mirror of
https://github.com/ekeeke/Genesis-Plus-GX.git
synced 2024-12-28 12:11:50 +01:00
dbc4d02fea
~fixed some memory leaks with Blip Buffer ~reverted previous change regarding frame synchronization: always use VSYNC when switched to 60hz (fix video issues in interlace mode 2) ~optimized cell rendering (less byte swapping) ~fixed savestate file version checking
1812 lines
48 KiB
C
1812 lines
48 KiB
C
/***************************************************************************************
|
|
* Genesis Plus
|
|
* Video Display Processor (Rendering)
|
|
*
|
|
* Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003 Charles Mac Donald (original code)
|
|
* Eke-Eke (2007,2008,2009), additional code & fixes for the GCN/Wii port
|
|
*
|
|
* 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
|
|
*
|
|
****************************************************************************************/
|
|
|
|
#include "shared.h"
|
|
#include "md_ntsc.h"
|
|
#include "sms_ntsc.h"
|
|
|
|
/*** NTSC Filters ***/
|
|
extern md_ntsc_t md_ntsc;
|
|
extern sms_ntsc_t sms_ntsc;
|
|
|
|
/* Look-up pixel table information */
|
|
#define LUT_MAX (5)
|
|
#define LUT_SIZE (0x10000)
|
|
|
|
/* Clip structure */
|
|
typedef struct
|
|
{
|
|
uint8 left;
|
|
uint8 right;
|
|
uint8 enable;
|
|
}clip_t;
|
|
|
|
#undef ALIGN_LONG
|
|
#ifdef ALIGN_LONG
|
|
/* Or change the names if you depend on these from elsewhere.. */
|
|
#undef READ_LONG
|
|
#undef WRITE_LONG
|
|
|
|
static __inline__ uint32 READ_LONG(void *address)
|
|
{
|
|
if ((uint32)address & 3)
|
|
{
|
|
#ifdef LSB_FIRST /* little endian version */
|
|
return ( *((uint8 *)address) +
|
|
(*((uint8 *)address+1) << 8) +
|
|
(*((uint8 *)address+2) << 16) +
|
|
(*((uint8 *)address+3) << 24) );
|
|
#else /* big endian version */
|
|
return ( *((uint8 *)address+3) +
|
|
(*((uint8 *)address+2) << 8) +
|
|
(*((uint8 *)address+1) << 16) +
|
|
(*((uint8 *)address) << 24) );
|
|
#endif /* LSB_FIRST */
|
|
}
|
|
else return *(uint32 *)address;
|
|
}
|
|
|
|
static __inline__ void WRITE_LONG(void *address, uint32 data)
|
|
{
|
|
if ((uint32)address & 3)
|
|
{
|
|
#ifdef LSB_FIRST
|
|
*((uint8 *)address) = data;
|
|
*((uint8 *)address+1) = (data >> 8);
|
|
*((uint8 *)address+2) = (data >> 16);
|
|
*((uint8 *)address+3) = (data >> 24);
|
|
#else
|
|
*((uint8 *)address+3) = data;
|
|
*((uint8 *)address+2) = (data >> 8);
|
|
*((uint8 *)address+1) = (data >> 16);
|
|
*((uint8 *)address) = (data >> 24);
|
|
#endif /* LSB_FIRST */
|
|
return;
|
|
}
|
|
else *(uint32 *)address = data;
|
|
}
|
|
|
|
#endif /* ALIGN_LONG */
|
|
|
|
|
|
/*
|
|
Two Pattern Attributes are written in VRAM as two 16bits WORD:
|
|
|
|
P = priority bit
|
|
C = color palette (2 bits)
|
|
V = Vertical Flip bit
|
|
H = Horizontal Flip bit
|
|
N = Pattern Number (11 bits)
|
|
|
|
MSB PCCVHNNN NNNNNNNN LSB PCCVHNNN NNNNNNNN LSB
|
|
PATTERN1 PATTERN2
|
|
|
|
Pattern attributes are read from VRAM as 32bits WORD like this:
|
|
|
|
LIT_ENDIAN: ATTR is MSB PCCVHNNN NNNNNNNN PCCVHNNN NNNNNNNN LSB
|
|
PATTERN2 PATTERN1
|
|
|
|
BIG_ENDIAN: ATTR is MSB PCCVHNNN NNNNNNNN PCCVHNNN NNNNNNNN LSB
|
|
PATTERN1 PATTERN2
|
|
|
|
|
|
Each Line Buffer written byte describe one pixel data like this:
|
|
|
|
msb SPppcccc lsb
|
|
|
|
with:
|
|
S = sprite data indicator (not written here)
|
|
P = priority bit (from Pattern Attribute)
|
|
p = color palette (from Pattern Attribute)
|
|
c = color data (from Pattern Cache)
|
|
|
|
|
|
A column is 2 patterns wide
|
|
A pattern is 8 pixels wide = 8 bytes = two 32 bits write
|
|
*/
|
|
|
|
/* Draw a single 8-pixel column */
|
|
/*
|
|
pattern cache is addressed like this: 00000VHN NNNNNNNN NNYYYXXX
|
|
with : Y = pattern row (1-8 lines)
|
|
X = pattern column (1-8 pixels)
|
|
V = Vertical Flip bit
|
|
H = Horizontal Flip bit
|
|
N = Pattern Number (1-2048)
|
|
*/
|
|
#ifdef ALIGN_LONG
|
|
#ifdef LSB_FIRST
|
|
#define DRAW_COLUMN(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF) << 6 | (LINE)]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
ATTR >>= 16; \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF) << 6 | (LINE)]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++;
|
|
#else
|
|
#define DRAW_COLUMN(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 29) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF0000) >> 10 | (LINE)]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF) << 6 | (LINE)]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++;
|
|
#endif
|
|
#else /* NOT ALIGNED */
|
|
#ifdef LSB_FIRST
|
|
#define DRAW_COLUMN(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF) << 6 | (LINE)]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex); \
|
|
ATTR >>= 16; \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF) << 6 | (LINE)]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex);
|
|
#else
|
|
#define DRAW_COLUMN(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 29) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF0000) >> 10 | (LINE)]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex); \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
src = (uint32 *)&bg_pattern_cache[(ATTR & 0x1FFF) << 6 | (LINE)]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex);
|
|
#endif
|
|
#endif /* ALIGN_LONG */
|
|
|
|
|
|
/* Draw a single 16-pixel column */
|
|
/*
|
|
pattern cache is addressed like this: 00000VHN NNNNNNNN NYYYYXXX
|
|
with : Y = pattern row (1-16 lines)
|
|
X = pattern column (1-8 pixels)
|
|
V = Vertical Flip bit
|
|
H = Horizontal Flip bit
|
|
N = Pattern Number (1-1024)
|
|
|
|
one pattern line is 8 pixels = 8 bytes = 2 * 32 bits
|
|
*/
|
|
#ifdef ALIGN_LONG
|
|
#ifdef LSB_FIRST
|
|
#define DRAW_COLUMN_IM2(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
offs = (ATTR & 0x03FF) << 7 | (ATTR & 0x1800) << 6 | (LINE); \
|
|
if(ATTR & 0x1000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
ATTR >>= 16; \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
offs = (ATTR & 0x03FF) << 7 | (ATTR & 0x1800) << 6 | (LINE); \
|
|
if(ATTR & 0x1000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++;
|
|
#else
|
|
#define DRAW_COLUMN_IM2(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 29) & 7]; \
|
|
offs = (ATTR & 0x03FF0000) >> 9 | (ATTR & 0x18000000) >> 10 | (LINE); \
|
|
if(ATTR & 0x10000000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
offs = (ATTR & 0x03FF) << 7 | (ATTR & 0x1800) << 6 | (LINE); \
|
|
if(ATTR & 0x1000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++; \
|
|
WRITE_LONG(dst, READ_LONG(src) | atex); \
|
|
dst++; \
|
|
src++;
|
|
#endif
|
|
#else /* NOT ALIGNED */
|
|
#ifdef LSB_FIRST
|
|
#define DRAW_COLUMN_IM2(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
offs = (ATTR & 0x03FF) << 7 | (ATTR & 0x1800) << 6 | (LINE); \
|
|
if(ATTR & 0x1000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex); \
|
|
ATTR >>= 16; \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
offs = (ATTR & 0x03FF) << 7 | (ATTR & 0x1800) << 6 | (LINE); \
|
|
if(ATTR & 0x1000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex);
|
|
#else
|
|
#define DRAW_COLUMN_IM2(ATTR, LINE) \
|
|
atex = atex_table[(ATTR >> 29) & 7]; \
|
|
offs = (ATTR & 0x03FF0000) >> 9 | (ATTR & 0x18000000) >> 10 | (LINE); \
|
|
if(ATTR & 0x10000000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex); \
|
|
atex = atex_table[(ATTR >> 13) & 7]; \
|
|
offs = (ATTR & 0x03FF) << 7 | (ATTR & 0x1800) << 6 | (LINE); \
|
|
if(ATTR & 0x1000) offs ^= 0x40; \
|
|
src = (uint32 *)&bg_pattern_cache[offs]; \
|
|
*dst++ = (*src++ | atex); \
|
|
*dst++ = (*src++ | atex);
|
|
#endif
|
|
#endif /* ALIGN_LONG */
|
|
|
|
/*
|
|
gcc complains about this:
|
|
*lb++ = table[(*lb << 8) |(*src++ | palette)];
|
|
.. claiming the result on lb is undefined.
|
|
So we manually advance lb and use constant offsets into the line buffer.
|
|
|
|
Added sprite collision detection
|
|
(check if non-transparent sprite data has been previously drawn)
|
|
*/
|
|
#define DRAW_SPRITE_TILE \
|
|
for(i=0; i<8; i++) \
|
|
{ \
|
|
if ((lb[i] & 0x80) && (lb[i] & 0x0F) && (src[i] & 0x0F)) status |= 0x20; \
|
|
lb[i] = table[(lb[i] << 8) |(src[i] | palette)]; \
|
|
}
|
|
|
|
|
|
/**************************************************/
|
|
/* Pixel creation macros */
|
|
/* Input is four bits each (R,G,B), 12 bits total */
|
|
/* Color range depends on the S/TE mode: */
|
|
/* */
|
|
/* normal mode : xxx0 (0-14) */
|
|
/* shadow mode : 0xxx (0-7) */
|
|
/* highlight mode: 1xxx (8-15) */
|
|
/* */
|
|
/* with xxx0 = original 4-bits CRAM value */
|
|
/**************************************************/
|
|
|
|
/* 5:6:5 RGB */
|
|
/* This RGB format uses 5 or 6bits color */
|
|
/* 4 bits color value need to be dithered */
|
|
/* to match the whole color range: */
|
|
/* */
|
|
/* R,B (5 bits) : yyyyy (0-31) */
|
|
/* G (6 bits) : yyyyyy (0-63) */
|
|
/* */
|
|
/* normal mode : xxx0 (0-14) */
|
|
/* 0000 -> 00000 (000000) */
|
|
/* 0010 -> 00100 (001000) */
|
|
/* 0100 -> 01000 (010001) */
|
|
/* 0110 -> 01100 (011001) */
|
|
/* 1000 -> 10001 (100010) */
|
|
/* 1010 -> 10101 (101010) */
|
|
/* 1100 -> 11001 (110011) */
|
|
/* 1110 -> 11101 (111011) */
|
|
/* */
|
|
/* shadow mode : 0xxx (0-7) */
|
|
/* 0000 -> 00000 (000000) */
|
|
/* 0001 -> 00010 (000100) */
|
|
/* 0010 -> 00100 (001000) */
|
|
/* 0011 -> 00110 (001100) */
|
|
/* 0100 -> 01000 (010001) */
|
|
/* 0101 -> 01010 (010101) */
|
|
/* 0110 -> 01100 (011001) */
|
|
/* 0111 -> 01110 (011101) */
|
|
/* */
|
|
/* highlight mode: 1xxx (8-15) */
|
|
/* 1000 -> 10001 (100010) */
|
|
/* 1001 -> 10011 (100110) */
|
|
/* 1010 -> 10101 (101010) */
|
|
/* 1011 -> 10111 (101110) */
|
|
/* 1100 -> 11001 (110011) */
|
|
/* 1101 -> 11011 (110111) */
|
|
/* 1110 -> 11101 (111011) */
|
|
/* 1111 -> 11111 (111111) */
|
|
/* */
|
|
/******************************************/
|
|
#define MAKE_PIXEL_16(r,g,b) (((r) << 12) | (((r) >> 3) << 12) | ((g) << 7) | (((g) >> 2) << 5) | ((b) << 1) | ((b) >> 3) )
|
|
|
|
#ifndef NGC
|
|
/* 8:8:8 RGB */
|
|
#define MAKE_PIXEL_32(r,g,b) ((r) << 20 | (g) << 12 | (b) << 4)
|
|
/* 5:5:5 RGB */
|
|
#define MAKE_PIXEL_15(r,g,b) ((r) << 11 | (g) << 6 | (b) << 1)
|
|
/* 3:3:2 RGB */
|
|
#define MAKE_PIXEL_8(r,g,b) ((r) << 5 | (g) << 2 | ((b) >> 1))
|
|
#endif
|
|
|
|
|
|
/* Clip data */
|
|
static clip_t clip[2];
|
|
|
|
/* Attribute expansion table */
|
|
static const uint32 atex_table[] = {
|
|
0x00000000, 0x10101010, 0x20202020, 0x30303030,
|
|
0x40404040, 0x50505050, 0x60606060, 0x70707070
|
|
};
|
|
|
|
/* Sprite name look-up table */
|
|
static uint8 name_lut[0x400];
|
|
|
|
struct
|
|
{
|
|
uint16 ypos;
|
|
uint16 xpos;
|
|
uint16 attr;
|
|
uint8 size;
|
|
uint8 index; // unused
|
|
} object_info[20];
|
|
|
|
/* Pixel look-up tables and table base address */
|
|
static uint8 *lut[5];
|
|
static uint8 *lut_base = NULL;
|
|
|
|
#ifndef NGC
|
|
/* 8-bit pixel remapping data */
|
|
static uint8 pixel_8[0x100];
|
|
static uint8 pixel_8_lut[3][0x200];
|
|
|
|
/* 15-bit pixel remapping data */
|
|
static uint16 pixel_15[0x100];
|
|
static uint16 pixel_15_lut[3][0x200];
|
|
|
|
/* 32-bit pixel remapping data */
|
|
static uint32 pixel_32[0x100];
|
|
static uint32 pixel_32_lut[3][0x200];
|
|
#endif
|
|
|
|
/* 16-bit pixel remapping data */
|
|
static uint16 pixel_16[0x100];
|
|
static uint16 pixel_16_lut[3][0x200];
|
|
|
|
/* Line buffers */
|
|
static uint8 tmp_buf[0x200]; /* Temporary buffer */
|
|
static uint8 bg_buf[0x200]; /* Merged background buffer */
|
|
static uint8 nta_buf[0x200]; /* Plane A / Window line buffer */
|
|
static uint8 ntb_buf[0x200]; /* Plane B line buffer */
|
|
static uint8 obj_buf[0x200]; /* Object layer line buffer */
|
|
|
|
/* Sprite line buffer data */
|
|
static uint32 object_index_count;
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Look-up table functions (handles priority between layers pixels) */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
static void palette_init(void)
|
|
{
|
|
int i;
|
|
|
|
for (i = 0; i < 0x200; i += 1)
|
|
{
|
|
int r, g, b;
|
|
|
|
r = (i >> 6) & 7;
|
|
g = (i >> 3) & 7;
|
|
b = (i >> 0) & 7;
|
|
|
|
#ifndef NGC
|
|
pixel_8_lut[0][i] = MAKE_PIXEL_8(r>>1,g>>1,b>>1);
|
|
pixel_8_lut[1][i] = MAKE_PIXEL_8(r,g,b);
|
|
pixel_8_lut[2][i] = MAKE_PIXEL_8((r>>1)|4,(g>>1)|4,(b>>1)|4);
|
|
|
|
pixel_15_lut[0][i] = MAKE_PIXEL_15(r,g,b);
|
|
pixel_15_lut[1][i] = MAKE_PIXEL_15(r<<1,g<<1,b<<1);
|
|
pixel_15_lut[2][i] = MAKE_PIXEL_15(r|8,g|8,b|8);
|
|
|
|
pixel_32_lut[0][i] = MAKE_PIXEL_32(r,g,b);
|
|
pixel_32_lut[1][i] = MAKE_PIXEL_32(r<<1,g<<1,b<<1);
|
|
pixel_32_lut[2][i] = MAKE_PIXEL_32(r|8,g|8,b|8);
|
|
#endif
|
|
|
|
pixel_16_lut[0][i] = MAKE_PIXEL_16(r,g,b);
|
|
pixel_16_lut[1][i] = MAKE_PIXEL_16(r<<1,g<<1,b<<1);
|
|
|
|
pixel_16_lut[2][i] = MAKE_PIXEL_16(r|8,g|8,b|8);
|
|
}
|
|
}
|
|
|
|
static void make_name_lut(void)
|
|
{
|
|
int col, row;
|
|
int vcol, vrow;
|
|
int width, height;
|
|
int flipx, flipy;
|
|
int i, name;
|
|
|
|
memset (name_lut, 0, sizeof (name_lut));
|
|
|
|
for (i = 0; i < 0x400; i += 1)
|
|
{
|
|
vcol = col = i & 3;
|
|
vrow = row = (i >> 2) & 3;
|
|
height = (i >> 4) & 3;
|
|
width = (i >> 6) & 3;
|
|
flipx = (i >> 8) & 1;
|
|
flipy = (i >> 9) & 1;
|
|
|
|
if(flipx) vcol = (width - col);
|
|
if(flipy) vrow = (height - row);
|
|
|
|
name = vrow + (vcol * (height + 1));
|
|
|
|
if ((row > height) || col > width) name = -1;
|
|
|
|
name_lut[i] = name;
|
|
}
|
|
}
|
|
|
|
/* Input (bx): d5-d0=color, d6=priority, d7=unused */
|
|
/* Input (ax): d5-d0=color, d6=priority, d7=unused */
|
|
/* Output: d5-d0=color, d6=priority, d7=unused */
|
|
static uint32 make_lut_bg(uint32 bx, uint32 ax)
|
|
{
|
|
int bf, bp, b;
|
|
int af, ap, a;
|
|
int x = 0;
|
|
int c;
|
|
|
|
bf = (bx & 0x7F);
|
|
bp = (bx >> 6) & 1;
|
|
b = (bx & 0x0F);
|
|
|
|
af = (ax & 0x7F);
|
|
ap = (ax >> 6) & 1;
|
|
a = (ax & 0x0F);
|
|
|
|
c = (ap ? (a ? af : (b ? bf : x)) : \
|
|
(bp ? (b ? bf : (a ? af : x)) : \
|
|
( (a ? af : (b ? bf : x)) )));
|
|
|
|
/* Strip palette bits from transparent pixels */
|
|
if((c & 0x0F) == 0x00) c &= 0xC0;
|
|
|
|
return (c);
|
|
}
|
|
|
|
|
|
/* Input (bx): d5-d0=color, d6=priority, d7=sprite pixel marker */
|
|
/* Input (sx): d5-d0=color, d6=priority, d7=unused */
|
|
/* Output: d5-d0=color, d6=zero, d7=sprite pixel marker */
|
|
static uint32 make_lut_obj(uint32 bx, uint32 sx)
|
|
{
|
|
int bf, bp, bs, b;
|
|
int sf, sp, s;
|
|
int c;
|
|
|
|
bf = (bx & 0x3F);
|
|
bs = (bx >> 7) & 1;
|
|
bp = (bx >> 6) & 1;
|
|
b = (bx & 0x0F);
|
|
|
|
sf = (sx & 0x3F);
|
|
sp = (sx >> 6) & 1;
|
|
s = (sx & 0x0F);
|
|
|
|
if(s == 0) return bx;
|
|
|
|
if(bs)
|
|
{
|
|
c = bf; /* previous sprite has higher priority */
|
|
}
|
|
else
|
|
{
|
|
c = (sp ? (s ? sf : bf) : \
|
|
(bp ? (b ? bf : (s ? sf : bf)) : \
|
|
(s ? sf : bf) ));
|
|
}
|
|
|
|
/* Strip palette bits from transparent pixels */
|
|
if((c & 0x0F) == 0x00) c &= 0xC0;
|
|
|
|
return (c | 0x80);
|
|
}
|
|
|
|
|
|
/* Input (bx): d5-d0=color, d6=priority, d7=unused */
|
|
/* Input (sx): d5-d0=color, d6=priority, d7=unused */
|
|
/* Output: d5-d0=color, d6=priority, d7=intensity select (half/normal) */
|
|
static uint32 make_lut_bg_ste(uint32 bx, uint32 ax)
|
|
{
|
|
int bf, bp, b;
|
|
int af, ap, a;
|
|
int gi;
|
|
int x = 0;
|
|
int c;
|
|
|
|
bf = (bx & 0x7F);
|
|
bp = (bx >> 6) & 1;
|
|
b = (bx & 0x0F);
|
|
|
|
af = (ax & 0x7F);
|
|
ap = (ax >> 6) & 1;
|
|
a = (ax & 0x0F);
|
|
|
|
gi = (ap | bp) ? 0x80 : 0x00;
|
|
|
|
c = (ap ? (a ? af : (b ? bf : x)) :
|
|
(bp ? (b ? bf : (a ? af : x)) : ((a ? af : (b ? bf : x)))));
|
|
|
|
c |= gi;
|
|
|
|
/* Strip palette bits from transparent pixels */
|
|
if((c & 0x0F) == 0x00) c &= 0xC0;
|
|
|
|
return (c);
|
|
}
|
|
|
|
|
|
/* Input (bx): d5-d0=color, d6=priority, d7=sprite pixel marker */
|
|
/* Input (sx): d5-d0=color, d6=priority, d7=unused */
|
|
/* Output: d5-d0=color, d6=priority, d7=sprite pixel marker */
|
|
static uint32 make_lut_obj_ste(uint32 bx, uint32 sx)
|
|
{
|
|
int bf, bs;
|
|
int sf;
|
|
int c;
|
|
|
|
bf = (bx & 0x7F);
|
|
bs = (bx >> 7) & 1;
|
|
sf = (sx & 0x7F);
|
|
|
|
if((sx & 0x0F) == 0) return bx;
|
|
|
|
c = (bs) ? bf : sf;
|
|
|
|
/* Strip palette bits from transparent pixels */
|
|
if((c & 0x0F) == 0x00) c &= 0xC0;
|
|
|
|
return (c | 0x80);
|
|
}
|
|
|
|
|
|
/* Input (bx): d5-d0=color, d6=priority, d7=intensity (half/normal) */
|
|
/* Input (sx): d5-d0=color, d6=priority, d7=sprite marker */
|
|
/* Output: d5-d0=color, d6=intensity (half/normal), d7=(double/invalid) */
|
|
static uint32 make_lut_bgobj_ste(uint32 bx, uint32 sx)
|
|
{
|
|
int c;
|
|
|
|
int bf = (bx & 0x3F);
|
|
int bp = (bx >> 6) & 1;
|
|
int bi = (bx & 0x80) ? 0x40 : 0x00;
|
|
int b = (bx & 0x0F);
|
|
|
|
int sf = (sx & 0x3F);
|
|
int sp = (sx >> 6) & 1;
|
|
int si = (sx & 0x40);
|
|
int s = (sx & 0x0F);
|
|
|
|
if(bi & 0x40) si |= 0x40;
|
|
|
|
if(sp)
|
|
{
|
|
if(s)
|
|
{
|
|
if((sf & 0x3E) == 0x3E)
|
|
{
|
|
if(sf & 1)
|
|
{
|
|
c = (bf | 0x00);
|
|
}
|
|
else
|
|
{
|
|
c = (bx & 0x80) ? (bf | 0x80) : (bf | 0x40);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if(sf == 0x0E || sf == 0x1E || sf == 0x2E)
|
|
{
|
|
c = (sf | 0x40);
|
|
}
|
|
else
|
|
{
|
|
c = (sf | si);
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
c = (bf | bi);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if(bp)
|
|
{
|
|
if(b)
|
|
{
|
|
c = (bf | bi);
|
|
}
|
|
else
|
|
{
|
|
if(s)
|
|
{
|
|
if((sf & 0x3E) == 0x3E)
|
|
{
|
|
if(sf & 1)
|
|
{
|
|
c = (bf | 0x00);
|
|
}
|
|
else
|
|
{
|
|
c = (bx & 0x80) ? (bf | 0x80) : (bf | 0x40);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if(sf == 0x0E || sf == 0x1E || sf == 0x2E)
|
|
{
|
|
c = (sf | 0x40);
|
|
}
|
|
else
|
|
{
|
|
c = (sf | si);
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
c = (bf | bi);
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if(s)
|
|
{
|
|
if((sf & 0x3E) == 0x3E)
|
|
{
|
|
if(sf & 1)
|
|
{
|
|
c = (bf | 0x00);
|
|
}
|
|
else
|
|
{
|
|
c = (bx & 0x80) ? (bf | 0x80) : (bf | 0x40);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if(sf == 0x0E || sf == 0x1E || sf == 0x2E)
|
|
{
|
|
c = (sf | 0x40);
|
|
}
|
|
else
|
|
{
|
|
c = (sf | si);
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
c = (bf | bi);
|
|
}
|
|
}
|
|
}
|
|
|
|
if((c & 0x0f) == 0x00) c &= 0xC0;
|
|
|
|
return (c);
|
|
}
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Color update functions */
|
|
/*--------------------------------------------------------------------------*/
|
|
#ifndef NGC
|
|
|
|
static void color_update_8(int index, uint16 data)
|
|
{
|
|
/* VDP Palette Selection bit */
|
|
/* color value is limited to 00X00X00X */
|
|
if (!(reg[0] & 4)) data &= 0x49;
|
|
|
|
if(reg[12] & 8)
|
|
{
|
|
pixel_8[0x00 | index] = pixel_8_lut[0][data];
|
|
pixel_8[0x40 | index] = pixel_8_lut[1][data];
|
|
pixel_8[0x80 | index] = pixel_8_lut[2][data];
|
|
}
|
|
else
|
|
{
|
|
uint8 temp = pixel_8_lut[1][data];
|
|
pixel_8[0x00 | index] = temp;
|
|
pixel_8[0x40 | index] = temp;
|
|
pixel_8[0x80 | index] = temp;
|
|
}
|
|
}
|
|
|
|
static void color_update_15(int index, uint16 data)
|
|
{
|
|
/* VDP Palette Selection bit */
|
|
/* color value is limited to 00X00X00X */
|
|
if (!(reg[0] & 4)) data &= 0x49;
|
|
|
|
if(reg[12] & 8)
|
|
{
|
|
pixel_15[0x00 | index] = pixel_15_lut[0][data];
|
|
pixel_15[0x40 | index] = pixel_15_lut[1][data];
|
|
pixel_15[0x80 | index] = pixel_15_lut[2][data];
|
|
}
|
|
else
|
|
{
|
|
uint16 temp = pixel_15_lut[1][data];
|
|
pixel_15[0x00 | index] = temp;
|
|
pixel_15[0x40 | index] = temp;
|
|
pixel_15[0x80 | index] = temp;
|
|
}
|
|
}
|
|
|
|
static void color_update_32(int index, uint16 data)
|
|
{
|
|
/* VDP Palette Selection bit */
|
|
/* color value is limited to 00X00X00X */
|
|
if (!(reg[0] & 4)) data &= 0x49;
|
|
|
|
if(reg[12] & 8)
|
|
{
|
|
pixel_32[0x00 | index] = pixel_32_lut[0][data];
|
|
pixel_32[0x40 | index] = pixel_32_lut[1][data];
|
|
pixel_32[0x80 | index] = pixel_32_lut[2][data];
|
|
}
|
|
else
|
|
{
|
|
uint32 temp = pixel_32_lut[1][data];
|
|
pixel_32[0x00 | index] = temp;
|
|
pixel_32[0x40 | index] = temp;
|
|
pixel_32[0x80 | index] = temp;
|
|
}
|
|
}
|
|
|
|
#endif
|
|
|
|
static void color_update_16(int index, uint16 data)
|
|
{
|
|
/* VDP Palette Selection bit */
|
|
/* color value is limited to 00X00X00X */
|
|
if (!(reg[0] & 4)) data &= 0x49;
|
|
|
|
if(reg[12] & 8)
|
|
{
|
|
pixel_16[0x00 | index] = pixel_16_lut[0][data];
|
|
pixel_16[0x40 | index] = pixel_16_lut[1][data];
|
|
pixel_16[0x80 | index] = pixel_16_lut[2][data];
|
|
}
|
|
else
|
|
{
|
|
uint16 temp = pixel_16_lut[1][data];
|
|
pixel_16[0x00 | index] = temp;
|
|
pixel_16[0x40 | index] = temp;
|
|
pixel_16[0x80 | index] = temp;
|
|
}
|
|
}
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Remap functions */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
#ifndef NGC
|
|
|
|
static inline void remap_8(uint8 *src, uint8 *dst, uint8 *table, int length)
|
|
{
|
|
int count;
|
|
for(count = 0; count < length; count += 1)
|
|
{
|
|
*dst++ = table[*src++];
|
|
}
|
|
}
|
|
|
|
static inline void remap_16(uint8 *src, uint16 *dst, uint16 *table, int length)
|
|
{
|
|
int count;
|
|
for(count = 0; count < length; count += 1)
|
|
{
|
|
*dst++ = table[*src++];
|
|
}
|
|
}
|
|
|
|
static inline void remap_32(uint8 *src, uint32 *dst, uint32 *table, int length)
|
|
{
|
|
int count;
|
|
for(count = 0; count < length; count += 1)
|
|
{
|
|
*dst++ = table[*src++];
|
|
}
|
|
}
|
|
|
|
#else
|
|
|
|
static inline void remap_texture(uint8 *src, uint16 *dst, uint32 tiles)
|
|
{
|
|
int count;
|
|
uint16 *table = pixel_16;
|
|
|
|
for(count = 0; count < tiles; count ++)
|
|
{
|
|
/* one tile is 4 pixels wide */
|
|
*dst++ = table[*src++];
|
|
*dst++ = table[*src++];
|
|
*dst++ = table[*src++];
|
|
*dst++ = table[*src++];
|
|
dst += 12;
|
|
}
|
|
}
|
|
#endif
|
|
|
|
|
|
static inline void merge(uint8 *srca, uint8 *srcb, uint8 *dst, uint8 *table, uint32 width)
|
|
{
|
|
int i;
|
|
for(i = 0; i < width; i += 1)
|
|
{
|
|
*dst++ = table[(*srcb++ << 8) | (*srca++)];
|
|
}
|
|
}
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Helper functions (cache update, hscroll, window clip) */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
static inline void update_bg_pattern_cache(uint32 index)
|
|
{
|
|
int i;
|
|
uint8 x, y, c;
|
|
uint16 name;
|
|
uint8 *dst;
|
|
uint32 bp;
|
|
#ifdef LSB_FIRST
|
|
uint8 shift_table[8] = {12, 8, 4, 0, 28, 24, 20, 16};
|
|
#else
|
|
uint8 shift_table[8] = {28, 24, 20, 16, 12, 8, 4, 0};
|
|
#endif
|
|
|
|
for(i = 0; i < index; i ++)
|
|
{
|
|
name = bg_name_list[i];
|
|
bg_name_list[i] = 0;
|
|
|
|
for(y = 0; y < 8; y ++)
|
|
{
|
|
if(bg_name_dirty[name] & (1 << y))
|
|
{
|
|
dst = &bg_pattern_cache[name << 6];
|
|
bp = *(uint32 *)&vram[(name << 5) | (y << 2)];
|
|
|
|
for(x = 0; x < 8; x ++)
|
|
{
|
|
c = (bp >> shift_table[x]) & 0x0F;
|
|
dst[0x00000 | (y << 3) | (x)] = (c); /* hf=0, vf=0: normal */
|
|
dst[0x20000 | (y << 3) | (x ^ 7)] = (c); /* hf=1, vf=0: horizontal flipped */
|
|
dst[0x40000 | ((y ^ 7) << 3) | (x)] = (c); /* hf=0, vf=1: vertical flipped */
|
|
dst[0x60000 | ((y ^ 7) << 3) | (x ^ 7)] = (c); /* hf=1, vf=1: horizontal & vertical flipped */
|
|
}
|
|
}
|
|
}
|
|
bg_name_dirty[name] = 0;
|
|
}
|
|
}
|
|
|
|
static inline uint32 get_hscroll(uint32 line)
|
|
{
|
|
switch(reg[11] & 3)
|
|
{
|
|
case 0: /* Full-screen */
|
|
return *(uint32 *)&vram[hscb];
|
|
|
|
case 1: /* First 8 lines */
|
|
return *(uint32 *)&vram[hscb + ((line & 7) << 2)];
|
|
|
|
case 2: /* Every 8 lines */
|
|
return *(uint32 *)&vram[hscb + ((line & ~7) << 2)];
|
|
|
|
default: /* Every line */
|
|
return *(uint32 *)&vram[hscb + (line << 2)];
|
|
}
|
|
}
|
|
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Layers render functions */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
static void render_bg(uint32 line, uint32 width)
|
|
{
|
|
uint32 column, atex, atbuf, *src, *dst;
|
|
|
|
/* common data */
|
|
uint32 xscroll = get_hscroll(line);
|
|
uint32 pf_col_mask = playfield_col_mask;
|
|
uint32 pf_row_mask = playfield_row_mask;
|
|
uint32 pf_shift = playfield_shift;
|
|
uint32 pf_y_mask = y_mask;
|
|
uint32 *vs = (uint32 *)&vsram[0];
|
|
|
|
/* B Plane */
|
|
uint8 *buf = ntb_buf;
|
|
uint32 start = 0;
|
|
uint32 end = width >> 4;
|
|
|
|
#ifdef LSB_FIRST
|
|
uint32 shift = (xscroll >> 16) & 0x0F;
|
|
uint32 index = pf_col_mask + 1 - ((xscroll >> 20) & pf_col_mask);
|
|
uint32 y_scroll = (line + ((vs[0] >> 16) & 0x3FF)) & pf_row_mask;
|
|
#else
|
|
uint32 shift = (xscroll & 0x0F);
|
|
uint32 index = pf_col_mask + 1 - ((xscroll >> 4) & pf_col_mask);
|
|
uint32 y_scroll = (line + (vs[0] & 0x3FF)) & pf_row_mask;
|
|
#endif
|
|
|
|
uint32 v_line = (y_scroll & 7) << 3;
|
|
uint32 *nt = (uint32 *)&vram[ntbb + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
|
|
if(shift)
|
|
{
|
|
dst = (uint32 *)&buf[0x10 + shift];
|
|
atbuf = nt[(index-1) & pf_col_mask];
|
|
DRAW_COLUMN(atbuf, v_line);
|
|
}
|
|
|
|
dst = (uint32 *)&buf[0x20 + shift];
|
|
for(column = 0; column < end; column ++, index ++)
|
|
{
|
|
atbuf = nt[index & pf_col_mask];
|
|
DRAW_COLUMN(atbuf, v_line)
|
|
}
|
|
|
|
/* Window and Plane A */
|
|
buf = nta_buf;
|
|
uint32 a = (reg[18] & 0x1F) << 3;
|
|
uint32 w = (reg[18] >> 7) & 1;
|
|
if (w == (line >= a))
|
|
{
|
|
/* Window takes up entire line */
|
|
a = 0;
|
|
w = 1;
|
|
}
|
|
else
|
|
{
|
|
/* Window and Plane A share the line */
|
|
a = clip[0].enable;
|
|
w = clip[1].enable;
|
|
}
|
|
|
|
/* Plane A */
|
|
if (a)
|
|
{
|
|
/* set for Plane A */
|
|
start = clip[0].left;
|
|
end = clip[0].right;
|
|
|
|
#ifdef LSB_FIRST
|
|
shift = (xscroll & 0x0F);
|
|
index = pf_col_mask + start + 1 - ((xscroll >> 4) & pf_col_mask);
|
|
y_scroll = (line + (vs[0] & 0x3FF)) & pf_row_mask;
|
|
#else
|
|
shift = (xscroll >> 16) & 0x0F;
|
|
index = pf_col_mask + start + 1 - ((xscroll >> 20) & pf_col_mask);
|
|
y_scroll = (line + ((vs[0] >> 16) & 0x3FF)) & pf_row_mask;
|
|
#endif
|
|
|
|
v_line = (y_scroll & 7) << 3;
|
|
nt = (uint32 *)&vram[ntab + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
|
|
if(shift)
|
|
{
|
|
dst = (uint32 *)&buf[0x10 + shift + (start<<4)];
|
|
|
|
/* Window bug */
|
|
if (start) atbuf = nt[index & pf_col_mask];
|
|
else atbuf = nt[(index-1) & pf_col_mask];
|
|
|
|
DRAW_COLUMN(atbuf, v_line);
|
|
}
|
|
|
|
dst = (uint32 *)&buf[0x20 + shift + (start<<4)];
|
|
for(column = start; column < end; column ++, index ++)
|
|
{
|
|
atbuf = nt[index & pf_col_mask];
|
|
DRAW_COLUMN(atbuf, v_line)
|
|
}
|
|
|
|
/* set for Window */
|
|
start = clip[1].left;
|
|
end = clip[1].right;
|
|
}
|
|
|
|
/* Window */
|
|
if (w)
|
|
{
|
|
v_line = (line & 7) << 3;
|
|
nt = (uint32 *)&vram[ntwb | ((line >> 3) << (6 + (reg[12] & 1)))];
|
|
dst = (uint32 *)&buf[0x20 + (start << 4)];
|
|
for(column = start; column < end; column ++)
|
|
{
|
|
atbuf = nt[column];
|
|
DRAW_COLUMN(atbuf, v_line)
|
|
}
|
|
}
|
|
}
|
|
|
|
static void render_bg_vs(uint32 line, uint32 width)
|
|
{
|
|
uint32 column, atex, atbuf, *src, *dst;
|
|
|
|
/* common data */
|
|
uint32 xscroll = get_hscroll(line);
|
|
uint32 pf_col_mask = playfield_col_mask;
|
|
uint32 pf_row_mask = playfield_row_mask;
|
|
uint32 pf_shift = playfield_shift;
|
|
uint32 pf_y_mask = y_mask;
|
|
uint32 *vs = (uint32 *)&vsram[0];
|
|
|
|
/* B Plane */
|
|
uint8 *buf = ntb_buf;
|
|
uint32 start = 0;
|
|
uint32 end = width >> 4;
|
|
|
|
#ifdef LSB_FIRST
|
|
uint32 shift = (xscroll >> 16) & 0x0F;
|
|
uint32 index = pf_col_mask + 1 - ((xscroll >> 20) & pf_col_mask);
|
|
#else
|
|
uint32 shift = (xscroll & 0x0F);
|
|
uint32 index = pf_col_mask + 1 - ((xscroll >> 4) & pf_col_mask);
|
|
#endif
|
|
|
|
uint32 y_scroll, v_line, *nt;
|
|
|
|
if(shift)
|
|
{
|
|
y_scroll = (line & pf_row_mask);
|
|
v_line = (y_scroll & 7) << 3;
|
|
nt = (uint32 *)&vram[ntbb + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
dst = (uint32 *)&buf[0x10 + shift];
|
|
atbuf = nt[(index-1) & pf_col_mask];
|
|
DRAW_COLUMN(atbuf, v_line);
|
|
}
|
|
|
|
dst = (uint32 *)&buf[0x20 + shift];
|
|
for(column = start; column < end; column ++, index ++)
|
|
{
|
|
#ifdef LSB_FIRST
|
|
y_scroll = (line + ((vs[column] >> 16) & 0x3FF)) & pf_row_mask;
|
|
#else
|
|
y_scroll = (line + (vs[column] & 0x3FF)) & pf_row_mask;
|
|
#endif
|
|
v_line = (y_scroll & 7) << 3;
|
|
nt = (uint32 *)&vram[ntbb + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
atbuf = nt[index & pf_col_mask];
|
|
DRAW_COLUMN(atbuf, v_line)
|
|
}
|
|
|
|
/* Window and Plane A */
|
|
buf = nta_buf;
|
|
uint32 a = (reg[18] & 0x1F) << 3;
|
|
uint32 w = (reg[18] >> 7) & 1;
|
|
if (w == (line >= a))
|
|
{
|
|
/* Window takes up entire line */
|
|
a = 0;
|
|
w = 1;
|
|
}
|
|
else
|
|
{
|
|
/* Window and Plane A share the line */
|
|
a = clip[0].enable;
|
|
w = clip[1].enable;
|
|
}
|
|
|
|
/* Plane A*/
|
|
if (a)
|
|
{
|
|
/* set for Plane A */
|
|
start = clip[0].left;
|
|
end = clip[0].right;
|
|
|
|
#ifdef LSB_FIRST
|
|
shift = (xscroll & 0x0F);
|
|
index = pf_col_mask + start + 1 - ((xscroll >> 4) & pf_col_mask);
|
|
#else
|
|
shift = (xscroll >> 16) & 0x0F;
|
|
index = pf_col_mask + start + 1 - ((xscroll >> 20) & pf_col_mask);
|
|
#endif
|
|
|
|
if(shift)
|
|
{
|
|
dst = (uint32 *)&buf[0x10 + shift + (start<<4)];
|
|
y_scroll = (line & pf_row_mask);
|
|
v_line = (y_scroll & 7) << 3;
|
|
nt = (uint32 *)&vram[ntab + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
|
|
/* Window bug */
|
|
if (start) atbuf = nt[index & pf_col_mask];
|
|
else atbuf = nt[(index-1) & pf_col_mask];
|
|
|
|
DRAW_COLUMN(atbuf, v_line);
|
|
}
|
|
|
|
dst = (uint32 *)&buf[0x20 + shift + (start<<4)];
|
|
for(column = start; column < end; column ++, index ++)
|
|
{
|
|
#ifdef LSB_FIRST
|
|
y_scroll = (line + (vs[column] & 0x3FF)) & pf_row_mask;
|
|
#else
|
|
y_scroll = (line + ((vs[column] >> 16) & 0x3FF)) & pf_row_mask;
|
|
#endif
|
|
v_line = (y_scroll & 7) << 3;
|
|
nt = (uint32 *)&vram[ntab + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
atbuf = nt[index & pf_col_mask];
|
|
DRAW_COLUMN(atbuf, v_line)
|
|
}
|
|
|
|
/* set for Window */
|
|
start = clip[1].left;
|
|
end = clip[1].right;
|
|
}
|
|
|
|
/* Window */
|
|
if (w)
|
|
{
|
|
v_line = (line & 7) << 3;
|
|
nt = (uint32 *)&vram[ntwb | ((line >> 3) << (6 + (reg[12] & 1)))];
|
|
dst = (uint32 *)&buf[0x20 + (start << 4)];
|
|
for(column = start; column < end; column ++)
|
|
{
|
|
atbuf = nt[column];
|
|
DRAW_COLUMN(atbuf, v_line)
|
|
}
|
|
}
|
|
}
|
|
|
|
static void render_bg_im2(uint32 line, uint32 width, uint32 odd)
|
|
{
|
|
uint32 column, atex, atbuf, offs, *src, *dst;
|
|
|
|
/* common data */
|
|
uint32 xscroll = get_hscroll(line);
|
|
uint32 pf_col_mask = playfield_col_mask;
|
|
uint32 pf_row_mask = playfield_row_mask;
|
|
uint32 pf_shift = playfield_shift;
|
|
uint32 pf_y_mask = y_mask;
|
|
uint32 *vs = (uint32 *)&vsram[0];
|
|
|
|
/* B Plane */
|
|
uint8 *buf = ntb_buf;
|
|
uint32 start = 0;
|
|
uint32 end = width >> 4;
|
|
|
|
#ifdef LSB_FIRST
|
|
uint32 shift = (xscroll >> 16) & 0x0F;
|
|
uint32 index = pf_col_mask + 1 - ((xscroll >> 20) & pf_col_mask);
|
|
uint32 y_scroll = (line + ((vs[0] >> 17) & 0x3FF)) & pf_row_mask; /* IM2 specific */
|
|
#else
|
|
uint32 shift = (xscroll & 0x0F);
|
|
uint32 index = pf_col_mask + 1 - ((xscroll >> 4) & pf_col_mask);
|
|
uint32 y_scroll = (line + ((vs[0] >> 1) & 0x3FF)) & pf_row_mask; /* IM2 specific */
|
|
#endif
|
|
|
|
uint32 v_line = (((y_scroll & 7) << 1) | odd) << 3; /* IM2 specific */
|
|
uint32 *nt = (uint32 *)&vram[ntbb + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
|
|
if(shift)
|
|
{
|
|
dst = (uint32 *)&buf[0x10 + shift];
|
|
atbuf = nt[(index-1) & pf_col_mask];
|
|
DRAW_COLUMN_IM2(atbuf, v_line); /* IM2 specific */
|
|
}
|
|
|
|
dst = (uint32 *)&buf[0x20 + shift];
|
|
for(column = 0; column < end; column ++, index ++)
|
|
{
|
|
atbuf = nt[index & pf_col_mask];
|
|
DRAW_COLUMN_IM2(atbuf, v_line) /* IM2 specific */
|
|
}
|
|
|
|
/* Window and Plane A */
|
|
buf = nta_buf;
|
|
uint32 a = (reg[18] & 0x1F) << 3;
|
|
uint32 w = (reg[18] >> 7) & 1;
|
|
if (w == (line >= a))
|
|
{
|
|
/* Window takes up entire line */
|
|
a = 0;
|
|
w = 1;
|
|
}
|
|
else
|
|
{
|
|
/* Window and Plane A share the line */
|
|
a = clip[0].enable;
|
|
w = clip[1].enable;
|
|
}
|
|
|
|
/* Plane A */
|
|
if (a)
|
|
{
|
|
/* set for Plane A */
|
|
start = clip[0].left;
|
|
end = clip[0].right;
|
|
|
|
#ifdef LSB_FIRST
|
|
shift = (xscroll & 0x0F);
|
|
index = pf_col_mask + start + 1 - ((xscroll >> 4) & pf_col_mask);
|
|
y_scroll = (line + ((vs[0] >> 1) & 0x3FF)) & pf_row_mask; /* IM2 specific */
|
|
#else
|
|
shift = (xscroll >> 16) & 0x0F;
|
|
index = pf_col_mask + start + 1 - ((xscroll >> 20) & pf_col_mask);
|
|
y_scroll = (line + ((vs[0] >> 17) & 0x3FF)) & pf_row_mask; /* IM2 specific */
|
|
#endif
|
|
|
|
v_line = (((y_scroll & 7) << 1) | odd) << 3; /* IM2 specific */
|
|
nt = (uint32 *)&vram[ntab + (((y_scroll >> 3) << pf_shift) & pf_y_mask)];
|
|
|
|
if(shift)
|
|
{
|
|
dst = (uint32 *)&buf[0x10 + shift + (start<<4)];
|
|
|
|
/* Window bug */
|
|
if (start) atbuf = nt[index & pf_col_mask];
|
|
else atbuf = nt[(index-1) & pf_col_mask];
|
|
|
|
DRAW_COLUMN_IM2(atbuf, v_line); /* IM2 specific */
|
|
}
|
|
|
|
dst = (uint32 *)&buf[0x20 + shift + (start<<4)];
|
|
for(column = start; column < end; column ++, index ++)
|
|
{
|
|
atbuf = nt[index & pf_col_mask];
|
|
DRAW_COLUMN_IM2(atbuf, v_line) /* IM2 specific */
|
|
}
|
|
|
|
/* set for Window */
|
|
start = clip[1].left;
|
|
end = clip[1].right;
|
|
}
|
|
|
|
/* Window */
|
|
if (w)
|
|
{
|
|
v_line = ((line & 7) << 1 | odd) << 3; /* IM2 specific */
|
|
nt = (uint32 *)&vram[ntwb | ((line >> 3) << (6 + (reg[12] & 1)))];
|
|
dst = (uint32 *)&buf[0x20 + (start << 4)];
|
|
for(column = start; column < end; column ++)
|
|
{
|
|
atbuf = nt[column];
|
|
DRAW_COLUMN_IM2(atbuf, v_line) /* IM2 specific */
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Object render functions */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
static int spr_over = 0;
|
|
|
|
static void render_obj(uint32 line, uint8 *buf, uint8 *table)
|
|
{
|
|
uint16 ypos;
|
|
uint16 attr;
|
|
uint16 xpos;
|
|
uint8 sizetab[] = {8, 16, 24, 32};
|
|
uint8 size;
|
|
uint8 *src;
|
|
|
|
int count,i;
|
|
int pixelcount = 0;
|
|
int width;
|
|
int height;
|
|
int v_line;
|
|
int column;
|
|
int left = 0x80;
|
|
int right = 0x80 + bitmap.viewport.w;
|
|
|
|
uint8 *s, *lb;
|
|
uint16 name, index;
|
|
uint8 palette;
|
|
|
|
int attr_mask, nt_row;
|
|
int mask = 0;
|
|
|
|
for(count = 0; count < object_index_count; count += 1)
|
|
{
|
|
xpos = object_info[count].xpos & 0x1ff;
|
|
|
|
/* sprite masking (ignore the 1st sprite) */
|
|
if (xpos) spr_over = 1;
|
|
else if (spr_over)
|
|
{
|
|
spr_over = 0;
|
|
mask = 1;
|
|
}
|
|
|
|
size = object_info[count].size & 0x0f;
|
|
width = sizetab[(size >> 2) & 3];
|
|
|
|
/* update pixel count (off-screen sprites included) */
|
|
pixelcount += xpos ? width : (width * 2);
|
|
|
|
if(((xpos + width) >= left) && (xpos < right) && !mask)
|
|
{
|
|
ypos = object_info[count].ypos;
|
|
attr = object_info[count].attr;
|
|
attr_mask = (attr & 0x1800);
|
|
|
|
height = sizetab[size & 3];
|
|
palette = (attr >> 9) & 0x70;
|
|
|
|
v_line = (line - ypos);
|
|
nt_row = (v_line >> 3) & 3;
|
|
v_line = (v_line & 7) << 3;
|
|
|
|
name = (attr & 0x07FF);
|
|
s = &name_lut[((attr >> 3) & 0x300) | (size << 4) | (nt_row << 2)];
|
|
|
|
lb = (uint8 *)&buf[0x20 + (xpos - 0x80)];
|
|
|
|
/* number of tiles to draw */
|
|
/* adjusted for sprite limit */
|
|
if (pixelcount > bitmap.viewport.w)
|
|
{
|
|
width -= (pixelcount - bitmap.viewport.w);
|
|
}
|
|
|
|
width >>= 3;
|
|
|
|
for(column = 0; column < width; column += 1, lb+=8)
|
|
{
|
|
index = attr_mask | ((name + s[column]) & 0x07FF);
|
|
src = &bg_pattern_cache[(index << 6) | (v_line)];
|
|
DRAW_SPRITE_TILE;
|
|
}
|
|
}
|
|
|
|
/* sprite limit (256 or 320 pixels) */
|
|
if (pixelcount >= bitmap.viewport.w)
|
|
{
|
|
spr_over = 1;
|
|
return;
|
|
}
|
|
}
|
|
|
|
spr_over = 0;
|
|
}
|
|
|
|
static void render_obj_im2(uint32 line, uint32 odd, uint8 *buf, uint8 *table)
|
|
{
|
|
uint16 ypos;
|
|
uint16 attr;
|
|
uint16 xpos;
|
|
uint8 sizetab[] = {8, 16, 24, 32};
|
|
uint8 size;
|
|
uint8 *src;
|
|
|
|
int count,i;
|
|
int pixelcount = 0;
|
|
int width;
|
|
int height;
|
|
int v_line;
|
|
int column;
|
|
int left = 0x80;
|
|
int right = 0x80 + bitmap.viewport.w;
|
|
|
|
uint8 *s, *lb;
|
|
uint16 name, index;
|
|
uint8 palette;
|
|
uint32 offs;
|
|
|
|
int attr_mask, nt_row;
|
|
int mask = 0;
|
|
|
|
for(count = 0; count < object_index_count; count += 1)
|
|
{
|
|
xpos = object_info[count].xpos & 0x1ff;
|
|
|
|
/* sprite masking (ignore the 1st sprite) */
|
|
if (xpos) spr_over = 1;
|
|
else if(spr_over)
|
|
{
|
|
spr_over = 0;
|
|
mask = 1;
|
|
}
|
|
|
|
size = object_info[count].size & 0x0f;
|
|
width = sizetab[(size >> 2) & 3];
|
|
|
|
/* update pixel count (off-screen sprites included) */
|
|
pixelcount += xpos ? width : (width * 2);
|
|
|
|
if(((xpos + width) >= left) && (xpos < right) && !mask)
|
|
{
|
|
ypos = object_info[count].ypos;
|
|
attr = object_info[count].attr;
|
|
attr_mask = (attr & 0x1800);
|
|
|
|
height = sizetab[size & 3];
|
|
palette = (attr >> 9) & 0x70;
|
|
|
|
v_line = (line - ypos);
|
|
nt_row = (v_line >> 3) & 3;
|
|
v_line = (((v_line & 7) << 1) | odd) << 3;
|
|
|
|
name = (attr & 0x03FF);
|
|
s = &name_lut[((attr >> 3) & 0x300) | (size << 4) | (nt_row << 2)];
|
|
|
|
lb = (uint8 *)&buf[0x20 + (xpos - 0x80)];
|
|
|
|
/* number of tiles to draw */
|
|
/* adjusted for sprite limit */
|
|
if (pixelcount > bitmap.viewport.w) width -= (pixelcount - bitmap.viewport.w);
|
|
width >>= 3;
|
|
|
|
for(column = 0; column < width; column += 1, lb+=8)
|
|
{
|
|
index = (name + s[column]) & 0x3ff;
|
|
offs = index << 7 | attr_mask << 6 | v_line;
|
|
if(attr & 0x1000) offs ^= 0x40;
|
|
src = &bg_pattern_cache[offs];
|
|
DRAW_SPRITE_TILE;
|
|
}
|
|
}
|
|
|
|
/* sprite limit (256 or 320 pixels) */
|
|
if (pixelcount >= bitmap.viewport.w)
|
|
{
|
|
spr_over = 1;
|
|
return;
|
|
}
|
|
}
|
|
|
|
spr_over = 0;
|
|
}
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Init, reset, shutdown routines */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
void render_init(void)
|
|
{
|
|
int bx, ax, i;
|
|
|
|
/* Allocate and align pixel look-up tables */
|
|
if (lut_base == NULL) lut_base = malloc ((LUT_MAX * LUT_SIZE) + LUT_SIZE);
|
|
lut[0] = (uint8 *) (((uint32) lut_base + LUT_SIZE) & ~(LUT_SIZE - 1));
|
|
for (i = 1; i < LUT_MAX; i += 1) lut[i] = lut[0] + (i * LUT_SIZE);
|
|
|
|
/* Make pixel look-up table data */
|
|
for (bx = 0; bx < 0x100; bx += 1)
|
|
for (ax = 0; ax < 0x100; ax += 1)
|
|
{
|
|
uint16 index = (bx << 8) | (ax);
|
|
lut[0][index] = make_lut_bg (bx, ax);
|
|
lut[1][index] = make_lut_obj (bx, ax);
|
|
lut[2][index] = make_lut_bg_ste (bx, ax);
|
|
lut[3][index] = make_lut_obj_ste (bx, ax);
|
|
lut[4][index] = make_lut_bgobj_ste (bx, ax);
|
|
}
|
|
|
|
/* Make pixel data tables */
|
|
palette_init();
|
|
|
|
/* Set up color update function */
|
|
#ifndef NGC
|
|
switch(bitmap.depth)
|
|
{
|
|
case 8: color_update = color_update_8; break;
|
|
case 15: color_update = color_update_15; break;
|
|
case 16: color_update = color_update_16; break;
|
|
case 32: color_update = color_update_32; break;
|
|
}
|
|
#else
|
|
color_update = color_update_16;
|
|
#endif
|
|
|
|
/* Make sprite name look-up table */
|
|
make_name_lut();
|
|
}
|
|
|
|
void render_reset(void)
|
|
{
|
|
/* Clear display bitmap */
|
|
memset(bitmap.data, 0, bitmap.pitch * bitmap.height);
|
|
|
|
memset(bg_buf, 0, sizeof(bg_buf));
|
|
memset(tmp_buf, 0, sizeof(tmp_buf));
|
|
memset(nta_buf, 0, sizeof(nta_buf));
|
|
memset(ntb_buf, 0, sizeof(ntb_buf));
|
|
memset(obj_buf, 0, sizeof(obj_buf));
|
|
|
|
#ifndef NGC
|
|
memset(&pixel_8, 0, sizeof(pixel_8));
|
|
memset(&pixel_15, 0, sizeof(pixel_15));
|
|
memset(&pixel_32, 0, sizeof(pixel_32));
|
|
#endif
|
|
memset(&pixel_16, 0, sizeof(pixel_16));
|
|
|
|
window_clip();
|
|
}
|
|
|
|
|
|
void render_shutdown(void)
|
|
{
|
|
if(lut_base) free(lut_base);
|
|
}
|
|
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Line render function */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
void render_line(uint32 line, uint32 overscan)
|
|
{
|
|
uint32 width = bitmap.viewport.w;
|
|
uint32 x_offset = bitmap.viewport.x;
|
|
|
|
/* display OFF */
|
|
if (reg[0] & 0x01) return;
|
|
|
|
/* background color (blanked display or vertical borders) */
|
|
if (!(reg[1] & 0x40) || overscan)
|
|
{
|
|
width += 2 * x_offset;
|
|
memset(&tmp_buf[0x20 - x_offset], 0x40, width);
|
|
}
|
|
else
|
|
{
|
|
uint8 *lb = tmp_buf;
|
|
|
|
/* update pattern generator */
|
|
if (bg_list_index)
|
|
{
|
|
update_bg_pattern_cache(bg_list_index);
|
|
bg_list_index = 0;
|
|
}
|
|
|
|
/* double-resolution mode */
|
|
if(im2_flag)
|
|
{
|
|
uint32 odd = odd_frame;
|
|
|
|
/* render BG layers */
|
|
render_bg_im2(line, width, odd);
|
|
|
|
if (reg[12] & 8)
|
|
{
|
|
/* Shadow & Highlight */
|
|
merge(&nta_buf[0x20], &ntb_buf[0x20], &bg_buf[0x20], lut[2], width);
|
|
memset(&obj_buf[0x20], 0, width);
|
|
if (object_index_count) render_obj_im2(line, odd, obj_buf, lut[3]);
|
|
merge(&obj_buf[0x20], &bg_buf[0x20], &lb[0x20], lut[4], width);
|
|
}
|
|
else
|
|
{
|
|
merge(&nta_buf[0x20], &ntb_buf[0x20], &lb[0x20], lut[0], width);
|
|
if (object_index_count) render_obj_im2(line, odd, lb, lut[1]);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
/* render BG layers */
|
|
if(reg[11] & 4) render_bg_vs(line, width);
|
|
else render_bg(line, width);
|
|
|
|
if(reg[12] & 8)
|
|
{
|
|
/* Shadow & Highlight */
|
|
merge(&nta_buf[0x20], &ntb_buf[0x20], &bg_buf[0x20], lut[2], width);
|
|
memset(&obj_buf[0x20], 0, width);
|
|
render_obj(line, obj_buf, lut[3]);
|
|
merge(&obj_buf[0x20], &bg_buf[0x20], &lb[0x20], lut[4], width);
|
|
}
|
|
else
|
|
{
|
|
merge(&nta_buf[0x20], &ntb_buf[0x20], &lb[0x20], lut[0], width);
|
|
render_obj(line, lb, lut[1]);
|
|
}
|
|
}
|
|
|
|
/* Mode 4 feature only (unemulated, no games rely on this) */
|
|
/*if(!(reg[1] & 0x04) && (reg[0] & 0x20)) memset(&lb[0x20], 0x40, 0x08);*/
|
|
|
|
/* borders */
|
|
if (x_offset)
|
|
{
|
|
memset(&lb[0x20 - x_offset], 0x40, x_offset);
|
|
memset(&lb[0x20 + width], 0x40, x_offset);
|
|
width += 2 * x_offset;
|
|
}
|
|
}
|
|
|
|
/* pixel color remapping */
|
|
remap_buffer(line,width);
|
|
}
|
|
|
|
void remap_buffer(uint32 line, uint32 width)
|
|
{
|
|
/* get line offset from framebuffer */
|
|
line = (line + bitmap.viewport.y) % lines_per_frame;
|
|
|
|
/* double resolution mode */
|
|
if (config.render && interlaced) line = (line * 2) + odd_frame;
|
|
|
|
/* NTSC Filter */
|
|
if (config.ntsc)
|
|
{
|
|
if (reg[12]&1) md_ntsc_blit(&md_ntsc, ( MD_NTSC_IN_T const * )pixel_16, tmp_buf+0x20-bitmap.viewport.x, width, line);
|
|
else sms_ntsc_blit(&sms_ntsc, ( SMS_NTSC_IN_T const * )pixel_16, tmp_buf+0x20-bitmap.viewport.x, width, line);
|
|
return;
|
|
}
|
|
|
|
#ifdef NGC
|
|
/* directly fill the RGB565 texture */
|
|
/* one tile is 32 byte = 4x4 pixels */
|
|
/* tiles are stored continuously in texture memory */
|
|
width = width >> 2;
|
|
int offset = ((width << 5) * (line >> 2)) + ((line & 3) * 8);
|
|
remap_texture(tmp_buf+0x20-bitmap.viewport.x, (uint16 *)(texturemem + offset), width);
|
|
|
|
#else
|
|
void *out =((void *)&bitmap.data[(line * bitmap.pitch)]);
|
|
switch(bitmap.depth)
|
|
{
|
|
case 8:
|
|
remap_8(tmp_buf+0x20-bitmap.viewport.x, (uint8 *)out, pixel_8, width);
|
|
break;
|
|
case 15:
|
|
remap_16(tmp_buf+0x20-bitmap.viewport.x, (uint16 *)out, pixel_15, width);
|
|
break;
|
|
case 16:
|
|
remap_16(tmp_buf+0x20-bitmap.viewport.x, (uint16 *)out, pixel_16, width);
|
|
break;
|
|
case 32:
|
|
remap_32(tmp_buf+0x20-bitmap.viewport.x, (uint32 *)out, pixel_32, width);
|
|
break;
|
|
}
|
|
#endif
|
|
}
|
|
|
|
/* Update Window Clipping (now only called on VDP registers change) */
|
|
void window_clip(void)
|
|
{
|
|
/* Window size and invert flags */
|
|
int hp = (reg[17] & 0x1f);
|
|
int hf = (reg[17] >> 7) & 1;
|
|
|
|
/* Display size */
|
|
int sw = bitmap.viewport.w >> 4;
|
|
|
|
/* Clear clipping data */
|
|
memset(&clip, 0, sizeof(clip));
|
|
|
|
/* Perform horizontal clipping; the results are applied in reverse
|
|
if the horizontal inversion flag is set */
|
|
int a = hf;
|
|
int w = hf ^ 1;
|
|
|
|
if(hp)
|
|
{
|
|
if(hp > sw)
|
|
{
|
|
/* Plane W takes up entire line */
|
|
clip[w].right = sw;
|
|
clip[w].enable = 1;
|
|
}
|
|
else
|
|
{
|
|
/* Window takes left side, Plane A takes right side */
|
|
clip[w].right = hp;
|
|
clip[a].left = hp;
|
|
clip[a].right = sw;
|
|
clip[0].enable = clip[1].enable = 1;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
/* Plane A takes up entire line */
|
|
clip[a].right = sw;
|
|
clip[a].enable = 1;
|
|
}
|
|
}
|
|
|
|
/*--------------------------------------------------------------------------*/
|
|
/* Sprites Parsing function */
|
|
/*--------------------------------------------------------------------------*/
|
|
|
|
void parse_satb(uint32 line)
|
|
{
|
|
uint8 sizetab[] = {8, 16, 24, 32};
|
|
uint32 link = 0;
|
|
uint32 count, ypos, size, height;
|
|
|
|
uint32 limit = (reg[12] & 1) ? 20 : 16;
|
|
uint32 total = limit << 2;
|
|
|
|
uint16 *p = (uint16 *) &vram[satb];
|
|
uint16 *q = (uint16 *) &sat[0];
|
|
|
|
object_index_count = 0;
|
|
|
|
for(count = 0; count < total; count += 1)
|
|
{
|
|
ypos = (q[link] >> im2_flag) & 0x1FF;
|
|
size = q[link + 1] >> 8;
|
|
height = sizetab[size & 3];
|
|
|
|
if((line >= ypos) && (line < (ypos + height)))
|
|
{
|
|
/* sprite limit (max. 16 or 20 sprites displayed per line) */
|
|
if(object_index_count == limit)
|
|
{
|
|
if(vint_pending == 0) status |= 0x40;
|
|
return;
|
|
}
|
|
|
|
// using xpos from internal satb stops sprite x
|
|
// scrolling in bloodlin.bin,
|
|
// but this seems to go against the test prog
|
|
object_info[object_index_count].attr = p[link + 2];
|
|
object_info[object_index_count].xpos = p[link + 3];
|
|
object_info[object_index_count].ypos = ypos;
|
|
object_info[object_index_count].size = size;
|
|
++object_index_count;
|
|
}
|
|
|
|
link = (q[link + 1] & 0x7F) << 2;
|
|
if(link == 0) break;
|
|
}
|
|
}
|