> One issue in comining up with the notion of a "WWAN device" is that > a typical WWAN device (especially USB ones) is not composed of just > a single function, but may have one or multiple network functions, > some TTYs for control, etc. These are actually not even seen by a > single driver in Linux, but various. An additional wrinkle is that > each of those drivers will not be aware of the others, and also be > a driver for a generic network or TTY device (for example), so by > itself it cannot even know it's dealing with a unified WWAN device. > > To still achieve a unified model, we allow each WWAN device to be > composed of "component devices". Each component device can offer a > certain subset of the overall functionality (which is shown in the > struct wwan_component_device_ops). This isn't implemented right now, > but ultimately it will allow us to have a "tentative" state, where > a number component drivers register their component, but the full > WWAN device is only formed if any one of them says that it indeed > knows for sure that it's a piece of a WWAN device, or perhaps by > some other heuristic. Hi Johannes Looking at it bottom up, is the WWAN device itself made up of multiple devices? Are the TTYs separate drivers to the packet moving engines? They have there own USB end points, and could just be standard CDC ACM? driver/base/component.c could be useful for bringing together these individual devices to form the whole WWAN device. This is often used for graphics drivers, where there can be i2c devices, display pipeline devices, acceleration drivers etc, which each probe separately, but need to be brought together to form a gpu driver as a whole. Plus you need to avoid confusion by not adding another "component framework" which means something totally different to the existing component framework. Andrew
