1, modify the DSI interface program, drive our own selected LED screen, this modification process, mainly modify which part of the content, each part of the modified content, under what directory;
2, modify the touchscreen interface, which part of the content should be modified during the modification process, and what contents of each part should be changed.
3, modify the Mipi interface protocol, we use the data flow pattern MIPI, which we use FPGA, and do not read and write interfaces to similar parameters to the ordinary camera; how to modify the program, and which contents under which directory to be modified separately!
4, how to compile and load its own chip driver, which is to put the contents in what directory, so that the system can automatically load.
1, what is the limit ability of TX2 hard coding and decoding? (HP)
2, 00_video_decode, the decoding is hardware decoding or software decoding. The ability of hardware decoding can solve several ways at the same time, and what format can be solved?
3, is 01_video_encode hard coded, and how many ways can it compile at the same time?
4, will CUDA encoding and decoding affect the rendering efficiency of GPU?
5, upload buffer through OpenGL and upload through EGLImageKHR. Which is faster?
6, the TX2 development board’s own camera can’t be retrieved correctly by V4L2 protocol. It will block. The demo can’t run normally. Is there a demo that can run?
7, how to change the logo of NVIDIA when starting up?