Cannot run AFI on F1 instance

0

Hi there,

I followed all the steps in the guideline provided by Xilinx and tried to run AFI on F1 instance (I followed https://github.com/Xilinx/SDAccel-Tutorials/blob/master/docs/aws-getting-started/CPP/STEP2.md)

When I ran ./<project_name.exe> <awsxclbin_file> , I first received

Segmentation fault (core dumped) error.

And for all runs after it, I received

\[XRT] ERROR: unable to issue xclExecBuf

../src/host.cpp:103 Error calling cl::Program program(context, devices, bins, NULL, &err), error code is: -6

Any help is appreciated!

질문됨 4년 전296회 조회
5개 답변
0
수락된 답변

Hello, by any chance have you tried clearing/resetting the FPGA with the 'fpga-clear-local-image' command, then rerunning the executable to reload it?

https://github.com/aws/aws-fpga/blob/master/sdk/userspace/fpga_mgmt_tools/README.md

The tutorial mentioned just runs the host executable, so there shouldn't be a need to add the *.awsxclbin as an argument as it should be pulled in by the host code.

답변함 4년 전
0

Another thing to note while clearing the slot is that you will have to restart the mpd service after doing that.

sudo systemctl restart mpd

This lets XRT load the default AFI to be able to bind the XOCL driver to the correct device ID.

-Deep

Deep_P
답변함 4년 전
0

When I ran this command, it gives an error.

답변함 4년 전
0

It needed the .awsxclbin, otherwise, it throws an error.

답변함 4년 전
0

We use Amazon FPGA Image (AFI) for the binary that is loaded on the FPGA.
It is the compiled FPGA code that is loaded into an FPGA in AWS for performing the Custom Logic (CL) function created by the developer. AFIs are maintained by AWS according and associated with the AWS account that created them. The AFI includes the CL and AWS FPGA Shell. An AFI ID is used to reference a particular AFI from an F1 instance.
An awsxclbin includes the AFI which makes it different from xclbin. Here is the step where it is created documented in our end-to-end guide:
https://github.com/aws/aws-fpga/tree/master/Vitis#createafi

AWS
답변함 4년 전

로그인하지 않았습니다. 로그인해야 답변을 게시할 수 있습니다.

좋은 답변은 질문에 명확하게 답하고 건설적인 피드백을 제공하며 질문자의 전문적인 성장을 장려합니다.

질문 답변하기에 대한 가이드라인

관련 콘텐츠