**• Hardware Platform (GPU) RTX 3060
**• DeepStream Version 7.0
• TensorRT Version
**• NVIDIA GPU Driver Version (valid for GPU only) CUDA12.1
• Issue Type (questions)
Does deepstream-app use NVDEC by default, or can it be enabled through the configuration file?
For deepstream-app sample app, if the input video/stream is H264/H265 encoded, NVDEC is used for decoding.
When I run deepstream-app, I get the following error messages:
Error while setting IOCTL
Invalid control
S_EXT_CTRLS for CUDA_GPU_ID failed
However, the program still runs normally and performs object detection. Does this mean that NVDEC is not actually being used?
Additionally, when I run the following command:
gst-launch-1.0 filesrc location=/sample/config/t2.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! fakesink
The pipeline gets stuck after displaying the following messages:
Setting pipeline to PAUSED ...
Pipeline is PREROLLING ...
Error while setting IOCTL
Invalid control
S_EXT_CTRLS for CUDA_GPU_ID failed
Error while setting IOCTL
Invalid control
S_EXT_CTRLS for CUDA_GPU_ID failed
What could be causing this issue?
Please follow the DeepStream 7.0 version platform compatibility. Installation — DeepStream documentation
I don’t find any problem with the pipeline you provided. Please follow Installation — DeepStream documentation 6.4 documentation to install DeepStream 7.0 correctly.
Thank you for your reply. I will test it and see.
