(cherry picked from commit b2790973)
(cherry picked from commit f8758da2)

(cherry picked from commit e650d87e)
(cherry picked from commit ca9c52ac)

How this worked before, I do not know.

Feature allows to exclude CUDA or OpenCL optimizations at all even CUDA is used
on build. Exclusion of CUDA or OpenCL cut unwanted dependencies.

Improve comments to indicate actual usage of WORDS_BIGENDIAN where
it is tested with #ifdef rather than #if

EMBED_CUDA and FORCE_EMBED_OPENCV flags added to cmake macro add_android_project;
INSTALL_CUDA_LIBRARIES option added to OpenCV.mk
opencv_dynamicuda library installation with enabled OPENCV_INSTALL_MODULES flag fixed;
CUDA initialization apportunity added to OpenCVLoader.initDebug();
Tutorial-4-CUDA sample reimplemented with static OpenCV and CUDA initialization.

It's not documented, and it does nothing unless CMake 2.0 compatibility
is enabled (and it isn't):

https://github.com/Kitware/CMake/blob/v2.6.0/Source/cmConfigureFileCommand.cxx

Added opencv_testing.sh.in file;
opencv_testing.sh installation guarded by OS check.

Env setup for testing package implemented using /etc/profile.d;
Variable with path for all native samples added;
Path for test binaries and test data updated.