From 705f5e5e155f6f280a360af220fc5b30cfcee702 Mon Sep 17 00:00:00 2001 From: Janne Grunau Date: Tue, 1 Dec 2015 13:37:41 +0100 Subject: arm64: port synth_filter_float_neon from arm ~25% faster dts decoding overall. The checkasm CPU cycles numbers are not that useful since synth_filter_float() calls FFTContext.imdct_half(). cortex-a57 cortex-a53 synth_filter_float_c: 1866.2 3490.9 synth_filter_float_neon: 915.0 1531.5 With fftc.imdct_half forced to imdct_half_neon: cortex-a57 cortex-a53 synth_filter_float_c: 1718.4 3025.3 synth_filter_float_neon: 926.2 1530.1 --- libavcodec/synth_filter.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'libavcodec/synth_filter.c') diff --git a/libavcodec/synth_filter.c b/libavcodec/synth_filter.c index d0ace4040e..708bd4e00c 100644 --- a/libavcodec/synth_filter.c +++ b/libavcodec/synth_filter.c @@ -60,6 +60,10 @@ av_cold void ff_synth_filter_init(SynthFilterContext *c) { c->synth_filter_float = synth_filter_float; - if (ARCH_ARM) ff_synth_filter_init_arm(c); - if (ARCH_X86) ff_synth_filter_init_x86(c); + if (ARCH_AARCH64) + ff_synth_filter_init_aarch64(c); + if (ARCH_ARM) + ff_synth_filter_init_arm(c); + if (ARCH_X86) + ff_synth_filter_init_x86(c); } -- cgit v1.2.3