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: Bojan Zivkovic (bojan@mips.com)
30 *
31 * Compute antialias function optimised for MIPS fixed-point architecture
32 *
33 * This file is part of FFmpeg.
34 *
35 * FFmpeg is free software; you can redistribute it and/or
36 * modify it under the terms of the GNU Lesser General Public
37 * License as published by the Free Software Foundation; either
38 * version 2.1 of the License, or (at your option) any later version.
39 *
40 * FFmpeg is distributed in the hope that it will be useful,
41 * but WITHOUT ANY WARRANTY; without even the implied warranty of
42 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
43 * Lesser General Public License for more details.
44 *
45 * You should have received a copy of the GNU Lesser General Public
46 * License along with FFmpeg; if not, write to the Free Software
47 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
48 */
49
50 /**
51 * @file
52 * Reference: libavcodec/mpegaudiodec.c
53 */
54
55 #ifndef AVCODEC_MIPS_COMPUTE_ANTIALIAS_FIXED_H
56 #define AVCODEC_MIPS_COMPUTE_ANTIALIAS_FIXED_H
57
58 #if HAVE_INLINE_ASM
compute_antialias_mips_fixed(MPADecodeContext * s,GranuleDef * g)59 static void compute_antialias_mips_fixed(MPADecodeContext *s,
60 GranuleDef *g)
61 {
62 const int32_t *csa;
63 int32_t *ptr;
64 int n, i;
65 int MAX_lo = 0xffffffff;
66
67 /* we antialias only "long" bands */
68 if (g->block_type == 2) {
69 if (!g->switch_point)
70 return;
71 /* XXX: check this for 8000Hz case */
72 n = 1;
73 } else {
74 n = SBLIMIT - 1;
75 }
76
77
78 ptr = g->sb_hybrid + 18;
79
80 for(i = n;i > 0;i--) {
81 int tmp0, tmp1, tmp2, tmp00, tmp11;
82 int temp_reg1, temp_reg2, temp_reg3, temp_reg4, temp_reg5, temp_reg6;
83 csa = &csa_table[0][0];
84
85 /**
86 * instructions are scheduled to minimize pipeline stall.
87 */
88 __asm__ volatile (
89 "lw %[tmp0], -1*4(%[ptr]) \n\t"
90 "lw %[tmp1], 0*4(%[ptr]) \n\t"
91 "lw %[temp_reg1], 0*4(%[csa]) \n\t"
92 "lw %[temp_reg2], 2*4(%[csa]) \n\t"
93 "add %[tmp2], %[tmp0], %[tmp1] \n\t"
94 "lw %[temp_reg3], 3*4(%[csa]) \n\t"
95 "mult $ac0, %[tmp2], %[temp_reg1] \n\t"
96 "mult $ac1, %[tmp2], %[temp_reg1] \n\t"
97 "lw %[tmp00], -2*4(%[ptr]) \n\t"
98 "lw %[tmp11], 1*4(%[ptr]) \n\t"
99 "lw %[temp_reg4], 4*4(%[csa]) \n\t"
100 "mtlo %[MAX_lo], $ac0 \n\t"
101 "mtlo $zero, $ac1 \n\t"
102 "msub $ac0, %[tmp1], %[temp_reg2] \n\t"
103 "madd $ac1, %[tmp0], %[temp_reg3] \n\t"
104 "add %[tmp2], %[tmp00], %[tmp11] \n\t"
105 "lw %[temp_reg5], 6*4(%[csa]) \n\t"
106 "mult $ac2, %[tmp2], %[temp_reg4] \n\t"
107 "mult $ac3, %[tmp2], %[temp_reg4] \n\t"
108 "mfhi %[temp_reg1], $ac0 \n\t"
109 "mfhi %[temp_reg2], $ac1 \n\t"
110 "lw %[temp_reg6], 7*4(%[csa]) \n\t"
111 "mtlo %[MAX_lo], $ac2 \n\t"
112 "msub $ac2, %[tmp11], %[temp_reg5] \n\t"
113 "mtlo $zero, $ac3 \n\t"
114 "madd $ac3, %[tmp00], %[temp_reg6] \n\t"
115 "sll %[temp_reg1], %[temp_reg1], 2 \n\t"
116 "sw %[temp_reg1], -1*4(%[ptr]) \n\t"
117 "mfhi %[temp_reg4], $ac2 \n\t"
118 "sll %[temp_reg2], %[temp_reg2], 2 \n\t"
119 "mfhi %[temp_reg5], $ac3 \n\t"
120 "sw %[temp_reg2], 0*4(%[ptr]) \n\t"
121 "lw %[tmp0], -3*4(%[ptr]) \n\t"
122 "lw %[tmp1], 2*4(%[ptr]) \n\t"
123 "lw %[temp_reg1], 8*4(%[csa]) \n\t"
124 "sll %[temp_reg4], %[temp_reg4], 2 \n\t"
125 "add %[tmp2], %[tmp0], %[tmp1] \n\t"
126 "sll %[temp_reg5], %[temp_reg5], 2 \n\t"
127 "mult $ac0, %[tmp2], %[temp_reg1] \n\t"
128 "mult $ac1, %[tmp2], %[temp_reg1] \n\t"
129 "sw %[temp_reg4], -2*4(%[ptr]) \n\t"
130 "sw %[temp_reg5], 1*4(%[ptr]) \n\t"
131 "lw %[temp_reg2], 10*4(%[csa]) \n\t"
132 "mtlo %[MAX_lo], $ac0 \n\t"
133 "lw %[temp_reg3], 11*4(%[csa]) \n\t"
134 "msub $ac0, %[tmp1], %[temp_reg2] \n\t"
135 "mtlo $zero, $ac1 \n\t"
136 "madd $ac1, %[tmp0], %[temp_reg3] \n\t"
137 "lw %[tmp00], -4*4(%[ptr]) \n\t"
138 "lw %[tmp11], 3*4(%[ptr]) \n\t"
139 "mfhi %[temp_reg1], $ac0 \n\t"
140 "lw %[temp_reg4], 12*4(%[csa]) \n\t"
141 "mfhi %[temp_reg2], $ac1 \n\t"
142 "add %[tmp2], %[tmp00], %[tmp11] \n\t"
143 "mult $ac2, %[tmp2], %[temp_reg4] \n\t"
144 "mult $ac3, %[tmp2], %[temp_reg4] \n\t"
145 "lw %[temp_reg5], 14*4(%[csa]) \n\t"
146 "lw %[temp_reg6], 15*4(%[csa]) \n\t"
147 "sll %[temp_reg1], %[temp_reg1], 2 \n\t"
148 "mtlo %[MAX_lo], $ac2 \n\t"
149 "msub $ac2, %[tmp11], %[temp_reg5] \n\t"
150 "mtlo $zero, $ac3 \n\t"
151 "madd $ac3, %[tmp00], %[temp_reg6] \n\t"
152 "sll %[temp_reg2], %[temp_reg2], 2 \n\t"
153 "sw %[temp_reg1], -3*4(%[ptr]) \n\t"
154 "mfhi %[temp_reg4], $ac2 \n\t"
155 "sw %[temp_reg2], 2*4(%[ptr]) \n\t"
156 "mfhi %[temp_reg5], $ac3 \n\t"
157 "lw %[tmp0], -5*4(%[ptr]) \n\t"
158 "lw %[tmp1], 4*4(%[ptr]) \n\t"
159 "lw %[temp_reg1], 16*4(%[csa]) \n\t"
160 "lw %[temp_reg2], 18*4(%[csa]) \n\t"
161 "add %[tmp2], %[tmp0], %[tmp1] \n\t"
162 "lw %[temp_reg3], 19*4(%[csa]) \n\t"
163 "mult $ac0, %[tmp2], %[temp_reg1] \n\t"
164 "mult $ac1, %[tmp2], %[temp_reg1] \n\t"
165 "sll %[temp_reg4], %[temp_reg4], 2 \n\t"
166 "sll %[temp_reg5], %[temp_reg5], 2 \n\t"
167 "sw %[temp_reg4], -4*4(%[ptr]) \n\t"
168 "mtlo %[MAX_lo], $ac0 \n\t"
169 "msub $ac0, %[tmp1], %[temp_reg2] \n\t"
170 "mtlo $zero, $ac1 \n\t"
171 "madd $ac1, %[tmp0], %[temp_reg3] \n\t"
172 "sw %[temp_reg5], 3*4(%[ptr]) \n\t"
173 "lw %[tmp00], -6*4(%[ptr]) \n\t"
174 "mfhi %[temp_reg1], $ac0 \n\t"
175 "lw %[tmp11], 5*4(%[ptr]) \n\t"
176 "mfhi %[temp_reg2], $ac1 \n\t"
177 "lw %[temp_reg4], 20*4(%[csa]) \n\t"
178 "add %[tmp2], %[tmp00], %[tmp11] \n\t"
179 "lw %[temp_reg5], 22*4(%[csa]) \n\t"
180 "mult $ac2, %[tmp2], %[temp_reg4] \n\t"
181 "mult $ac3, %[tmp2], %[temp_reg4] \n\t"
182 "lw %[temp_reg6], 23*4(%[csa]) \n\t"
183 "sll %[temp_reg1], %[temp_reg1], 2 \n\t"
184 "sll %[temp_reg2], %[temp_reg2], 2 \n\t"
185 "mtlo %[MAX_lo], $ac2 \n\t"
186 "msub $ac2, %[tmp11], %[temp_reg5] \n\t"
187 "mtlo $zero, $ac3 \n\t"
188 "madd $ac3, %[tmp00], %[temp_reg6] \n\t"
189 "sw %[temp_reg1], -5*4(%[ptr]) \n\t"
190 "sw %[temp_reg2], 4*4(%[ptr]) \n\t"
191 "mfhi %[temp_reg4], $ac2 \n\t"
192 "lw %[tmp0], -7*4(%[ptr]) \n\t"
193 "mfhi %[temp_reg5], $ac3 \n\t"
194 "lw %[tmp1], 6*4(%[ptr]) \n\t"
195 "lw %[temp_reg1], 24*4(%[csa]) \n\t"
196 "lw %[temp_reg2], 26*4(%[csa]) \n\t"
197 "add %[tmp2], %[tmp0], %[tmp1] \n\t"
198 "lw %[temp_reg3], 27*4(%[csa]) \n\t"
199 "mult $ac0, %[tmp2], %[temp_reg1] \n\t"
200 "mult $ac1, %[tmp2], %[temp_reg1] \n\t"
201 "sll %[temp_reg4], %[temp_reg4], 2 \n\t"
202 "sll %[temp_reg5], %[temp_reg5], 2 \n\t"
203 "sw %[temp_reg4], -6*4(%[ptr]) \n\t"
204 "mtlo %[MAX_lo], $ac0 \n\t"
205 "msub $ac0, %[tmp1], %[temp_reg2] \n\t"
206 "mtlo $zero, $ac1 \n\t"
207 "madd $ac1, %[tmp0], %[temp_reg3] \n\t"
208 "sw %[temp_reg5], 5*4(%[ptr]) \n\t"
209 "lw %[tmp00], -8*4(%[ptr]) \n\t"
210 "mfhi %[temp_reg1], $ac0 \n\t"
211 "lw %[tmp11], 7*4(%[ptr]) \n\t"
212 "mfhi %[temp_reg2], $ac1 \n\t"
213 "lw %[temp_reg4], 28*4(%[csa]) \n\t"
214 "add %[tmp2], %[tmp00], %[tmp11] \n\t"
215 "lw %[temp_reg5], 30*4(%[csa]) \n\t"
216 "mult $ac2, %[tmp2], %[temp_reg4] \n\t"
217 "mult $ac3, %[tmp2], %[temp_reg4] \n\t"
218 "lw %[temp_reg6], 31*4(%[csa]) \n\t"
219 "sll %[temp_reg1], %[temp_reg1], 2 \n\t"
220 "sll %[temp_reg2], %[temp_reg2], 2 \n\t"
221 "mtlo %[MAX_lo], $ac2 \n\t"
222 "msub $ac2, %[tmp11], %[temp_reg5] \n\t"
223 "mtlo $zero, $ac3 \n\t"
224 "madd $ac3, %[tmp00], %[temp_reg6] \n\t"
225 "sw %[temp_reg1], -7*4(%[ptr]) \n\t"
226 "sw %[temp_reg2], 6*4(%[ptr]) \n\t"
227 "mfhi %[temp_reg4], $ac2 \n\t"
228 "mfhi %[temp_reg5], $ac3 \n\t"
229 "sll %[temp_reg4], %[temp_reg4], 2 \n\t"
230 "sll %[temp_reg5], %[temp_reg5], 2 \n\t"
231 "sw %[temp_reg4], -8*4(%[ptr]) \n\t"
232 "sw %[temp_reg5], 7*4(%[ptr]) \n\t"
233
234 : [tmp0] "=&r" (tmp0), [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2),
235 [tmp00] "=&r" (tmp00), [tmp11] "=&r" (tmp11),
236 [temp_reg1] "=&r" (temp_reg1), [temp_reg2] "=&r" (temp_reg2),
237 [temp_reg3] "=&r" (temp_reg3), [temp_reg4] "=&r" (temp_reg4),
238 [temp_reg5] "=&r" (temp_reg5), [temp_reg6] "=&r" (temp_reg6)
239 : [csa] "r" (csa), [ptr] "r" (ptr),
240 [MAX_lo] "r" (MAX_lo)
241 : "memory", "hi", "lo", "$ac1hi", "$ac1lo", "$ac2hi", "$ac2lo",
242 "$ac3hi", "$ac3lo"
243 );
244
245 ptr += 18;
246 }
247 }
248 #define compute_antialias compute_antialias_mips_fixed
249 #endif /* HAVE_INLINE_ASM */
250
251 #endif /* AVCODEC_MIPS_COMPUTE_ANTIALIAS_FIXED_H */
252