Apple-2gs-SW
/
iigs-game-engine
mirror of https://github.com/lscharen/iigs-game-engine.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
iigs-game-engine/src/SpriteRender.s

326 lines
7.4 KiB
ArmAsm
Raw Normal View History

Rough outline of streamlined sprite subsystem * Split the creation of the sprite stamps from adding the sprites themselves. This allows for 48 stamps that can be pre-rendered and quickly reassigned to sprites for animations. * Inlined all calls to PushDirtyTile. This both removed significant overhead from calling the small function and, since almost all callers we checking multiple tiles, we were able to avoid incrementing the count each time and just add a single incrments at the end. * Switched from recording each tile that a sprite intersects with each from to only recording the top-left tile and the overlap size. This reduced overhead for larger sprites and removed the needs for an end-of-list marker. * Much more aggressive caching of Sprite and Tile Store values in order to streamline the inner tile dispatch routines. * Moving TileStore and Sprites (and other supporting data structures) into a separate data bank. Needed just for size purposes and provide micro-optimizations by opening up the use of abs,y addressing modes. * Revamped multi-sprite rendering code to avoid the need to copy any masks and all stacked sprites can be drawn via a sequence of and [addrX],y; ora (addrX),y where addrX is set once per tile. * General streamlining to reduct overhead. This work was focused on removing as much per-tile overhead as possible.
2022-04-20 12:43:16 +00:00
; Alternate entry point that takes arguments in registers instead of using a _Sprite
; record
;
; Y = VBUFF address
; X = Tile Data address
; A = Sprite Flags
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
DISP_VFLIP equ $0004 ; hard code these because they are internal values
DISP_HFLIP equ $0002
DISP_MASK equ $0018 ; Preserve the size bits
Rough outline of streamlined sprite subsystem * Split the creation of the sprite stamps from adding the sprites themselves. This allows for 48 stamps that can be pre-rendered and quickly reassigned to sprites for animations. * Inlined all calls to PushDirtyTile. This both removed significant overhead from calling the small function and, since almost all callers we checking multiple tiles, we were able to avoid incrementing the count each time and just add a single incrments at the end. * Switched from recording each tile that a sprite intersects with each from to only recording the top-left tile and the overlap size. This reduced overhead for larger sprites and removed the needs for an end-of-list marker. * Much more aggressive caching of Sprite and Tile Store values in order to streamline the inner tile dispatch routines. * Moving TileStore and Sprites (and other supporting data structures) into a separate data bank. Needed just for size purposes and provide micro-optimizations by opening up the use of abs,y addressing modes. * Revamped multi-sprite rendering code to avoid the need to copy any masks and all stacked sprites can be drawn via a sequence of and [addrX],y; ora (addrX),y where addrX is set once per tile. * General streamlining to reduct overhead. This work was focused on removing as much per-tile overhead as possible.
2022-04-20 12:43:16 +00:00
_DrawSpriteStamp
sty tmp1
stx tmp2
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
xba
Rough outline of streamlined sprite subsystem * Split the creation of the sprite stamps from adding the sprites themselves. This allows for 48 stamps that can be pre-rendered and quickly reassigned to sprites for animations. * Inlined all calls to PushDirtyTile. This both removed significant overhead from calling the small function and, since almost all callers we checking multiple tiles, we were able to avoid incrementing the count each time and just add a single incrments at the end. * Switched from recording each tile that a sprite intersects with each from to only recording the top-left tile and the overlap size. This reduced overhead for larger sprites and removed the needs for an end-of-list marker. * Much more aggressive caching of Sprite and Tile Store values in order to streamline the inner tile dispatch routines. * Moving TileStore and Sprites (and other supporting data structures) into a separate data bank. Needed just for size purposes and provide micro-optimizations by opening up the use of abs,y addressing modes. * Revamped multi-sprite rendering code to avoid the need to copy any masks and all stacked sprites can be drawn via a sequence of and [addrX],y; ora (addrX),y where addrX is set once per tile. * General streamlining to reduct overhead. This work was focused on removing as much per-tile overhead as possible.
2022-04-20 12:43:16 +00:00
and #DISP_MASK ; dispatch to all of the different orientations
sta tmp3
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
phb
pea #^tiledata ; Set the bank to the tile data
plb
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
; X = sprite ID
; Y = Tile Data
; A = VBUFF address
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
ldx tmp3
ldy tmp2
lda tmp1
jsr _DrawSprite
lda tmp3
More tweaks and fixes * Adjusted tile calculation when drawing sprites * Fixed incorred draw order in 16x16 sprite * Added more early-out checks to save work
2022-02-21 21:45:11 +00:00
ora #DISP_HFLIP
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
tax
ldy tmp2
lda tmp1
clc
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
adc #3*4
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
jsr _DrawSprite
lda tmp3
More tweaks and fixes * Adjusted tile calculation when drawing sprites * Fixed incorred draw order in 16x16 sprite * Added more early-out checks to save work
2022-02-21 21:45:11 +00:00
ora #DISP_VFLIP
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
tax
ldy tmp2
lda tmp1
clc
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
adc #6*4
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
jsr _DrawSprite
lda tmp3
Fix the mask values when enumerating through the sprite sheet
2022-02-18 22:20:07 +00:00
ora #DISP_HFLIP+DISP_VFLIP
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
tax
ldy tmp2
lda tmp1
clc
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
adc #9*4
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
jsr _DrawSprite
; Restore bank
plb ; pop extra byte
plb
rts
;
; X = _Sprites array offset
_DrawSprite
jmp (draw_sprite,x)
draw_sprite dw draw_8x8,draw_8x8h,draw_8x8v,draw_8x8hv
dw draw_8x16,draw_8x16h,draw_8x16v,draw_8x16hv
dw draw_16x8,draw_16x8h,draw_16x8v,draw_16x8hv
dw draw_16x16,draw_16x16h,draw_16x16v,draw_16x16hv
dw :rtn,:rtn,:rtn,:rtn ; hidden bit is set
dw :rtn,:rtn,:rtn,:rtn
dw :rtn,:rtn,:rtn,:rtn
dw :rtn,:rtn,:rtn,:rtn
:rtn rts
draw_8x8
draw_8x8h
tax
jmp _DrawTile8x8
draw_8x8v
draw_8x8hv
tax
jmp _DrawTile8x8V
draw_8x16
draw_8x16h
tax
jsr _DrawTile8x8
clc
txa
adc #{8*SPRITE_PLANE_SPAN}
tax
tya
adc #{128*32} ; 32 tiles to the next vertical one, each tile is 128 bytes
tay
jmp _DrawTile8x8
draw_8x16v
draw_8x16hv
tax
jsr _DrawTile8x8V
clc
txa
adc #{8*SPRITE_PLANE_SPAN}
tax
tya
adc #{128*32}
tay
jmp _DrawTile8x8V
draw_16x8
tax
jsr _DrawTile8x8
clc
txa
adc #4
tax
tya
adc #128 ; Next tile is 128 bytes away
tay
jmp _DrawTile8x8
draw_16x8h
clc
tax
tya
pha
adc #128
tay
jsr _DrawTile8x8
txa
adc #4
tax
ply
jmp _DrawTile8x8
draw_16x8v
tax
jsr _DrawTile8x8V
clc
txa
adc #4
tax
tya
adc #128
tay
jmp _DrawTile8x8V
draw_16x8hv
clc
tax
tya
pha
adc #128
tay
jsr _DrawTile8x8V
txa
adc #4
tax
ply
jmp _DrawTile8x8V
Fix several bugs; biggest one in the SpriteStamp routine
2022-06-01 03:53:33 +00:00
; X = sprite ID
; Y = Tile Data
; A = VBUFF address
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
draw_16x16
clc
tax
jsr _DrawTile8x8
txa
adc #4
tax
tya
adc #128
tay
jsr _DrawTile8x8
txa
adc #{8*SPRITE_PLANE_SPAN}-4
tax
tya
adc #{128*{32-1}}
tay
jsr _DrawTile8x8
txa
adc #4
tax
tya
adc #128
tay
jmp _DrawTile8x8
draw_16x16h
clc
tax
tya
More tweaks and fixes * Adjusted tile calculation when drawing sprites * Fixed incorred draw order in 16x16 sprite * Added more early-out checks to save work
2022-02-21 21:45:11 +00:00
adc #64
Checkpoint for sprite rewrite All of the sprite rendering has been deferred down to the level of the tile drawing. Sprites are no longer drawn/erased, but instead a sprite sheet is generated in AddSprite and referenced by the renderer. Because there is no longer a single off-screen buffer that holds a copy of all the rendered sprites, the TileStore size must be expanded to hold a reference to the sprite data address fo each tile. This increase in data structure size require the TileStore to be put into its own bank and appropriate code restructuring. The benefits to the rewrite are significant: 1. Sprites are never drawn/erased off-screen. They are only ever drawn directly to the screen or play field. 2. The concept of "damaged" sprites is gone. Every dirty tile automatically renders just to portion of a sprite that it intersects. These two properties result in a substantial increase in throughput.
2022-02-18 18:12:32 +00:00
pha
adc #128
tay
jsr _DrawTile8x8
txa
adc #4
tax
ply
jsr _DrawTile8x8
txa
adc #{8*SPRITE_PLANE_SPAN}-4
tax
tya
adc #{128*32}
pha
adc #128
tay
jsr _DrawTile8x8
txa
adc #4
tax
ply
jmp _DrawTile8x8
draw_16x16v
clc
tax
tya
pha ; store some copies
phx
pha
adc #{128*32}
tay
jsr _DrawTile8x8V
txa
adc #{8*SPRITE_PLANE_SPAN}
tax
ply
jsr _DrawTile8x8V
pla
adc #4
tax
lda 1,s
adc #{128*{32+1}}
tay
jsr _DrawTile8x8V
txa
adc #{8*SPRITE_PLANE_SPAN}
tax
pla
adc #128
tay
jmp _DrawTile8x8V
draw_16x16hv
clc
tax
tya
pha
adc #128+{128*32} ; Bottom-right source to top-left
tay
jsr _DrawTile8x8V
txa
adc #4
tax
lda 1,s
adc #{128*32}
tay
jsr _DrawTile8x8V
txa
adc #{8*SPRITE_PLANE_SPAN}-4
tax
lda 1,s
adc #128
tay
jsr _DrawTile8x8V
txa
adc #4
tax
ply
jmp _DrawTile8x8V
; X = sprite vbuff address
; Y = tile data pointer
_DrawTile8x8
_CopyTile8x8
]line equ 0
lup 8
lda: tiledata+32+{]line*4},y
stal spritemask+{]line*SPRITE_PLANE_SPAN},x
lda: tiledata+{]line*4},y
stal spritedata+{]line*SPRITE_PLANE_SPAN},x
lda: tiledata+32+{]line*4}+2,y
stal spritemask+{]line*SPRITE_PLANE_SPAN}+2,x
lda: tiledata+{]line*4}+2,y
stal spritedata+{]line*SPRITE_PLANE_SPAN}+2,x
]line equ ]line+1
--^
rts
_DrawTile8x8V
_CopyTile8x8V
]line equ 0
lup 8
lda: tiledata+32+{{7-]line}*4},y
stal spritemask+{]line*SPRITE_PLANE_SPAN},x
lda: tiledata+{{7-]line}*4},y
stal spritedata+{]line*SPRITE_PLANE_SPAN},x
lda: tiledata+32+{{7-]line}*4}+2,y
stal spritemask+{]line*SPRITE_PLANE_SPAN}+2,x
lda: tiledata+{{7-]line}*4}+2,y
stal spritedata+{]line*SPRITE_PLANE_SPAN}+2,x
]line equ ]line+1
--^
rts