You could write an OMX source/sink adapter MIO. This is how the OMX decoders are integrated into OpenCore.
Data tunneling is achieved using an opaque pointer (I think it's called "private_platform" that is passed between the decoder and MIO by the framework. You'll see evidence of this in android_surface_output.cpp where some of the hardware specific code from the G1 leaked into the plaform code. In the future, we plan to have a better abstraction layer for this. The recommended way to handle this is using a pmem or ashmem heap instead of passing our physical addresses. For example, the decoder may require physical memory addresses. Those addresses are allocated in the pmem driver and mapped to virtual addresses in the application processor. When the MIO gets the pmem heap, it can pass the fd and offset into the video display driver where the physical memory address can be retrieved. On Dec 11, 3:39 am, Sachin <[email protected]> wrote: > Hi, > i have 2 queries about Android framework. > > 1) Can we use OMX source and sink components instead of MIO components > for capturing / rendering ? > Does it have any impact on performance or conformance? > > 2) Is it feasible to achieve the "data tunneling" in Android framework > as supported by standard OMX components? > If yes can this tunnel include source/sink, processing components to > directly capture / render data ? > If this is not directly supported in the current framework, how can > this be achieved ? > > thanks a lot, > Sachin --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "android-framework" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-framework?hl=en -~----------~----~----~----~------~----~------~--~---
