search

csp256 / cudaLATCH

GPU implementation of LATCH descriptor & matcher.
40 stars 12 forks source link

Error compiling min.cpp #4

Open tzumby opened 8 years ago

tzumby commented 8 years ago

Hi @csp256, I tried compiling the min.cpp file and I'm getting this error. Did you come across anything like this before ?

$ make min nvcc -c -lineinfo -Xptxas -v -use_fast_math -O3 -o latch.o latch.cu -gencode arch=compute_52,code=sm_52 ptxas info : 0 bytes gmem, 6144 bytes cmem[3] ptxas info : Compiling entry function 'Z5latchPKfPjiiS0' for 'sm_52' ptxas info : Function properties for Z5latchPKfPjiiS0 0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads ptxas info : Used 32 registers, 17816 bytes smem, 360 bytes cmem[0], 2 textures nvcc -c -lineinfo -Xptxas -v -use_fast_math -O3 -o bitMatcher.o bitMatcher.cu -gencode arch=compute_52,code=sm_52 ptxas info : 0 bytes gmem ptxas info : Compiling entry function '_Z8bitMatchPKjS0_Piii' for 'sm_52' ptxas info : Function properties for _Z8bitMatchPKjS0_Piii 8 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads ptxas info : Used 64 registers, 32768 bytes smem, 352 bytes cmem[0] g++ pkg-config --cflags opencv min.cpp latch.o bitMatcher.o -I/usr/local/cuda-7.5/include/ -L/usr/local/cuda/lib64 -lcuda -lcudart -L/usr/local/lib -lopencv_stitching -lopencv_superres -lopencv_videostab -lopencv_aruco -lopencv_bgsegm -lopencv_bioinspired -lopencv_ccalib -lopencv_dnn -lopencv_dpm -lopencv_fuzzy -lopencv_line_descriptor -lopencv_optflow -lopencv_plot -lopencv_reg -lopencv_saliency -lopencv_stereo -lopencv_structured_light -lopencv_rgbd -lopencv_surface_matching -lopencv_tracking -lopencv_datasets -lopencv_text -lopencv_face -lopencv_xfeatures2d -lopencv_shape -lopencv_video -lopencv_ximgproc -lopencv_calib3d -lopencv_features2d -lopencv_flann -lopencv_xobjdetect -lopencv_objdetect -lopencv_ml -lopencv_xphoto -lopencv_highgui -lopencv_videoio -lopencv_imgcodecs -lopencv_photo -lopencv_imgproc -lopencvcore -o min min.cpp: In function ‘int main(int, char*)’: min.cpp:61:12: error: conflicting declaration ‘float_ d_K’ float _d_K, d_mask; ^ min.cpp:60:10: error: ‘dK’ has a previous declaration as ‘int d_K’ int _d_K, d_M1, d_M2; ^ min.cpp:76:23: error: ‘h_mask’ was not declared in this scope initMask(&d_mask, hmask); ^ min.cpp:92:103: error: cannot convert ‘int’ to ‘float’ for argument ‘4’ to ‘void latch(cv::Mat, unsigned char, sizet, float, unsigned int, int, int, float, std::vectorcv::KeyPoint, float_, cudaEvent_t)’ latch( img1g, d_I, pitch, h_K1, d_D1, &numKP1, maxKP, d_K, &keypoints1, dmask, latchFinishedEvent); ^ min.cpp:93:103: error: cannot convert ‘int’ to ‘float’ for argument ‘4’ to ‘void latch(cv::Mat, unsigned char, sizet, float, unsigned int, int, int, float, std::vectorcv::KeyPoint, float_, cudaEvent_t)’ latch( img2g, d_I, pitch, h_K2, d_D2, &numKP2, maxKP, d_K, &keypoints2, d_mask, latchFinishedEvent); ^ make: _* [min] Error 1

~/cudaLATCH$ nvcc -V nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2015 NVIDIA Corporation Built on Tue_Aug_11_14:27:32_CDT_2015 Cuda compilation tools, release 7.5, V7.5.17

csp256 commented 8 years ago

Oh yeah min is broken. I'm on vacation. Matthew's interface is far more user friendly. It'd pretty much have to be.

On Jul 9, 2016 6:42 PM, "tzumby" notifications@github.com wrote:

Hi @csp256 https://github.com/csp256, I tried compiling the min.cpp file and I'm getting this error. Did you come across anything like this before ?

$ make min nvcc -c -lineinfo -Xptxas -v -use_fast_math -O3 -o latch.o latch.cu -gencode arch=compute_52,code=sm_52 ptxas info : 0 bytes gmem, 6144 bytes cmem[3] ptxas info : Compiling entry function 'Z5latchPKfPjiiS0' for 'sm_52' ptxas info : Function properties for Z5latchPKfPjiiS0 0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads ptxas info : Used 32 registers, 17816 bytes smem, 360 bytes cmem[0], 2 textures nvcc -c -lineinfo -Xptxas -v -use_fast_math -O3 -o bitMatcher.o bitMatcher.cu -gencode arch=compute_52,code=sm_52 ptxas info : 0 bytes gmem ptxas info : Compiling entry function '_Z8bitMatchPKjS0_Piii' for 'sm_52' ptxas info : Function properties for _Z8bitMatchPKjS0_Piii 8 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads ptxas info : Used 64 registers, 32768 bytes smem, 352 bytes cmem[0] g++ pkg-config --cflags opencv min.cpp latch.o bitMatcher.o -I/usr/local/cuda-7.5/include/ -L/usr/local/cuda/lib64 -lcuda -lcudart -L/usr/local/lib -lopencv_stitching -lopencv_superres -lopencv_videostab -lopencv_aruco -lopencv_bgsegm -lopencv_bioinspired -lopencv_ccalib -lopencv_dnn -lopencv_dpm -lopencv_fuzzy -lopencv_line_descriptor -lopencv_optflow -lopencv_plot -lopencv_reg -lopencv_saliency -lopencv_stereo -lopencv_structured_light -lopencv_rgbd -lopencv_surface_matching -lopencv_tracking -lopencv_datasets -lopencv_text -lopencv_face -lopencv_xfeatures2d -lopencv_shape -lopencv_video -lopencv_ximgproc -lopencv_calib3d -lopencv_features2d -lopencv_flann -lopencv_xobjdetect -lopencv_objdetect -lopencv_ml -lopencv_xphoto -lopencv_highgui -lopencv_videoio -lopencv_imgcodecs -lopencv_photo -lopencv_imgproc -lopencvcore -o min min.cpp: In function ‘int main(int, char* )’: min.cpp:61:12: error: conflicting declaration ‘float_ d_K’ float

_d_K, *d_mask; ^ min.cpp:60:10: error: ‘dK’ has a previous declaration as ‘int d_K’ int

_d_K, d_M1, d_M2; ^ min.cpp:76:23: error: ‘h_mask’ was not declared in this scope initMask(&d_mask, hmask); ^ min.cpp:92:103: error: cannot convert ‘int’ to ‘float’ for argument ‘4’ to ‘void latch(cv::Mat, unsigned char, sizet, float, unsigned int, int, int, float, std::vectorcv::KeyPoint, float_, cudaEvent_t)’ latch( img1g, d_I, pitch, h_K1, d_D1, &numKP1, maxKP, d_K, &keypoints1, dmask, latchFinishedEvent); ^ min.cpp:93:103: error: cannot convert ‘int’ to ‘float’ for argument ‘4’ to ‘void latch(cv::Mat, unsigned char, sizet, float, unsigned int, int, int, float, std::vectorcv::KeyPoint, float

*, cudaEvent_t)’ latch( img2g, d_I, pitch, h_K2, d_D2, &numKP2, maxKP,

d_K, &keypoints2, d_mask, latchFinishedEvent); ^ make: *\ [min] Error 1

~/cudaLATCH$ nvcc -V nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2015 NVIDIA Corporation Built on Tue_Aug_11_14:27:32_CDT_2015 Cuda compilation tools, release 7.5, V7.5.17

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/csp256/cudaLATCH/issues/4, or mute the thread https://github.com/notifications/unsubscribe/AIIMxZ1cZ1DpzbfZn2I4zJ__wK7YAhZcks5qUDHRgaJpZM4JIufD .