GPUassert: an illegal memory access was encountered src/modules/NvMultiObjectTracker/context.cpp 197

[chinthy2]

Please provide complete information as applicable to your setup.

• Hardware Platform - DGPU (T4)
• DeepStream Version - 6.0.1
• Issue Type - Bug
• How to reproduce the issue ?

Ran a custom deepstream pipeline for 2 weeks. Then the program crashed with following log.
Custom pipeline - rtspsrc ! nvstreammux ! nvinfer ! nvtracker ! fakesink
Used nvdcf tracker configurations. Crash log is attached here.

"log":"ERROR: nvdsinfer_context_impl.cpp:341 Failed to make stream wait on event, cuda err_no:700, err_str:cudaErrorIllegalAddress"
"log":"ERROR: nvdsinfer_context_impl.cpp:1619 Preprocessor transform input data failed., nvinfer error:NVDSINFER_CUDA_ERROR"
"log":"339:15:08.523733669 \u001B[332m    1\u001B[00m 0x7f38bc002f70 \u001B[33;01mWARN   \u001B[00m \u001B[00m             nvinfer gstnvinfer.cpp:1324:gst_nvinfer_input_queue_loop:<primary-nvinference-engine>\u001B[00m error: Failed to queue input batch for inferencing"
"log":"[0x00007f3b827f4700][error] Error from element: primary-nvinference-engine:Failed to queue input batch for inferencing"
"log":"[0x00007f3b827f4700][error] Error details: gstnvinfer.cpp(1324): gst_nvinfer_input_queue_loop (): /GstPipeline:main-pipeline/GstNvInfer:primary-nvinference-engine"
"log":"GPUassert: an illegal memory access was encountered src/modules/NvMultiObjectTracker/context.cpp 197"
"log":"terminate called after throwing an instance of 'std::invalid_argument'"

[kesong]

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

Can you have a try with the latest DeepStream 6.2? Can you check if there is any memory leak?