[mythtv-users] OT: unreliable firmware load after cold boot

R. G. Newbury newbury at mandamus.org
Wed Jul 18 19:01:57 UTC 2012


On 07/18/2012 02:08 PM, Devin Heitmueller wrote:
> On Wed, Jul 18, 2012 at 2:03 PM, R. G. Newbury<newbury at mandamus.org>  wrote:
>> But the problem is with the required sub-modules which are not (visibly)
>> listed as dependencies.
>
> This is actually the expected behavior.  Unlike many drivers, tuner
> and demodulator drivers which are subdevices of a bridge don't use the
> standard module dependency mechanism.  Otherwise it would result in
> *every* tuner and demodulator driver for a given bridge to be
> required, which for many PCI or USB bridge can be in the dozens.
>
> For example, if you have an em28xx based product, there can be a
> couple of dozen bridges or demodulators that can be invoked because
> there are so many different em28xx based products that have different
> tuners/demods.  The dependencies are resolved automatically at runtime
> by the tuner-core and dvb framework once the actual product model is
> determined.  This avoids having to load modules for every single tuner
> and demodulator that could *possibly* be used by that driver.

Which of course makes perfect and reasonable sense. The problem is how 
to work around a situation where the invocation is not happening 
correctly. I presume that there are a series of pokes and peeks, but 
they are internal to the module code and not exposed. We really need to 
latch onto something exposed, and unique.

And it may not exist.

And as you say, there are lots of hardware items which use different 
chipsets internally, while declaring that they are twins to a different 
model. Or they just report themselves wrongly anyway!

Geoff



More information about the mythtv-users mailing list