Pipeline set state meet Segmentation fault (core dumped) and app crashed

DeepStream and Gstreamer CRASH when release memory.

• Hardware Platform = Jetson Xavier Nx
• DeepStream Version = 5.10
• JetPack Version =4.5.1
• TensorRT Version 7.1.3.0
• Issue Type BUGS

Hello everybody, I am a beginner of DS5 and Gst. I writed a project with c++ in Qt creater.

I was trying to Launch four different pipelines in four different subThread,By this time,the pipelines works great.

When i tries to release memory and destroy pipeline by:

gst_element_set_state (Cam1.Pipeline, GST_STATE_NULL);
gst_object_unref (GST_OBJECT (Cam1.Pipeline));
g_source_remove  (Cam1.bus_watch_id);
g_main_loop_unref (Cam1.Loop);

At First,it works,but after i relaunch this pipeline and destroy several times,Exactly 5 Times, The app will crash after “gst_element_set_state (Cam1.Pipeline, GST_STATE_NULL);”

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:00:30.007696820 20033   0x7edc0740c0 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:00:30.007966007 20033   0x7edc0740c0 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:00:30.036346390 20033   0x7edc0740c0 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/config_infer_primary_yolov5s.txt sucessfully
Decodebin child added: source
Decodebin child added: decodebin3
QThread::wait: Thread tried to wait on itself
Decodebin child added: mpegpsdemux3
Decodebin child added: multiqueue3
Decodebin child added: h264parse3
Decodebin child added: capsfilter3
Decodebin child added: nvv4l2decoder3
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
In cb_newpad
IOU Tracker Init with threshold 0.100000
Segmentation fault (core dumped)

Does the pipeline launch and destroy frequently could cause memory leaks or NULL ptr?

How to solve it?

That depends on your code.

Please check with your code first. If you find any clue shows the crash or memory leak is caused by deepstream elements, you can tell us the details or give us a simple sample to reproduce the problem.

Thanks for reply. I got new information about the crash.I will show below.

I create pipeline just like deepstream-app-test3:

bool CDlg_GeneralDisplay::InitPipeLine(CamPipeline *CamConfig){

    int current_device = -1;
    cudaGetDevice(&current_device);
    struct cudaDeviceProp prop;
    cudaGetDeviceProperties(&prop, current_device);                                   //Nidia官方CUDA代码,用来识别设备是不是Jetson

    //    CamConfig->Loop         = NULL;                                                   //初始化变量
    CamConfig->Pipeline     = NULL;
    CamConfig->streammux    = NULL;
    CamConfig->sink         = NULL;
    CamConfig->pgie         = NULL;
    CamConfig->queue1       = NULL;
    CamConfig->queue2       = NULL;
    CamConfig->queue3       = NULL;
    CamConfig->queue4       = NULL;
    CamConfig->queue5       = NULL;

    CamConfig->nvvidconv    = NULL;
    CamConfig->nvosd        = NULL;
    CamConfig->transform    = NULL;
    CamConfig->nvtracker    = NULL;
    CamConfig->bus          = NULL;
    CamConfig->osd_sink_pad = NULL;

//    CamConfig->Loop = g_main_loop_new (NULL, FALSE);                                  //标准的Gsteamer初始化流程 在此初始化主循环

    CamConfig->Pipeline = gst_pipeline_new ("PipeLine");                              //给创建的管道命名

    CamConfig->streammux = gst_element_factory_make ("nvstreammux", "stream-muxer");  //创建nvstreammux混合器用于Batch推理或者混流

    if (!CamConfig->Pipeline || !CamConfig->streammux){                               //判断是否创建成功,不成功就返回且弹窗
        QMessageBox::warning(this, tr("Warning"), tr("Failed to create Streammux & Pipeline."),QMessageBox::Ok);
        return false;
    }

    gst_bin_add (GST_BIN (CamConfig->Pipeline), CamConfig->streammux);                //把Streammux添加到Bin箱柜

    CamConfig->source_bin = create_source_bin (0, CamConfig->SourceUrl);              //创建视频源头箱柜

    if (!CamConfig->source_bin) {
        QMessageBox::warning(this, tr("Warning"), tr("Failed to create Source_Bin."),QMessageBox::Ok);
        return false;
    }

    gst_bin_add (GST_BIN (CamConfig->Pipeline), CamConfig->source_bin);               //添加视频源头箱柜到管道

    g_snprintf (CamConfig->pad_name, 15, "sink_%u", 0);
    CamConfig->sinkpad = gst_element_get_request_pad (CamConfig->streammux, CamConfig->pad_name);
                                                                                      //获取指定element中的指定pad的名称
    if (!CamConfig->sinkpad) {
        QMessageBox::warning(this, tr("Warning"), tr("Streammux request sink pad failed. "),QMessageBox::Ok);
        return false;
    }

    CamConfig->srcpad = gst_element_get_static_pad (CamConfig->source_bin, "src");    //获取指定element中的指定pad的名称
    if (!CamConfig->srcpad) {
        QMessageBox::warning(this, tr("Warning"), tr("Failed to get src pad of source bin."),QMessageBox::Ok);
        return false;
    }

    if (gst_pad_link (CamConfig->srcpad, CamConfig->sinkpad) != GST_PAD_LINK_OK) {    //如果mux组和src组连接失败就报错
        QMessageBox::warning(this, tr("Warning"), tr("Failed to link source bin to stream muxer."),QMessageBox::Ok);
        return false;
    }

    gst_object_unref (CamConfig->srcpad);
    gst_object_unref (CamConfig->sinkpad);

    CamConfig->pgie = gst_element_factory_make ("nvinfer", "primary-nvinference-engine"); //设置Nvinfer推理组件

    CamConfig->nvtracker = gst_element_factory_make ("nvtracker", "tracker");             //设置Nvtracker跟踪组件

    CamConfig->queue1 = gst_element_factory_make ("queue", "queue1");                     //每两个组件之间设置缓冲队列
    CamConfig->queue2 = gst_element_factory_make ("queue", "queue2");
    CamConfig->queue3 = gst_element_factory_make ("queue", "queue3");
    CamConfig->queue4 = gst_element_factory_make ("queue", "queue4");
    CamConfig->queue5 = gst_element_factory_make ("queue", "queue5");

    CamConfig->nvvidconv = gst_element_factory_make ("nvvideoconvert", "nvvideo-converter"); //NV12转化RGB插件

    CamConfig->nvosd = gst_element_factory_make ("nvdsosd", "nv-onscreendisplay");           //NVOSD绘制插件

    if(prop.integrated) {
      CamConfig->transform = gst_element_factory_make ("nvegltransform", "nvegl-transform"); //如果是Jetson硬件就额外创建转化插件
    }

    CamConfig->sink = gst_element_factory_make ("nveglglessink", "nvvideo-renderer");        //显示窗口插件

    if (!CamConfig->pgie || !CamConfig->nvtracker || !CamConfig->nvvidconv || !CamConfig->nvosd || !CamConfig->sink) {
        QMessageBox::warning(this, tr("Warning"), tr("One of the deepstream elements could not be created."),QMessageBox::Ok);
        return false;
    }

    if(!CamConfig->transform && prop.integrated) {
        QMessageBox::warning(this, tr("Warning"), tr("One jetson plantform element could not be created. Exiting."),QMessageBox::Ok);
        return false;
    }
                                                                                             //以下均为Deeptream插件详细配置代码
    g_object_set (G_OBJECT (CamConfig->streammux), "batch-size", 1, NULL);

    g_object_set (G_OBJECT (CamConfig->streammux), "width", MUXER_OUTPUT_WIDTH, "height",
        MUXER_OUTPUT_HEIGHT,
        "batched-push-timeout", MUXER_BATCH_TIMEOUT_USEC, NULL);

    g_object_set (G_OBJECT (CamConfig->pgie),
        "config-file-path", "./Model/PrimaryInfer/yolov5s.txt", NULL);

    g_object_get (G_OBJECT (CamConfig->pgie), "batch-size", &CamConfig->pgie_batch_size, NULL);
    if (CamConfig->pgie_batch_size != 1) {
      g_object_set (G_OBJECT (CamConfig->pgie), "batch-size", 1, NULL);
    }

    g_object_set (G_OBJECT (CamConfig->nvosd), "process-mode", OSD_PROCESS_MODE,
        "display-text", OSD_DISPLAY_TEXT, NULL);

    g_object_set (G_OBJECT (CamConfig->sink), "qos", 0, NULL);

    if (!set_tracker_properties(CamConfig->nvtracker)) {
        QMessageBox::warning(this, tr("Warning"), tr("Failed to set tracker properties."),QMessageBox::Ok);
        return false;
    }

    CamConfig->bus = gst_pipeline_get_bus (GST_PIPELINE (CamConfig->Pipeline));
    CamConfig->bus_watch_id = gst_bus_add_watch (CamConfig->bus, bus_call, CamConfig->Loop);
    gst_object_unref (CamConfig->bus);

    if(prop.integrated) {
      gst_bin_add_many (GST_BIN (CamConfig->Pipeline), 
          CamConfig->queue1,     CamConfig->pgie,      CamConfig->queue2, 
          CamConfig->nvtracker,  CamConfig->queue3,    CamConfig->nvvidconv, 
          CamConfig->queue4,     CamConfig->nvosd,     CamConfig->queue5, 
          CamConfig->transform,  CamConfig->sink,      NULL);
      if (!gst_element_link_many (
          CamConfig->streammux,  CamConfig->queue1,    CamConfig->pgie,   
          CamConfig->queue2,     CamConfig->nvtracker, CamConfig->queue3, 
          CamConfig->nvvidconv,  CamConfig->queue4,    CamConfig->nvosd,  
          CamConfig->queue5,     CamConfig->transform, CamConfig->sink, 
          NULL)) {
        QMessageBox::warning(this, tr("Warning"), tr("Elements could not be linked."),QMessageBox::Ok);
        return false;
      }
    }else {
      gst_bin_add_many (GST_BIN (CamConfig->Pipeline), 
          CamConfig->queue1,     CamConfig->pgie,      CamConfig->queue2, 
          CamConfig->nvtracker,  CamConfig->queue3,    CamConfig->nvvidconv, 
          CamConfig->queue4,     CamConfig->nvosd,     CamConfig->queue5, 
          CamConfig->sink,      NULL);
    
      if (!gst_element_link_many (
          CamConfig->streammux,  CamConfig->queue1,    CamConfig->pgie,   
          CamConfig->queue2,     CamConfig->nvtracker, CamConfig->queue3, 
          CamConfig->nvvidconv,  CamConfig->queue4,    CamConfig->nvosd,  
          CamConfig->queue5,     CamConfig->sink,      NULL)) {
        QMessageBox::warning(this, tr("Warning"), tr("Elements could not be linked."),QMessageBox::Ok);
        return false;
      }
    }

    gst_video_overlay_set_window_handle (GST_VIDEO_OVERLAY (CamConfig->sink), CamConfig->Window_ID);

    CamConfig->osd_sink_pad = gst_element_get_static_pad (CamConfig->nvosd, "sink");
    if (!CamConfig->osd_sink_pad)
      QMessageBox::warning(this, tr("Warning"), tr("Unable to get sink pad."),QMessageBox::Ok);
    else
      gst_pad_add_probe (CamConfig->osd_sink_pad, GST_PAD_PROBE_TYPE_BUFFER,
          tiler_src_pad_buffer_probe, NULL, NULL);
    gst_object_unref (CamConfig->osd_sink_pad);

    return true;
}

And i destroy pipeline by the code below:

    gst_element_set_state (CamConfig.Pipeline, GST_STATE_NULL);
    gst_object_unref (GST_OBJECT (CamConfig.Pipeline));
    g_source_remove  (CamConfig.bus_watch_id);

I have used QT to debug for hour and hours,finally found all of the crash caused by "gst_element_set_state (CamConfig.Pipeline, GST_STATE_NULL)"

At first i think maybe my code cause memory leaks or used NULL ptr.

But after enabled the GST_DEBUG = 2,I found the crash caused by nveglglessink!

Here is the debug info.

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:08.974516488 26221   0x7e501c0ca0 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:08.974846281 26221   0x7e501c0ca0 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:08.981265619 26221   0x7e501c0ca0 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:08.982921589 26221   0x7e501c0ca0 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin8
0:01:08.984058999 26221   0x7e501c0ca0 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux8
Decodebin child added: multiqueue8
Decodebin child added: h264parse8
Decodebin child added: capsfilter8
Decodebin child added: nvv4l2decoder8
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:09.120146855 26221   0x7f180045e0 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder8:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:09.120258855 26221   0x7f180045e0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ee64d4380 Failed to determine interlace mode
0:01:09.120341928 26221   0x7f180045e0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ee64d4380 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:09.226783370 26221   0x7f180045e0 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder8:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:09.226907467 26221   0x7f180045e0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ee64d4380 Failed to determine interlace mode
0:01:09.226986891 26221   0x7f180045e0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ee64d4380 Failed to determine interlace mode
In cb_newpad
0:01:09.230545200 26221   0x7f180045e0 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder8> Duration invalid, not setting latency
0:01:09.232970356 26221   0x7f180045e0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder8:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:09.256720088 26221   0x7e400274a0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder8:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:09.685287752 26221   0x7e400274a0 WARN            videodecoder gstvideodecoder.c:2780:gst_video_decoder_prepare_finish_frame:<nvv4l2decoder8> decreasing timestamp (0:01:03.573044444 < 0:01:04.093044444)

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:11.297992172 26221   0x558e1b1a30 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:11.298219180 26221   0x558e1b1a30 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:11.321387120 26221   0x558e1b1a30 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:11.324113908 26221   0x558e1b1a30 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin9
0:01:11.325442646 26221   0x558e1b1a30 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux9
Decodebin child added: multiqueue9
Decodebin child added: h264parse9
Decodebin child added: capsfilter9
Decodebin child added: nvv4l2decoder9
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:11.523274021 26221   0x7f2402bd40 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder9:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:11.523459941 26221   0x7f2402bd40 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed0024ef0 Failed to determine interlace mode
0:01:11.523585157 26221   0x7f2402bd40 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed0024ef0 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:11.634951056 26221   0x7f2402bd40 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder9:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:11.635080304 26221   0x7f2402bd40 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed0024ef0 Failed to determine interlace mode
0:01:11.635242512 26221   0x7f2402bd40 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed0024ef0 Failed to determine interlace mode
In cb_newpad
0:01:11.638527541 26221   0x7f2402bd40 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder9> Duration invalid, not setting latency
0:01:11.655704303 26221   0x7f2402bd40 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder9:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:11.689382627 26221   0x558e672280 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder9:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:12.247621786 26221   0x558e672280 WARN            videodecoder gstvideodecoder.c:2780:gst_video_decoder_prepare_finish_frame:<nvv4l2decoder9> decreasing timestamp (0:01:03.933044444 < 0:01:04.493044444)

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:13.803584361 26221   0x7f1c0022d0 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:13.803820489 26221   0x7f1c0022d0 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:13.827969966 26221   0x7f1c0022d0 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:13.829987185 26221   0x7f1c0022d0 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin10
0:01:13.832171669 26221   0x7f1c0022d0 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux10
Decodebin child added: multiqueue10
Decodebin child added: h264parse10
Decodebin child added: capsfilter10
Decodebin child added: nvv4l2decoder10
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:13.960961946 26221   0x7ed50e8990 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder10:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:13.961118042 26221   0x7ed50e8990 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ea0004970 Failed to determine interlace mode
0:01:13.961229082 26221   0x7ed50e8990 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ea0004970 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:14.067455677 26221   0x7ed50e8990 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder10:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:14.067592029 26221   0x7ed50e8990 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ea0004970 Failed to determine interlace mode
0:01:14.067662526 26221   0x7ed50e8990 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ea0004970 Failed to determine interlace mode
In cb_newpad
0:01:14.070979299 26221   0x7ed50e8990 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder10> Duration invalid, not setting latency
0:01:14.084732536 26221   0x7ed50e8990 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder10:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:14.104897847 26221   0x7f240035e0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder10:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:14.396197685 26221   0x7f240035e0 WARN            videodecoder gstvideodecoder.c:2780:gst_video_decoder_prepare_finish_frame:<nvv4l2decoder10> decreasing timestamp (0:01:03.573044444 < 0:01:04.133044444)

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:16.368090980 26221   0x7ecc053460 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:16.368412708 26221   0x7ecc053460 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:16.376804465 26221   0x7ecc053460 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:16.378339572 26221   0x7ecc053460 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin11
0:01:16.381341528 26221   0x7ecc053460 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux11
Decodebin child added: multiqueue11
Decodebin child added: h264parse11
Decodebin child added: capsfilter11
Decodebin child added: nvv4l2decoder11
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:16.528598522 26221   0x7f240428f0 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder11:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:16.528709498 26221   0x7f240428f0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7eb4025530 Failed to determine interlace mode
0:01:16.528789754 26221   0x7f240428f0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7eb4025530 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:16.635618174 26221   0x7f240428f0 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder11:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:16.636056831 26221   0x7f240428f0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7eb4025530 Failed to determine interlace mode
0:01:16.636333727 26221   0x7f240428f0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7eb4025530 Failed to determine interlace mode
In cb_newpad
0:01:16.639409060 26221   0x7f240428f0 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder11> Duration invalid, not setting latency
0:01:16.640286789 26221   0x7f240428f0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder11:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:16.660486340 26221   0x7f240408f0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder11:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:17.018512425 26221   0x7f240408f0 WARN            videodecoder gstvideodecoder.c:2780:gst_video_decoder_prepare_finish_frame:<nvv4l2decoder11> decreasing timestamp (0:01:03.773044444 < 0:01:04.293044444)

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:18.405993786 26221   0x558e42f390 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:18.406256314 26221   0x558e42f390 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:18.441083888 26221   0x558e42f390 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:18.443200595 26221   0x558e42f390 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin12
0:01:18.444353301 26221   0x558e42f390 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux12
Decodebin child added: multiqueue12
Decodebin child added: h264parse12
Decodebin child added: capsfilter12
Decodebin child added: nvv4l2decoder12
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:18.594367259 26221   0x7f24003540 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder12:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:18.594499547 26221   0x7f24003540 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed8359900 Failed to determine interlace mode
0:01:18.594618844 26221   0x7f24003540 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed8359900 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:18.701709504 26221   0x7f24003540 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder12:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:18.701823904 26221   0x7f24003540 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed8359900 Failed to determine interlace mode
0:01:18.701975392 26221   0x7f24003540 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ed8359900 Failed to determine interlace mode
In cb_newpad
0:01:18.706191495 26221   0x7f24003540 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder12> Duration invalid, not setting latency
0:01:18.719482523 26221   0x7f24003540 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder12:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:18.736721590 26221   0x7f1c004990 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder12:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:19.019754120 26221   0x7f1c004990 WARN            videodecoder gstvideodecoder.c:2780:gst_video_decoder_prepare_finish_frame:<nvv4l2decoder12> decreasing timestamp (0:01:03.693044444 < 0:01:04.213044444)

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:20.511287802 26221   0x7ed5380870 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:20.511468762 26221   0x7ed5380870 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:20.547316369 26221   0x7ed5380870 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:20.549630805 26221   0x7ed5380870 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin13
0:01:20.551393592 26221   0x7ed5380870 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux13
Decodebin child added: multiqueue13
Decodebin child added: h264parse13
Decodebin child added: capsfilter13
Decodebin child added: nvv4l2decoder13
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:20.712526639 26221   0x7e3c003ed0 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder13:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:20.712649775 26221   0x7e3c003ed0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7f203bb320 Failed to determine interlace mode
0:01:20.712735375 26221   0x7e3c003ed0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7f203bb320 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:20.820233141 26221   0x7e3c003ed0 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder13:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:20.820378005 26221   0x7e3c003ed0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7f203bb320 Failed to determine interlace mode
0:01:20.820482805 26221   0x7e3c003ed0 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7f203bb320 Failed to determine interlace mode
In cb_newpad
0:01:20.823534330 26221   0x7e3c003ed0 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder13> Duration invalid, not setting latency
0:01:20.835069451 26221   0x7e3c003ed0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder13:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:20.852149958 26221   0x7f1c004ad0 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder13:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:21.152636307 26221   0x7f1c004ad0 WARN            videodecoder gstvideodecoder.c:2780:gst_video_decoder_prepare_finish_frame:<nvv4l2decoder13> decreasing timestamp (0:01:03.573044444 < 0:01:04.093044444)

Using winsys: x11 
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.1/lib/libnvds_mot_iou.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
gstnvtracker: Past frame output is OFF
0:01:22.401582482 26221   0x7f2800fc40 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 1]: deserialized trt engine from :/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
INFO: [Implicit Engine Info]: layers num: 2
0   INPUT  kFLOAT data            3x640x640       
1   OUTPUT kFLOAT prob            6001x1x1        

0:01:22.401757906 26221   0x7f2800fc40 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-nvinference-engine> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 1]: Use deserialized engine model: /home/jetson/Desktop/Project/build-AI_System-Desktop-Debug/Model/PrimaryInfer/yolov5s_fp16.engine
0:01:22.426900505 26221   0x7f2800fc40 INFO                 nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus:<primary-nvinference-engine> [UID 1]: Load new model:./Model/PrimaryInfer/yolov5s.txt sucessfully
0:01:22.428516347 26221   0x7f2800fc40 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: source
Decodebin child added: decodebin14
0:01:22.429479421 26221   0x7f2800fc40 WARN                 basesrc gstbasesrc.c:3583:gst_base_src_start_complete:<source> pad not activated yet
Decodebin child added: mpegpsdemux14
Decodebin child added: multiqueue14
Decodebin child added: h264parse14
Decodebin child added: capsfilter14
Decodebin child added: nvv4l2decoder14
Opening in BLOCKING MODE
Opening in BLOCKING MODE 
0:01:22.571218231 26221   0x7f24042800 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder14:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:22.571315159 26221   0x7f24042800 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ec83632a0 Failed to determine interlace mode
0:01:22.571484375 26221   0x7f24042800 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ec83632a0 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261 
NVMEDIA: Reading vendor.tegra.display-size : status: 6 
NvMMLiteBlockCreate : Block : BlockType = 261 
0:01:22.683602339 26221   0x7f24042800 WARN                    v4l2 gstv4l2object.c:4434:gst_v4l2_object_probe_caps:<nvv4l2decoder14:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:01:22.683720356 26221   0x7f24042800 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ec83632a0 Failed to determine interlace mode
0:01:22.683805828 26221   0x7f24042800 WARN                    v4l2 gstv4l2object.c:2375:gst_v4l2_object_add_interlace_mode:0x7ec83632a0 Failed to determine interlace mode
In cb_newpad
0:01:22.686379432 26221   0x7f24042800 WARN            v4l2videodec gstv4l2videodec.c:1673:gst_v4l2_video_dec_decide_allocation:<nvv4l2decoder14> Duration invalid, not setting latency
0:01:22.701327423 26221   0x7f24042800 WARN          v4l2bufferpool gstv4l2bufferpool.c:1065:gst_v4l2_buffer_pool_start:<nvv4l2decoder14:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:01:22.726185669 26221   0x7f24003f20 WARN          v4l2bufferpool gstv4l2bufferpool.c:1512:gst_v4l2_buffer_pool_dqbuf:<nvv4l2decoder14:pool:src> Driver should never set v4l2_buffer.field to ANY
IOU Tracker Init with threshold 0.100000
0:01:22.852577703 26221   0x7ec0019990 WARN           nveglglessink gsteglglessink.c:443:gst_egl_image_buffer_pool_alloc_buffer: Fallback memory allocation

** (AI_System:26221): WARNING **: 23:41:09.290: Use gst_egl_image_allocator_alloc() to allocate from this allocator
0:01:22.854362666 26221   0x7ec0019990 WARN              GST_BUFFER gstbuffer.c:907:gst_buffer_new_allocate: failed to allocate 8294400 bytes
0:01:22.854476938 26221   0x7ec0019990 WARN               videopool gstvideopool.c:272:video_buffer_pool_alloc:<eglimagebufferpool14> can't create memory
0:01:22.854515018 26221   0x7ec0019990 WARN              bufferpool gstbufferpool.c:310:do_alloc_buffer:<eglimagebufferpool14> alloc function failed
0:01:22.854553002 26221   0x7ec0019990 WARN              bufferpool gstbufferpool.c:343:default_start:<eglimagebufferpool14> failed to allocate buffer
0:01:22.854579882 26221   0x7ec0019990 ERROR             bufferpool gstbufferpool.c:564:gst_buffer_pool_set_active:<eglimagebufferpool14> start failed
0:01:22.854734058 26221   0x7ec0019990 WARN           basetransform gstbasetransform.c:1670:default_prepare_output_buffer:<nvegl-transform> error: failed to activate bufferpool
0:01:22.854784811 26221   0x7ec0019990 WARN           basetransform gstbasetransform.c:1670:default_prepare_output_buffer:<nvegl-transform> error: failed to activate bufferpool
0:01:22.855079819 26221   0x7ec0019990 WARN           basetransform gstbasetransform.c:2159:default_generate_output:<nvegl-transform> could not get buffer from pool: error
realloc(): invalid next size
Aborted (core dumped)
root@jetson-desktop:/home/jetson/Desktop/Project/build-AI_System-Desktop-Debug# ^C

After init and destroy the pipeline again and again,I don‘t know why the nveglglessink raised a “Fallback memory allocation“. And then the GST_BUFFER failed to allocate 8294400 bytes memories——about 7.9 mb.

My device is Xavier NX, I found the memory only used 6.3 GB when the app running. Still have 1.7GB unused.
(Got the memory info by Jtop.)

So it seemd like the crash caused by nveglglessink. How to solve it?