Hello, I want to know the difference between tx2i and TX2, especially the external interface. We have completed the program of TX2 and FPGA chip PCIe interface, and hope to adjust the program to adapt to the PCIe interface between tx2i and FPGA chip;
In addition, the system versions of tx2i and TX2 are the same;
However, during the test, it is found that the FPGA can receive tx2i transmission request (data length required to be transmitted, transmission to the destination address, etc.), but tx2i can not receive data. So I hope to get your help
However, during the test, it is found that the FPGA can receive tx2i transmission request (data length required to be transmitted, transmission to the destination address, etc.), but tx2i can not receive data. So I hope to get your help
By this, I understand that you are able to get the PCIe device (FPGA here) enumerated in the TX2i based system also. Please confirm.
Also, are you using the exact same SW package on TX2 and TX2i ? If not already, please try that.
Also, what exactly is the transmission request that TX2i is receiving? I mean is it a PCIe interrupt sent by FPGA that it is going to send the data soon?
thank you for your answer ; I confirm get the PCIe device . I confirm that the TX2 and TX2I systems are identical. My program architecture is as follows: FPGA as slave device, GPU as master device; GPU send request (address, data length and transfer request) , FPGA receive the request and send the convention data to GPU, GPU periodically query data. The problem now is that the FPGA acknowledges the request and sends the data to the GPU. But the GPU didn’t get the data.
How exactly is GPU coming into the picture? Which GPU is being referred here? is it the iGPU (Internal GPU of Tegra) or you connected any external GPU to Tegra over a different PCIe slot?
Is there a device driver running working in the host system to talk to FPGA? Are you using the exact same device driver in both TX2 and TX2i?
The GPU adopts the TX2I core card produced by nvidia. The FPGA is the acquisition card designed by us. The GPU and FPGA are connected through a special slot. Install the same system Ubuntu (version 16.04) on both TX2I and TX2. So I think the driver is the same
I’m lost w.r.t the platform here.
What exactly you meant by ‘GPU adopts the TX2i core card’?
Is this GPU also Nvidia produced?
What is the protocol/interface being used between GPU and FPGA?
If it is possible, could you please give a block diagram of your system?
First
this GPU also Nvidia produced;TX2i
Second
PCIE is the communication interface between GPU and FPGA；
Third
Site restrictions. I can’t upload pictures.
Could you give me your email address so I can send you a detailed description of the problem。
This doesn’t even look like a standard PCIe device driver. I’m afraid I can’t really help much here. I think you should write a proper PCIe device driver.
Looking at the code, I don’t see you are setting bus master enable it in the command register.
Dear Vidyas
Can you give me some information? Mainly the difference between TX2 and Tx2i. In particular, PCIe differences.or you can provide TX2 communication with the FPGA driver? We adapted the GPU program according to the program you provided.
