From 44a0036d10579ed91e48df24859e54b08a582742 Mon Sep 17 00:00:00 2001 From: Christophe Gisquet Date: Sat, 24 Nov 2012 23:15:07 +0100 Subject: x86: sbrdsp: Implement SSE sum64x5 698 to 174 cycles on Arrandale. Unrolling is a 6 cycles gain. Signed-off-by: Diego Biurrun --- libavcodec/x86/sbrdsp_init.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'libavcodec/x86/sbrdsp_init.c') diff --git a/libavcodec/x86/sbrdsp_init.c b/libavcodec/x86/sbrdsp_init.c index 51c4bd4a16..264b455b3e 100644 --- a/libavcodec/x86/sbrdsp_init.c +++ b/libavcodec/x86/sbrdsp_init.c @@ -25,6 +25,7 @@ #include "libavcodec/sbrdsp.h" float ff_sbr_sum_square_sse(float (*x)[2], int n); +void ff_sbr_sum64x5_sse(float *z); void ff_sbr_hf_g_filt_sse(float (*Y)[2], const float (*X_high)[40][2], const float *g_filt, int m_max, intptr_t ixh); void ff_sbr_hf_gen_sse(float (*X_high)[2], const float (*X_low)[2], @@ -37,6 +38,7 @@ void ff_sbrdsp_init_x86(SBRDSPContext *s) if (EXTERNAL_SSE(mm_flags)) { s->sum_square = ff_sbr_sum_square_sse; + s->sum64x5 = ff_sbr_sum64x5_sse; s->hf_g_filt = ff_sbr_hf_g_filt_sse; s->hf_gen = ff_sbr_hf_gen_sse; } -- cgit v1.2.3