On Friday 26 March 2004 15:14, Peter Urbanec wrote: > Andrew de Quincey wrote: > >On Friday 26 March 2004 14:33, Klaus Schmidinger wrote: > >>If it's not hard to do in userspace, why is it so hard in the driver? > >>Why should every application implement this by itself? I'd say this > >> should be done by the driver, and the applications shall exchange > >> complete TPDU's with the driver. > > > >IMO, the driver should be as simple as possible. I don't think the CAM > >interface benefits speed-wise from the reassembly being in the driver, so, > >apart from saving a little bit of extra code in userspace I can't see any > >good reason for the reassembly being in the kernel. Each app already has > > to implement all the other layers of the EN50221 control protocol stack > > itself; why not this extra bit as well? > > If I can chime in with a purely design based rationale... > > Why only think of this from the point of driver vs. application? I agree > with Andrew that the driver should be as simple as possible. The less > code one has to run in the kernel context the better. However, I don't > particularly like the idea of each application having to reimplement the > same basic code - someone is likely to get it wrong. > > How about adding another layer? Take all this from the driver, but > provide a library implementation that applications can dynamically link > to to get the reassembly for free. That way you get the benefit of both > worlds. You keep the driver simple and the application developer doesn't > have to reinvent the wheel.
Yeah, I agree, thats *exactly* the way to do it. -- Info: To unsubscribe send a mail to [EMAIL PROTECTED] with "unsubscribe linux-dvb" as subject.
