On Tue, May 20, 2014 at 08:25:41AM +0200, Andrzej Pietrasiewicz wrote: > Hi Peter, > > W dniu 20.05.2014 01:38, Peter Chen pisze: > >On Mon, May 19, 2014 at 05:40:29PM +0200, Andrzej Pietrasiewicz wrote: > >>Hi Peter, > >> > >>W dniu 16.05.2014 11:00, Peter Chen pisze: > >>>Hi Felipe & Alan, > >>> > >>>To continue with topic discussed at > >>>http://www.spinics.net/lists/linux-usb/msg105279.html, > >>>I implement the gadget bus to bind udc to gadget driver follow > >>>most ideas from your two. > >>> > >> > >>While the idea is interesting I think some aspects of the implementation > >>must be thought over. > > > >Hi Andrzej, thanks for running tests and giving comments. > > > > > <snip> > > > > >I intended to run more gadget (including gadgetfs) when I make v2 patch. > > Please note that gadgetfs is something totally different than configfs. > > Gadgetfs is a tool to build gadget's functionality entirely in userspace, > that is to handle setup requests and so on, all in userspace. > > Configfs is used to compose gadgets at runtime, as opposed to the traditional > (legacy) way of composing the gadgets which is to define configurations and > select functions to be provided by writing code of a module. For example, > in g_ether there are two configurations and 4 functions: ECM, ECM subset, > EEM and RNDIS. Configuration @ index 2 contains RNDIS and configuration > @ index 1 contains one of the others. This is all hardcoded in ether.c. > With configfs you create gadgets of available _kernel_ components at runtime. > Thanks Andrzej, I just typed wrongly, it should be "configfs". Yes, configfs is a magic feature, esp we can support all gadget functions in future, then, we can unity the gadget usage between android kernel and upstream kernel. -- Best Regards, Peter Chen -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
