Lower memory usage for printing buffer by 1/4th
Fixes https://github.com/jeffman/Mother2GbaTranslation/issues/131 The loading process now takes 1/10th more time
This commit is contained in:
Lorenzo Carletti
parent
188ac3ae75
commit
0e25bdf8da
109
src/c/vwf.c
109
src/c/vwf.c
|
|
@ -71,10 +71,9 @@ byte reduce_bit_depth(int row, int foregroundRow)
|
|||
|
||||
//The order is swapped in order to make this faster
|
||||
//Doing the bottom tile directly saves some cycles
|
||||
void reduce_bit_depth_sp(int* TileRows, int* bufferValues)
|
||||
void reduce_bit_depth_sp(int* TileRows, int* bufferValues, int* bottomBufferValues)
|
||||
{
|
||||
int* bottomTileRows = TileRows + (0x20 * 8);
|
||||
int* bottomBufferValues = bufferValues + 0x40;
|
||||
const int andValue = 0x11111111;
|
||||
|
||||
//First value
|
||||
|
|
@ -478,6 +477,19 @@ byte print_character_with_callback(byte chr, int x, int y, int font, int foregro
|
|||
return virtualWidth;
|
||||
}
|
||||
|
||||
int convert_tile_address_buffer(int tile)
|
||||
{
|
||||
//Allows using tiles with different size to make the buffer smaller
|
||||
tile -= *tile_offset;
|
||||
int inner_line = tile & 0x3F;
|
||||
int lines = (tile >> 6) << 5;
|
||||
int returned_address = ((*tile_offset) * TILESET_OFFSET_BUFFER_MULTIPLIER) + (lines * DEFAULT_DOUBLE_TILE_HEIGHT);
|
||||
|
||||
if(inner_line >= 0x20)
|
||||
return returned_address + (0x20 * 8) + ((inner_line & 0x1F) * (DEFAULT_DOUBLE_TILE_HEIGHT - 8));
|
||||
return returned_address + (inner_line * 8);
|
||||
}
|
||||
|
||||
byte print_character_with_callback_1bpp_buffer(byte chr, int x, int y, byte *dest, int (*getTileCallback)(int, int), int font,
|
||||
unsigned short *tilemapPtr, int tilemapWidth, byte doubleTileHeight)
|
||||
{
|
||||
|
|
@ -525,7 +537,7 @@ byte print_character_with_callback_1bpp_buffer(byte chr, int x, int y, byte *des
|
|||
|
||||
for (int row = 0; row < 8; row++)
|
||||
{
|
||||
unsigned short canvasRow = dest[(realTileIndex * 8) + ((row + offsetY - sumRemoved) & 7)] + (dest[(realTileIndexRight * 8) + ((row + offsetY - sumRemoved) & 7)] << 8);
|
||||
unsigned short canvasRow = dest[convert_tile_address_buffer(realTileIndex) + ((row + offsetY - sumRemoved) & 7)] + (dest[convert_tile_address_buffer(realTileIndexRight) + ((row + offsetY - sumRemoved) & 7)] << 8);
|
||||
unsigned short glyphRow = glyphRows[row + (dTileY * 8 * tileWidth) + (dTileX * 8)] << offsetX;
|
||||
|
||||
unsigned short tmpCanvasRow = canvasRow;
|
||||
|
|
@ -534,8 +546,8 @@ byte print_character_with_callback_1bpp_buffer(byte chr, int x, int y, byte *des
|
|||
if(!availableSwap && ((row + offsetY) >> 3) == 1 && canvasRow != tmpCanvasRow) //This changed the canvas, then it's useful... IF it's the extra vertical tile
|
||||
useful = true;
|
||||
|
||||
dest[(realTileIndex * 8) + ((row + offsetY - sumRemoved) & 7)] = canvasRow;
|
||||
dest[(realTileIndexRight * 8) + ((row + offsetY - sumRemoved) & 7)] = canvasRow >> 8;
|
||||
dest[convert_tile_address_buffer(realTileIndex) + ((row + offsetY - sumRemoved) & 7)] = canvasRow;
|
||||
dest[convert_tile_address_buffer(realTileIndexRight) + ((row + offsetY - sumRemoved) & 7)] = canvasRow >> 8;
|
||||
|
||||
if((row + offsetY - sumRemoved) == limit)
|
||||
{
|
||||
|
|
@ -1738,8 +1750,9 @@ void clear_tile_buffer(int x, int y, byte* dest)
|
|||
{
|
||||
// Clear pixels
|
||||
int tileIndex = get_tile_number_buffer(x, y) + *tile_offset;
|
||||
int *destTileInt = (int*)&dest[(tileIndex) * 8];
|
||||
int *destTileInt = (int*)&dest[convert_tile_address_buffer(tileIndex)];
|
||||
destTileInt[0] = 0;
|
||||
if(((tileIndex - *tile_offset) & 0x20) == 0)
|
||||
destTileInt[1] = 0;
|
||||
|
||||
// Reset the tilemap (e.g. get rid of equip or SMAAAASH!! tiles)
|
||||
|
|
@ -1939,6 +1952,27 @@ void printCashWindow()
|
|||
m2_sub_d3c50();
|
||||
}
|
||||
|
||||
void eb_cartridge_palette_change(bool background)
|
||||
{
|
||||
unsigned short *paletteDest = (unsigned short*)0x5000040;
|
||||
if(background)
|
||||
{
|
||||
if(BUILD_PALETTE)
|
||||
{
|
||||
//Makes the game do the palette work. Copy the result in a bin file and use that instead in order to make the swap fast
|
||||
unsigned short palettes[0x50];
|
||||
cpuset(paletteDest, palettes, 0x50);
|
||||
for(int i = 0; i < 5; i++)
|
||||
m12_dim_palette(&palettes[i * 0x10], 0x10, 0x800);
|
||||
cpuset(palettes, paletteDest, 0x50);
|
||||
}
|
||||
else
|
||||
cpuset(m12_cartridge_palettes_dimmed, paletteDest, 0x50);
|
||||
}
|
||||
else
|
||||
cpuset(&m12_cartridge_palettes[0x20], paletteDest, 0x50);
|
||||
}
|
||||
|
||||
// x, y, width: tile coordinates
|
||||
void print_blankstr_buffer(int x, int y, int width, byte *dest)
|
||||
{
|
||||
|
|
@ -2008,6 +2042,7 @@ void load_pixels_overworld_buffer()
|
|||
int tile = *tile_offset;
|
||||
byte* buffer = (byte*)(OVERWORLD_BUFFER - (tile * TILESET_OFFSET_BUFFER_MULTIPLIER));
|
||||
int* topBufferValues = (int*)(&buffer[tile * 8]);
|
||||
int* bottomBufferValues = topBufferValues + 0x40;
|
||||
int* topTilePointer;
|
||||
int nextValue = 0x20;
|
||||
int i = 0;
|
||||
|
|
@ -2021,24 +2056,28 @@ void load_pixels_overworld_buffer()
|
|||
if(i == nextValue)
|
||||
{
|
||||
nextValue += 0x20;
|
||||
topBufferValues += 0x40;
|
||||
topBufferValues = bottomBufferValues;
|
||||
bottomBufferValues += 0x40;
|
||||
}
|
||||
i++;
|
||||
//Using "reduce_bit_depth_sp" reduced the total amount of cycles from 300k to 162k
|
||||
reduce_bit_depth_sp(topTilePointer, topBufferValues);
|
||||
reduce_bit_depth_sp(topTilePointer, topBufferValues, bottomBufferValues);
|
||||
topTilePointer += 8;
|
||||
topBufferValues += 2;
|
||||
bottomBufferValues++;
|
||||
while(remainingTiles > 0)
|
||||
{
|
||||
if(i == nextValue)
|
||||
{
|
||||
nextValue += 0x20;
|
||||
topBufferValues += 0x40;
|
||||
topBufferValues = bottomBufferValues;
|
||||
bottomBufferValues += 0x40;
|
||||
}
|
||||
i++;
|
||||
reduce_bit_depth_sp(topTilePointer, topBufferValues);
|
||||
reduce_bit_depth_sp(topTilePointer, topBufferValues, bottomBufferValues);
|
||||
topTilePointer += 8;
|
||||
topBufferValues += 2;
|
||||
bottomBufferValues++;
|
||||
remainingTiles--;
|
||||
}
|
||||
}
|
||||
|
|
@ -2078,7 +2117,7 @@ void store_pixels_overworld_buffer(int totalYs)
|
|||
if(i == nextValue)
|
||||
{
|
||||
nextValue += 0x20;
|
||||
topBufferValues += 0x40;
|
||||
topBufferValues += 0x20;
|
||||
bottomBufferValues += 0x40;
|
||||
}
|
||||
i++;
|
||||
|
|
@ -2099,7 +2138,6 @@ void store_pixels_overworld_buffer(int totalYs)
|
|||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x10) & 0xFF];
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x18) & 0xFF];
|
||||
//Since those are unused
|
||||
bottomBufferValues++;
|
||||
bottomTilePointer += 4;
|
||||
/* The game doesn't use these
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(second_half >> 0) & 0xFF];
|
||||
|
|
@ -2113,7 +2151,7 @@ void store_pixels_overworld_buffer(int totalYs)
|
|||
if(i == nextValue)
|
||||
{
|
||||
nextValue += 0x20;
|
||||
topBufferValues += 0x40;
|
||||
topBufferValues += 0x20;
|
||||
bottomBufferValues += 0x40;
|
||||
}
|
||||
i++;
|
||||
|
|
@ -2134,7 +2172,6 @@ void store_pixels_overworld_buffer(int totalYs)
|
|||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x10) & 0xFF];
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x18) & 0xFF];
|
||||
//Since those are unused
|
||||
bottomBufferValues++;
|
||||
bottomTilePointer += 4;
|
||||
/* The game doesn't use these
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(second_half >> 0) & 0xFF];
|
||||
|
|
@ -2147,27 +2184,6 @@ void store_pixels_overworld_buffer(int totalYs)
|
|||
}
|
||||
}
|
||||
|
||||
void eb_cartridge_palette_change(bool background)
|
||||
{
|
||||
unsigned short *paletteDest = (unsigned short*)0x5000040;
|
||||
if(background)
|
||||
{
|
||||
if(BUILD_PALETTE)
|
||||
{
|
||||
//Makes the game do the palette work. Copy the result in a bin file and use that instead in order to make the swap fast
|
||||
unsigned short palettes[0x50];
|
||||
cpuset(paletteDest, palettes, 0x50);
|
||||
for(int i = 0; i < 5; i++)
|
||||
m12_dim_palette(&palettes[i * 0x10], 0x10, 0x800);
|
||||
cpuset(palettes, paletteDest, 0x50);
|
||||
}
|
||||
else
|
||||
cpuset(m12_cartridge_palettes_dimmed, paletteDest, 0x50);
|
||||
}
|
||||
else
|
||||
cpuset(&m12_cartridge_palettes[0x20], paletteDest, 0x50);
|
||||
}
|
||||
|
||||
void store_pixels_overworld_buffer_totalTiles(int totalTiles)
|
||||
{
|
||||
int tile = *tile_offset;
|
||||
|
|
@ -2200,7 +2216,7 @@ void store_pixels_overworld_buffer_totalTiles(int totalTiles)
|
|||
if(i == nextValue)
|
||||
{
|
||||
nextValue += 0x20;
|
||||
topBufferValues += 0x40;
|
||||
topBufferValues += 0x20;
|
||||
bottomBufferValues += 0x40;
|
||||
}
|
||||
i++;
|
||||
|
|
@ -2221,7 +2237,6 @@ void store_pixels_overworld_buffer_totalTiles(int totalTiles)
|
|||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x10) & 0xFF];
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x18) & 0xFF];
|
||||
//Since those are unused
|
||||
bottomBufferValues++;
|
||||
bottomTilePointer += 4;
|
||||
/* The game doesn't use these
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(second_half >> 0) & 0xFF];
|
||||
|
|
@ -2235,7 +2250,7 @@ void store_pixels_overworld_buffer_totalTiles(int totalTiles)
|
|||
if(i == nextValue)
|
||||
{
|
||||
nextValue += 0x20;
|
||||
topBufferValues += 0x40;
|
||||
topBufferValues += 0x20;
|
||||
bottomBufferValues += 0x40;
|
||||
}
|
||||
i++;
|
||||
|
|
@ -2256,7 +2271,6 @@ void store_pixels_overworld_buffer_totalTiles(int totalTiles)
|
|||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x10) & 0xFF];
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(first_half >> 0x18) & 0xFF];
|
||||
//Since those are unused
|
||||
bottomBufferValues++;
|
||||
bottomTilePointer += 4;
|
||||
/* The game doesn't use these
|
||||
*(bottomTilePointer++) = bits_to_nybbles_pointer[(second_half >> 0) & 0xFF];
|
||||
|
|
@ -2280,10 +2294,23 @@ void copy_tile_buffer(int xSource, int ySource, int xDest, int yDest, byte *dest
|
|||
{
|
||||
int sourceTileIndex = get_tile_number_buffer(xSource, ySource) + *tile_offset;
|
||||
int destTileIndex = get_tile_number_buffer(xDest, yDest) + *tile_offset;
|
||||
int* sourceTile = (int*)&dest[sourceTileIndex * 8];
|
||||
int* destTile = (int*)&dest[destTileIndex * 8];
|
||||
int* sourceTile = (int*)&dest[convert_tile_address_buffer(sourceTileIndex)];
|
||||
int* destTile = (int*)&dest[convert_tile_address_buffer(destTileIndex)];
|
||||
|
||||
//Copy the first part, no matter what
|
||||
destTile[0] = sourceTile[0];
|
||||
//Handle the 4 different cases
|
||||
if(((sourceTileIndex - *tile_offset) & 0x20) == 0)
|
||||
{
|
||||
if(((destTileIndex - *tile_offset) & 0x20) == 0)
|
||||
destTile[1] = sourceTile[1];
|
||||
}
|
||||
else
|
||||
{
|
||||
if(((destTileIndex - *tile_offset) & 0x20) == 0)
|
||||
destTile[1] = 0;
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
// x,y: tile coordinates
|
||||
|
|
|
|||
|
|
@ -8,6 +8,7 @@
|
|||
#define FALSE 0
|
||||
|
||||
#define TILESET_OFFSET_BUFFER_MULTIPLIER 0x8
|
||||
#define DEFAULT_DOUBLE_TILE_HEIGHT 0xC
|
||||
#define CHAR_OFFSET 0x50
|
||||
#define CHAR_END 0x60
|
||||
#define YOUWON_START 0x64
|
||||
|
|
@ -41,7 +42,7 @@ int count_pixels_to_tiles(byte *str, int length, int startingPos);
|
|||
int count_pixels_to_tiles_normal_string(byte *str, int startingPos);
|
||||
int expand_bit_depth(byte row, byte foreground);
|
||||
byte reduce_bit_depth(int row, int foregroundRow);
|
||||
void reduce_bit_depth_sp(int* TileRows, int* bufferValues);
|
||||
void reduce_bit_depth_sp(int* TileRows, int* bufferValues, int* bottomBufferValues);
|
||||
byte print_character(byte chr, int x, int y);
|
||||
byte print_character_formatted(byte chr, int x, int y, int font, int foreground);
|
||||
byte print_character_to_window(byte chr, WINDOW* window);
|
||||
|
|
|
|||
Loading…
Reference in New Issue