nVidia encoder crash after several re-launches of the pipeline

Autonomous Machines Jetson & Embedded Systems Jetson AGX Xavier
#1

We use nVidia elements (such as nvv4l2h265enc and nvv4l2decoder) in our sending and receiving pipelines inside the application. After several re-connections (and therefore re-launches of the pipeline) we might get out application crashed with the following output:

===== NVMEDIA: NVENC =====

GStreamer-CRITICAL **: 13:53:12.889: gst_poll_write_control: assertion 'set != NULL' failed

GStreamer-CRITICAL **: 13:53:12.889: gst_poll_write_control: assertion 'set != NULL' failed
Opening channel /dev/nvhost-nvenc1 failed
NVMEDIA: NvMMLiteVideoEncBlockOpen: 501: - Could not get NVENC Channel handle for inst 1
NvMMLiteBlockCreate : Block : BlockType = 8 
NvVideoEnc: NvMMLiteOpenBlock failed
Function v4l2_nvmmlite_init_encoder_params: failed
The process was ended forcefully.

Is it something that is fixed in the latest version of JP or maybe the fix is going to be provided soon?

#3

Hi,
We have not seen error like this? Which release do you use? Are you able to try Jetpack 4.6?

Nvv4l2decoder sometimes fails to negotiate with downstream after several pipeline re-launches
#4

It is with 4.5 definitely. Will also check 4.6

#6

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