[deb_ffmpeg.git] / ffmpeg / libavcodec / mips / compute_antialias_float.h

/*
 * Copyright (c) 2012
 *      MIPS Technologies, Inc., California.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the MIPS Technologies, Inc., nor the names of its
 *    contributors may be used to endorse or promote products derived from
 *    this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE MIPS TECHNOLOGIES, INC. ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE MIPS TECHNOLOGIES, INC. BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 * Author:  Bojan Zivkovic (bojan@mips.com)
 *
 * Compute antialias function optimised for MIPS floating-point architecture
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * Reference: libavcodec/mpegaudiodec.c
 */

#ifndef AVCODEC_MIPS_COMPUTE_ANTIALIAS_FLOAT_H
#define AVCODEC_MIPS_COMPUTE_ANTIALIAS_FLOAT_H

#if HAVE_INLINE_ASM
static void compute_antialias_mips_float(MPADecodeContext *s,
                                        GranuleDef *g)
{
    float *ptr, *ptr_end;
    float *csa = &csa_table[0][0];
    /* temporary variables */
    float in1, in2, in3, in4, in5, in6, in7, in8;
    float out1, out2, out3, out4;

    ptr = g->sb_hybrid + 18;
    /* we antialias only "long" bands */
    if (g->block_type == 2) {
        if (!g->switch_point)
            return;
        /* XXX: check this for 8000Hz case */
        ptr_end = ptr + 18;
    } else {
        ptr_end = ptr + 558;
    }

    /**
    * instructions are scheduled to minimize pipeline stall.
    */

    __asm__ volatile (
        "compute_antialias_float_loop%=:                                \t\n"
        "lwc1    %[in1],  -1*4(%[ptr])                                  \t\n"
        "lwc1    %[in2],  0(%[csa])                                     \t\n"
        "lwc1    %[in3],  1*4(%[csa])                                   \t\n"
        "lwc1    %[in4],  0(%[ptr])                                     \t\n"
        "lwc1    %[in5],  -2*4(%[ptr])                                  \t\n"
        "lwc1    %[in6],  4*4(%[csa])                                   \t\n"
        "mul.s   %[out1], %[in1],  %[in2]                               \t\n"
        "mul.s   %[out2], %[in1],  %[in3]                               \t\n"
        "lwc1    %[in7],  5*4(%[csa])                                   \t\n"
        "lwc1    %[in8],  1*4(%[ptr])                                   \t\n"
        "nmsub.s %[out1], %[out1], %[in3], %[in4]                       \t\n"
        "madd.s  %[out2], %[out2], %[in2], %[in4]                       \t\n"
        "mul.s   %[out3], %[in5],  %[in6]                               \t\n"
        "mul.s   %[out4], %[in5],  %[in7]                               \t\n"
        "lwc1    %[in1],  -3*4(%[ptr])                                  \t\n"
        "swc1    %[out1], -1*4(%[ptr])                                  \t\n"
        "swc1    %[out2], 0(%[ptr])                                     \t\n"
        "nmsub.s %[out3], %[out3], %[in7], %[in8]                       \t\n"
        "madd.s  %[out4], %[out4], %[in6], %[in8]                       \t\n"
        "lwc1    %[in2],  8*4(%[csa])                                   \t\n"
        "swc1    %[out3], -2*4(%[ptr])                                  \t\n"
        "swc1    %[out4], 1*4(%[ptr])                                   \t\n"
        "lwc1    %[in3],  9*4(%[csa])                                   \t\n"
        "lwc1    %[in4],  2*4(%[ptr])                                   \t\n"
        "mul.s   %[out1], %[in1],  %[in2]                               \t\n"
        "lwc1    %[in5],  -4*4(%[ptr])                                  \t\n"
        "lwc1    %[in6],  12*4(%[csa])                                  \t\n"
        "mul.s   %[out2], %[in1],  %[in3]                               \t\n"
        "lwc1    %[in7],  13*4(%[csa])                                  \t\n"
        "nmsub.s %[out1], %[out1], %[in3], %[in4]                       \t\n"
        "lwc1    %[in8],  3*4(%[ptr])                                   \t\n"
        "mul.s   %[out3], %[in5],  %[in6]                               \t\n"
        "madd.s  %[out2], %[out2], %[in2], %[in4]                       \t\n"
        "mul.s   %[out4], %[in5],  %[in7]                               \t\n"
        "swc1    %[out1], -3*4(%[ptr])                                  \t\n"
        "lwc1    %[in1],  -5*4(%[ptr])                                  \t\n"
        "nmsub.s %[out3], %[out3], %[in7], %[in8]                       \t\n"
        "swc1    %[out2], 2*4(%[ptr])                                   \t\n"
        "madd.s  %[out4], %[out4], %[in6], %[in8]                       \t\n"
        "lwc1    %[in2],  16*4(%[csa])                                  \t\n"
        "lwc1    %[in3],  17*4(%[csa])                                  \t\n"
        "swc1    %[out3], -4*4(%[ptr])                                  \t\n"
        "lwc1    %[in4],  4*4(%[ptr])                                   \t\n"
        "swc1    %[out4], 3*4(%[ptr])                                   \t\n"
        "mul.s   %[out1], %[in1],  %[in2]                               \t\n"
        "mul.s   %[out2], %[in1],  %[in3]                               \t\n"
        "lwc1    %[in5],  -6*4(%[ptr])                                  \t\n"
        "lwc1    %[in6],  20*4(%[csa])                                  \t\n"
        "lwc1    %[in7],  21*4(%[csa])                                  \t\n"
        "nmsub.s %[out1], %[out1], %[in3], %[in4]                       \t\n"
        "madd.s  %[out2], %[out2], %[in2], %[in4]                       \t\n"
        "lwc1    %[in8],  5*4(%[ptr])                                   \t\n"
        "mul.s   %[out3], %[in5],  %[in6]                               \t\n"
        "mul.s   %[out4], %[in5],  %[in7]                               \t\n"
        "swc1    %[out1], -5*4(%[ptr])                                  \t\n"
        "swc1    %[out2], 4*4(%[ptr])                                   \t\n"
        "lwc1    %[in1],  -7*4(%[ptr])                                  \t\n"
        "nmsub.s %[out3], %[out3], %[in7], %[in8]                       \t\n"
        "madd.s  %[out4], %[out4], %[in6], %[in8]                       \t\n"
        "lwc1    %[in2],  24*4(%[csa])                                  \t\n"
        "lwc1    %[in3],  25*4(%[csa])                                  \t\n"
        "lwc1    %[in4],  6*4(%[ptr])                                   \t\n"
        "swc1    %[out3], -6*4(%[ptr])                                  \t\n"
        "swc1    %[out4], 5*4(%[ptr])                                   \t\n"
        "mul.s   %[out1], %[in1],  %[in2]                               \t\n"
        "lwc1    %[in5],  -8*4(%[ptr])                                  \t\n"
        "mul.s   %[out2], %[in1],  %[in3]                               \t\n"
        "lwc1    %[in6],  28*4(%[csa])                                  \t\n"
        "lwc1    %[in7],  29*4(%[csa])                                  \t\n"
        "nmsub.s %[out1], %[out1], %[in3], %[in4]                       \t\n"
        "lwc1    %[in8],  7*4(%[ptr])                                   \t\n"
        "madd.s  %[out2], %[out2], %[in2], %[in4]                       \t\n"
        "mul.s   %[out3], %[in5],  %[in6]                               \t\n"
        "mul.s   %[out4], %[in5],  %[in7]                               \t\n"
        "swc1    %[out1], -7*4(%[ptr])                                  \t\n"
        "swc1    %[out2], 6*4(%[ptr])                                   \t\n"
        "addiu   %[ptr],  %[ptr],  72                                   \t\n"
        "nmsub.s %[out3], %[out3], %[in7], %[in8]                       \t\n"
        "madd.s  %[out4], %[out4], %[in6], %[in8]                       \t\n"
        "swc1    %[out3], -26*4(%[ptr])                                 \t\n"
        "swc1    %[out4], -11*4(%[ptr])                                 \t\n"
        "bne     %[ptr],  %[ptr_end],  compute_antialias_float_loop%=   \t\n"

        : [ptr] "+r" (ptr),
          [in1] "=&f" (in1), [in2] "=&f" (in2),
          [in3] "=&f" (in3), [in4] "=&f" (in4),
          [in5] "=&f" (in5), [in6] "=&f" (in6),
          [in7] "=&f" (in7), [in8] "=&f" (in8),
          [out1] "=&f" (out1), [out2] "=&f" (out2),
          [out3] "=&f" (out3), [out4] "=&f" (out4)
        : [csa] "r" (csa), [ptr_end] "r" (ptr_end)
        : "memory"
    );
}
#define compute_antialias compute_antialias_mips_float
#endif /* HAVE_INLINE_ASM */

#endif /* AVCODEC_MIPS_COMPUTE_ANTIALIAS_FLOAT_H */
Commit	Line	Data
2ba45a60 DM	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 floating-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_FLOAT_H
	56	#define AVCODEC_MIPS_COMPUTE_ANTIALIAS_FLOAT_H
	57
	58	#if HAVE_INLINE_ASM
	59	static void compute_antialias_mips_float(MPADecodeContext *s,
	60	GranuleDef *g)
	61	{
	62	float ptr, ptr_end;
	63	float *csa = &csa_table[0][0];
	64	/* temporary variables */
65	float in1, in2, in3, in4, in5, in6, in7, in8;
66	float out1, out2, out3, out4;
67
68	ptr = g->sb_hybrid + 18;
69	/* we antialias only "long" bands */
70	if (g->block_type == 2) {
71	if (!g->switch_point)
72	return;
73	/* XXX: check this for 8000Hz case */
74	ptr_end = ptr + 18;
75	} else {
76	ptr_end = ptr + 558;
77	}
78
79	/**
80	* instructions are scheduled to minimize pipeline stall.
81	*/
82
83	__asm__ volatile (
84	"compute_antialias_float_loop%=: \t\n"
85	"lwc1 %[in1], -1*4(%[ptr]) \t\n"
86	"lwc1 %[in2], 0(%[csa]) \t\n"
87	"lwc1 %[in3], 1*4(%[csa]) \t\n"
88	"lwc1 %[in4], 0(%[ptr]) \t\n"
89	"lwc1 %[in5], -2*4(%[ptr]) \t\n"
90	"lwc1 %[in6], 4*4(%[csa]) \t\n"
91	"mul.s %[out1], %[in1], %[in2] \t\n"
92	"mul.s %[out2], %[in1], %[in3] \t\n"
93	"lwc1 %[in7], 5*4(%[csa]) \t\n"
94	"lwc1 %[in8], 1*4(%[ptr]) \t\n"
95	"nmsub.s %[out1], %[out1], %[in3], %[in4] \t\n"
96	"madd.s %[out2], %[out2], %[in2], %[in4] \t\n"
97	"mul.s %[out3], %[in5], %[in6] \t\n"
98	"mul.s %[out4], %[in5], %[in7] \t\n"
99	"lwc1 %[in1], -3*4(%[ptr]) \t\n"
100	"swc1 %[out1], -1*4(%[ptr]) \t\n"
101	"swc1 %[out2], 0(%[ptr]) \t\n"
102	"nmsub.s %[out3], %[out3], %[in7], %[in8] \t\n"
103	"madd.s %[out4], %[out4], %[in6], %[in8] \t\n"
104	"lwc1 %[in2], 8*4(%[csa]) \t\n"
105	"swc1 %[out3], -2*4(%[ptr]) \t\n"
106	"swc1 %[out4], 1*4(%[ptr]) \t\n"
107	"lwc1 %[in3], 9*4(%[csa]) \t\n"
108	"lwc1 %[in4], 2*4(%[ptr]) \t\n"
109	"mul.s %[out1], %[in1], %[in2] \t\n"
110	"lwc1 %[in5], -4*4(%[ptr]) \t\n"
111	"lwc1 %[in6], 12*4(%[csa]) \t\n"
112	"mul.s %[out2], %[in1], %[in3] \t\n"
113	"lwc1 %[in7], 13*4(%[csa]) \t\n"
114	"nmsub.s %[out1], %[out1], %[in3], %[in4] \t\n"
115	"lwc1 %[in8], 3*4(%[ptr]) \t\n"
116	"mul.s %[out3], %[in5], %[in6] \t\n"
117	"madd.s %[out2], %[out2], %[in2], %[in4] \t\n"
118	"mul.s %[out4], %[in5], %[in7] \t\n"
119	"swc1 %[out1], -3*4(%[ptr]) \t\n"
120	"lwc1 %[in1], -5*4(%[ptr]) \t\n"
121	"nmsub.s %[out3], %[out3], %[in7], %[in8] \t\n"
122	"swc1 %[out2], 2*4(%[ptr]) \t\n"
123	"madd.s %[out4], %[out4], %[in6], %[in8] \t\n"
124	"lwc1 %[in2], 16*4(%[csa]) \t\n"
125	"lwc1 %[in3], 17*4(%[csa]) \t\n"
126	"swc1 %[out3], -4*4(%[ptr]) \t\n"
127	"lwc1 %[in4], 4*4(%[ptr]) \t\n"
128	"swc1 %[out4], 3*4(%[ptr]) \t\n"
129	"mul.s %[out1], %[in1], %[in2] \t\n"
130	"mul.s %[out2], %[in1], %[in3] \t\n"
131	"lwc1 %[in5], -6*4(%[ptr]) \t\n"
132	"lwc1 %[in6], 20*4(%[csa]) \t\n"
133	"lwc1 %[in7], 21*4(%[csa]) \t\n"
134	"nmsub.s %[out1], %[out1], %[in3], %[in4] \t\n"
135	"madd.s %[out2], %[out2], %[in2], %[in4] \t\n"
136	"lwc1 %[in8], 5*4(%[ptr]) \t\n"
137	"mul.s %[out3], %[in5], %[in6] \t\n"
138	"mul.s %[out4], %[in5], %[in7] \t\n"
139	"swc1 %[out1], -5*4(%[ptr]) \t\n"
140	"swc1 %[out2], 4*4(%[ptr]) \t\n"
141	"lwc1 %[in1], -7*4(%[ptr]) \t\n"
142	"nmsub.s %[out3], %[out3], %[in7], %[in8] \t\n"
143	"madd.s %[out4], %[out4], %[in6], %[in8] \t\n"
144	"lwc1 %[in2], 24*4(%[csa]) \t\n"
145	"lwc1 %[in3], 25*4(%[csa]) \t\n"
146	"lwc1 %[in4], 6*4(%[ptr]) \t\n"
147	"swc1 %[out3], -6*4(%[ptr]) \t\n"
148	"swc1 %[out4], 5*4(%[ptr]) \t\n"
149	"mul.s %[out1], %[in1], %[in2] \t\n"
150	"lwc1 %[in5], -8*4(%[ptr]) \t\n"
151	"mul.s %[out2], %[in1], %[in3] \t\n"
152	"lwc1 %[in6], 28*4(%[csa]) \t\n"
153	"lwc1 %[in7], 29*4(%[csa]) \t\n"
154	"nmsub.s %[out1], %[out1], %[in3], %[in4] \t\n"
155	"lwc1 %[in8], 7*4(%[ptr]) \t\n"
156	"madd.s %[out2], %[out2], %[in2], %[in4] \t\n"
157	"mul.s %[out3], %[in5], %[in6] \t\n"
158	"mul.s %[out4], %[in5], %[in7] \t\n"
159	"swc1 %[out1], -7*4(%[ptr]) \t\n"
160	"swc1 %[out2], 6*4(%[ptr]) \t\n"
161	"addiu %[ptr], %[ptr], 72 \t\n"
162	"nmsub.s %[out3], %[out3], %[in7], %[in8] \t\n"
163	"madd.s %[out4], %[out4], %[in6], %[in8] \t\n"
164	"swc1 %[out3], -26*4(%[ptr]) \t\n"
165	"swc1 %[out4], -11*4(%[ptr]) \t\n"
166	"bne %[ptr], %[ptr_end], compute_antialias_float_loop%= \t\n"
167
168	: [ptr] "+r" (ptr),
169	[in1] "=&f" (in1), [in2] "=&f" (in2),
170	[in3] "=&f" (in3), [in4] "=&f" (in4),
171	[in5] "=&f" (in5), [in6] "=&f" (in6),
172	[in7] "=&f" (in7), [in8] "=&f" (in8),
173	[out1] "=&f" (out1), [out2] "=&f" (out2),
174	[out3] "=&f" (out3), [out4] "=&f" (out4)
175	: [csa] "r" (csa), [ptr_end] "r" (ptr_end)
176	: "memory"
177	);
178	}
179	#define compute_antialias compute_antialias_mips_float
180	#endif /* HAVE_INLINE_ASM */
181
182	#endif /* AVCODEC_MIPS_COMPUTE_ANTIALIAS_FLOAT_H */