backup
/
gps
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
gps/GPSResources/tcpmp/common/softidct/mcomp_mips64.c

625 lines
12 KiB
C
Executable File
Raw Blame History

/*****************************************************************************
*
* This program is free software ; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* $Id: mcomp_mips64.c 284 2005-10-04 08:54:26Z picard $
*
* The Core Pocket Media Player
* Copyright (c) 2004-2005 Gabor Kovacs
*
****************************************************************************/
#include "../common.h"
#include "softidct.h"
#if defined(MIPS64)
// important: disable interrupts before using 64bit registers (but not too long, it could freeze)
// important: $8 can't be used as 64bit register (trashed by some kernel routine)
// $8 src end pointer
// $4 src pointer
// $5 dst pointer
// $6 src stride
// $7 dst stride
// $2,$9 first item lower 8 bytes (in two forms)
// $10,$11 first item upper 8 bytes (in two forms) - for 16x16 macroblocks
// $12,$13 second item lower 8 bytes (in two forms)
// $14,$15 second item upper 8 bytes (in two forms) - for 16x16 macroblocks
// $24 0x0101 0101 0101 0101 - for non horver
// $25 0xFEFE FEFE FEFE FEFE - for non horver
// $24 rounding - for horver
// $25 temporary - for 16x16 horver
// $3 0x0303 0303 0303 0303 - for horver
// $1 0xFCFC FCFC FCFC FCFC - for horver
// $16 temporary - for 16x16 horver (must be saved/restored)
#define SET_SRCEND8 \
"sll $8,$6,3;" \
"addu $8,$4,$8;"
#define SET_SRCEND16 \
"sll $8,$6,4;" \
"addu $8,$4,$8;"
#define SET_MASKS \
"li $24,0x01010101;" \
"dsll $25,$24,32;" \
"or $24,$24,$25;" \
"nor $25,$24,$0;"
#define SET_MASKS2 \
".set noat;" \
"li $3,0x03030303;" \
"dsll $1,$3,32;" \
"or $3,$3,$1;" \
"nor $1,$3,$0;"
#define LOAD_FIRST8(ofs) \
"uld $2, " #ofs "($4);" \
"and $9,$2,$25;" \
"dsrl $9,$9,1;"
#define LOAD_FIRST16(ofs) \
"uld $2, " #ofs "($4);" \
"uld $10," #ofs "+8($4);"\
"and $9,$2,$25;" \
"and $11,$10,$25;" \
"dsrl $9,$9,1;" \
"dsrl $11,$11,1;"
#define LOAD_SECOND8(ofs) \
"uld $12," #ofs "($4);" \
"and $13,$12,$25;" \
"dsrl $13,$13,1;"
#define LOAD_SECOND16(ofs) \
"uld $12," #ofs "($4);" \
"uld $14," #ofs "+8($4);"\
"and $13,$12,$25;" \
"and $15,$14,$25;" \
"dsrl $13,$13,1;" \
"dsrl $15,$15,1;"
#define LOAD_FIRST8_HV \
"uld $2,0($4);" \
"uld $9,1($4);" \
"and $10,$2,$1;" \
"and $11,$9,$1;" \
"and $2,$2,$3;" \
"and $9,$9,$3;" \
"dsrl $10,$10,2;" \
"dsrl $11,$11,2;" \
"daddu $2,$2,$9;" \
"daddu $9,$10,$11;"
#define LOAD_FIRST16_HV \
"uld $2,0($4);" \
"uld $9,1($4);" \
"and $16,$2,$1;" \
"and $25,$9,$1;" \
"and $2,$2,$3;" \
"and $9,$9,$3;" \
"dsrl $16,$16,2;" \
"dsrl $25,$25,2;" \
"daddu $2,$2,$9;" \
"daddu $9,$16,$25;" \
\
"uld $10,8($4);" \
"uld $11,9($4);" \
"and $16,$10,$1;" \
"and $25,$11,$1;" \
"and $10,$10,$3;" \
"and $11,$11,$3;" \
"dsrl $16,$16,2;" \
"dsrl $25,$25,2;" \
"daddu $10,$10,$11;" \
"daddu $11,$16,$25;"
#define LOAD_SECOND8_HV \
"uld $12,0($4);" \
"uld $13,1($4);" \
"and $14,$12,$1;" \
"and $15,$13,$1;" \
"and $12,$12,$3;" \
"and $13,$13,$3;" \
"dsrl $14,$14,2;" \
"dsrl $15,$15,2;" \
"daddu $12,$12,$13;" \
"daddu $13,$14,$15;"
#define LOAD_SECOND16_HV \
"uld $12,0($4);" \
"uld $13,1($4);" \
"and $16,$12,$1;" \
"and $25,$13,$1;" \
"and $12,$12,$3;" \
"and $13,$13,$3;" \
"dsrl $16,$16,2;" \
"dsrl $25,$25,2;" \
"daddu $12,$12,$13;" \
"daddu $13,$16,$25;" \
\
"uld $14,8($4);" \
"uld $15,9($4);" \
"and $16,$14,$1;" \
"and $25,$15,$1;" \
"and $14,$14,$3;" \
"and $15,$15,$3;" \
"dsrl $16,$16,2;" \
"dsrl $25,$25,2;" \
"daddu $14,$14,$15;" \
"daddu $15,$16,$25;"
#define AVG8 \
"or $2,$2,$12;" \
"and $2,$2,$24;" \
"daddu $2,$2,$9;" \
"daddu $2,$2,$13;"
#define AVG16 \
"or $2,$2,$12;" \
"or $10,$10,$14;" \
"and $2,$2,$24;" \
"and $10,$10,$24;" \
"daddu $2,$2,$9;" \
"daddu $10,$10,$11;" \
"daddu $2,$2,$13;" \
"daddu $10,$10,$15;"
#define AVGROUND8 \
"and $2,$2,$12;" \
"and $2,$2,$24;" \
"daddu $2,$2,$9;" \
"daddu $2,$2,$13;"
#define AVGROUND16 \
"and $2,$2,$12;" \
"and $10,$10,$14;" \
"and $2,$2,$24;" \
"and $10,$10,$24;" \
"daddu $2,$2,$9;" \
"daddu $10,$10,$11;" \
"daddu $2,$2,$13;" \
"daddu $10,$10,$15;"
#define SWAPSET8 \
"move $2,$12;" \
"move $9,$13;"
#define SWAPSET16 \
"move $2,$12;" \
"move $9,$13;" \
"move $10,$14;" \
"move $11,$15;"
#define WRITE8 \
"sdr $2,0($5);" \
"addu $5,$5,$7;"
#define WRITE16 \
"sdr $2,0($5);" \
"sdr $10,8($5);" \
"addu $5,$5,$7;"
#define SAVE \
"addiu $sp,$sp,-4;" \
"sw $16,0(sp);"
#define RESTORE \
"lw $16,0(sp);" \
"addiu $sp,$sp,4;"
#ifdef MIPSVR41XX
//cache without loading
#define CACHE16 \
".set noreorder;" \
"cache 13,0($5);" \
".set reorder;"
#else
#define CACHE16
#endif
void CopyBlock(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm( SET_SRCEND8
"loop:"
"uld $2,0($4);"
"addu $4,$4,$6;"
"sdr $2,0($5);"
"addu $5,$5,$7;"
"bne $4,$8,loop;");
}
void CopyBlockHor(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND8
SET_MASKS
"loophor:"
LOAD_FIRST8(0)
LOAD_SECOND8(1)
"addu $4,$4,$6;"
AVG8
WRITE8
"bne $4,$8,loophor;");
}
void CopyBlockHorRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND8
SET_MASKS
"loophorround:"
LOAD_FIRST8(0)
LOAD_SECOND8(1)
"addu $4,$4,$6;"
AVGROUND8
WRITE8
"bne $4,$8,loophorround;");
}
void CopyBlockVer(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND8
SET_MASKS
LOAD_FIRST8(0)
"loopver:"
"addu $4,$4,$6;"
LOAD_SECOND8(0)
AVG8
WRITE8
SWAPSET8
"bne $4,$8,loopver;");
}
void CopyBlockVerRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND8
SET_MASKS
LOAD_FIRST8(0)
"loopverround:"
"addu $4,$4,$6;"
LOAD_SECOND8(0)
AVGROUND8
WRITE8
SWAPSET8
"bne $4,$8,loopverround;");
}
void CopyBlockHorVer(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND8
SET_MASKS2
"dsll $24,$3,1;"
"and $24,$24,$3;" // 0x0202 0202 0202 0202
//preprocessing
LOAD_FIRST8_HV
"loophorver:"
"addu $4,$4,$6;"
LOAD_SECOND8_HV
"daddu $2,$2,$12;"
"daddu $9,$9,$13;"
"daddu $2,$2,$24;"
"and $2,$2,$1;"
"dsrl $2,$2,2;"
"daddu $2,$2,$9;"
WRITE8
SWAPSET8
"bne $4,$8,loophorver;");
}
void CopyBlockHorVerRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND8
SET_MASKS2
"dsrl $24,$3,1;"
"and $24,$24,$3;" // 0x0101 0101 0101 0101
//preprocessing
LOAD_FIRST8_HV
"loophorverround:"
"addu $4,$4,$6;"
LOAD_SECOND8_HV
"daddu $2,$2,$12;"
"daddu $9,$9,$13;"
"daddu $2,$2,$24;"
"and $2,$2,$1;"
"dsrl $2,$2,2;"
"daddu $2,$2,$9;"
WRITE8
SWAPSET8
"bne $4,$8,loophorverround;");
}
void CopyMBlock(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm( SET_SRCEND16
"loopm:"
CACHE16
"uld $2,0($4);"
"uld $10,8($4);"
"addu $4,$4,$6;"
WRITE16
"bne $4,$8,loopm;");
}
void CopyMBlockHor(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND16
SET_MASKS
"loopmhor:"
LOAD_FIRST16(0)
LOAD_SECOND16(1)
"addu $4,$4,$6;"
CACHE16
AVG16
WRITE16
"bne $4,$8,loopmhor;");
}
void CopyMBlockHorRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND16
SET_MASKS
"loopmhorround:"
LOAD_FIRST16(0)
LOAD_SECOND16(1)
"addu $4,$4,$6;"
CACHE16
AVGROUND16
WRITE16
"bne $4,$8,loopmhorround;");
}
void CopyMBlockVer(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND16
SET_MASKS
LOAD_FIRST16(0)
"loopmver:"
"addu $4,$4,$6;"
LOAD_SECOND16(0)
CACHE16
AVG16
WRITE16
SWAPSET16
"bne $4,$8,loopmver;"
);
}
void CopyMBlockVerRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SET_SRCEND16
SET_MASKS
LOAD_FIRST16(0)
"loopmverround:"
"addu $4,$4,$6;"
LOAD_SECOND16(0)
CACHE16
AVGROUND16
WRITE16
SWAPSET16
"bne $4,$8,loopmverround;");
}
void CopyMBlockHorVer(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SAVE
SET_SRCEND16
SET_MASKS2
"dsll $24,$3,1;"
"and $24,$24,$3;" // 0x0202 0202 0202 0202
//preprocessing
LOAD_FIRST16_HV
"loopmhorver:"
"addu $4,$4,$6;"
LOAD_SECOND16_HV
CACHE16);
__asm ( "daddu $2,$2,$12;"
"daddu $9,$9,$13;"
"daddu $10,$10,$14;"
"daddu $11,$11,$15;"
"daddu $2,$2,$24;"
"daddu $10,$10,$24;"
"and $2,$2,$1;"
"and $10,$10,$1;"
"dsrl $2,$2,2;"
"dsrl $10,$10,2;"
"daddu $2,$2,$9;"
"daddu $10,$10,$11;"
WRITE16
SWAPSET16
"bne $4,$8,loopmhorver;"
RESTORE);
}
void CopyMBlockHorVerRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm ( SAVE
SET_SRCEND16
SET_MASKS2
"dsrl $24,$3,1;"
"and $24,$24,$3;" // 0x0101 0101 0101 0101
//preprocessing
LOAD_FIRST16_HV
"loopmhorverround:"
"addu $4,$4,$6;");
__asm ( LOAD_SECOND16_HV
CACHE16
"daddu $2,$2,$12;"
"daddu $9,$9,$13;"
"daddu $10,$10,$14;"
"daddu $11,$11,$15;"
"daddu $2,$2,$24;"
"daddu $10,$10,$24;"
"and $2,$2,$1;"
"and $10,$10,$1;"
"dsrl $2,$2,2;"
"dsrl $10,$10,2;"
"daddu $2,$2,$9;"
"daddu $10,$10,$11;"
WRITE16
SWAPSET16
"bne $4,$8,loopmhorverround;"
RESTORE);
}
void AddBlock8x8(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm( SET_SRCEND8
SET_MASKS
"loopadd:"
"ldr $2,0($4);"
"addu $4,$4,$6;"
"ldr $9,0($5);"
"and $11,$2,$25;"
"or $2,$2,$9;"
"and $2,$2,$24;"
"dsrl $11,$11,1;"
"daddu $2,$2,$11;"
"and $9,$9,$25;"
"dsrl $9,$9,1;"
"daddu $2,$2,$9;"
"sdr $2,0($5);"
"addu $5,$5,$7;"
"bne $4,$8,loopadd;");
}
void AddBlock16x16(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride)
{
__asm( SET_SRCEND16
SET_MASKS
"loopadd16:"
"ldr $2,0($4);"
"ldr $10,8($4);"
#ifdef MIPSVR41XX
".set noreorder;"
"cache 17,0($4);" // hit invalidate (lose changes)
".set reorder;"
#endif
"addu $4,$4,$6;"
"ldr $9,0($5);"
"and $11,$2,$25;"
"or $2,$2,$9;"
"and $2,$2,$24;"
"dsrl $11,$11,1;"
"daddu $2,$2,$11;"
"and $9,$9,$25;"
"dsrl $9,$9,1;"
"daddu $2,$2,$9;"
"ldr $11,8($5);"
"and $9,$10,$25;"
"or $10,$10,$11;"
"and $10,$10,$24;"
"dsrl $9,$9,1;"
"daddu $10,$10,$9;"
"and $11,$11,$25;"
"dsrl $11,$11,1;"
"daddu $10,$10,$11;"
"sdr $2,0($5);"
"sdr $10,8($5);"
"addu $5,$5,$7;"
"bne $4,$8,loopadd16;");
}
#endif
Reference in New Issue View Git Blame Copy Permalink