jafet-chaves

libgstc_socket test is failing on version 0.12.0

2

System specs: * x86_64 * Ubuntu 18.04 * GCC 7.5 * Python 3.6.12 Below is some reference log: ````make check Making check in gstd make[1]: Entering directory '/home/jafet/work/devdirs/ridgerun/gstd-1.x/gstd' make[1]: Nothing...

I got the following memory leak while testing with these simple commands: `G_SLICE=always-malloc G_DEBUG=gc-friendly valgrind --leak-check=full --leak-resolution=high --num-callers=20 --trace-children=yes --suppressions=/home/nvidia/massif/gst.supp --suppressions=/home/nvidia/massif/glib.supp gstd -D` ``` gstd-client pipeline_create pipe_test videotestsrc ! fakesink...

Any other [hardware option ](https://github.com/microsoft/onnxruntime/blob/master/BUILD.md#openvino) for OpenVINO backend is not working with current models provided by RIdgerun. Here is a sample pipeline to reproduce the issue: ``` gst-launch-1.0 filesrc location=Test_benchmark_video.mp4...

Forum participation

1

- [x] Intel OpenVINO - [x] Intel Movidius - Myriad - [x] Arm Compute Library - [ ] Google Coral - EdgeTPU - [x] TensorFlow - [x] TFLite

R2Inference does not support models with multiple output tensors at this point in time.

This breaks TensorRT for example, or in general, any engine that uses memory not accesible by the CPU _Originally posted by @michaelgruner in https://github.com/RidgeRun/r2inference/pull/103#discussion_r496159589_

There is currently a limitation with the IFrame interface in the sense that it doesn't allow to pass a DataType. Basically no matter the type of pointer that one passes...

These are the commands to replicate: `./autogen.sh --disable-docs --enable-tflite` `make` `sudo make install` `cd examples/external` `make` Which gives: ``` /usr/local/lib/libr2inference-0.0.so: undefined reference to 'pthread_create' /usr/local/lib/libr2inference-0.0.so: undefined reference to 'pthread_join' collect2:...