AR# 72958

2019.1 Zynq UltraScale+ MPSoC VCU - Why does the VCU Decoder crash when trying to decode two reduced latency streams?

描述

Why does the VCU Decoder crash when trying to decode two reduced latency streams?

gst-launch-1.0 -e filesrc location=/media/card/1080p50_low-delay_gdr.mkv ! matroskademux ! tee name=t t. ! h265parse ! video/x-h265, alignment=au ! omxh265dec low-latency=true internal-entropy-buffers=5 ! queue max-size-bytes=0 leaky=2 ! fakesink t. ! h265parse ! video/x-h265, alignment=au ! omxh265dec low-latency=true internal-entropy-buffers=5 ! queue max-size-bytes=0 leaky=2 ! fakesink t. ! h265parse ! video/x-h265, alignment=au ! omxh265dec low-latency=true internal-entropy-buffers=5 ! queue max-size-bytes=0 leaky=2 ! fakesink t. ! h265parse ! video/x-h265, alignment=au ! omxh265dec low-latency=true internal-entropy-buffers=5 !
queue max-size-bytes=0 leaky=2 ! fakesink


Error from the log:

Setting pipeline to PAUSED ...
Pipeline is PREROLLING ...
Redistribute latency...
Redistribute latency...
Redistribute latency...
Redistribute latency...
[ 351.885781] al5d a0120000.al5d: VCU: unavailable resources or wrong configuration
[ 351.886215] al5d a0120000.al5d: VCU: unavailable resources or wrong configuration
[ 351.893279] al5d a0120000.al5d: Channel wasn't created.
/!\ decoder: hardware doesn't have enough resources (142)[ 351.905988] al5d a0120000.al5d: Channel wasn't created.
/!\ decoder: hardware doesn't have enough resources (142)
[E] [TreatSetStateCommand]: OMX_ALG_ErrorChannelResourceUnavailable
ERROR: from element /GstPipeline:pipeline0/GstOMXH265Dec-omxh265dec:omxh265dec-omxh265dec2: GStreamer encountered a general supporting library error.

解决方案

This is a known issue with the Zynq UltraScale+ MPSoC - LogiCORE H.264/H.265 Video Codec Unit (VCU) that can occur when trying to use more than two instances of the decoder engine.

The software is artificially limiting the number of instances. See the below list of solutions to work around this issue.

  • 2018.2 - Users must upgrade to the 2019.1 version (with patches) or version 2019.2 and later
  • 2019.1 - Users can download the PetaLinux Recipes and Patch files from (Xilinx Answer 72324) to work around this issue
  • 2019.2 - This issue is resolved in the 2019.2 release and later versions

链接问答记录

主要问答记录

AR# 72958
日期 11/07/2019
状态 Active
Type 已知问题
器件
Tools
IP