Freezes with Pipeline is PREROLLED ... Caught SIGSEGV

[gsrujana]

gst-launch-1.0 filesrc location=$VIDEO_FILE ! decodebin ! videoconvert ! videoscale ! queue ! net.sink_model inceptionv4 name=net model-location=$MODEL_LOCATION backend=tensorflow backend::input-layer=$INPUT_LAYER backend::output-layer=$OUTPUT_LAYER Setting pipeline to PAUSED ... 2019-04-02 13:16:53.378880: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA Pipeline is PREROLLING ... Redistribute latency... Pipeline is PREROLLED ... Caught SIGSEGV Setting pipeline to PLAYING ... New clock: GstSystemClock Redistribute latency... Spinning. Please run 'gdb gst-launch-1.0 3791' to continue debugging, Ctrl-C to quit, or Ctrl-\ to dump core. ^Chandling interrupt. Interrupt: Stopping pipeline ... Execution ended after 0:04:14.362928966 Setting pipeline to PAUSED ... Setting pipeline to READY ... on using gstreamer inference with gshark tracer.. Ubuntu 16.04

[michaelgruner]

@GFallasRR @migueltaylor can you take a look at this? It looks the particularity of this pipeline is that there are no src pads in the inference element.

[GFallasRR]

I tested the same pipeline and it ran at my side, please check commands below.

export VIDEO_FILE=/home/cat.mp4 export MODEL_LOCATION=/home/graph_inceptionv4_tensorflow.pb export INPUT_LAYER='input' export OUTPUT_LAYER='InceptionV4/Logits/Predictions' #The pipeline does not have any display element, adding GST_DEBUG to check the output GST_DEBUG=inception:6 gst-launch-1.0 filesrc location=$VIDEO_FILE ! decodebin ! videoconvert ! videoscale ! queue ! net.sink_model inceptionv4 name=net model-location=$MODEL_LOCATION backend=tensorflow backend::input-layer=$INPUT_LAYER backend::output-layer=$OUTPUT_LAYER

Output

gfallas@gfallas-G5-5587:/home$ GST_DEBUG=inception:6 gst-launch-1.0 filesrc location=$VIDEO_FILE ! decodebin ! videoconvert ! videoscale ! queue ! net.sink_model inceptionv4 name=net model-location=$MODEL_LOCATION backend=tensorflow backend::input-layer=$INPUT_LAYER backend::output-layer=$OUTPUT_LAYER Setting pipeline to PAUSED ... 2019-04-12 15:05:36.272577: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA 0:00:01.188643347 1816 0xd68a70 INFO inceptionv4 gstinceptionv4.c:308:gst_inceptionv4_start: Starting Inception v4 Pipeline is PREROLLING ... Redistribute latency... Pipeline is PREROLLED ... Setting pipeline to PLAYING ... New clock: GstSystemClock Redistribute latency... 0:00:01.228488636 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:208:gst_inceptionv4_preprocess: Preprocess 0:00:02.812904929 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:277:gst_inceptionv4_postprocess: Postprocess 0:00:02.812928382 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:298:gst_inceptionv4_postprocess: Highest probability is label 282 : (0.636030) 0:00:02.812971662 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:208:gst_inceptionv4_preprocess: Preprocess 0:00:02.996602922 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:277:gst_inceptionv4_postprocess: Postprocess 0:00:02.996624197 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:298:gst_inceptionv4_postprocess: Highest probability is label 282 : (0.654386) 0:00:02.996658732 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:208:gst_inceptionv4_preprocess: Preprocess 0:00:03.176918040 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:277:gst_inceptionv4_postprocess: Postprocess 0:00:03.176939625 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:298:gst_inceptionv4_postprocess: Highest probability is label 282 : (0.659639) 0:00:03.176974896 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:208:gst_inceptionv4_preprocess: Preprocess 0:00:03.358122337 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:277:gst_inceptionv4_postprocess: Postprocess 0:00:03.358160043 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:298:gst_inceptionv4_postprocess: Highest probability is label 282 : (0.657630) 0:00:03.358195696 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:208:gst_inceptionv4_preprocess: Preprocess 0:00:03.545497339 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:277:gst_inceptionv4_postprocess: Postprocess 0:00:03.545518816 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:298:gst_inceptionv4_postprocess: Highest probability is label 282 : (0.658413) 0:00:03.545551458 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:208:gst_inceptionv4_preprocess: Preprocess 0:00:03.727754331 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:277:gst_inceptionv4_postprocess: Postprocess 0:00:03.727775046 1816 0x9d4000 LOG inceptionv4 gstinceptionv4.c:298:gst_inceptionv4_postprocess: Highest probability is label 282 : (0.654718)

@gsrujana can you share a log from this issue? More information about your system will help to solve this problem, as example:

• GStreamer version
• TensorFlow version
• Branch tested
• Log files
• Hardware specifications