📄 newspr2_16.s
字号:
// See newspr2.s for more comments... // 16bpp functions#include "asmdefs.inc"CODE_SEG//void Draw64x64_Mapped_16(UINT8 *SPR, int x, int y, UINT8 *cmap) // In the 16bpp and 32bpp versions, I need one more register// I can't just swap parts of ax as in the 8bpp version.// The best choice seems to be bx so far. It will force to process// only 4 pixels at a time, but it isn't a big deal ! #define HANDLE_4_PIXELS( base ) \ movl base(%esi),%eax ;\ ;\ /* 3 instructions for 1 pixel... is there any way to do better ? */ ;\ movb %al,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,base*2(%edi) ;\ ;\ movb %ah,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,base*2+2(%edi) ;\ ;\ bswap %eax /* is it faster than reading twice ax ??? probably... */;\ movb %al,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,base*2+4(%edi) ;\ ;\ movb %ah,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,base*2+6(%edi)// FlipY : just invert the order for edi#define HANDLE_4_PIXELS_FLIPY( base ) \ movl base(%esi),%eax ;\ ;\ /* 3 instructions for 1 pixel... is there any way to do better ? */ ;\ movb %al,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,126-base*2(%edi) ;\ ;\ movb %ah,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,124-base*2(%edi) ;\ ;\ bswap %eax /* is it faster than reading twice ax ??? probably... */;\ movb %al,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,122-base*2(%edi) ;\ ;\ movb %ah,%dl ;\ movw (%ecx,%edx,2),%bx ;\ movw %bx,120-base*2(%edi) FUNC( Draw64x64_Mapped_16 ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_00: movl 28(%esp),%eax // x movl 36(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x (doubled in 16bpp) xorl %edx,%edx movl $64,%ebp // Tile Height jmp loop0 // Make sure we are aligned....align 8 loop0: HANDLE_4_PIXELS(0) HANDLE_4_PIXELS(4) HANDLE_4_PIXELS(8) HANDLE_4_PIXELS(12) HANDLE_4_PIXELS(16) HANDLE_4_PIXELS(20) HANDLE_4_PIXELS(24) HANDLE_4_PIXELS(28) HANDLE_4_PIXELS(32) HANDLE_4_PIXELS(36) HANDLE_4_PIXELS(40) HANDLE_4_PIXELS(44) HANDLE_4_PIXELS(48) HANDLE_4_PIXELS(52) HANDLE_4_PIXELS(56) HANDLE_4_PIXELS(60) addl $64,%esi /* Next Tile Line */ addl $0xDEADBEEF,%edi // Next Screen Linebitw_00: decl %ebp jne loop0 popl %ecx popl %ebx popl %esi popl %edi popl %ebp retFUNC( Draw64x64_Mapped_16_FlipY ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y sall $2,%eax movl 0xDEADBEEF(%eax),%edi // Goto the end...blin_01: movl 28(%esp),%eax // x movl 36(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x (doubled in 16bpp) xorl %edx,%edx movl $64,%ebp // Tile Height jmp loop01 // Make sure we are aligned....align 8 loop01: HANDLE_4_PIXELS_FLIPY(0) HANDLE_4_PIXELS_FLIPY(4) HANDLE_4_PIXELS_FLIPY(8) HANDLE_4_PIXELS_FLIPY(12) HANDLE_4_PIXELS_FLIPY(16) HANDLE_4_PIXELS_FLIPY(20) HANDLE_4_PIXELS_FLIPY(24) HANDLE_4_PIXELS_FLIPY(28) HANDLE_4_PIXELS_FLIPY(32) HANDLE_4_PIXELS_FLIPY(36) HANDLE_4_PIXELS_FLIPY(40) HANDLE_4_PIXELS_FLIPY(44) HANDLE_4_PIXELS_FLIPY(48) HANDLE_4_PIXELS_FLIPY(52) HANDLE_4_PIXELS_FLIPY(56) HANDLE_4_PIXELS_FLIPY(60) addl $64,%esi /* Next Tile Line */ addl $0xDEADBEEF,%edi // Next Screen Linebitw_01: decl %ebp jne loop01 popl %ecx popl %ebx popl %esi popl %edi popl %ebp retFUNC( Draw64x64_Mapped_16_FlipX ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y movl %esi,%edx sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_02: movl 28(%esp),%eax // x movl 36(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x (doubled in 16bpp) xorl %edx,%edx movl $64,%ebp // Tile Height jmp loop02 // Make sure we are aligned....align 8 loop02: HANDLE_4_PIXELS(0) HANDLE_4_PIXELS(4) HANDLE_4_PIXELS(8) HANDLE_4_PIXELS(12) HANDLE_4_PIXELS(16) HANDLE_4_PIXELS(20) HANDLE_4_PIXELS(24) HANDLE_4_PIXELS(28) HANDLE_4_PIXELS(32) HANDLE_4_PIXELS(36) HANDLE_4_PIXELS(40) HANDLE_4_PIXELS(44) HANDLE_4_PIXELS(48) HANDLE_4_PIXELS(52) HANDLE_4_PIXELS(56) HANDLE_4_PIXELS(60) addl $64,%esi /* Next Tile Line */ subl $0xDEADBEEF,%edi // Prev Screen Linebitw_02: decl %ebp jne loop02 popl %ecx popl %ebx popl %esi popl %edi popl %ebp retFUNC( Draw64x64_Mapped_16_FlipXY ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y movl %esi,%edx sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_03: movl 28(%esp),%eax // x movl 36(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x (doubled in 16bpp) xorl %edx,%edx movl $64,%ebp // Tile Height jmp loop03 // Make sure we are aligned....align 8 loop03: HANDLE_4_PIXELS_FLIPY(0) HANDLE_4_PIXELS_FLIPY(4) HANDLE_4_PIXELS_FLIPY(8) HANDLE_4_PIXELS_FLIPY(12) HANDLE_4_PIXELS_FLIPY(16) HANDLE_4_PIXELS_FLIPY(20) HANDLE_4_PIXELS_FLIPY(24) HANDLE_4_PIXELS_FLIPY(28) HANDLE_4_PIXELS_FLIPY(32) HANDLE_4_PIXELS_FLIPY(36) HANDLE_4_PIXELS_FLIPY(40) HANDLE_4_PIXELS_FLIPY(44) HANDLE_4_PIXELS_FLIPY(48) HANDLE_4_PIXELS_FLIPY(52) HANDLE_4_PIXELS_FLIPY(56) HANDLE_4_PIXELS_FLIPY(60) addl $64,%esi /* Next Tile Line */ subl $0xDEADBEEF,%edi // Prev Screen Linebitw_03: decl %ebp jne loop03 popl %ecx popl %ebx popl %esi popl %edi popl %ebp ret/************************************************************Notice : for now, I don't translate the Trans version of the64x64 functions. I am not sure they will be usefull...************************************************************/ // These are functions called by the f3 games (not all of them)// They basically draw a sprite of 8 pixels wide and with an arbitrary heigh // void Draw8xH_Trans_Packed_Mapped_Column(UINT8 *SPR, int x, int y, int height, UINT8 *cmap)// What was supposed to happen has finally happened : I am out of registers!// I need to create a counter in memory instead of using ebp for the height// of the sprite. cx will be used for reading the palette...counter: .long 0 FUNC( Draw8xH_Trans_Packed_Mapped_Column_16 ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_20: movl 28(%esp),%eax // +x movl 40(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x doubled in 16bpp xorl %edx,%edx // Tsss 2 instructions for memory -> memory transfer... movl 36(%esp),%eax // Tile Height movl %eax,counter jmp loop20 // Alignement.align 8loop20: movl (%esi),%eax testl %eax,%eax // Skip Blank Lines jz 8f movl %eax,%ebx andl $0x0F0F0F0F,%eax shrl $4,%ebx andl $0x0F0F0F0F,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,2(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,6(%edi)7: testb %bh,%bh jz 7f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,4(%edi)7: shr $16,%eax shr $16,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,10(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,8(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,14(%edi)7: testb %bh,%bh jz 8f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,12(%edi)8: addl $4,%esi /* Next Tile Line */ addl $0xdeadbeef,%edi /* Next Screen Line */bitw_20: decl counter jne loop20 popl %ecx popl %ebx popl %esi popl %edi popl %ebp ret//void Draw8xH_Trans_Packed_Mapped_FlipY_Column(UINT8 *SPR, int x, int y, int height, UINT8 *cmap)FUNC( Draw8xH_Trans_Packed_Mapped_Column_16_FlipY ) pushl %ebp pushl %edi pushl %esi pushl %ebx push %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_21: movl 28(%esp),%eax // +x movl 40(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x doubled in 16bpp xorl %edx,%edx movl 36(%esp),%eax // Tile Height movl %eax,counter jmp loop21 // Alignement.align 8loop21: movl (%esi),%eax testl %eax,%eax // Skip Blank Lines jz 8f movl %eax,%ebx andl $0x0F0F0F0F,%eax shrl $4,%ebx andl $0x0F0F0F0F,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,12(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,14(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,8(%edi)7: testb %bh,%bh jz 7f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,10(%edi)7: shr $16,%eax shr $16,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,4(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,6(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,(%edi)7: testb %bh,%bh jz 8f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,2(%edi)8: addl $4,%esi /* Next Tile Line */ addl $0xdeadbeef,%edi /* Next Screen Line */bitw_21: decl counter jne loop21 popl %ecx popl %ebx popl %esi popl %edi popl %ebp ret/* The same as the normal Mapped_Column function but : begins at the last line and goes back to the begining (subl) *///void Draw8xH_Trans_Packed_Mapped_FlipX_Column(UINT8 *SPR, int x, int y, int height, UINT8 *cmap)FUNC( Draw8xH_Trans_Packed_Mapped_Column_16_FlipX ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y movl 36(%esp),%eax // Tile Height movl %eax,counter addl %ebp,%eax // tile height sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_22: movl 28(%esp),%eax // +x movl 40(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x doubled in 16bpp xorl %edx,%edx jmp loop22 // Alignement.align 8loop22: movl (%esi),%eax testl %eax,%eax // Skip Blank Lines jz 8f movl %eax,%ebx andl $0x0F0F0F0F,%eax shrl $4,%ebx andl $0x0F0F0F0F,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,2(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,6(%edi)7: testb %bh,%bh jz 7f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,4(%edi)7: shr $16,%eax shr $16,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,10(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,8(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,14(%edi)7: testb %bh,%bh jz 8f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,12(%edi)8: addl $4,%esi /* Next Tile Line */ subl $0xdeadbeef,%edi /* Previous Screen Line */bitw_22: decl counter jne loop22 popl %ecx popl %ebx popl %esi popl %edi popl %ebp ret/* The 2 previous functions combined : starts at last line and goes back x offsets are taken backward (from 7 to 0) *///void Draw8xH_Trans_Packed_Mapped_FlipXY_Column(UINT8 *SPR, int x, int y, int height, UINT8 *cmap)FUNC( Draw8xH_Trans_Packed_Mapped_Column_16_FlipXY ) pushl %ebp pushl %edi pushl %esi pushl %ebx pushl %ecx movl 24(%esp),%esi // source movl 32(%esp),%eax // y movl 36(%esp),%eax // Tile Height movl %eax,counter addl %ebp,%eax // tile height sall $2,%eax movl 0xDEADBEEF(%eax),%ediblin_23: movl 28(%esp),%eax // +x movl 40(%esp),%ebp // cmap leal (%edi,%eax,2),%edi // x doubled in 16bpp xorl %edx,%edx jmp loop23 // Alignement.align 8loop23: movl (%esi),%eax testl %eax,%eax // Skip Blank Lines jz 8f movl %eax,%ebx andl $0x0F0F0F0F,%eax shrl $4,%ebx andl $0x0F0F0F0F,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,12(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,14(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,8(%edi)7: testb %bh,%bh jz 7f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,10(%edi)7: shr $16,%eax shr $16,%ebx testb %al,%al jz 7f movb %al,%dl movw (%ebp,%edx,2),%cx movw %cx,4(%edi)7: testb %bl,%bl jz 7f movb %bl,%dl movw (%ebp,%edx,2),%cx movw %cx,6(%edi)7: testb %ah,%ah jz 7f movb %ah,%dl movw (%ebp,%edx,2),%cx movw %cx,(%edi)7: testb %bh,%bh jz 8f movb %bh,%dl movw (%ebp,%edx,2),%cx movw %cx,2(%edi)8: addl $4,%esi /* Next Tile Line */ subl $0xdeadbeef,%edi /* Previous Screen Line */bitw_23: decl counter jne loop23 popl %ecx popl %ebx popl %esi popl %edi popl %ebp retFUNC(init_newspr2asm_16) movl GLOBL(GameBitmap),%eax // Adress of the GameBitmap's structure movl (%eax),%eax // Width addl %eax,%eax // double for 16bpp movl %eax,bitw_00-4 movl %eax,bitw_01-4 movl %eax,bitw_02-4 movl %eax,bitw_03-4 //movl %eax,bitw_10-4 //movl %eax,bitw_11-4 //movl %eax,bitw_12-4 //movl %eax,bitw_13-4 movl %eax,bitw_20-4 movl %eax,bitw_21-4 movl %eax,bitw_22-4 movl %eax,bitw_23-4 movl GLOBL(GameBitmap),%eax addl $64,%eax // Line 0 movl %eax,blin_00-4 movl %eax,blin_01-4 //movl %eax,blin_10-4 //movl %eax,blin_11-4 movl %eax,blin_20-4 movl %eax,blin_21-4 pushl %eax addl $63*4,%eax // Line 63 !!! movl %eax,blin_02-4 movl %eax,blin_03-4 //movl %eax,blin_12-4 //movl %eax,blin_13-4 popl %eax subl $1*4,%eax // This time line 7 really! movl %eax,blin_22-4 movl %eax,blin_23-4 ret ⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -