1 /*
2 * Copyright (c) 2012
3 * MIPS Technologies, Inc., California.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * 3. Neither the name of the MIPS Technologies, Inc., nor the names of its
14 * contributors may be used to endorse or promote products derived from
15 * this software without specific prior written permission.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE MIPS TECHNOLOGIES, INC. ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE MIPS TECHNOLOGIES, INC. BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 *
29 * Author: Nedeljko Babic (nbabic@mips.com)
30 *
31 * adaptive and fixed codebook vector operations for ACELP-based codecs
32 * optimized for MIPS
33 *
34 * This file is part of FFmpeg.
35 *
36 * FFmpeg is free software; you can redistribute it and/or
37 * modify it under the terms of the GNU Lesser General Public
38 * License as published by the Free Software Foundation; either
39 * version 2.1 of the License, or (at your option) any later version.
40 *
41 * FFmpeg is distributed in the hope that it will be useful,
42 * but WITHOUT ANY WARRANTY; without even the implied warranty of
43 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
44 * Lesser General Public License for more details.
45 *
46 * You should have received a copy of the GNU Lesser General Public
47 * License along with FFmpeg; if not, write to the Free Software
48 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
49 */
50
51 /**
52 * @file
53 * Reference: libavcodec/acelp_vectors.c
54 */
55 #include "config.h"
56 #include "libavcodec/acelp_vectors.h"
57 #include "libavutil/mips/asmdefs.h"
58
59 #if HAVE_INLINE_ASM
60 #if !HAVE_MIPS32R6 && !HAVE_MIPS64R6
ff_weighted_vector_sumf_mips(float * out,const float * in_a,const float * in_b,float weight_coeff_a,float weight_coeff_b,int length)61 static void ff_weighted_vector_sumf_mips(
62 float *out, const float *in_a, const float *in_b,
63 float weight_coeff_a, float weight_coeff_b, int length)
64 {
65 const float *a_end = in_a + length;
66
67 /* loop unrolled two times */
68 __asm__ volatile (
69 "blez %[length], ff_weighted_vector_sumf_end%= \n\t"
70
71 "ff_weighted_vector_sumf_madd%=: \n\t"
72 "lwc1 $f0, 0(%[in_a]) \n\t"
73 "lwc1 $f3, 4(%[in_a]) \n\t"
74 "lwc1 $f1, 0(%[in_b]) \n\t"
75 "lwc1 $f4, 4(%[in_b]) \n\t"
76 "mul.s $f2, %[weight_coeff_a], $f0 \n\t"
77 "mul.s $f5, %[weight_coeff_a], $f3 \n\t"
78 "madd.s $f2, $f2, %[weight_coeff_b], $f1 \n\t"
79 "madd.s $f5, $f5, %[weight_coeff_b], $f4 \n\t"
80 PTR_ADDIU "%[in_a],8 \n\t"
81 PTR_ADDIU "%[in_b],8 \n\t"
82 "swc1 $f2, 0(%[out]) \n\t"
83 "swc1 $f5, 4(%[out]) \n\t"
84 PTR_ADDIU "%[out], 8 \n\t"
85 "bne %[in_a], %[a_end], ff_weighted_vector_sumf_madd%= \n\t"
86
87 "ff_weighted_vector_sumf_end%=: \n\t"
88
89 : [out] "+r" (out), [in_a] "+r" (in_a), [in_b] "+r" (in_b)
90 : [weight_coeff_a] "f" (weight_coeff_a),
91 [weight_coeff_b] "f" (weight_coeff_b),
92 [length] "r" (length), [a_end]"r"(a_end)
93 : "$f0", "$f1", "$f2", "$f3", "$f4", "$f5", "memory"
94 );
95 }
96 #endif /* !HAVE_MIPS32R6 && !HAVE_MIPS64R6 */
97 #endif /* HAVE_INLINE_ASM */
98
ff_acelp_vectors_init_mips(ACELPVContext * c)99 void ff_acelp_vectors_init_mips(ACELPVContext *c)
100 {
101 #if HAVE_INLINE_ASM
102 #if !HAVE_MIPS32R6 && !HAVE_MIPS64R6
103 c->weighted_vector_sumf = ff_weighted_vector_sumf_mips;
104 #endif
105 #endif
106 }
107