virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-18 03:19:31 +02:00
Code Issues Releases Activity
cf52e6d012
FFmpeg/libavcodec/mips/compute_antialias_float.h

185 lines
10 KiB
C
Raw Normal View History

Optimization of MP3 decoders for MIPS MP3 fixed and floating point decoders are optimized for MIPS architecture. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-02 15:20:24 +03:00
/*
* 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
mips: port optimizations to mips n64 This mainly consists of replacing all the pointer arithmatic 'addiu' instructions with PTR_ADDIU which will handle the differences in pointer sizes when compiled on 64 bit mips systems. The header asmdefs.h contains the PTR_ macros which expend to the correct mips instructions to manipulate registers containing pointers. Signed-off-by: James Cowgill <james410@cowgill.org.uk> Reviewed-by: Nedeljko Babic <Nedeljko.Babic@imgtec.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2015-03-05 19:40:15 +02:00
#include "libavutil/mips/asmdefs.h"
mips: Add dependencies on HAVE_INLINE_ASM Add dependencies on HAVE_INLINE_ASM for files and parts of code where it is necessary. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-11-09 14:45:40 +03:00
#if HAVE_INLINE_ASM
Optimization of MP3 decoders for MIPS MP3 fixed and floating point decoders are optimized for MIPS architecture. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-02 15:20:24 +03:00
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.
*/
replace all __volatile__ by volatile __volatile__ can cause problems with some compilers and volatile is a standard keyword. Found-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-11 04:50:34 +03:00
__asm__ volatile (
Optimization of MP3 decoders for MIPS MP3 fixed and floating point decoders are optimized for MIPS architecture. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-02 15:20:24 +03:00
"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"
mips: port optimizations to mips n64 This mainly consists of replacing all the pointer arithmatic 'addiu' instructions with PTR_ADDIU which will handle the differences in pointer sizes when compiled on 64 bit mips systems. The header asmdefs.h contains the PTR_ macros which expend to the correct mips instructions to manipulate registers containing pointers. Signed-off-by: James Cowgill <james410@cowgill.org.uk> Reviewed-by: Nedeljko Babic <Nedeljko.Babic@imgtec.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2015-03-05 19:40:15 +02:00
PTR_ADDIU "%[ptr],%[ptr], 72 \t\n"
Optimization of MP3 decoders for MIPS MP3 fixed and floating point decoders are optimized for MIPS architecture. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-02 15:20:24 +03:00
"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)
mips: inline assembly clobber list polishing List of clobbered registers fixed and added where it is lacking. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2013-02-28 17:57:09 +03:00
: "memory"
Optimization of MP3 decoders for MIPS MP3 fixed and floating point decoders are optimized for MIPS architecture. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-02 15:20:24 +03:00
);
}
#define compute_antialias compute_antialias_mips_float
mips: Add dependencies on HAVE_INLINE_ASM Add dependencies on HAVE_INLINE_ASM for files and parts of code where it is necessary. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-11-09 14:45:40 +03:00
#endif /* HAVE_INLINE_ASM */
Optimization of MP3 decoders for MIPS MP3 fixed and floating point decoders are optimized for MIPS architecture. Signed-off-by: Nedeljko Babic <nbabic@mips.com> Reviewed-by: Vitor Sessak <vitor1001@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-07-02 15:20:24 +03:00
#endif /* AVCODEC_MIPS_COMPUTE_ANTIALIAS_FLOAT_H */
Reference in New Issue Copy Permalink