06-23-2015 05:11 PM
Hi everybody,this is my current situation: I have realized a custom FPGA component, which exposes two Avalon MM Slave Interfaces (one to write data to it and one to read data from it). Then, I have a Linux application running on the HPS Arm CPU, which has to transfer inputs to the FPGA component and to read back results from it after its computation. To maximize performances, I need to use a SGDMA to manage transfers between the HPS and the FPGA. This SGDMA should be programmed by the Linux application, which controls the whole dataflow. I have really no idea about how to use the SGDMA, since I cannot find any good documentation on the web. It looks like a pretty easy approach to me and I'm sure that also other people have met the need to transfer data from the HPS to the FPGA. Can anyone give me a hint? Any explanation, any tutorial or link? It would be really appreciated. Thanks in advance!
06-24-2015 08:23 AM
06-25-2015 10:30 PM
Ki skbeh,Thanks for your answer. I've already seen this documentation. My question was more about the whole "architecture": how to allocate some memory space from Linux to make it accessible from the SGDMA for example and how to get the physical address to program the SGDMA. SGDMA programming details shouldn't be a problem. Thanks anyway.