ref: ee0efbb921061aa956ef6df8f1415f1fd603b808
dir: /codec/encoder/core/mips/score_mmi.c/
/*!
* \copy
* Copyright (c) 2009-2018, Cisco Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* * 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.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "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
* COPYRIGHT HOLDER OR CONTRIBUTORS 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.
*
*
* \file score_mmi.c
*
* \brief Loongson optimization
*
* \date 21/07/2018 Created
*
*************************************************************************************
*/
#include <stdint.h>
#include "asmdefs_mmi.h"
unsigned char nozero_count_table[] __attribute__((aligned(16))) = {
0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4, 1, 2, 2, 3, 2, 3, 3, 4,
2, 3, 3, 4, 3, 4, 4, 5, 1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6, 1, 2, 2, 3, 2, 3, 3, 4,
2, 3, 3, 4, 3, 4, 4, 5, 2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6, 3, 4, 4, 5, 4, 5, 5, 6,
4, 5, 5, 6, 5, 6, 6, 7, 1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6, 2, 3, 3, 4, 3, 4, 4, 5,
3, 4, 4, 5, 4, 5, 5, 6, 3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6, 3, 4, 4, 5, 4, 5, 5, 6,
4, 5, 5, 6, 5, 6, 6, 7, 3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
4, 5, 5, 6, 5, 6, 6, 7, 5, 6, 6, 7, 6, 7, 7, 8};
int32_t WelsGetNoneZeroCount_mmi(int16_t *level) {
int ret_val = 0;
__asm__ volatile(
".set arch=loongson3a \n\t"
"gslqc1 $f2, $f0, 0x0(%[level]) \n\t"
"gslqc1 $f6, $f4, 0x10(%[level]) \n\t"
"xor $f8, $f8, $f8 \n\t"
"pcmpeqh $f0, $f0, $f8 \n\t"
"pcmpeqh $f2, $f2, $f8 \n\t"
"pcmpeqh $f4, $f4, $f8 \n\t"
"pcmpeqh $f6, $f6, $f8 \n\t"
"packsshb $f4, $f4, $f6 \n\t"
"packsshb $f6, $f0, $f2 \n\t"
"pmovmskb $f0, $f4 \n\t"
"pmovmskb $f2, $f6 \n\t"
"dmfc1 $8, $f0 \n\t"
"dmfc1 $9, $f2 \n\t"
"xor $8, 0xFF \n\t"
"xor $9, 0xFF \n\t"
PTR_ADDU "$10, $8, %[nozero_count_table] \n\t"
"lbu $8, 0x0($10) \n\t"
PTR_ADDU "$10, $9, %[nozero_count_table] \n\t"
"lbu $9, 0x0($10) \n\t"
PTR_ADDU "%[ret_val], $8, $9 \n\t"
: [ret_val] "=r"((int)ret_val)
: [level] "r"((unsigned char *)level),
[nozero_count_table] "r"((unsigned char *)nozero_count_table)
: "memory", "$8", "$9", "$10", "$f0", "$f2", "$f4", "$f6", "$f8"
);
return ret_val;
}
void WelsScan4x4DcAc_mmi(int16_t level[16], int16_t *pDct) {
BACKUP_REG;
__asm__ volatile(
".set arch=loongson3a \n\t"
"gslqc1 $f2, $f0, 0x0(%[pDct]) \n\t"
"gslqc1 $f6, $f4, 0x10(%[pDct]) \n\t"
"dli $8, 0x3 \n\t"
"dmtc1 $8, $f22 \n\t"
"dli $8, 0x2 \n\t"
"dmtc1 $8, $f24 \n\t"
"dli $8, 0x1 \n\t"
"dmtc1 $8, $f26 \n\t"
"dmtc1 $0, $f28 \n\t"
"pextrh $f18, $f2, $f22 \n\t"
"pextrh $f20, $f4, $f24 \n\t"
"pextrh $f16, $f2, $f26 \n\t"
"pinsrh_2 $f4, $f4, $f18 \n\t"
"pinsrh_3 $f2, $f2, $f16 \n\t"
"pextrh $f18, $f4, $f28 \n\t"
"pinsrh_1 $f2, $f2, $f18 \n\t"
"pinsrh_0 $f4, $f4, $f20 \n\t"
"dli $8, 0x93 \n\t"
"dmtc1 $8, $f22 \n\t"
"dli $8, 0x39 \n\t"
"dmtc1 $8, $f24 \n\t"
"punpckhwd $f10, $f0, $f2 \n\t"
"punpcklwd $f8, $f0, $f2 \n\t"
"punpckhwd $f14, $f4, $f6 \n\t"
"punpcklwd $f12, $f4, $f6 \n\t"
"mov.d $f0, $f8 \n\t"
"pshufh $f2, $f10, $f22 \n\t"
"pshufh $f4, $f12, $f24 \n\t"
"mov.d $f6, $f14 \n\t"
"gssqc1 $f2, $f0, 0x0(%[level]) \n\t"
"gssqc1 $f6, $f4, 0x10(%[level]) \n\t"
:
: [level] "r"((short *)level), [pDct] "r"((short *)pDct)
: "memory", "$8", "$f0", "$f2", "$f4", "$f6", "$f8", "$f10", "$f12",
"$f14", "$f16", "$f18", "$f20", "$f22", "$f24", "$f26", "$f28"
);
RECOVER_REG;
}
void WelsScan4x4Ac_mmi(int16_t *zig_value, int16_t *pDct) {
BACKUP_REG;
__asm__ volatile(
".set arch=loongson3a \n\t"
"gslqc1 $f2, $f0, 0x0(%[pDct]) \n\t"
"gslqc1 $f6, $f4, 0x10(%[pDct]) \n\t"
"mov.d $f8, $f2 \n\t"
"mov.d $f2, $f4 \n\t"
"mov.d $f10, $f6 \n\t"
"mov.d $f12, $f2 \n\t"
"punpckhwd $f2, $f0, $f8 \n\t"
"punpcklwd $f0, $f0, $f8 \n\t"
"punpckhwd $f14, $f12, $f10 \n\t"
"punpcklwd $f12, $f12, $f10 \n\t"
"dmtc1 $0, $f20 \n\t"
"dli $8, 0x10 \n\t"
"dmtc1 $8, $f22 \n\t"
"dli $8, 0x30 \n\t"
"dmtc1 $8, $f24 \n\t"
"dli $8, 0x3 \n\t"
"dmtc1 $8, $f26 \n\t"
"dli $8, 0x93 \n\t"
"dmtc1 $8, $f28 \n\t"
"dli $8, 0x39 \n\t"
"dmtc1 $8, $f30 \n\t"
"pextrh $f16, $f0, $f26 \n\t"
"pextrh $f18, $f2, $f26 \n\t"
"pinsrh_3 $f2, $f2, $f16 \n\t"
"pextrh $f16, $f14, $f20 \n\t"
"pinsrh_0 $f14, $f14, $f18 \n\t"
"pextrh $f18, $f12, $f20 \n\t"
"pinsrh_0 $f12, $f12, $f16 \n\t"
"pinsrh_3 $f0, $f0, $f18 \n\t"
"mov.d $f4, $f0 \n\t"
"pshufh $f6, $f2, $f28 \n\t"
"pshufh $f8, $f12, $f30 \n\t"
"mov.d $f10, $f14 \n\t"
"mov.d $f12, $f8 \n\t"
"mov.d $f14, $f10 \n\t"
"dsrl $f4, $f4, $f22 \n\t"
"pinsrh_3 $f4, $f4, $f6 \n\t"
"dsrl $f6, $f6, $f22 \n\t"
"dsll $f14, $f12, $f24 \n\t"
"xor $f12, $f12, $f12 \n\t"
"or $f4, $f4, $f12 \n\t"
"or $f6, $f6, $f14 \n\t"
"dsrl $f8, $f8, $f22 \n\t"
"pinsrh_3 $f8, $f8, $f10 \n\t"
"dsrl $f10, $f10, $f22 \n\t"
"gssqc1 $f6, $f4, 0x0(%[zig_value]) \n\t"
"gssqc1 $f10, $f8, 0x10(%[zig_value]) \n\t"
:
: [zig_value] "r"((short *)zig_value), [pDct] "r"((short *)pDct)
: "memory", "$8", "$f0", "$f2", "$f4", "$f6", "$f8", "$f10", "$f12",
"$f14", "$f16", "$f18", "$f20", "$f22", "$f24", "$f26", "$f28", "$f30"
);
RECOVER_REG;
}
unsigned char i_ds_table[]__attribute__((aligned(16))) = {
3, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
unsigned char high_mask_table[]__attribute__((aligned(16))) = {
0, 0, 0, 3, 0, 2, 3, 6, 0, 2,
2, 5, 3, 5, 6, 9, 0, 1, 2, 5,
2, 4, 5, 8, 3, 5, 5, 8, 6, 8,
9,12, 0, 1, 1, 4, 2, 4, 5, 8,
2, 4, 4, 7, 5, 7, 8,11, 3, 4,
5, 8, 5, 7, 8,11, 6, 8, 8,11,
9,11,12,15, 0, 1, 1, 4, 1, 3,
4, 7, 2, 4, 4, 7, 5, 7, 8,11,
2, 3, 4, 7, 4, 6, 7,10, 5, 7,
7,10, 8,10,11,14, 3, 4, 4, 7,
5, 7, 8,11, 5, 7, 7,10, 8,10,
11,14, 6, 7, 8,11, 8,10,11,14,
9,11,11,14,12,14,15,18, 0, 0,
1, 4, 1, 3, 4, 7, 1, 3, 3, 6,
4, 6, 7,10, 2, 3, 4, 7, 4, 6,
7,10, 5, 7, 7,10, 8,10,11,14,
2, 3, 3, 6, 4, 6, 7,10, 4, 6,
6, 9, 7, 9,10,13, 5, 6, 7,10,
7, 9,10,13, 8,10,10,13,11,13,
14,17, 3, 4, 4, 7, 4, 6, 7,10,
5, 7, 7,10, 8,10,11,14, 5, 6,
7,10, 7, 9,10,13, 8,10,10,13,
11,13,14,17, 6, 7, 7,10, 8,10,
11,14, 8,10,10,13,11,13,14,17,
9,10,11,14,11,13,14,17,12,14,
14,17,15,17,18,21};
unsigned char low_mask_table[]__attribute__((aligned(16))) = {
0, 3, 2, 6, 2, 5, 5, 9, 1, 5,
4, 8, 5, 8, 8,12, 1, 4, 4, 8,
4, 7, 7,11, 4, 8, 7,11, 8,11,
11,15, 1, 4, 3, 7, 4, 7, 7,11,
3, 7, 6,10, 7,10,10,14, 4, 7,
7,11, 7,10,10,14, 7,11,10,14,
11,14,14,18, 0, 4, 3, 7, 3, 6,
6,10, 3, 7, 6,10, 7,10,10,14,
3, 6, 6,10, 6, 9, 9,13, 6,10,
9,13,10,13,13,17, 4, 7, 6,10,
7,10,10,14, 6,10, 9,13,10,13,
13,17, 7,10,10,14,10,13,13,17,
10,14,13,17,14,17,17,21, 0, 3,
3, 7, 3, 6, 6,10, 2, 6, 5, 9,
6, 9, 9,13, 3, 6, 6,10, 6, 9,
9,13, 6,10, 9,13,10,13,13,17,
3, 6, 5, 9, 6, 9, 9,13, 5, 9,
8,12, 9,12,12,16, 6, 9, 9,13,
9,12,12,16, 9,13,12,16,13,16,
16,20, 3, 7, 6,10, 6, 9, 9,13,
6,10, 9,13,10,13,13,17, 6, 9,
9,13, 9,12,12,16, 9,13,12,16,
13,16,16,20, 7,10, 9,13,10,13,
13,17, 9,13,12,16,13,16,16,20,
10,13,13,17,13,16,16,20,13,17,
16,20,17,20,20,24};
int32_t WelsCalculateSingleCtr4x4_mmi(int16_t *pDct) {
int32_t iSingleCtr = 0;
__asm__ volatile(
".set arch=loongson3a \n\t"
"gslqc1 $f2, $f0, 0x0(%[pDct]) \n\t"
"gslqc1 $f6, $f4, 0x10(%[pDct]) \n\t"
"packsshb $f0, $f0, $f2 \n\t"
"packsshb $f2, $f4, $f6 \n\t"
"xor $f10, $f10, $f10 \n\t"
"xor $f8, $f8, $f8 \n\t"
"pcmpeqb $f0, $f0, $f8 \n\t"
"pcmpeqb $f2, $f2, $f8 \n\t"
"pmovmskb $f10, $f0 \n\t"
"pmovmskb $f12, $f2 \n\t"
"punpcklbh $f10, $f10, $f12 \n\t"
"dmfc1 $12, $f10 \n\t"
"dli $8, 0xffff \n\t"
"xor $12, $12, $8 \n\t"
"xor %[pDct], %[pDct], %[pDct] \n\t"
"dli $8, 0x80 \n\t"
"dli $9, 0x7 \n\t"
"dli $10, 0x100 \n\t"
"dli $11, 0x8 \n\t"
"1: \n\t"
"and $13, $12, $8 \n\t"
"bnez $13, 2f \n\t"
"nop \n\t"
"daddiu $9, -0x1 \n\t"
"dsrl $8, 1 \n\t"
"bnez $9, 1b \n\t"
"nop \n\t"
"2: \n\t"
"and $13, $12, $10 \n\t"
"bnez $13, 3f \n\t"
"nop \n\t"
"daddiu $11, 0x1 \n\t"
"dsll $10, 1 \n\t"
"daddiu $13, $11, -0x10 \n\t"
"bltz $13, 2b \n\t"
"nop \n\t"
"3: \n\t"
"dsubu $11, $11, $9 \n\t"
"daddiu $11, -0x1 \n\t"
PTR_ADDU "$8, %[i_ds_table], $11 \n\t"
"lb $10, 0x0($8) \n\t"
PTR_ADDU "%[pDct], %[pDct], $10 \n\t"
"move $11, $12 \n\t"
"dli $10, 0xff \n\t"
"and $12, $10 \n\t"
"dsrl $11, 0x8 \n\t"
"and $11, $10 \n\t"
PTR_ADDU "$8, %[low_mask_table], $12 \n\t"
"lb $10, 0x0($8) \n\t"
PTR_ADDU "%[pDct], %[pDct], $10 \n\t"
PTR_ADDU "$8, %[high_mask_table], $11 \n\t"
"lb $10, 0x0($8) \n\t"
PTR_ADDU "%[iSingleCtr], %[pDct], $10 \n\t"
: [iSingleCtr] "=r"(iSingleCtr)
: [pDct] "r"((short *)pDct),
[i_ds_table] "r"((unsigned char *)i_ds_table),
[high_mask_table] "r"((unsigned char *)high_mask_table),
[low_mask_table] "r"((unsigned char *)low_mask_table)
: "memory", "$8", "$9", "$10", "$11", "$12", "$13", "$f0", "$f2", "$f4",
"$f6", "$f8", "$f10", "$f12"
);
return iSingleCtr;
}