* raise an error when wrong bit depth passed
  * raise an build error when wrong depth is specified for cvtScaleHalf_
  * remove unnecessary safe check in cvtScaleHalf_
  * use intrinsic instead of direct pointer access
  * update the explanation

  * add feature of Fp16 on GPU (cudev)
  * add test
  * leave template function as unimplemented to raise error

  * make sure that CV_FP16 has the correct meaning
  * check FP16 feature correctly

amends work done in pull request #6976.

When using OCL, the results of goodFeaturesToTrack() vary slightly from
run to run. This appears to be because the order of the results from
the findCorners kernel depends on thread execution and the sorting
function that is used at the end to rank the features only enforces are
partial sort order.

This does not materially impact the quality of the results, but it
makes it hard to build regression tests and generally introduces noise
into the system that should be avoided.

An easy fix is to change the sort function to enforce a total sort on
the features, even in cases where the match quality is exactly the same
for two features.

    Non  existence of _FPU_EXTENDED definition on powerpc64 (like it is for apple and arm)
    prevent compilation when WITH_CUDA is set. Adding powerpc64 as case to not use these
    definitions

        modifié:         modules/cudalegacy/test/TestHaarCascadeApplication.cpp
        modifié:         modules/cudalegacy/test/test_precomp.hpp
Signed-off-by: Thierry Fauck <tfauck@free.fr>

  * NAN is not defined on some platform
  * follow the comment on PR

 * test build on VS2012

  * NAN is not defined on some platform