neon variance: process 4x blocks
authorJohann <johannkoenig@google.com>
Tue, 2 May 2017 14:31:05 +0000 (07:31 -0700)
committerJohann <johannkoenig@google.com>
Thu, 18 May 2017 00:35:01 +0000 (17:35 -0700)
commit7b742da63e4b829ba013670dd838d263f5df8956
tree223562c9789d8fbcbc3b71fba016da4b99eeacc1
parent2057d3ef757a18e6bb005812a9912748ae4c7610
neon variance: process 4x blocks

Continue processing sets of 16 values. Plenty of improvement for 4x8
(doubles the speed) but only about 30% for 4x4.

BUG=webm:1422

Change-Id: Ib8dd96f75d474f0348800271d11e58356b620905
test/variance_test.cc
vpx_dsp/arm/mem_neon.h
vpx_dsp/arm/variance_neon.c
vpx_dsp/vpx_dsp_rtcd_defs.pl