Problem writing to video device

[Ralph Metzler <rjkm@netcologne.de>]

stefan@pettersson.tc writes:
 > I am trying to make a video recorder for my Siemens Fujitsu DVB card, using 
 > 
 > driver version 0.05. The problem occurs when I try to playback a previous 
 > 
 > recorded file and write from my memory buffer to the videodevice. The problem is 
 > 
 > that the process hangs after a couple of seconds (sometimes longer and sometimes 
 > 
 > shorter time.) After a while it hangs almost immediately and then I get no 
 > 
 > picture at all. After having reinstalled the driver modules I can run it for a 
 > 
 > couple of seconds again and so on. I have traced the problem and it hangs in the 
 > 
 > write(buffer => device) command. But I am stuck with the problem - I cannot 
 > 
 > figure out why it happens.
 > 
 > 
 > 
 > I am running Red Hat linux and a kernel version 2.2.14. I am also using the 
 > 
 > i2c-core.c file from the last driver release as described in the news file.
 > 
 > 
 > 
 > To test my hardware and the driver, I tried to run the gVideo application. It 
 > 
 > works pretty well, also the playback function, but sometimes it also fails. Then 
 > 
 > I get a black screen which remains even if I quit the program and start it over 
 > 
 > again. The problem remains until I have reinstalled the driver modules into the 
 > 
 > kernel and/or rebooted the computer. (And sometimes the computer hang when I try 
 > 
 > to remove/install the modules)
 > 
 > 
 > 
 > After a long time of debugging I am starting to getting pretty mad. Therefore I 
 > 
 > would like to ask if there is someone knowing anything about this problem - if I 
 > 
 > am doing anything completly wrong or if there is some problem with the driver 
 > 
 > itself. I would appreciate if someone knowing anything about the problem would 
 > 
 > like to please tell me somethings about it...  
 > 


Please wait until friday for the new driver release.
The driver as well as the firmware will be much more stable.

I donīt know if your specific problem will be solved, but we changed
the whole resource locking mechanism inside the driver.
The old code caused a lot of symptoms like you describe, even total
system locks on SMP machines. This should no longer happen.


Ralph