By using AWS re:Post, you agree to the Terms of Use

Question: Error at the end of building SDAccel helloworld project

0

Hi,

I am trying to use AWS F1 to run the first helloworld project of SDAccel. And I followed the tutorial in the following link:https://github.com/Xilinx/SDAccel-Tutorials/blob/master/docs/aws-getting-started/RTL/STEP1.md

And when I come to the last step which is:
cd $SDACCEL_DIR/examples/xilinx/getting_started/host/helloworld_ocl/
sudo sh
source /opt/xilinx/xrt/setup.sh
./helloworld

I got an error. When I directly ran the ./helloworld command it shows :
Usage: ./helloworld <XCLBIN File>

Then I try the ./helloworld vector_addition.hw.xilinx_aws-vu9p-f1-04261818_dynamic_5_0.awsxclbin

And the result is:
ERROR: xclProbe-scan failed at fpga_pci_get_all_slot_specs
xclProbe found 0 FPGA slots with xocl driver running
XRT build version: 2.2.0
Build hash: e21b8a5b208618834760593bbb15063f7e399642
Build date: 2019-09-06 20:10:43
Git branch: 2019.1
PID: 27040
UID: 0
[Sat Nov 16 03:45:05 2019]
HOST: ip-172-31-56-40.us-west-2.compute.internal
EXE: /home/centos/src/project_data/aws-fpga/SDAccel/examples/xilinx_2019.1/getting_started/hello_world/helloworld_ocl/helloworld
[XRT] ERROR: No devices found
ERROR: xclProbe-scan failed at fpga_pci_get_all_slot_specs
xclProbe found 0 FPGA slots with xocl driver running
[XRT] ERROR: No devices found
ERROR: xclProbe-scan failed at fpga_pci_get_all_slot_specs
xclProbe found 0 FPGA slots with xocl driver running
[XRT] ERROR: No devices found
../../..//libs/xcl2/xcl2.cpp:40 Error calling err = cl::Platform::get(&platforms), error code is: -1001

I searched on forum and ran cat *.awsxclbin | strings | grep agfi | xargs fpga-load-local-image -S0 -I

And the error showed that:
Error: (20) pci-device-missing
Unable to locate a PCI device or resource for communicating with the
FPGA API. This can happen if the FPGA has stopped behaving correctly
and the instance will need to be stopped and restarted. This can also
happen if the tools are run on a system with no AWS FPGA attached.

I wonder why this will happen and whether it is because I fail to give any credentials to the user.
My environment is t2.2xlarge instance in us-west-2.

All the former results seems correct and got
"State": {
"Code": "available"
}, while running the aws ec2 describe-fpga-images --fpga-image-ids <AFI ID>.

Hope someone can help.

Thanks!

asked 3 years ago24 views
2 Answers
0

Hello

My environment is t2.2xlarge instance in us-west-2.

To run on real FPGA hardware, you need to use an F1 type of instances, not T2.

answered 3 years ago
0

Hi,

Okay... Not a good news, but thanks a lot for your help! Your answer totally covered my concern. Thanks again!

answered 3 years ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions