[mythtv-users] Nova T-500 on Lucid stopped working

R. G. Newbury newbury at mandamus.org
Fri Aug 6 20:54:36 UTC 2010


On 08/06/2010 08:00 AM, me wrote:
> This card doesn't seem to appear in lsusb or lspci after updating from
> 2.6.32-23 to 2.6.32-24 - although there were other problems in the
> update so maybe something went horribly wrong in the process. Can
> someone verify what should be appearing with lsusb and lspci? I have
> never had a hardware failure before and it seems unlikely that a dual
> tuner would be completely dead. dmesg|grep -i dvb only shows the following:
>
> [  792.049374] usbcore: registered new interface driver dvb_usb_dib0700
>
> Any ideas.... I missed the game and I don't want the lame off the shelf
> set top box...
>

Does it show up in the output of hal-device? You should direct the 
output ('> junk') to a file: it's big.

Then match the hardware info which is reported (/dev/input/eventx or 
whatever) with the output of lspci or lsusb.

It should also show up in the output of Xorg.0.log, if hal-device sees 
it, and *hald/haldaemon is trying to use evdev with it as if it is a 
keyboard or mouse. This appears to be due to some recent changes in the 
kernel. Its got me on the change from 2.6.32.10-76 to 2.6.32.11-99.


It appears that the newer hal, in its attempts to completely insulate us 
from dealing with the hardware, and make hot-plugging of all things 
work, creates another /dev socket as if the hardware were attached at 
that point. However the hardware continues to be handled by its usual 
/dev point (/dev/video etc).

Except: that hal and evdev may be a little too eager to grab stuff, and 
ATM it appears that hal + evdev may have a problem with USB devices not 
being handled properly. ATM my mythbox gives no keyboard output unless I 
kill haldaemon completely (the alleged xorg.conf 'fix' does not work).

Weirdly, this is entirely an X level problem: the keyboard works if 
booted to a console level, and goes away when startx is run. (The mouse 
continues to work. So messages in dmesg and the top of Xorg.0.log are 
misleading.

So boot into a console level (3 not 5), and test your hardware from that 
level. Delete the Xorg.0.log file and run startx. If the hardware goes 
away, then kill X (I ssh in from my laptop and do a 'killall X' since I 
cannot use Ctrl-Alt-Backspace with no working keyboard!) Examine the 
Xorg.0.log file and the tail of dmesg. Then try again after stopping 
hald/haldaemon whichever it is on your box.

There are supposed to be ways to fix this with files in 
/etc/hal/fdi/policy or .../fdi/preprobe but so far I can only plead the 
Schulz defence: I know nothing!.

              R. Geoffrey Newbury			
            Barrister and Solicitor
       Suite 106, 150 Lakeshore Road West
          Mississauga, Ontario, L5H 3R2

         o905-271-9600 f905-271-1638
           newbury at mandamus.org


More information about the mythtv-users mailing list