Nveglglessink doesn't work with GStreamer tracing subsystem

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU): GeForce 4090
• DeepStream Version: 6.4
• JetPack Version (valid for Jetson only): NA
• TensorRT Version: 8.6
• NVIDIA GPU Driver Version (valid for GPU only): 535
• Issue Type( questions, new requirements, bugs): question
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing): please see, below, the details
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description): NA

I try to collect latency using GStreamer tracing subsystem (GST_TRACERS="latency(flags=pipeline+element)" GST_DEBUG=GST_TRACER:7) for deepstream-image-meta-test sample app, but the pipeline blocks indefinitely. I replace nveglglessink with fakesink and the pipeline doesn’t block anymore. The modified code is bellow (please change the file extension from .txt to .c):

deepstream_image_meta_test.txt (23.4 KB)

I found this topic that mentions that nveglglessink is deprecated from DeepStream 6.2, what is the new plugin that replace nveglglessink? May be it works with GStreamer tracing subsystem.

3

Could you try to get the latency by reffering to our FAQ?

It’s deprecated on Jetson board. For DGPU, we have always supported this plugin.

4

@yuweiw

When I use GStreamer tracing subsystem, I have to replace nveglglessink is replaced with fakesink, otherwise, the pipeline will block indefinitely. In diff to enable latency measurement from the link you shared, nveglglessink is also replaced with fakesink, does that mean nvds_measure_buffer_latency also doesn’t support nveglglessink?

5

No. It supports nveglglessink.

Can you describe your operation steps in detail?

6

@yuweiw

I build and run the deepstream-image-meta-test sample:

make CUDA_VER=12.2
GST_TRACERS="latency(flags=pipeline+element)" GST_DEBUG=GST_TRACER:7 GST_DEBUG_FILE=/tmp/log.txt ./deepstream-image-meta-test 0 file:///home/ubuntu/video1.mp4

The pipeline blocks indefinitely, replacing nveglglessink with fakesink unblocks the pipeline.

Here is the output before the blocking happens:

WARNING: Overriding infer-config batch-size (2) with number of sources (1)
Now playing...
INFO: ../nvdsinfer/nvdsinfer_model_builder.cpp:610 [Implicit Engine Info]: layers num: 3
0   INPUT  kFLOAT input_1         3x544x960       
1   OUTPUT kFLOAT output_bbox/BiasAdd 16x34x60        
2   OUTPUT kFLOAT output_cov/Sigmoid 4x34x60         

Running...

(deepstream-image-meta-test:3748740): GStreamer-CRITICAL **: 09:34:33.150: gst_debug_log_valist: assertion 'category != NULL' failed

(deepstream-image-meta-test:3748740): GStreamer-CRITICAL **: 09:34:33.150: gst_debug_log_valist: assertion 'category != NULL' failed

(deepstream-image-meta-test:3748740): GStreamer-CRITICAL **: 09:34:33.150: gst_debug_log_valist: assertion 'category != NULL' failed

(deepstream-image-meta-test:3748740): GStreamer-CRITICAL **: 09:34:33.150: gst_debug_log_valist: assertion 'category != NULL' failed
Frame Number = 0 Number of objects = 15 Vehicle Count = 10 Person Count = 5
Frame Number = 1 Number of objects = 16 Vehicle Count = 10 Person Count = 6
Frame Number = 2 Number of objects = 14 Vehicle Count = 9 Person Count = 5
Frame Number = 3 Number of objects = 15 Vehicle Count = 9 Person Count = 6
Frame Number = 4 Number of objects = 15 Vehicle Count = 8 Person Count = 7
7

As I attached before, please refer to our FAQ to get the latency in DeepStream.

8

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.