Re: [Em28xx] ehci_hcd map_single: unable to map unsafe buffer on a standard NSLU2

Thierry Merle a écrit :
> Markus Rechberger a écrit :
>> On 9/20/07, Thierry Merle <[email protected]> wrote:
>>  
>>> Hi Markus,
>>>
>>> To sum up:
>>> - on David's NSLU2, his tuner makes dma errors in ehci_hcd module.
>>> - on my NSLU2, his tuner if I plug his tuner directly on the NSLU2 with
>>> the v4l-dvb-experimental tree, I get a i2c error.
>>> With a hub plugging the device freezes the kernel but I have not wired
>>> the serial line to see a kernel oops).
>>> Attached is the /var/log/messages corresponding to my try.
>>>
>>> The NSLU2 is a based on a IXP420 CPU (ARMv5TE architecture). The
>>> internal USB driver is not used, on the PCI bus a NEC USB chip is 
>>> connected.
>>> See http://www.nslu2-linux.org/wiki/Info/CPUOverview for all infos.
>>>
>>> On the debian-arm ML, Lennart Sorensen invoked a CPU usage problem but
>>> he was talking about the MPEG-TS stream processing.
>>> My cinergyT2 can stream at 33% CPU user consumption on the NSLU2.
>>>
>>>     
>>
>> Sep 20 23:05:56 shagrat kernel: Em28xx: Initialized (Em28xx Audio
>> Extension) extension
>> Sep 20 23:05:57 shagrat kernel: em2880-dvb.c: DVB Init
>> Sep 20 23:05:57 shagrat kernel: Loading base firmware: 
>> xc3028_8MHz_init0.i2c.fw
>> Sep 20 23:05:57 shagrat kernel: xc3028-tuner.c: Unable to load firmware
>> Sep 20 23:05:57 shagrat kernel: xc3028-tuner.c: ** PLEASE HAVE A LOOK 
>> AT **
>> Sep 20 23:05:57 shagrat kernel: xc3028-tuner.c:
>> http://linuxtv.org/v4lwiki/index.php/Talk:Em2880#Firmware
>>
>> is udev started or even the right version of udev installed?
>> otherwise you won't be able to load the firmware..
>>
>> Markus
>>
>>   
OK sorry, I just realized that the /lib/firmware were not present.
I think the filesystem has not been flushed when the NSLU2 freezed last 
time. This corrupted the v4l-dvb-experimental tree and the /lib/firmware 
directory.

Now the driver loads correctly as before, but retrying to scan makes 
kernel freeze again. Good to see that I have a "stable" behaviour ;)
Since I do not have a serial console I can only suppose that this is a 
kernel panic...
I will try to wire it but it will take time to do it properly.

This is not a power drain problem since I have the same behaviour with 
or without a auto-powered hub.

Thierry

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel