DeHalo_alpha: Fix 32-bit float support

Keep up with the changes in DFTTest

Fix KNLMeansCL

Cosmetics

Toon: Remove unnecessary workaround for AWarpSharp2

It was added when AWarpSharp2 supported only 8 bit depth years ago.

Add MCTemporalDenoise function (no PP)

Add 'opencl' parameter to daa, santiag and QTGMC

MinBlur: Remove workaround for CTMF in r=2 when input is 16 bits

GSMC: Use proper scaling in luma-adaptative mask

Overlay: Fix variable referenced before assignment error