* re-enabled NEON routines for arm64

backport 997b7b18

Due to size limit of shared memory, histogram is built on
the global memory for CV_16UC1 case.

The amount of memory needed for building histogram is:

    65536 * 4byte = 256KB

and shared memory limit is 48KB typically.

Added test cases for CV_16UC1 and various clip limits.
Added perf tests for CV_16UC1 on both CPU and CUDA code.

There was also a bug in CV_8UC1 case when redistributing
"residual" clipped pixels. Adding the test case where clip
limit is 5.0 exposes this bug.

instead of get_local_size(0) and dynamic local memory allocation

* Optical Flow rework to use wide universal intrinsics

* remove if (useSIMD) check as review requested

YUV42x color conversions rewritten to wide intrinsics (#13708)

* a*b+c -> fma

* YUV420sp2RGB initially vectorized

* shorter var names

* loops by 4

* yuv420p2rgb vectorized

* yuv422toRGB vectorized

* reg arrays

* rgb2yuv420 vectorized

* warnings fixed

* try to fix align error

SVM sigmoid kernel fix (issue #13621) (#13718)

* Added test for sigmoid case for retrieving support vectors

* undo unhelpful test

* add test for sigmoid SVM with data that is easily separable into two concentric circles

* Update sigmoid kernel to use tanh(gamma * <x, y> + coef0) instead of -tanh(gamma * <x, y> + coef0)

* remove unnecessary constraint on coef0

* cleanup

* fixing inappropriate use of doubles

* Add f to float literal

* replace CV_Assert with ASSERT_EQ where appropriate

* spatialGradient() reworked to use wide universal intrinsics

* Moved row pointers inside loops