Re: [FFADO-devel] MOTU and the new device cache code

[Jonathan W. <jw...@ph...>]

Hi Pieter

> >> The idea was that loadFromCache takes the ConfigRom and checks if there 
> >> is a directory named after the GUID. I thought that should be enough but 
> >> as you say it is not safe to do so.
> > 
> > If all it did is check the ConfigRom then it would be fine.  However, the
> > fact that I got the "send oops" message from libiec61883 showed that
> > somewhere an AVC command is being attempted.  It's the AVC command which
> > trips up non-AVC things like the MOTU.
> 
> I would propose a slightly different approach to this probing + cache 
> loading:
> :
> If we put this is IAvDevice:
> 
> IAvDevice::loadFromCache() {
>    return false;
> }
> 
> and put the current loadFromCache in the BeBoB class, we avoid this 
> situation.

Agreed.  At the time I'll admit I was after a quick fix so I could continue
on some other tests I wanted to do.

> It would be good though if we could eliminate all explicit references to 
> device types from the code, and let it be handled by the OO mechanisms. 
> There are some places where we can't avoid it, but currently that's only 
> getDriverForDevice(), and it would be nice to keep it that way.

Sure.

> I think the root of the problem is that the AvDevice is casted to a 
> BeBoB::AvDevice no matter whether it is one or not.

Yes - as I mentioned, whatever reinterpret_cast was thought to be doing is
clearly not happening.

> This is IMHO due to the reinterpret_cast, that should be a dynamic_cast:
> 
> "reinterpret_cast converts any pointer type to any other pointer type, 
> even of unrelated classes. The operation result is a simple binary copy 
> of the value from one pointer to the other. All pointer conversions are 
> allowed: neither the content pointed nor the pointer type itself is 
> checked."
> 
> "dynamic_cast can be used only with pointers and references to objects. 
> Its purpose is to ensure that the result of the type conversion is a 
> valid complete object of the requested class."
> 
> dynamic_cast sets the pointer to NULL when it fails, which is the 
> behavior we want.

I agree - dynamic_cast is the behaviour implied by the code.

Regards
  jonathan