3 * MIPS Technologies, Inc., California.
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
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.
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
29 * Author: Branimir Vasic (bvasic@mips.com)
30 * Author: Zoran Lukic (zoranl@mips.com)
32 * This file is part of FFmpeg.
34 * FFmpeg is free software; you can redistribute it and/or
35 * modify it under the terms of the GNU Lesser General Public
36 * License as published by the Free Software Foundation; either
37 * version 2.1 of the License, or (at your option) any later version.
39 * FFmpeg is distributed in the hope that it will be useful,
40 * but WITHOUT ANY WARRANTY; without even the implied warranty of
41 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
42 * Lesser General Public License for more details.
44 * You should have received a copy of the GNU Lesser General Public
45 * License along with FFmpeg; if not, write to the Free Software
46 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
51 * Reference: libavutil/float_dsp.c
55 #include "libavutil/float_dsp.h"
57 #if HAVE_INLINE_ASM && HAVE_MIPSFPU
58 static void vector_fmul_mips(float *dst, const float *src0, const float *src1,
64 for (i = 0; i < len; i++)
65 dst[i] = src0[i] * src1[i];
67 float *d = (float *)dst;
68 float *d_end = d + len;
69 float *s0 = (float *)src0;
70 float *s1 = (float *)src1;
72 float src0_0, src0_1, src0_2, src0_3;
73 float src1_0, src1_1, src1_2, src1_3;
77 "lwc1 %[src0_0], 0(%[s0]) \n\t"
78 "lwc1 %[src1_0], 0(%[s1]) \n\t"
79 "lwc1 %[src0_1], 4(%[s0]) \n\t"
80 "lwc1 %[src1_1], 4(%[s1]) \n\t"
81 "lwc1 %[src0_2], 8(%[s0]) \n\t"
82 "lwc1 %[src1_2], 8(%[s1]) \n\t"
83 "lwc1 %[src0_3], 12(%[s0]) \n\t"
84 "lwc1 %[src1_3], 12(%[s1]) \n\t"
85 "mul.s %[src0_0], %[src0_0], %[src1_0] \n\t"
86 "mul.s %[src0_1], %[src0_1], %[src1_1] \n\t"
87 "mul.s %[src0_2], %[src0_2], %[src1_2] \n\t"
88 "mul.s %[src0_3], %[src0_3], %[src1_3] \n\t"
89 "swc1 %[src0_0], 0(%[d]) \n\t"
90 "swc1 %[src0_1], 4(%[d]) \n\t"
91 "swc1 %[src0_2], 8(%[d]) \n\t"
92 "swc1 %[src0_3], 12(%[d]) \n\t"
93 "addiu %[s0], %[s0], 16 \n\t"
94 "addiu %[s1], %[s1], 16 \n\t"
95 "addiu %[d], %[d], 16 \n\t"
96 "bne %[d], %[d_end], 1b \n\t"
98 : [src0_0]"=&f"(src0_0), [src0_1]"=&f"(src0_1),
99 [src0_2]"=&f"(src0_2), [src0_3]"=&f"(src0_3),
100 [src1_0]"=&f"(src1_0), [src1_1]"=&f"(src1_1),
101 [src1_2]"=&f"(src1_2), [src1_3]"=&f"(src1_3),
102 [d]"+r"(d), [s0]"+r"(s0), [s1]"+r"(s1)
109 static void vector_fmul_window_mips(float *dst, const float *src0,
110 const float *src1, const float *win, int len)
114 * variables used in inline assembler
116 float * dst_i, * dst_j, * dst_i2, * dst_j2;
117 float temp, temp1, temp2, temp3, temp4, temp5, temp6, temp7;
123 for (i = -len, j = len - 1; i < 0; i += 8, j -= 8) {
128 dst_i2 = dst + i + 4;
129 dst_j2 = dst + j - 4;
132 "mul.s %[temp], %[s1], %[wi] \n\t"
133 "mul.s %[temp1], %[s1], %[wj] \n\t"
134 "mul.s %[temp2], %[s11], %[wi1] \n\t"
135 "mul.s %[temp3], %[s11], %[wj1] \n\t"
137 "msub.s %[temp], %[temp], %[s0], %[wj] \n\t"
138 "madd.s %[temp1], %[temp1], %[s0], %[wi] \n\t"
139 "msub.s %[temp2], %[temp2], %[s01], %[wj1] \n\t"
140 "madd.s %[temp3], %[temp3], %[s01], %[wi1] \n\t"
142 "swc1 %[temp], 0(%[dst_i]) \n\t" /* dst[i] = s0*wj - s1*wi; */
143 "swc1 %[temp1], 0(%[dst_j]) \n\t" /* dst[j] = s0*wi + s1*wj; */
144 "swc1 %[temp2], 4(%[dst_i]) \n\t" /* dst[i+1] = s01*wj1 - s11*wi1; */
145 "swc1 %[temp3], -4(%[dst_j]) \n\t" /* dst[j-1] = s01*wi1 + s11*wj1; */
147 "mul.s %[temp4], %[s12], %[wi2] \n\t"
148 "mul.s %[temp5], %[s12], %[wj2] \n\t"
149 "mul.s %[temp6], %[s13], %[wi3] \n\t"
150 "mul.s %[temp7], %[s13], %[wj3] \n\t"
152 "msub.s %[temp4], %[temp4], %[s02], %[wj2] \n\t"
153 "madd.s %[temp5], %[temp5], %[s02], %[wi2] \n\t"
154 "msub.s %[temp6], %[temp6], %[s03], %[wj3] \n\t"
155 "madd.s %[temp7], %[temp7], %[s03], %[wi3] \n\t"
157 "swc1 %[temp4], 8(%[dst_i]) \n\t" /* dst[i+2] = s02*wj2 - s12*wi2; */
158 "swc1 %[temp5], -8(%[dst_j]) \n\t" /* dst[j-2] = s02*wi2 + s12*wj2; */
159 "swc1 %[temp6], 12(%[dst_i]) \n\t" /* dst[i+2] = s03*wj3 - s13*wi3; */
160 "swc1 %[temp7], -12(%[dst_j]) \n\t" /* dst[j-3] = s03*wi3 + s13*wj3; */
161 : [temp]"=&f"(temp), [temp1]"=&f"(temp1), [temp2]"=&f"(temp2),
162 [temp3]"=&f"(temp3), [temp4]"=&f"(temp4), [temp5]"=&f"(temp5),
163 [temp6]"=&f"(temp6), [temp7]"=&f"(temp7)
164 : [dst_j]"r"(dst_j), [dst_i]"r" (dst_i),
165 [s0] "f"(src0[i]), [wj] "f"(win[j]), [s1] "f"(src1[j]),
166 [wi] "f"(win[i]), [s01]"f"(src0[i + 1]),[wj1]"f"(win[j - 1]),
167 [s11]"f"(src1[j - 1]), [wi1]"f"(win[i + 1]), [s02]"f"(src0[i + 2]),
168 [wj2]"f"(win[j - 2]), [s12]"f"(src1[j - 2]),[wi2]"f"(win[i + 2]),
169 [s03]"f"(src0[i + 3]), [wj3]"f"(win[j - 3]), [s13]"f"(src1[j - 3]),
175 "mul.s %[temp], %[s1], %[wi] \n\t"
176 "mul.s %[temp1], %[s1], %[wj] \n\t"
177 "mul.s %[temp2], %[s11], %[wi1] \n\t"
178 "mul.s %[temp3], %[s11], %[wj1] \n\t"
180 "msub.s %[temp], %[temp], %[s0], %[wj] \n\t"
181 "madd.s %[temp1], %[temp1], %[s0], %[wi] \n\t"
182 "msub.s %[temp2], %[temp2], %[s01], %[wj1] \n\t"
183 "madd.s %[temp3], %[temp3], %[s01], %[wi1] \n\t"
185 "swc1 %[temp], 0(%[dst_i2]) \n\t" /* dst[i] = s0*wj - s1*wi; */
186 "swc1 %[temp1], 0(%[dst_j2]) \n\t" /* dst[j] = s0*wi + s1*wj; */
187 "swc1 %[temp2], 4(%[dst_i2]) \n\t" /* dst[i+1] = s01*wj1 - s11*wi1; */
188 "swc1 %[temp3], -4(%[dst_j2]) \n\t" /* dst[j-1] = s01*wi1 + s11*wj1; */
190 "mul.s %[temp4], %[s12], %[wi2] \n\t"
191 "mul.s %[temp5], %[s12], %[wj2] \n\t"
192 "mul.s %[temp6], %[s13], %[wi3] \n\t"
193 "mul.s %[temp7], %[s13], %[wj3] \n\t"
195 "msub.s %[temp4], %[temp4], %[s02], %[wj2] \n\t"
196 "madd.s %[temp5], %[temp5], %[s02], %[wi2] \n\t"
197 "msub.s %[temp6], %[temp6], %[s03], %[wj3] \n\t"
198 "madd.s %[temp7], %[temp7], %[s03], %[wi3] \n\t"
200 "swc1 %[temp4], 8(%[dst_i2]) \n\t" /* dst[i+2] = s02*wj2 - s12*wi2; */
201 "swc1 %[temp5], -8(%[dst_j2]) \n\t" /* dst[j-2] = s02*wi2 + s12*wj2; */
202 "swc1 %[temp6], 12(%[dst_i2]) \n\t" /* dst[i+2] = s03*wj3 - s13*wi3; */
203 "swc1 %[temp7], -12(%[dst_j2]) \n\t" /* dst[j-3] = s03*wi3 + s13*wj3; */
205 [temp1]"=&f"(temp1), [temp2]"=&f"(temp2), [temp3]"=&f"(temp3),
206 [temp4]"=&f"(temp4), [temp5]"=&f"(temp5), [temp6]"=&f"(temp6),
207 [temp7] "=&f" (temp7)
208 : [dst_j2]"r"(dst_j2), [dst_i2]"r"(dst_i2),
209 [s0] "f"(src0[i + 4]), [wj] "f"(win[j - 4]), [s1] "f"(src1[j - 4]),
210 [wi] "f"(win[i + 4]), [s01]"f"(src0[i + 5]),[wj1]"f"(win[j - 5]),
211 [s11]"f"(src1[j - 5]), [wi1]"f"(win[i + 5]), [s02]"f"(src0[i + 6]),
212 [wj2]"f"(win[j - 6]), [s12]"f"(src1[j - 6]),[wi2]"f"(win[i + 6]),
213 [s03]"f"(src0[i + 7]), [wj3]"f"(win[j - 7]), [s13]"f"(src1[j - 7]),
219 #endif /* HAVE_INLINE_ASM && HAVE_MIPSFPU */
221 void ff_float_dsp_init_mips(AVFloatDSPContext *fdsp) {
222 #if HAVE_INLINE_ASM && HAVE_MIPSFPU
223 fdsp->vector_fmul = vector_fmul_mips;
224 fdsp->vector_fmul_window = vector_fmul_window_mips;
225 #endif /* HAVE_INLINE_ASM && HAVE_MIPSFPU */