I want to develop algorithm LBP on TDA4VM.
How should I load data on C71x DSP or C66x DSP? Is the data loaded with DMA set by me per cycle? Or, the data is automatically loaded by DMA?
That means: Before calculation for every cycle , I should load the corresponding data into L1 cache through DMA(so to main memory and L2)?
Is there an example about TIOVX that show the whole process included load data(from sensor and local memory), data transfer, calculation of per cycle, get final result ?
For C71x DSP, is there a user manual to introduce the DSP (include cache, instruction, corePac, Optimizing Compiler)?