Mailing List archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[linux-dvb] Re: PROC_FS patch and other stuff



Dennis Noordsij wrote:
>>What about a writeable complete register dump for each dvb_i2c device?
>>Then you don't have to reimplement this for every new frontend driver.
> 
> 
> Sounds interesting. But right now I don't need to change any registers on the 
> device directly, but rather specify that when new frontend parameters are 
> set, register values XXX are used to program the frontend demodulator. I.e. 
> the register values used by the driver need to be modified, so that the 
> driver will use the new values instead of the default (as opposed to changing 
> a register value, then having the driver write the default one again in the 
> next frontend change).
> 
> I hope that made sense :-) 
> 
> Sorry if I misunderstood you.

no, you didn't - your approach is just a little bit more highlevel...


>>Yes, to make it simple you can add a two proc files: one that takes a
>>binary dvb_frontend_parameters struct and one that takes one in ascii
>>format like generated by the scan utility and czap/tzap.
>>The dump and parser code you can take from czap.c, tzap.c and scan/dump.c
> 
> 
> In addition to the existing ones, or to replace them?

They could replace these procfs entries that mapped directly to members 
of the dvb_frontend_parameter struct.


> Sounds like a good idea, just wasn't sure if you'd be happy with the added 
> complexity inside the kernel (i.e. parsers and pre/post processors). But I 
> will add them :)

I want to make all this code conditional too - things like this don't 
have to do anything on production machines. They are debug tools in my mind.

Holger




-- 
Info:
To unsubscribe send a mail to listar@linuxtv.org with "unsubscribe linux-dvb" as subject.



Home | Main Index | Thread Index