=~=~=~=~=~=~=~=~=~=~=~= PuTTY log 2017.08.29 20:36:08 =~=~=~=~=~=~=~=~=~=~=~=

d1230@linse3:~/no_backup/d1230/TensorRT-2.1.2/bin> cuda-memcheck  ./sample_int8 imagenet 
========= CUDA-MEMCHECK

INT8 run:1 batches of size 64 starting at 0
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (63,0,0) in block (87,0,0)
=========     Address 0x1050d7e6988 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (62,0,0) in block (87,0,0)
=========     Address 0x1050d7e6978 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (61,0,0) in block (87,0,0)
=========     Address 0x1050d7e6968 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (60,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e98 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (59,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e88 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (58,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e78 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (57,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e68 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (56,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e58 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (55,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e48 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (54,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e38 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (53,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e28 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (52,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e18 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (51,0,0) in block (87,0,0)
=========     Address 0x1050d7e5e08 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (50,0,0) in block (87,0,0)
=========     Address 0x1050d7e5df8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (49,0,0) in block (87,0,0)
=========     Address 0x1050d7e5de8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (48,0,0) in block (87,0,0)
=========     Address 0x1050d7e5dd8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (47,0,0) in block (87,0,0)
=========     Address 0x1050d7e5dc8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (46,0,0) in block (87,0,0)
=========     Address 0x1050d7e5db8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (45,0,0) in block (87,0,0)
=========     Address 0x1050d7e5da8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (44,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d98 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (43,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d88 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (42,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d78 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (41,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d68 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (40,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d58 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (39,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d48 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (38,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d38 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (37,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d28 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (36,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d18 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (35,0,0) in block (87,0,0)
=========     Address 0x1050d7e5d08 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (34,0,0) in block (87,0,0)
=========     Address 0x1050d7e5cf8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (33,0,0) in block (87,0,0)
=========     Address 0x1050d7e5ce8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Invalid __global__ read of size 4
=========     at 0x00000608 in trtwell_scudnn_128x32_relu_interior_nn
=========     by thread (32,0,0) in block (87,0,0)
=========     Address 0x1050d7e5cd8 is out of bounds
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204205]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x15645]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 [0x156d7]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaLaunchKernel + 0x1e1) [0x43a01]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN3dit12ImplicitGemm5doRunINS_24ImplicitGemmKernelParamsILi512EEEEEiRKNS_20ConvolutionRunConfigEP11CUstream_st + 0x110) [0x436f20]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn20WinogradConvActLayer7executeERKNS0_13CommonContextE + 0xb08) [0x3715a8]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2aa) [0x31fb1a]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn15calibrateEngineERNS_15IInt8CalibratorERNS_11ICudaEngineERSt13unordered_mapISsfSt4hashISsESt8equal_toISsESaISt4pairIKSsfEEEb + 0x3b7) [0x34e8e7]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17builder11buildEngineERNS_21CudaEngineBuildConfigERKNS_5cudnn15HardwareContextERKNS_7NetworkE + 0xd3d) [0x30e62d]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer17Builder15buildCudaEngineERNS_18INetworkDefinitionE + 0x11) [0x378eb1]
=========     Host Frame:./sample_int8 [0x66f5]
=========     Host Frame:./sample_int8 [0x75ec]
=========     Host Frame:./sample_int8 [0x85bf]
=========     Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xf5) [0x206e5]
=========     Host Frame:./sample_int8 [0x5a89]
=========
========= Program hit cudaErrorLaunchFailure (error 4) due to "unspecified launch failure" on CUDA API call to cudaDeviceSynchronize. 
=========     Saved host backtrace up to driver entry point at error
=========     Host Frame:/usr/lib64/libcuda.so.1 [0x2ef423]
=========     Host Frame:/usr/local/cuda-8.0/lib64/libcudart.so.8.0 (cudaDeviceSynchronize + 0x166) [0x334a6]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinfer.so.3 (_ZN8nvinfer15cudnn16ExecutionContext7executeEiPPv + 0x2c6) [0x31fb36]
=========     Host Frame:/home/d1230/no_backup/d1230/TensorRT-2.1.2/lib/libnvinf