[VTA Tutorial] deploy_classification.py segmentation fault

I follow the VTA tutorial Deploy Pretrained Vision Model from MxNet on VTA to deploy MxNet model into PYNQ-Z2. After running the python script:

python3 ./deploy_classification.py

Segmentation fault 11 error raised, and here is the output from terminal:

Reconfigured FPGA and RPC runtime in 2.95s!
"-target" is deprecated, use "-mtriple" instead.
"-target" is deprecated, use "-mtriple" instead.
"-target" is deprecated, use "-mtriple" instead.
"-target" is deprecated, use "-mtriple" instead.
"-target" is deprecated, use "-mtriple" instead.
"-target" is deprecated, use "-mtriple" instead.

Segmentation fault: 11


Segmentation fault: 11


Segmentation fault: 11


Segmentation fault: 11

Stack trace:
  [bt] (0) /home/sunnychen/.local/lib/python3.8/site-packages/mxnet/libmxnet.so(+0x3c27360) [0x7fa8d94a2360]
Stack trace:
  [bt] (0) /home/sunnychen/.local/lib/python3.8/site-packages/mxnet/libmxnet.so(+0x3c27360) [0x7fa8d94a2360]
Stack trace:
  [bt] (0) /home/sunnychen/.local/lib/python3.8/site-packages/mxnet/libmxnet.so(+0x3c27360) [0x7fa8d94a2360]
Segmentation fault (core dumped)

I located the error is taking place at line 198 executing relay.build():

graph, lib, params = relay.build(
                relay_prog, target=target,
                params=params, target_host=env.target_host)

I tried to replace the device target from vta to arm_cpu and it works well. My system’s info:

OS: Manjaro Linux 20.0.3
Python version: 3.8.0
MxNet version: 1.6.0

Thanks in advance for your help.

Same issue here using Pytorch and Ultra96 (or simulation). The following error is raised:

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

It was working previously with commit afaa9e492.

Any update on this issue ?

I also experienced a same problem when I ran autotune with pynq-z1. Maybe you can try to build mxnet from source.

The answer to this issue is here: