I am using A10 SOCFPGA , where Linux runs on SOC ARM（dual-A9）.
In my application, there‘re bulks of memory copying from one place to another place each 1ms，currently I use memcpy() which occupy CPU heavily.
Thus I want to use ARM DMA to do this, to avoid use cpu memcpy() . The question is " HOW _TO DO THIS " ? is there any code example ?
Thanks for help
You may find design example in our design store, unfortunately there is no exact design example that you are requesting.
However, there is a Data Mover which utilize the msGDMA, that uses the ARM Cortex A9, you may check it out here:
In RocketBoards, I was told here is an very similar example code ( https://github.com/robertofem/CycloneVSoC-examples/tree/master/Linux-modules/DMA_PL330_LKM_basic ),
Thus I'm trying to check this example and update to my A10SOC LINUX.