In the Dev branch it is stated that after building gpgpusim, it is possible to run cuda based applications by chnaging the cuda library paths to gpgusim library folder. Also, the GPU config file should be in the working directory.
Well, I did that but the application says no cuda device found! I also ran deviceQuery and it also said there is no cuda.
mahmood@u1604:~/gpgpu-sim_distribution$ echo $LD_LIBRARY_PATH
mahmood@u1604:~/gpgpu-sim_distribution$ source setup_environment release
GPGPU-Sim version 3.2.2 (build ) WARNING ** GPGPU-Sim version 3.2.2 not fully tested with CUDA version 7.5 (please see README)
configured with GPUWattch.
setup_environment succeeded
mahmood@u1604:~/gpgpu-sim_distribution$ echo $LD_LIBRARY_PATH
/home/mahmood/gpgpu-sim_distribution/lib/gcc-4.8.5/cuda-7050/release:
mahmood@u1604:~/gpgpu-sim_distribution$ ls /home/mahmood/gpgpu-sim_distribution/lib/gcc-4.8.5/cuda-7050/release
libcudart.so libcudart.so.4 libcudart.so.6.0 libcudart.so.8.0
libcudart.so.2 libcudart.so.5.0 libcudart.so.6.5
libcudart.so.3 libcudart.so.5.5 libcudart.so.7.5
mahmood@u1604:~/gunrock/build/bin$ ls
bc chesapeake.mtx hits shared_lib_bc shared_lib_example sssp
bfs config_quadro_islip.icnt pr shared_lib_bfs shared_lib_pr topk
cc gpgpusim.config salsa shared_lib_cc shared_lib_sssp wtf
mahmood@u1604:~/gunrock/build/bin$ ./bfs market chesapeake.mtx --src=0 --unidirected
Loading Matrix-market coordinate-formatted graph ...
Reading from chesapeake.mtx:
Parsing MARKET COO format (39 nodes, 340 directed edges)... Done parsing (0s).
Converting 39 vertices, 340 directed edges (unordered tuples) to CSR format...
Done converting (0s).
Degree Histogram (39 vertices, 340 edges):
Degree 0: 0 (0.00%)
Degree 2^0: 0 (0.00%)
Degree 2^1: 1 (2.56%)
Degree 2^2: 22 (56.41%)
Degree 2^3: 13 (33.33%)
Degree 2^4: 2 (5.13%)
Degree 2^5: 1 (2.56%)
Converting 39 vertices, 340 directed edges (unordered tuples) to CSR format...
Done converting (0s).
Source vertex: 0
Using 1 GPU: [ 0 ].
[/home/mahmood/gunrock/gunrock/util/info.cuh, 470 @ gpu 32764] cudaGetDevice failed (CUDA error 38: no CUDA-capable device is detected)
[/home/mahmood/gunrock/gunrock/util/test_utils.cu, 61 @ gpu 0] cudaSetDevice failed. (CUDA error 38: no CUDA-capable device is detected)
[/home/mahmood/gunrock/gunrock/util/info.cuh, 484 @ gpu 0] cudaStreamCreate failed. (CUDA error 38: no CUDA-capable device is detected)
CODE REQUESTED INVALID CUDA DEVICE -2050039464
mahmood@u1604:~/NVIDIA_CUDA-7.5_Samples/1_Utilities/deviceQuery$ make
"/usr/local/cuda-7.5"/bin/nvcc -ccbin g++ -I../../common/inc -m64 -gencode arch=compute_20,code=sm_20 -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_37,code=sm_37 -gencode arch=compute_50,code=sm_50 -gencode arch=compute_52,code=sm_52 -gencode arch=compute_52,code=compute_52 -o deviceQuery.o -c deviceQuery.cpp
"/usr/local/cuda-7.5"/bin/nvcc -ccbin g++ -m64 -gencode arch=compute_20,code=sm_20 -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_37,code=sm_37 -gencode arch=compute_50,code=sm_50 -gencode arch=compute_52,code=sm_52 -gencode arch=compute_52,code=compute_52 -o deviceQuery deviceQuery.o
mkdir -p ../../bin/x86_64/linux/release
cp deviceQuery ../../bin/x86_64/linux/release
mahmood@u1604:~/NVIDIA_CUDA-7.5_Samples/1_Utilities/deviceQuery$ ../../bin/x86_64/linux/release/deviceQuery
../../bin/x86_64/linux/release/deviceQuery Starting...
CUDA Device Query (Runtime API) version (CUDART static linking)
cudaGetDeviceCount returned 38
-> no CUDA-capable device is detected
Result = FAIL