tree 33ebeccead7560c96914f8f8cd48428590264e12
parent 1f1d6fcb3aca5b6942da121c4e16a8423a7e0f66
author Jonathan Wright <jonathan.wright@arm.com> 1684938390 +0100
committer James Zern <jzern@google.com> 1685485209 +0000

Optimize Neon implementation of aom_int_pro_row

Unroll to operate on 4 rows per iteration. Also increase the number
of accumulators from 2 to 4, removing a bottleneck.

Change-Id: I4149f7f026ac15d5eee676b7e1b65ea873716d76