Error in prepare.sh Jetson and Jetson_trt on last commit

[edsalazarx]

*[ULTALPR_SDK INFO]:
** This application is used to generate optimized NVIDIA TensorRT models **

*[COMPV INFO]: [UltAlprSdkEngine] Call: optimizeTRT
*[COMPV INFO]: [CompVBase] Initializing [base] modules (v 1.0.0, nt 1)...
*[COMPV INFO]: [CompVBase] sizeof(compv_scalar_t)= #8
*[COMPV INFO]: [CompVBase] sizeof(float)= #4
*[COMPV INFO]: Initializing window registery
*[COMPV INFO]: [ImageDecoder] Initializing image decoder...
*[COMPV INFO]: [CompVCpu] H: '', S: '', M: '', MN: 'ARMv8 Processor rev 1 (v8l)'
*[COMPV INFO]: [CompVBase] CPU features: [arm];[arm64];neon;neon_fma;vfpv4;
*[COMPV INFO]: [CompVBase] CPU cores: #4
*[COMPV INFO]: [CompVBase] CPU cache1: line size: #64B, size :#0KB
*[COMPV INFO]: [CompVBase] CPU Phys RAM size: #3964GB
*[COMPV INFO]: [CompVBase] CPU endianness: LITTLE
*[COMPV INFO]: [CompVBase] Binary type: AArch64
*[COMPV INFO]: [CompVBase] Intrinsic enabled
*[COMPV INFO]: [CompVBase] Assembler enabled
*[COMPV INFO]: [CompVBase] Code built with option /arch:NEON
*[COMPV INFO]: [CompVBase] OS name: Jetson_TFTRT
*[COMPV INFO]: [CompVBase] Math Fast Trig.: true
*[COMPV INFO]: [CompVBase] Math Fixed Point: true
*[COMPV INFO]: [CompVMathExp] Init
*[COMPV INFO]: [CompVBase] Default alignment: #32
*[COMPV INFO]: [CompVBase] Best alignment: #32
*[COMPV INFO]: [CompVBase] Heap limit: #262144KB (#256MB)
*[COMPV INFO]: [CompVParallel] Initializing [parallel] module...
*[COMPV INFO]: [CompVParallel] [Parallel] module initialized
*[COMPV INFO]: [CompVBase] [Base] modules initialized
*[COMPV INFO]: [UltAlprSdkTRT] Optimizing models in [../../../assets/models.tensorrt/] folder...
*[COMPV INFO]: [FileUtils] Loading files in ../../../assets/models.tensorrt/ ...
*[COMPV INFO]: [CompVSharedLib] Loaded shared lib: /home/homelock/Downloads/ultimateALPR-SDK/binaries/jetson/aarch64/libultimatePluginTensorRT.so
*[COMPV INFO]: [UltAlprSdkTRT] **** Optimizing ultimateALPR-SDK_klass_lpci.desktop.model.tensorrt.doubango [3.3]: 1/6 ****
*[COMPV INFO]: /!\ Code in file '/home/ultimate/compv/base/compv_mem.cxx' in function 'CompVMemCopy_C' starting at line #956: Not optimized -> No SIMD implementation found. On ARM consider http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.faqs/ka13544.html
*[PLUGIN_TENSORRT INFO]: [TensorRT Optimizer] Loading TensorRT local plugins...
*[PLUGIN_TENSORRT INFO]: [TensorRT Optimizer] Loading TensorRT local plugins done.
*[PLUGIN_TENSORRT INFO]: [TensorRT Optimizer] Starting to parse the ONNX/UFF data...
*[PLUGIN_TENSORRT INFO]: [TensorRT Optimizer] Parsing the ONNX/UFF data is done.
*[PLUGIN_TENSORRT INFO]: [TensorRT Optimizer] Starting to build the CUDA engine...
***[PLUGIN_TENSORRT ERROR]: function: "log()"
file: "/home/projects/ultimateTRT/pluginTensorRT/source/plugin_tensorrt_inference_engine.cxx"
line: "33"
message: [TensorRT Inference] From logger: /home/jenkins/workspace/TensorRT/helpers/rel-7.1/L1_Nightly_Internal/build/source/rtSafe/resources.h (460) - Cuda Error in loadKernel: 702 (the launch timed out and was terminated)
***[PLUGIN_TENSORRT ERROR]: function: "log()"
file: "/home/projects/ultimateTRT/pluginTensorRT/source/plugin_tensorrt_inference_engine.cxx"
line: "33"
message: [TensorRT Inference] From logger: ../rtSafe/safeRuntime.cpp (32) - Cuda Error in free: 702 (the launch timed out and was terminated)
terminate called after throwing an instance of 'nvinfer1::CudaError'
what(): std::exception
./prepare.sh: line 5: 18725 Aborted sudo ./trt_optimizer --assets ../../../assets

[DoubangoTelecom]

Are you using jetpack version 4.4.1? I see "nigthy" in your version. If you're using the rigth jetpack version, share nvidia-smi output

[DoubangoTelecom]

Other recommendations:

• use binaries in jetson folder instead of jetson_tftrt.
• Disconnect the monitor and connect to the device via putty
• put the device in maximum power mode
• use power jack instead of USB charger
  I'm giving these recommendations because based on https://forums.developer.nvidia.com/t/fatal-error-the-launch-timed-out-and-was-terminated/42339 it may be your GPU is too busy with other tasks like displaying the desktop (ui)

[edsalazarx]

Previus commit generare a models optimized wihout problem, y try it with jetson_trft and the same problem

[DoubangoTelecom]

Previus commit generare a models optimized wihout problem, y try it with jetson_trft and the same problem

You're not answering our question about jetpack version. Also, previous commits didn't change jetson at all, they were about Linux x86, android and Raspberry. Not a single jetson file was changed in the previous commits. You should double check your device.

[DoubangoTelecom]

Last commit that changed jetson was on February. We have more that 10k devices using the implementation on jetson. Few chances that no one noticed the issue for 5 months. Again, check your device, answer to our questions about jetpack and nvidia-smi