Do avx2/hbd scaling*grain multiplication in 16bit instead of 32bit

approved this merge request

added 2 commits

• 90dad3ee - 1 commit from branch videolan:master
• a8b13fc1 - Do avx2/hbd scaling*grain multiplication in 16bit instead of 32bit

Compare with previous version

enabled an automatic merge when the pipeline for a8b13fc1 succeeds

merged

changed milestone to %0.9.1

added performance label

added x86 label