|
Título: 16x16 Sprite mirroring Publicado por: jltursan en 02 de Octubre de 2006, 07:10:21 pm Mensaje original de Robsy
Hola. Aqui os dejo unas rutinas que os pueden ayudar. Estan muy lejos de ser "optimas" pero funcionan bien, que es de lo que se trata. No usan ningun buffer en RAM aparte de la pila, claro. Por eso alguna parte es un poco "oscura". Salud y saludos. Código: ;---------------------------------------------- ; SPRITE MIRRORING ; Eduardo A. Robsy Petrus [22/04/2005] ; (c) Karoshi Corporation, 2005 ;---------------------------------------------- ; Horizontal and vertical mirroring routines ; for 16x16 sized sprites in VRAM. ;---------------------------------------------- ; - Does not use temporary RAM, only the stack. ; - Total assembled size: 110 bytes. ; - Requires the .BIOS directive to be used and ; BIOS routines available in page 0. ; - Requires SPRTBL label to point to VRAM ; sprite pattern table (BIOS default: 3800h). ;---------------------------------------------- ; PROs: Reliable, does not use a RAM buffer. ; CONs: Not very fast, not very compact. ;---------------------------------------------- ;---------------------------------------------- SPRITE_H_MIRROR: ; Horizontal sprite mirroring ;---------------------------------------------- ; Input: A=original sprite number ; B=mirrored sprite number ; Output: none ; Modifies: AF,BC,DE,HL,EI ;---------------------------------------------- ; Calculate original sprite address call MULT32 ld a,b ld bc,SPRTBL add hl,bc ; Store original address in DE ex de,hl ; Calculate mirrored sprite address call MULT32 add hl,bc ; Store in HL and save into the stack push hl ; Adjust HL to second part ld bc,16 add hl,bc ; Copy and invert second half call @@COPY16 ; Copy and invert first half pop hl call @@COPY16 ; Done ret ; Copy and inverts 16 bytes of VRAM (DE->HL) @@COPY16: ld b,16 ; Copy loop @@COPY: ; Read byte ex de,hl call RDVRM ; Invert byte push bc push hl ld b,8 @@INVERT: rla ld l,a ld a,c rra ld c,a ld a,l djnz @@INVERT ld a,c pop hl pop bc ; Write byte ex de,hl call WRTVRM ; Increase pointers inc de inc hl djnz @@COPY ; Done ret ;---------------------------------------------- ;---------------------------------------------- SPRITE_V_MIRROR: ; Vertical sprite mirroring ;---------------------------------------------- ; Input: A=original sprite number ; B=mirrored sprite number ; Output: none ; Modifies: AF,BC,DE,HL,EI ;---------------------------------------------- ; Original address ; (ORIGINAL_NUM*32)+SPRITE_TABLE call MULT32 ld a,b ld bc,SPRTBL add hl,bc ; Store original address in DE ex de,hl call MULT32 add hl,bc ; Push into stack push hl ld bc,15 add hl,bc call @@COPY16 pop hl ld bc,31 add hl,bc call @@COPY16 ret ; DE->HL 16 bytes, VRAM @@COPY16: ld b,16 @@COPY: ex de,hl call RDVRM ex de,hl call WRTVRM inc de dec hl djnz @@COPY ret ;---------------------------------------------- ;---------------------------------------------- MULT32: ; Support function ;---------------------------------------------- ; Input: A=value ; Output: HL=value*32 ; Modifies: HL ;---------------------------------------------- ld l,a ld h,0 add hl,hl add hl,hl add hl,hl add hl,hl add hl,hl ret ;---------------------------------------------- Con la mejora de Ricardo Bittencourt ;) : Citar MULT32 is very, very slow. Use this instead: Código: MULT32: RRCA RRCA LD L,A AND 31 LD H,A XOR L LD L,A RET |