Working environment:
- Ubuntu 14.04 with stock dev. tools (CMake 3.4.3, GCC 4.8.4 ..)
- OpenCV 3.1.0
- CUDA 7.5
OpenCV builds and installs OK build with the following cmake configuration and all "normal" functionality works as expected
cmake -D CMAKE_BUILD_TYPE=DEBUG -D CMAKE_INSTALL_PREFIX=/usr/local -D WITH_CUDA=ON -D ENABLE_FAST_MATH=1 -D CUDA_FAST_MATH=1 -D WITH_CUBLAS=1 -D OPENCV_EXTRA_MODULES_PATH=../../opencv_contrib-3.1.0/modules -D PYTHON_INCLUDE_DIR=$(python-config --prefix)/include/python2.7 -D BUILD_TESTS=OFF -D BUILD_PERF_TESTS=OFF -D BUILD_DOCS=OFF -D BUILD_EXAMPLES=ON ..
OpenCV configuration reported by cmake:
-- General configuration for OpenCV 3.1.0 =====================================
-- Version control: d097d6d
--
-- Platform:
-- Host: Linux 4.2.0-27-generic x86_64
-- CMake: 3.4.3
-- CMake generator: Unix Makefiles
-- CMake build tool: /usr/bin/make
-- Configuration: DEBUG
--
-- C/C++:
-- Built as dynamic libs?: YES
-- C++ Compiler: /usr/bin/c++ (ver 4.8.4)
-- C++ flags (Release): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wno-narrowing -Wno-delete-non-virtual-dtor -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffast-math -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -fvisibility-inlines-hidden -O3 -DNDEBUG -DNDEBUG
-- C++ flags (Debug): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wno-narrowing -Wno-delete-non-virtual-dtor -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffast-math -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -fvisibility-inlines-hidden -g -O0 -DDEBUG -D_DEBUG
-- C Compiler: /usr/bin/cc
-- C flags (Release): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wno-narrowing -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffast-math -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -O3 -DNDEBUG -DNDEBUG
-- C flags (Debug): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wno-narrowing -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffast-math -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -g -O0 -DDEBUG -D_DEBUG
-- Linker flags (Release):
-- Linker flags (Debug):
-- Precompiled headers: YES
-- Extra dependencies: /usr/lib/x86_64-linux-gnu/libpng.so /usr/lib/x86_64-linux-gnu/libz.so /usr/lib/x86_64-linux-gnu/libtiff.so /usr/lib/x86_64-linux-gnu/libjasper.so /usr/lib/x86_64-linux-gnu/libjpeg.so gtk-x11-2.0 gdk-x11-2.0 atk-1.0 gio-2.0 pangoft2-1.0 pangocairo-1.0 gdk_pixbuf-2.0 cairo pango-1.0 fontconfig gobject-2.0 freetype gthread-2.0 glib-2.0 dc1394 v4l1 v4l2 avcodec avformat avutil swscale dl m pthread rt cudart nppc nppi npps cublas cufft -L/usr/local/cuda/lib64
-- 3rdparty dependencies: libwebp IlmImf libprotobuf
--
-- OpenCV modules:
-- To be built: cudev core cudaarithm flann imgproc ml reg surface_matching video cudabgsegm cudafilters cudaimgproc cudawarping dnn fuzzy imgcodecs photo shape videoio cudacodec highgui objdetect plot xobjdetect xphoto bgsegm bioinspired dpm face features2d line_descriptor saliency text calib3d ccalib cudafeatures2d cudalegacy cudaobjdetect cudaoptflow cudastereo datasets rgbd stereo structured_light superres tracking videostab xfeatures2d ximgproc aruco optflow stitching python2
-- Disabled: world contrib_world
-- Disabled by dependency: -
-- Unavailable: java python3 ts viz cvv hdf matlab sfm
--
-- GUI:
-- QT: NO
-- GTK+ 2.x: YES (ver 2.24.23)
-- GThread : YES (ver 2.40.2)
-- GtkGlExt: NO
-- OpenGL support: NO
-- VTK support: NO
--
-- Media I/O:
-- ZLib: /usr/lib/x86_64-linux-gnu/libz.so (ver 1.2.8)
-- JPEG: /usr/lib/x86_64-linux-gnu/libjpeg.so (ver )
-- WEBP: build (ver 0.3.1)
-- PNG: /usr/lib/x86_64-linux-gnu/libpng.so (ver 1.2.50)
-- TIFF: /usr/lib/x86_64-linux-gnu/libtiff.so (ver 42 - 4.0.3)
-- JPEG 2000: /usr/lib/x86_64-linux-gnu/libjasper.so (ver 1.900.1)
-- OpenEXR: build (ver 1.7.1)
-- GDAL: NO
--
-- Video I/O:
-- DC1394 1.x: NO
-- DC1394 2.x: YES (ver 2.2.1)
-- FFMPEG: YES
-- codec: YES (ver 54.35.0)
-- format: YES (ver 54.20.4)
-- util: YES (ver 52.3.0)
-- swscale: YES (ver 2.1.1)
-- resample: NO
-- gentoo-style: YES
-- GStreamer: NO
-- OpenNI: NO
-- OpenNI PrimeSensor Modules: NO
-- OpenNI2: NO
-- PvAPI: NO
-- GigEVisionSDK: NO
-- UniCap: NO
-- UniCap ucil: NO
-- V4L/V4L2: Using libv4l1 (ver 1.0.1) / libv4l2 (ver 1.0.1)
-- XIMEA: NO
-- Xine: NO
-- gPhoto2: NO
--
-- Parallel framework: pthreads
--
-- Other third-party libraries:
-- Use IPP: 9.0.1 [9.0.1]
-- at: /home/developer/projects/opencv/3rdparty/ippicv/unpack/ippicv_lnx
-- Use IPP Async: NO
-- Use VA: NO
-- Use Intel VA-API/OpenCL: NO
-- Use Eigen: NO
-- Use Cuda: YES (ver 7.5)
-- Use OpenCL: YES
-- Use custom HAL: NO
--
-- NVIDIA CUDA
-- Use CUFFT: YES
-- Use CUBLAS: YES
-- USE NVCUVID: NO
-- NVIDIA GPU arch: 20 21 30 35
-- NVIDIA PTX archs: 30
-- Use fast math: YES
--
-- OpenCL:
-- Version: dynamic
-- Include path: /home/developer/projects/opencv/3rdparty/include/opencl/1.2
-- Use AMDFFT: NO
-- Use AMDBLAS: NO
--
-- Python 2:
-- Interpreter: /usr/bin/python2.7 (ver 2.7.6)
-- Libraries: /usr/lib/x86_64-linux-gnu/libpython2.7.so (ver 2.7.6)
-- numpy: /usr/lib/python2.7/dist-packages/numpy/core/include (ver 1.8.2)
-- packages path: lib/python2.7/dist-packages
--
-- Python 3:
-- Interpreter: /usr/bin/python3.4 (ver 3.4.3)
--
-- Python (for build): /usr/bin/python2.7
--
-- Java:
-- ant: NO
-- JNI: NO
-- Java wrappers: NO
-- Java tests: NO
--
-- Matlab: Matlab not found or implicitly disabled
--
-- Tests and samples:
-- Tests: NO
-- Performance tests: NO
-- C/C++ Examples: YES
--
-- Install path: /usr/local
--
-- cvconfig.h is in: /home/developer/projects/opencv/build
-- -----------------------------------------------------------------
However any attempt to call opencv gpu accelerated CUDA functionality results in segmentation faults. None of the sample projects in samples/gpu will run properly. It seems that non-trivial CUDA functions seg fault within the initial function call, however have observed a couple of trivial CUDA functions such as getCudaEnabledDeviceCount() and printShortCudaDeviceInfo() do appear to execute OK and return sensible data, however subsequent opencv calls (not necessarily CUDA functions) will subsequently seg fault.
I have tried building and running a couple of CUDA library samples from /usr/local/cuda-7.5/samples (1_utilities/deviceQuery, 1_utilities/bandwithTest) and they seem fine.
Simple program that exhibits problem:
makefile
CFLAGS = `pkg-config --cflags opencv`
LIBS = `pkg-config --libs opencv`
fail : fail.cpp
g++ $(CFLAGS) $< $(LIBS) -o $@
fail.cpp
#include "opencv2/cvconfig.h"
#include "opencv2/core.hpp"
#include "opencv2/cudaarithm.hpp"
using namespace std;
using namespace cv;
using namespace cv::cuda;
int main(int argc, const char* argv[])
{
int i = getCudaEnabledDeviceCount();
Mat src(1000, 1000, CV_32F);
Mat dst;
printf("got to here #1\n");
cv::transpose(src, dst);
printf("got to here #2\n");
return 0;
}
Program output:
got to here #1
Segmentation fault (core dumped)
Comment out the getCudaEnabledDeviceCount() statement and it runs to completion as expected.
Sort of smells like a calling convention mismatch between the opencv and CUDA libraries or something similar, but would have thought this should "just work"...
[edit]
with core dumps enabled get the following call stack in gdb...
developer@odin:~/projects/temp/opencv_CUDA$ gdb -c core -e fail
GNU gdb (Ubuntu 7.7.1-0ubuntu5~14.04.2) 7.7.1
Copyright (C) 2014 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word".
warning: exec file is newer than core file.
[New LWP 28690]
[New LWP 28691]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `./fail'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 __GI___pthread_mutex_lock (mutex=0x0) at ../nptl/pthread_mutex_lock.c:66
66 ../nptl/pthread_mutex_lock.c: No such file or directory.
(gdb) bt
#0 __GI___pthread_mutex_lock (mutex=0x0) at ../nptl/pthread_mutex_lock.c:66
#1 0x00007f9e85782008 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#2 0x00007f9e85836671 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#3 0x00007f9e858367e5 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#4 0x00007f9e85787cb4 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#5 0x00007f9e857894e7 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#6 0x00007f9e8575cc66 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#7 0x00007f9e8565bf3d in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#8 0x00007f9e8565bed8 in ?? () from /usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1
#9 0x00007f9e85fe2022 in ?? () from /usr/local/cuda/lib64/libOpenCL.so
#10 0x00007f9e85fe3d42 in ?? () from /usr/local/cuda/lib64/libOpenCL.so
#11 0x00007f9e85fe34d0 in clGetPlatformIDs () from /usr/local/cuda/lib64/libOpenCL.so
#12 0x00007f9e895cea83 in (anonymous namespace)::opencl_fn3<58, int, unsigned int, _cl_platform_id**, unsigned int*>::switch_fn (p1=0, p2=0x0,
p3=0x7ffe56410aac) at /home/developer/projects/opencv/modules/core/src/opencl/runtime/autogenerated/opencl_core_impl.hpp:127
#13 0x00007f9e8965c452 in cv::ocl::haveOpenCL () at /home/developer/projects/opencv/modules/core/src/ocl.cpp:1466
#14 0x00007f9e8965c4af in cv::ocl::useOpenCL () at /home/developer/projects/opencv/modules/core/src/ocl.cpp:1487
#15 0x00007f9e896a6962 in cv::transpose (_src=..., _dst=...) at /home/developer/projects/opencv/modules/core/src/matrix.cpp:3235
#16 0x0000000000400f3c in ?? ()
#17 0x00007ffe56410e88 in ?? ()
#18 0x0000000156410df8 in ?? ()
#19 0x00007ffe56410dd0 in ?? ()
#20 0x000000018a755d48 in ?? ()
#21 0x0000000001010000 in ?? ()
#22 0x00007ffe56410cd0 in ?? ()
#23 0x0000000000000000 in ?? ()
