Optimize Neon implementation of vpx_int_pro_col
authorJonathan Wright <jonathan.wright@arm.com>
Tue, 30 May 2023 14:22:04 +0000 (15:22 +0100)
committerJonathan Wright <jonathan.wright@arm.com>
Wed, 31 May 2023 13:30:02 +0000 (14:30 +0100)
commitc738e87f27ef8e12dd28b9052f446a5f69abf3c9
treec24856549456db876b489192cd3b47b7e43c4272
parent99522d307ccef8b53d373beab8c5b6bf997ca4ef
Optimize Neon implementation of vpx_int_pro_col

Use widening pairwise addition instructions to halve the number of
additions required.

Change-Id: I0307a3b65e50d2b1ae582938bc5df9c2b21df734
vpx_dsp/arm/avg_neon.c