: configuration #1 chosen from 1 choice
>Linux video capture interface: v2.00
>gspca: main v0.2.8 registered
>gspca: probing 045e:00f7
>gspca: probe ok
>gspca: probing 045e:00f7
>gspca: probing 045e:00f7
>usbcore: registered new interface driver sonixj
>sonixj: v0.2.7 registered
>usbcore: registered new interface driver snd-usb-audio
>uhci_hcd 0000:00:1d.1: host system error, PCI problems?
>uhci_hcd 0000:00:1d.1: host controller halted, very bad!
>uhci_hcd 0000:00:1d.1: HC died; cleaning up
>usb 3-2: USB disconnect, address 2
>gspca: disconnect complete
	[snip]
>Afterward the device (and or the connected bus) seems to be completely down and
>doesn't appear in the lsusb output...

Hello Felix,

In your traces, I see a probe problem.

Normally, when you plug the webcam, the usb driver sets a probe
request in the system and wakes 'hal'which loads the specific driver
(here, 'sonixj'). At loading time the driver ('sonixj') registers
its probe and disconnect entries. Once done, it writes
'sonixj: v0... registered'. During the registration, the probe entry
is called (message 'gspca: probing xxx:xxx'). This one does a lot of
things (many USB exchanges) and ends with the device creation and
the message 'gspca: probe ok'.

In your trace, the probe entry is called 3 times. The first time
ends correctly (the device is fully created), but it seems the
probe request is activated again. These times, some error occurs
(I don't know which) and these requests fail. So, the device stay
ignored...

As the main driver (gspca_main) does not keep any trace of the
events, it cannot protect itself against these calls.

I could have a look at the kernel, but I am a bit busy. If anybody
has an idea...

Waiting for a fix of this problem, you may try to load the
subdriver before plugging the webcam (at system start time).

>I can provide some USB traffic captures if you'd like to take a look at.
> The version of the sonixj
>module is 0.2.7 - think you forgot to update. ?

No: there was no change in this subdriver.

Cheers.

-- 
Ken ar c'hentañ	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Spca50x-devs mailing list
Spca50x-devs@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/spca50x-devs