Re: mga_vid with 2.6.24?

ael wrote:
> Attila Kinali wrote:
[..snip..]
>>>>> It failed with a complaint that
>>>>> 'struct kobject' has no member named 'name'
>>>> Yes, old cruft failing on us.
>>>>
>>>> svn up
>>> Built new version. I am having a small problem with udev creating
>>> /dev/mga_vid/0 and a symlink to /dev/mga0 while my (debian etch) mplayer
>>>  is looking for /dev/mga and failing by hitting the directory.
> 
> Typos above :-) I meant /dev/mga_vid0 and /dev/mga_vid0.
> 
>> mplayer checks in ./configure whether /dev/mga_vid exists,
>> if it does (x)mga is compiled. At runtime it tries to access
>> /dev/mga_div. This looks like you have not properly configured
>> your udev subsystem. Have you read README.udev?
> 
> Yes. I was using that rule and I  am now using the debian mga_vid rule
> which seems to be the same.
> 
>>> I have pulled down the svn version of mplayer and am now compiling that
>>> just in case it is old mplayer code. Otherwise I will have to start
>>> debugging udev configuration and rules, I suppose. Is this a known issue?
>> Nope, i use udev at home and it works fine.
> 
> Ok, Thanks for the reply. I will reach for udevmonitor and so on if need
> be.

Just for the record, compiling and using the latest mplayer fixed that
problem. It may be that the debian etch package has tweaked the
configuration somehow.

ael