/* * 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) * * MPEG Audio decoder optimized 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/mpegaudiodsp_template.c *            libavcodec/dct32.c */ #include <string.h> #include "config.h"#include "libavutil/mips/asmdefs.h"#include "libavcodec/mpegaudiodsp.h" #if HAVE_INLINE_ASM && HAVE_MIPSFPU#if !HAVE_MIPS32R6 && !HAVE_MIPS64R6 static void ff_mpadsp_apply_window_mips_float(float *synth_buf, float *window,                               int *dither_state, float *samples, ptrdiff_t incr){    register const float *w, *w2, *p;    int j;    float *samples2;    float sum, sum2;    /* temporary variables */    int incr1 = incr << 2;    int t_sample;    float in1, in2, in3, in4, in5, in6, in7, in8;    float *p2;     /* copy to avoid wrap */    memcpy(synth_buf + 512, synth_buf, 32 * sizeof(*synth_buf));     /**    * instructions are scheduled to minimize pipeline stall.    * use of round_sample function from the original code is    * changed with appropriate assembly instructions.    */     __asm__ volatile (        "lwc1    %[sum],      0(%[dither_state])                            \t\n"        "sll     %[t_sample], %[incr1],     5                               \t\n"        "sub     %[t_sample], %[t_sample],  %[incr1]                        \n\t"        "li      %[j],        4                                             \t\n"        "lwc1    %[in1],      0(%[window])                                  \t\n"        "lwc1    %[in2],      16*4(%[synth_buf])                            \t\n"        "sw      $zero,       0(%[dither_state])                            \t\n"        "lwc1    %[in3],      64*4(%[window])                               \t\n"        "lwc1    %[in4],      80*4(%[synth_buf])                            \t\n"        PTR_ADDU "%[samples2],%[samples],   %[t_sample]                     \t\n"        "madd.s  %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in5],      128*4(%[window])                              \t\n"        "lwc1    %[in6],      144*4(%[synth_buf])                           \t\n"        "lwc1    %[in7],      192*4(%[window])                              \t\n"        "madd.s  %[sum],      %[sum],       %[in3], %[in4]                  \t\n"        "lwc1    %[in8],      208*4(%[synth_buf])                           \t\n"        "lwc1    %[in1],      256*4(%[window])                              \t\n"        "lwc1    %[in2],      272*4(%[synth_buf])                           \t\n"        "madd.s  %[sum],      %[sum],       %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      320*4(%[window])                              \t\n"        "lwc1    %[in4],      336*4(%[synth_buf])                           \t\n"        "lwc1    %[in5],      384*4(%[window])                              \t\n"        "madd.s  %[sum],      %[sum],       %[in7], %[in8]                  \t\n"        "lwc1    %[in6],      400*4(%[synth_buf])                           \t\n"        "lwc1    %[in7],      448*4(%[window])                              \t\n"        "lwc1    %[in8],      464*4(%[synth_buf])                           \t\n"        "madd.s  %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in1],      32*4(%[window])                               \t\n"        "lwc1    %[in2],      48*4(%[synth_buf])                            \t\n"        "madd.s  %[sum],      %[sum],       %[in3], %[in4]                  \t\n"        "lwc1    %[in3],      96*4(%[window])                               \t\n"        "lwc1    %[in4],      112*4(%[synth_buf])                           \t\n"        "madd.s  %[sum],      %[sum],       %[in5], %[in6]                  \t\n"        "lwc1    %[in5],      160*4(%[window])                              \t\n"        "lwc1    %[in6],      176*4(%[synth_buf])                           \t\n"        "madd.s  %[sum],      %[sum],       %[in7], %[in8]                  \t\n"        "lwc1    %[in7],      224*4(%[window])                              \t\n"        "lwc1    %[in8],      240*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in1],      288*4(%[window])                              \t\n"        "lwc1    %[in2],      304*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in3], %[in4]                  \t\n"        "lwc1    %[in3],      352*4(%[window])                              \t\n"        "lwc1    %[in4],      368*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in5], %[in6]                  \t\n"        "lwc1    %[in5],      416*4(%[window])                              \t\n"        "lwc1    %[in6],      432*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in7], %[in8]                  \t\n"        "lwc1    %[in7],      480*4(%[window])                              \t\n"        "lwc1    %[in8],      496*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        PTR_ADDU "%[w],       %[window],    4                               \t\n"        "nmsub.s %[sum],      %[sum],       %[in3], %[in4]                  \t\n"        PTR_ADDU "%[w2],      %[window],    124                             \t\n"        PTR_ADDIU "%[p],      %[synth_buf], 68                              \t\n"        PTR_ADDIU "%[p2],     %[synth_buf], 188                             \t\n"        "nmsub.s %[sum],      %[sum],       %[in5], %[in6]                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in7], %[in8]                  \t\n"        "swc1    %[sum],      0(%[samples])                                 \t\n"        PTR_ADDU "%[samples], %[samples],   %[incr1]                        \t\n"         /* calculate two samples at the same time to avoid one memory           access per two sample */         "ff_mpadsp_apply_window_loop%=:                                     \t\n"        "lwc1    %[in1],      0(%[w])                                       \t\n"        "lwc1    %[in2],      0(%[p])                                       \t\n"        "lwc1    %[in3],      0(%[w2])                                      \t\n"        "lwc1    %[in4],      64*4(%[w])                                    \t\n"        "lwc1    %[in5],      64*4(%[p])                                    \t\n"        "lwc1    %[in6],      64*4(%[w2])                                   \t\n"        "mul.s   %[sum],      %[in1],       %[in2]                          \t\n"        "mul.s   %[sum2],     %[in2],       %[in3]                          \t\n"        "lwc1    %[in1],      128*4(%[w])                                   \t\n"        "lwc1    %[in2],      128*4(%[p])                                   \t\n"        "madd.s  %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "nmadd.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      128*4(%[w2])                                  \t\n"        "lwc1    %[in4],      192*4(%[w])                                   \t\n"        "madd.s  %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in5],      192*4(%[p])                                   \t\n"        "lwc1    %[in6],      192*4(%[w2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        "lwc1    %[in1],      256*4(%[w])                                   \t\n"        "lwc1    %[in2],      256*4(%[p])                                   \t\n"        "madd.s  %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      256*4(%[w2])                                  \t\n"        "lwc1    %[in4],      320*4(%[w])                                   \t\n"        "madd.s  %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in5],      320*4(%[p])                                   \t\n"        "lwc1    %[in6],      320*4(%[w2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        "lwc1    %[in1],      384*4(%[w])                                   \t\n"        "lwc1    %[in2],      384*4(%[p])                                   \t\n"        "madd.s  %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      384*4(%[w2])                                  \t\n"        "lwc1    %[in4],      448*4(%[w])                                   \t\n"        "madd.s  %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in5],      448*4(%[p])                                   \t\n"        "lwc1    %[in6],      448*4(%[w2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        "madd.s  %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "lwc1    %[in1],      32*4(%[w])                                    \t\n"        "lwc1    %[in2],      0(%[p2])                                      \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      32*4(%[w2])                                   \t\n"        "lwc1    %[in4],      96*4(%[w])                                    \t\n"        "lwc1    %[in5],      64*4(%[p2])                                   \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in6],      96*4(%[w2])                                   \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        "lwc1    %[in1],      160*4(%[w])                                   \t\n"        "nmsub.s %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "lwc1    %[in2],      128*4(%[p2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      160*4(%[w2])                                  \t\n"        "lwc1    %[in4],      224*4(%[w])                                   \t\n"        "lwc1    %[in5],      192*4(%[p2])                                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in6],      224*4(%[w2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        "lwc1    %[in1],      288*4(%[w])                                   \t\n"        "nmsub.s %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "lwc1    %[in2],      256*4(%[p2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      288*4(%[w2])                                  \t\n"        "lwc1    %[in4],      352*4(%[w])                                   \t\n"        "lwc1    %[in5],      320*4(%[p2])                                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in6],      352*4(%[w2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        "lwc1    %[in1],      416*4(%[w])                                   \t\n"        "nmsub.s %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        "lwc1    %[in2],      384*4(%[p2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "lwc1    %[in3],      416*4(%[w2])                                  \t\n"        "lwc1    %[in4],      480*4(%[w])                                   \t\n"        "lwc1    %[in5],      448*4(%[p2])                                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in6],      480*4(%[w2])                                  \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in2], %[in3]                  \t\n"        PTR_ADDIU "%[w],      %[w],         4                               \t\n"        "nmsub.s %[sum],      %[sum],       %[in4], %[in5]                  \t\n"        PTR_ADDIU "%[w2],     %[w2],        -4                              \t\n"        "nmsub.s %[sum2],     %[sum2],      %[in5], %[in6]                  \t\n"        "addu    %[j],        %[j],         4                               \t\n"        PTR_ADDIU "%[p],      4                                             \t\n"        "swc1    %[sum],      0(%[samples])                                 \t\n"        PTR_ADDIU "%[p2],     -4                                            \t\n"        "swc1    %[sum2],     0(%[samples2])                                \t\n"        PTR_ADDU "%[samples], %[samples],   %[incr1]                        \t\n"        PTR_SUBU "%[samples2],%[samples2],  %[incr1]                        \t\n"        "bne     %[j],        64,           ff_mpadsp_apply_window_loop%=   \t\n"         "lwc1    %[in1],      48*4(%[window])                               \t\n"        "lwc1    %[in2],      32*4(%[synth_buf])                            \t\n"        "lwc1    %[in3],      112*4(%[window])                              \t\n"        "lwc1    %[in4],      96*4(%[synth_buf])                            \t\n"        "lwc1    %[in5],      176*4(%[window])                              \t\n"        "lwc1    %[in6],      160*4(%[synth_buf])                           \t\n"        "mul.s   %[sum],      %[in1],       %[in2]                          \t\n"        "lwc1    %[in7],      240*4(%[window])                              \t\n"        "lwc1    %[in8],      224*4(%[synth_buf])                           \t\n"        "lwc1    %[in1],      304*4(%[window])                              \t\n"        "nmadd.s %[sum],      %[sum],       %[in3], %[in4]                  \t\n"        "lwc1    %[in2],      288*4(%[synth_buf])                           \t\n"        "lwc1    %[in3],      368*4(%[window])                              \t\n"        "lwc1    %[in4],      352*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in5], %[in6]                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in7], %[in8]                  \t\n"        "lwc1    %[in5],      432*4(%[window])                              \t\n"        "lwc1    %[in6],      416*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in1], %[in2]                  \t\n"        "lwc1    %[in7],      496*4(%[window])                              \t\n"        "lwc1    %[in8],      480*4(%[synth_buf])                           \t\n"        "nmsub.s %[sum],      %[sum],       %[in3], %[in4]                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in5], %[in6]                  \t\n"        "nmsub.s %[sum],      %[sum],       %[in7], %[in8]                  \t\n"        "swc1    %[sum],      0(%[samples])                                 \t\n"         : [sum] "=&f" (sum), [sum2] "=&f" (sum2),          [w2] "=&r" (w2),   [w] "=&r" (w),          [p] "=&r" (p), [p2] "=&r" (p2), [j] "=&r" (j),          [samples] "+r" (samples), [samples2] "=&r" (samples2),          [in1] "=&f" (in1), [in2] "=&f" (in2),          [in3] "=&f" (in3), [in4] "=&f" (in4),          [in5] "=&f" (in5), [in6] "=&f" (in6),          [in7] "=&f" (in7), [in8] "=&f" (in8),          [t_sample] "=&r" (t_sample)        : [synth_buf] "r" (synth_buf), [window] "r" (window),          [dither_state] "r" (dither_state), [incr1] "r" (incr1)        : "memory"    );} static void ff_dct32_mips_float(float *out, const float *tab){    float val0 , val1 , val2 , val3 , val4 , val5 , val6 , val7,          val8 , val9 , val10, val11, val12, val13, val14, val15,          val16, val17, val18, val19, val20, val21, val22, val23,          val24, val25, val26, val27, val28, val29, val30, val31;    float fTmp1, fTmp2, fTmp3, fTmp4, fTmp5, fTmp6, fTmp8, fTmp9;    float f1, f2, f3, f4, f5, f6, f7;     f1 = 0.50241928618815570551;    f2 = 0.50060299823519630134;    f3 = 10.19000812354805681150;    f4 = 5.10114861868916385802;    f5 = 0.67480834145500574602;    f6 = 0.74453627100229844977;    f7 = 0.50979557910415916894;    /**    * instructions are scheduled to minimize pipeline stall.    */    __asm__ volatile (        "lwc1       %[fTmp1],       0*4(%[tab])                             \n\t"        "lwc1       %[fTmp2],       31*4(%[tab])                            \n\t"        "lwc1       %[fTmp3],       15*4(%[tab])                            \n\t"        "lwc1       %[fTmp4],       16*4(%[tab])                            \n\t"        "add.s      %[fTmp5],       %[fTmp1],       %[fTmp2]                \n\t"        "sub.s      %[fTmp8],       %[fTmp1],       %[fTmp2]                \n\t"        "add.s      %[fTmp6],       %[fTmp3],       %[fTmp4]                \n\t"        "sub.s      %[fTmp9],       %[fTmp3],       %[fTmp4]                \n\t"        "mul.s      %[fTmp8],       %[fTmp8],       %[f2]                   \n\t"        "add.s      %[val0],        %[fTmp5],       %[fTmp6]                \n\t"        "sub.s      %[val15],       %[fTmp5],       %[fTmp6]                \n\t"        "lwc1       %[fTmp1],       7*4(%[tab])                             \n\t"        "lwc1       %[fTmp2],       24*4(%[tab])                            \n\t"        "madd.s     %[val16],       %[fTmp8],       %[fTmp9],   %[f3]       \n\t"        "nmsub.s    %[val31],       %[fTmp8],       %[fTmp9],   %[f3]       \n\t"        "mul.s      %[val15],       %[val15],       %[f1]                   \n\t"        "lwc1       %[fTmp3],       8*4(%[tab])                             \n\t"        "lwc1       %[fTmp4],       23*4(%[tab])                            \n\t"        "add.s      %[fTmp5],       %[fTmp1],       %[fTmp2]                \n\t"        "mul.s      %[val31],       %[val31],       %[f1]                   \n\t"        "sub.s      %[fTmp8],       %[fTmp1],       %[fTmp2]                \n\t"        "add.s      %[fTmp6],       %[fTmp3],       %[fTmp4]                \n\t"        "sub.s      %[fTmp9],       %[fTmp3],       %[fTmp4]                \n\t"        "add.s      %[val7],        %[fTmp5],       %[fTmp6]                \n\t"        "sub.s      %[val8],        %[fTmp5],       %[fTmp6]                \n\t"        "mul.s      %[fTmp8],       %[fTmp8],       %[f5]                   \n\t"        "sub.s      %[fTmp2],       %[val0],        %[val7]                 \n\t"        "mul.s      %[val8],        %[val8],        %[f4]                   \n\t"        "madd.s     %[val23],       %[fTmp8],       %[fTmp9],   %[f6]       \n\t"        "nmsub.s    %[val24],       %[fTmp8],       %[fTmp9],   %[f6]       \n\t"        "add.s      %[val0],        %[val0],        %[val7]                 \n\t"        "mul.s      %[val7],        %[f7],          %[fTmp2]                \n\t"        "sub.s      %[fTmp2],       %[val15],       %[val8]                 \n\t"        "add.s      %[val8],        %[val15],       %[val8]                 \n\t"        "mul.s      %[val24],       %[val24],       %[f4]                   \n\t"        "sub.s      %[fTmp3],       %[val16],       %[val23]                \n\t"        "add.s      %[val16],       %[val16],       %[val23]                \n\t"        "mul.s      %[val15],       %[f7],          %[fTmp2]                \n\t"        "sub.s      %[fTmp4],       %[val31],       %[val24]                \n\t"        "mul.s      %[val23],       %[f7],          %[fTmp3]                \n\t"        "add.s      %[val24],       %[val31],       %[val24]                \n\t"        "mul.s      %[val31],       %[f7],          %[fTmp4]                \n\t"         : [fTmp1]  "=&f" (fTmp1),  [fTmp2] "=&f" (fTmp2), [fTmp3] "=&f" (fTmp3),          [fTmp4]  "=&f" (fTmp4),  [fTmp5] "=&f" (fTmp5), [fTmp6] "=&f" (fTmp6),          [fTmp8] "=&f" (fTmp8), [fTmp9] "=&f" (fTmp9),          [val0]  "=&f" (val0),  [val7]  "=&f" (val7),          [val8]  "=&f" (val8),  [val15] "=&f" (val15),          [val16] "=&f" (val16), [val23] "=&f" (val23),          [val24] "=&f" (val24), [val31] "=&f" (val31)        : [tab] "r" (tab), [f1]"f"(f1), [f2]"f"(f2), [f3]"f"(f3),          [f4]"f"(f4), [f5]"f"(f5), [f6]"f"(f6), [f7]"f"(f7)        : "memory"    );     f1 = 0.64682178335999012954;    f2 = 0.53104259108978417447;    f3 = 1.48416461631416627724;    f4 = 0.78815462345125022473;    f5 = 0.55310389603444452782;    f6 = 1.16943993343288495515;    f7 = 2.56291544774150617881;    __asm__ volatile (        "lwc1       %[fTmp1],       3*4(%[tab])                             \n\t"        "lwc1       %[fTmp2],       28*4(%[tab])                            \n\t"        "lwc1       %[fTmp3],       12*4(%[tab])                            \n\t"        "lwc1       %[fTmp4],       19*4(%[tab])                            \n\t"        "add.s      %[fTmp5],       %[fTmp1],       %[fTmp2]                \n\t"        "sub.s      %[fTmp8],       %[fTmp1],       %[fTmp2]                \n\t"        "add.s      %[fTmp6],       %[fTmp3],       %[fTmp4]                \n\t"        "sub.s      %[fTmp9],       %[fTmp3],       %[fTmp4]                \n\t"        "mul.s      %[fTmp8],       %[fTmp8],       %[f2]                   \n\t"        "add.s      %[val3],        %[fTmp5],       %[fTmp6]                \n\t"        "sub.s      %[val12],       %[fTmp5],       %[fTmp6]                \n\t"        "lwc1       %[fTmp1],       4*4(%[tab])                             \n\t"        "lwc1       %[fTmp2],       27*4(%[tab])                            \n\t"        "madd.s     %[val19],       %[fTmp8],       %[fTmp9],   %[f3]       \n\t"        "nmsub.s    %[val28],       %[fTmp8],       %[fTmp9],   %[f3]       \n\t"        "mul.s      %[val12],       %[val12],       %[f1]                   \n\t"        "lwc1       %[fTmp3],       11*4(%[tab])                            \n\t"        "lwc1       %[fTmp4],       20*4(%[tab])                            \n\t"        "add.s      %[fTmp5],       %[fTmp1],       %[fTmp2]                \n\t"        "mul.s      %[val28],       %[val28],       %[f1]                   \n\t"        "sub.s      %[fTmp8],       %[fTmp1],       %[fTmp2]                \n\t"        "add.s      %[fTmp6],       %[fTmp3],       %[fTmp4]                \n\t"        "sub.s      %[fTmp9],       %[fTmp3],       %[fTmp4]                \n\t"        "mul.s      %[fTmp8],       %[fTmp8],       %[f5]                   \n\t"        "add.s      %[val4],        %[fTmp5],       %[fTmp6]                \n\t"        "sub.s      %[val11],       %[fTmp5],       %[fTmp6]                \n\t"        "madd.s     %[val20],       %[fTmp8],       %[fTmp9],   %[f6]       \n\t"        "nmsub.s    %[val27],       %[fTmp8],       %[fTmp9],   %[f6]       \n\t"        "mul.s      %[val11],       %[val11],       %[f4]                   \n\t"        "sub.s      %[fTmp2],       %[val3],        %[val4]                 \n\t"        "add.s      %[val3],        %[val3],        %[val4]                 \n\t"        "sub.s      %[fTmp4],       %[val19],       %[val20]                \n\t"        "mul.s      %[val27],       %[val27],       %[f4]                   \n\t"        "sub.s      %[fTmp3],       %[val12],       %[val11]                \n\t"        "mul.s      %[val4],        %[f7],          %[fTmp2]                \n\t"        "add.s      %[val11],       %[val12],       %[val11]                \n\t"        "add.s      %[val19],       %[val19],       %[val20]                \n\t"        "mul.s      %[val20],       %[f7],          %[fTmp4]                \n\t"        "mul.s      %[val12],       %[f7],          %[fTmp3]                \n\t"        "sub.s      %[fTmp2],       %[val28],       %[val27]                \n\t"        "add.s      %[val27],       %[val28],       %[val27]                \n\t"        "mul.s      %[val28],       %[f7],          %[fTmp2]                \n\t"         : [fTmp1]  "=&f" (fTmp1),  [fTmp2]  "=&f" (fTmp2), [fTmp3] "=&f" (fTmp3),          [fTmp4]  "=&f" (fTmp4),  [fTmp5]  "=&f" (fTmp5), [fTmp6] "=&f" (fTmp6),          [fTmp8]  "=&f" (fTmp8), [fTmp9] "=&f" (fTmp9),          [val3]  "=&f" (val3),  [val4]  "=&f" (val4),          [val11] "=&f" (val11), [val12] "=&f" (val12),          [val19] "=&f" (val19), [val20] "=&f" (val20),          [val27] "=&f" (val27), [val28] "=&f" (val28)        : [tab] "r" (tab), [f1]"f"(f1), [f2]"f"(f2), [f3]"f"(f3),          [f4]"f"(f4), [f5]"f"(f5), [f6]"f"(f6), [f7]"f"(f7)        : "memory"    );     f1 = 0.54119610014619698439;    __asm__ volatile (        "sub.s      %[fTmp2],       %[val0],        %[val3]                 \n\t"        "add.s      %[val0],        %[val0],        %[val3]                 \n\t"        "sub.s      %[fTmp3],       %[val7],        %[val4]                 \n\t"        "add.s      %[val4],        %[val7],        %[val4]                 \n\t"        "sub.s      %[fTmp4],       %[val8],        %[val11]                \n\t"        "mul.s      %[val3],        %[f1],          %[fTmp2]                \n\t"        "add.s      %[val8],        %[val8],        %[val11]                \n\t"        "mul.s      %[val7],        %[f1],          %[fTmp3]                \n\t"        "sub.s      %[fTmp2],       %[val15],       %[val12]                \n\t"        "mul.s      %[val11],       %[f1],          %[fTmp4]                \n\t"        "add.s      %[val12],       %[val15],       %[val12]                \n\t"        "mul.s      %[val15],       %[f1],          %[fTmp2]                \n\t"         : [val0]  "+&f" (val0),   [val3] "+&f" (val3),          [val4]  "+&f" (val4),   [val7] "+&f" (val7),          [val8]  "+&f" (val8),   [val11] "+&f" (val11),          [val12] "+&f" (val12),  [val15] "+&f" (val15),          [fTmp2] "=&f" (fTmp2),          [fTmp3] "=&f" (fTmp3), [fTmp4] "=&f" (fTmp4)        : [f1] "f" (f1)    );     __asm__ volatile (        "sub.s      %[fTmp2],       %[val16],       %[val19]                \n\t"        "add.s      %[val16],       %[val16],       %[val19]                \n\t"        "sub.s      %[fTmp3],       %[val23],       %[val20]                \n\t"        "add.s      %[val20],       %[val23],       %[val20]                \n\t"        "sub.s      %[fTmp4],       %[val24],       %[val27]                \n\t"        "mul.s      %[val19],       %[f1],          %[fTmp2]                \n\t"        "add.s      %[val24],       %[val24],       %[val27]                \n\t"        "mul.s      %[val23],       %[f1],          %[fTmp3]                \n\t"        "sub.s      %[fTmp2],       %[val31],       %[val28]                \n\t"        "mul.s      %[val27],       %[f1],          %[fTmp4]                \n\t"        "add.s      %[val28],       %[val31],       %[val28]                \n\t"        "mul.s      %[val31],       %[f1],          %[fTmp2]                \n\t"         : [fTmp2] "=&f" (fTmp2), [fTmp3] "=&f" (fTmp3), [fTmp4] "=&f" (fTmp4),          [val16] "+&f" (val16), [val19] "+&f" (val19), [val20] "+&f" (val20),          [val23] "+&f" (val23), [val24] "+&f" (val24), [val27] "+&f" (val27),          [val28] "+&f" (val28), [val31] "+&f" (val31)        : [f1] "f" (f1)    );     f1 = 0.52249861493968888062;    f2 = 0.50547095989754365998;    f3 = 3.40760841846871878570;    f4 = 1.72244709823833392782;    f5 = 0.62250412303566481615;    f6 = 0.83934964541552703873;    f7 = 0.60134488693504528054;    __asm__ volatile (        "lwc1       %[fTmp1],       1*4(%[tab])                             \n\t"        "lwc1       %[fTmp2],       30*4(%[tab])                            \n\t"        "lwc1       %[fTmp3],       14*4(%[tab])                            \n\t"        "lwc1       %[fTmp4],       17*4(%[tab])                            \n\t"        "add.s      %[fTmp5],       %[fTmp1],       %[fTmp2]                \n\t"        "sub.s      %[fTmp8],       %[fTmp1],       %[fTmp2]                \n\t"        "add.s      %[fTmp6],       %[fTmp3],       %[fTmp4]                \n\t"        "sub.s      %[fTmp9],       %[fTmp3],       %[fTmp4]                \n\t"        "mul.s      %[fTmp8],       %[fTmp8],       %[f2]                   \n\t"        "add.s      %[val1],        %[fTmp5],       %[fTmp6]                \n\t"        "sub.s      %[val14],       %[fTmp5],       %[fTmp6]                \n\t"        "lwc1       %[fTmp1],       6*4(%[tab])                             \n\t"        "lwc1       %[fTmp2],       25*4(%[tab])                            \n\t"        "madd.s     %[val17],       %[fTmp8],       %[fTmp9],   %[f3]       \n\t"        "nmsub.s    %[val30],       %[fTmp8],       %[fTmp9],   %[f3]       \n\t"        "mul.s      %[val14],       %[val14],       %[f1]                   \n\t"        "lwc1       %[fTmp3],       9*4(%[tab])                             \n\t"        "lwc1       %[fTmp4],       22*4(%[tab])                            \n\t"        "add.s      %[fTmp5],       %[fTmp1],       %[fTmp2]                \n\t"        "mul.s      %[val30],       %[val30],       %[f1]                   \n\t"        "sub.s      %[fTmp8],       %[fTmp1],       %[fTmp2]                \n\t"        "add.s      %[fTmp6],       %[fTmp3],       %[fTmp4]                \n\t"        "sub.s      %[fTmp9],       %[fTmp3],       %[fTmp4]                \n\t"        "add.s      %[val6],        %[fTmp5],       %[fTmp6]                \n\t"        "sub.s      %[val9],        %[fTmp5],       %[fTmp6]                \n\t"

Lines 1–500 of 1258. Load more