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 * Authors:  Darko Laus      (darko@mips.com)
30 *           Djordje Pesut   (djordje@mips.com)
31 *           Mirjana Vulin   (mvulin@mips.com)
32 *
33 * AAC Spectral Band Replication decoding functions optimized for MIPS
34 *
35 * This file is part of FFmpeg.
36 *
37 * FFmpeg is free software; you can redistribute it and/or
38 * modify it under the terms of the GNU Lesser General Public
39 * License as published by the Free Software Foundation; either
40 * version 2.1 of the License, or (at your option) any later version.
41 *
42 * FFmpeg is distributed in the hope that it will be useful,
43 * but WITHOUT ANY WARRANTY; without even the implied warranty of
44 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
45 * Lesser General Public License for more details.
46 *
47 * You should have received a copy of the GNU Lesser General Public
48 * License along with FFmpeg; if not, write to the Free Software
49 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
50 */
51
52/**
53 * @file
54 * Reference: libavcodec/aacdec.c
55 */
56
57#ifndef AVCODEC_MIPS_AACDEC_MIPS_H
58#define AVCODEC_MIPS_AACDEC_MIPS_H
59
60#include "libavcodec/aac.h"
61#include "libavutil/mips/asmdefs.h"
62
63#if HAVE_INLINE_ASM && HAVE_MIPSFPU
64#if !HAVE_MIPS32R6 && !HAVE_MIPS64R6
65static inline float *VMUL2_mips(float *dst, const float *v, unsigned idx,
66                           const float *scale)
67{
68    float temp0, temp1, temp2;
69    int temp3, temp4;
70    float *ret;
71
72    __asm__ volatile(
73        "andi    %[temp3],  %[idx],       0x0F         \n\t"
74        "andi    %[temp4],  %[idx],       0xF0         \n\t"
75        "sll     %[temp3],  %[temp3],     2            \n\t"
76        "srl     %[temp4],  %[temp4],     2            \n\t"
77        "lwc1    %[temp2],  0(%[scale])                \n\t"
78        "lwxc1   %[temp0],  %[temp3](%[v])             \n\t"
79        "lwxc1   %[temp1],  %[temp4](%[v])             \n\t"
80        "mul.s   %[temp0],  %[temp0],     %[temp2]     \n\t"
81        "mul.s   %[temp1],  %[temp1],     %[temp2]     \n\t"
82        PTR_ADDIU "%[ret],  %[dst],       8            \n\t"
83        "swc1    %[temp0],  0(%[dst])                  \n\t"
84        "swc1    %[temp1],  4(%[dst])                  \n\t"
85
86        : [temp0]"=&f"(temp0), [temp1]"=&f"(temp1),
87          [temp2]"=&f"(temp2), [temp3]"=&r"(temp3),
88          [temp4]"=&r"(temp4), [ret]"=&r"(ret)
89        : [idx]"r"(idx), [scale]"r"(scale), [v]"r"(v),
90          [dst]"r"(dst)
91        : "memory"
92    );
93    return ret;
94}
95
96static inline float *VMUL4_mips(float *dst, const float *v, unsigned idx,
97                           const float *scale)
98{
99    int temp0, temp1, temp2, temp3;
100    float temp4, temp5, temp6, temp7, temp8;
101    float *ret;
102
103    __asm__ volatile(
104        "andi    %[temp0],  %[idx],       0x03        \n\t"
105        "andi    %[temp1],  %[idx],       0x0C        \n\t"
106        "andi    %[temp2],  %[idx],       0x30        \n\t"
107        "andi    %[temp3],  %[idx],       0xC0        \n\t"
108        "sll     %[temp0],  %[temp0],     2           \n\t"
109        "srl     %[temp2],  %[temp2],     2           \n\t"
110        "srl     %[temp3],  %[temp3],     4           \n\t"
111        "lwc1    %[temp4],  0(%[scale])               \n\t"
112        "lwxc1   %[temp5],  %[temp0](%[v])            \n\t"
113        "lwxc1   %[temp6],  %[temp1](%[v])            \n\t"
114        "lwxc1   %[temp7],  %[temp2](%[v])            \n\t"
115        "lwxc1   %[temp8],  %[temp3](%[v])            \n\t"
116        "mul.s   %[temp5],  %[temp5],     %[temp4]    \n\t"
117        "mul.s   %[temp6],  %[temp6],     %[temp4]    \n\t"
118        "mul.s   %[temp7],  %[temp7],     %[temp4]    \n\t"
119        "mul.s   %[temp8],  %[temp8],     %[temp4]    \n\t"
120        PTR_ADDIU "%[ret],  %[dst],       16          \n\t"
121        "swc1    %[temp5],  0(%[dst])                 \n\t"
122        "swc1    %[temp6],  4(%[dst])                 \n\t"
123        "swc1    %[temp7],  8(%[dst])                 \n\t"
124        "swc1    %[temp8],  12(%[dst])                \n\t"
125
126        : [temp0]"=&r"(temp0), [temp1]"=&r"(temp1),
127          [temp2]"=&r"(temp2), [temp3]"=&r"(temp3),
128          [temp4]"=&f"(temp4), [temp5]"=&f"(temp5),
129          [temp6]"=&f"(temp6), [temp7]"=&f"(temp7),
130          [temp8]"=&f"(temp8), [ret]"=&r"(ret)
131        : [idx]"r"(idx), [scale]"r"(scale), [v]"r"(v),
132          [dst]"r"(dst)
133        : "memory"
134    );
135    return ret;
136}
137
138static inline float *VMUL2S_mips(float *dst, const float *v, unsigned idx,
139                            unsigned sign, const float *scale)
140{
141    int temp0, temp1, temp2, temp3, temp4, temp5;
142    float temp6, temp7, temp8, temp9;
143    float *ret;
144
145    __asm__ volatile(
146        "andi    %[temp0],  %[idx],       0x0F       \n\t"
147        "andi    %[temp1],  %[idx],       0xF0       \n\t"
148        "lw      %[temp4],  0(%[scale])              \n\t"
149        "srl     %[temp2],  %[sign],      1          \n\t"
150        "sll     %[temp3],  %[sign],      31         \n\t"
151        "sll     %[temp2],  %[temp2],     31         \n\t"
152        "sll     %[temp0],  %[temp0],     2          \n\t"
153        "srl     %[temp1],  %[temp1],     2          \n\t"
154        "lwxc1   %[temp8],  %[temp0](%[v])           \n\t"
155        "lwxc1   %[temp9],  %[temp1](%[v])           \n\t"
156        "xor     %[temp5],  %[temp4],     %[temp2]   \n\t"
157        "xor     %[temp4],  %[temp4],     %[temp3]   \n\t"
158        "mtc1    %[temp5],  %[temp6]                 \n\t"
159        "mtc1    %[temp4],  %[temp7]                 \n\t"
160        "mul.s   %[temp8],  %[temp8],     %[temp6]   \n\t"
161        "mul.s   %[temp9],  %[temp9],     %[temp7]   \n\t"
162        PTR_ADDIU "%[ret],  %[dst],       8          \n\t"
163        "swc1    %[temp8],  0(%[dst])                \n\t"
164        "swc1    %[temp9],  4(%[dst])                \n\t"
165
166        : [temp0]"=&r"(temp0), [temp1]"=&r"(temp1),
167          [temp2]"=&r"(temp2), [temp3]"=&r"(temp3),
168          [temp4]"=&r"(temp4), [temp5]"=&r"(temp5),
169          [temp6]"=&f"(temp6), [temp7]"=&f"(temp7),
170          [temp8]"=&f"(temp8), [temp9]"=&f"(temp9),
171          [ret]"=&r"(ret)
172        : [idx]"r"(idx), [scale]"r"(scale), [v]"r"(v),
173          [dst]"r"(dst), [sign]"r"(sign)
174        : "memory"
175    );
176    return ret;
177}
178
179static inline float *VMUL4S_mips(float *dst, const float *v, unsigned idx,
180                            unsigned sign, const float *scale)
181{
182    int temp0, temp1, temp2, temp3, temp4;
183    float temp10, temp11, temp12, temp13, temp14, temp15, temp16, temp17;
184    float *ret;
185    unsigned int mask = 1U << 31;
186
187    __asm__ volatile(
188        "lw      %[temp0],   0(%[scale])               \n\t"
189        "andi    %[temp1],  %[idx],       0x03         \n\t"
190        "andi    %[temp2],  %[idx],       0x0C         \n\t"
191        "andi    %[temp3],  %[idx],       0x30         \n\t"
192        "andi    %[temp4],  %[idx],       0xC0         \n\t"
193        "sll     %[temp1],  %[temp1],     2            \n\t"
194        "srl     %[temp3],  %[temp3],     2            \n\t"
195        "srl     %[temp4],  %[temp4],     4            \n\t"
196        "lwxc1   %[temp10],  %[temp1](%[v])            \n\t"
197        "lwxc1   %[temp11],  %[temp2](%[v])            \n\t"
198        "lwxc1   %[temp12],  %[temp3](%[v])            \n\t"
199        "lwxc1   %[temp13],  %[temp4](%[v])            \n\t"
200        "and     %[temp1],   %[sign],      %[mask]     \n\t"
201        "srl     %[temp2],   %[idx],       12          \n\t"
202        "srl     %[temp3],   %[idx],       13          \n\t"
203        "srl     %[temp4],   %[idx],       14          \n\t"
204        "andi    %[temp2],   %[temp2],     1           \n\t"
205        "andi    %[temp3],   %[temp3],     1           \n\t"
206        "andi    %[temp4],   %[temp4],     1           \n\t"
207        "sllv    %[sign],    %[sign],      %[temp2]    \n\t"
208        "xor     %[temp1],   %[temp0],     %[temp1]    \n\t"
209        "and     %[temp2],   %[sign],      %[mask]     \n\t"
210        "mtc1    %[temp1],   %[temp14]                 \n\t"
211        "xor     %[temp2],   %[temp0],     %[temp2]    \n\t"
212        "sllv    %[sign],    %[sign],      %[temp3]    \n\t"
213        "mtc1    %[temp2],   %[temp15]                 \n\t"
214        "and     %[temp3],   %[sign],      %[mask]     \n\t"
215        "sllv    %[sign],    %[sign],      %[temp4]    \n\t"
216        "xor     %[temp3],   %[temp0],     %[temp3]    \n\t"
217        "and     %[temp4],   %[sign],      %[mask]     \n\t"
218        "mtc1    %[temp3],   %[temp16]                 \n\t"
219        "xor     %[temp4],   %[temp0],     %[temp4]    \n\t"
220        "mtc1    %[temp4],   %[temp17]                 \n\t"
221        "mul.s   %[temp10],  %[temp10],    %[temp14]   \n\t"
222        "mul.s   %[temp11],  %[temp11],    %[temp15]   \n\t"
223        "mul.s   %[temp12],  %[temp12],    %[temp16]   \n\t"
224        "mul.s   %[temp13],  %[temp13],    %[temp17]   \n\t"
225        PTR_ADDIU "%[ret],   %[dst],       16          \n\t"
226        "swc1    %[temp10],  0(%[dst])                 \n\t"
227        "swc1    %[temp11],  4(%[dst])                 \n\t"
228        "swc1    %[temp12],  8(%[dst])                 \n\t"
229        "swc1    %[temp13],  12(%[dst])                \n\t"
230
231        : [temp0]"=&r"(temp0), [temp1]"=&r"(temp1),
232          [temp2]"=&r"(temp2), [temp3]"=&r"(temp3),
233          [temp4]"=&r"(temp4), [temp10]"=&f"(temp10),
234          [temp11]"=&f"(temp11), [temp12]"=&f"(temp12),
235          [temp13]"=&f"(temp13), [temp14]"=&f"(temp14),
236          [temp15]"=&f"(temp15), [temp16]"=&f"(temp16),
237          [temp17]"=&f"(temp17), [ret]"=&r"(ret),
238          [sign]"+r"(sign)
239        : [idx]"r"(idx), [scale]"r"(scale), [v]"r"(v),
240          [dst]"r"(dst), [mask]"r"(mask)
241        : "memory"
242    );
243    return ret;
244}
245
246#define VMUL2 VMUL2_mips
247#define VMUL4 VMUL4_mips
248#define VMUL2S VMUL2S_mips
249#define VMUL4S VMUL4S_mips
250#endif /* !HAVE_MIPS32R6 && !HAVE_MIPS64R6 */
251#endif /* HAVE_INLINE_ASM && HAVE_MIPSFPU */
252
253#endif /* AVCODEC_MIPS_AACDEC_MIPS_H */
254