From cc9f28c2ef0162b6f9e888aab046646745a01c48 Mon Sep 17 00:00:00 2001
From: Jean-Marc Valin <jmvalin@jmvalin.ca>
Date: Wed, 24 Oct 2012 01:49:50 -0400
Subject: [PATCH] Optimizations to comb_filter()
---
celt/celt.c | 42 +++++++++++++++++++++++++++---------------
1 file changed, 27 insertions(+), 15 deletions(-)
diff --git a/celt/celt.c b/celt/celt.c
index a055052fe..257199dbd 100644
--- a/celt/celt.c
+++ b/celt/celt.c
@@ -549,6 +549,7 @@ static void comb_filter(opus_val32 *y, opus_val32 *x, int T0, int T1, int N,
int i;
/* printf ("%d %d %f %f\n", T0, T1, g0, g1); */
opus_val16 g00, g01, g02, g10, g11, g12;
+ opus_val32 x0, x1, x2, x3, x4;
static const opus_val16 gains[3][3] = {
{QCONST16(0.3066406250f, 15), QCONST16(0.2170410156f, 15), QCONST16(0.1296386719f, 15)},
{QCONST16(0.4638671875f, 15), QCONST16(0.2680664062f, 15), QCONST16(0.f, 15)},
@@ -559,30 +560,41 @@ static void comb_filter(opus_val32 *y, opus_val32 *x, int T0, int T1, int N,
g10 = MULT16_16_Q15(g1, gains[tapset1][0]);
g11 = MULT16_16_Q15(g1, gains[tapset1][1]);
g12 = MULT16_16_Q15(g1, gains[tapset1][2]);
+ x1 = x[-T1+1];
+ x2 = x[-T1 ];
+ x3 = x[-T1-1];
+ x4 = x[-T1-2];
for (i=0;i<overlap;i++)
{
opus_val16 f;
+ x0=x[i-T1+2];
f = MULT16_16_Q15(window[i],window[i]);
y[i] = x[i]
+ MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g00),x[i-T0])
- + MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g01),x[i-T0-1])
- + MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g01),x[i-T0+1])
- + MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g02),x[i-T0-2])
- + MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g02),x[i-T0+2])
- + MULT16_32_Q15(MULT16_16_Q15(f,g10),x[i-T1])
- + MULT16_32_Q15(MULT16_16_Q15(f,g11),x[i-T1-1])
- + MULT16_32_Q15(MULT16_16_Q15(f,g11),x[i-T1+1])
- + MULT16_32_Q15(MULT16_16_Q15(f,g12),x[i-T1-2])
- + MULT16_32_Q15(MULT16_16_Q15(f,g12),x[i-T1+2]);
+ + MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g01),ADD32(x[i-T0+1],x[i-T0-1]))
+ + MULT16_32_Q15(MULT16_16_Q15((Q15ONE-f),g02),ADD32(x[i-T0+2],x[i-T0-2]))
+ + MULT16_32_Q15(MULT16_16_Q15(f,g10),x2)
+ + MULT16_32_Q15(MULT16_16_Q15(f,g11),ADD32(x1,x3))
+ + MULT16_32_Q15(MULT16_16_Q15(f,g12),ADD32(x0,x4));
+ x4=x3;
+ x3=x2;
+ x2=x1;
+ x1=x0;
}
- for (i=overlap;i<N;i++)
+ /* OPT: For machines where the movs are costly, unroll by 5 */
+ for (;i<N;i++)
+ {
+ x0=x[i-T1+2];
y[i] = x[i]
- + MULT16_32_Q15(g10,x[i-T1])
- + MULT16_32_Q15(g11,x[i-T1-1])
- + MULT16_32_Q15(g11,x[i-T1+1])
- + MULT16_32_Q15(g12,x[i-T1-2])
- + MULT16_32_Q15(g12,x[i-T1+2]);
+ + MULT16_32_Q15(g10,x2)
+ + MULT16_32_Q15(g11,ADD32(x1,x3))
+ + MULT16_32_Q15(g12,ADD32(x0,x4));
+ x4=x3;
+ x3=x2;
+ x2=x1;
+ x1=x0;
+ }
}
static const signed char tf_select_table[4][8] = {
--
GitLab