WebSVN – Kolibri OS – Diff – /programs/demos/3DS/BUMP_CAT.INC


;SIZE_X equ 350
;SIZE_Y equ 350
ROUND equ 8
;TEX_X equ 512
;TEX_Y equ 512
;TEXTURE_SIZE EQU (512*512)-1
;TEX_SHIFT EQU 9
CATMULL_SHIFT equ 8
;TEXTURE_SIZE EQU (TEX_X * TEX_Y)-1
;Ext = NON
;MMX = 1
;NON = 0
;------- Big thanks to Majuma (www.majuma.xt.pl) for absolutely great---
;------- DOS 13h mode demos --------------------------------------------
;------- Procedure draws bump triangle using Catmull Z-buffer algorithm-
;------- (Z coordinate interpolation)-----------------------------------
bump_triangle_z:
;------------------in - eax - x1 shl 16 + y1 -----------
;---------------------- ebx - x2 shl 16 + y2 -----------
;---------------------- ecx - x3 shl 16 + y3 -----------
;---------------------- edx - pointer to bump map ------
;---------------------- esi - pointer to environment map
;---------------------- edi - pointer to screen buffer--
;---------------------- stack : bump coordinates--------
;----------------------         environment coordinates-
;----------------------         Z position coordinates--
;----------------------         pointer io Z buffer-----
;-- Z-buffer - filled with coordinates as dword --------
;-- (Z coor. as word) shl CATMULL_SHIFT ----------------
.b_x1	equ ebp+4   ; procedure don't save registers !!!
.b_y1	equ ebp+6   ; each coordinate as word
.b_x2	equ ebp+8
.b_y2	equ ebp+10
.b_x3	equ ebp+12
.b_y3	equ ebp+14
.e_x1	equ ebp+16
.e_y1	equ ebp+18
.e_x2	equ ebp+20
.e_y2	equ ebp+22
.e_x3	equ ebp+24
.e_y3	equ ebp+26
.z1	equ word[ebp+28]
.z2	equ word[ebp+30]
.z3	equ word[ebp+32]
.z_buff equ dword[ebp+34]	; pointer to Z-buffer
 
 
.t_bmap equ dword[ebp-4]	; pointer to bump map
.t_emap equ dword[ebp-8]	; pointer to e. map
.x1	equ word[ebp-10]
.y1	equ word[ebp-12]
.x2	equ word[ebp-14]
.y2	equ word[ebp-16]
.x3	equ word[ebp-18]
.y3	equ word[ebp-20]
 
.dx12  equ dword[ebp-24]
.dz12  equ	[ebp-28]
.dbx12 equ dword[ebp-32]
.dby12 equ	[ebp-36]
.dex12 equ dword[ebp-40]
.dey12 equ	[ebp-44]
 
.dx13  equ dword[ebp-48]
.dz13  equ	[ebp-52]
.dbx13 equ dword[ebp-56]
.dby13 equ	[ebp-60]
.dex13 equ dword[ebp-64]
.dey13 equ	[ebp-68]
 
.dx23  equ dword[ebp-72]
.dz23  equ	[ebp-76]
.dbx23 equ dword[ebp-80]
.dby23 equ	[ebp-84]
.dex23 equ dword[ebp-88]
.dey23 equ	[ebp-92]
 
.cx1   equ dword[ebp-96]   ; current variables
.cz1   equ	[ebp-100]
.cx2   equ dword[ebp-104]
.cz2   equ	[ebp-108]
.cbx1  equ dword[ebp-112]
.cby1  equ	[ebp-116]
.cbx2  equ dword[ebp-120]
.cby2  equ	[ebp-124]
.cex1  equ dword[ebp-128]
.cey1  equ	[ebp-132]
.cex2  equ dword[ebp-136]
.cey2  equ	[ebp-140]
 
       mov     ebp,esp
       push    edx	  ; store bump map
       push    esi	  ; store e. map
     ; sub     esp,120
 .sort3:		  ; sort triangle coordinates...
       cmp     ax,bx
       jle     .sort1
       xchg    eax,ebx
       mov     edx,dword[.b_x1]
       xchg    edx,dword[.b_x2]
       mov     dword[.b_x1],edx
       mov     edx,dword[.e_x1]
       xchg    edx,dword[.e_x2]
       mov     dword[.e_x1],edx
       mov     dx,.z1
       xchg    dx,.z2
       mov     .z1,dx
 .sort1:
       cmp	bx,cx
       jle	.sort2
       xchg	ebx,ecx
       mov	edx,dword[.b_x2]
       xchg	edx,dword[.b_x3]
       mov	dword[.b_x2],edx
       mov	edx,dword[.e_x2]
       xchg	edx,dword[.e_x3]
       mov	dword[.e_x2],edx
       mov     dx,.z2
       xchg    dx,.z3
       mov     .z2,dx
       jmp	.sort3
 .sort2:
       push	eax	; store triangle coords in variables
       push	ebx
       push	ecx
 
	 mov	  edx,80008000h  ; eax,ebx,ecx are ANDd together into edx which means that
	 and	  edx,ebx	 ; if *all* of them are negative a sign flag is raised
	 and	  edx,ecx
	 and	  edx,eax
	 test	  edx,80008000h  ; Check both X&Y at once
	 jne	  .loop23_done
    ;   mov     edx,eax         ; eax,ebx,ecx are ORd together into edx which means that
    ;   or      edx,ebx         ; if any *one* of them is negative a sign flag is raised
    ;   or      edx,ecx
    ;   test    edx,80000000h   ; Check only X
    ;   jne     .loop23_done
 
    ;   cmp     .x1,SIZE_X    ; {
    ;   jg      .loop23_done
    ;   cmp     .x2,SIZE_X     ; This can be optimized with effort
    ;   jg      .loop23_done
    ;   cmp     .x3,SIZE_X
    ;   jg      .loop23_done    ; {
 
 
       mov	bx,.y2	     ; calc delta 12
       sub	bx,.y1
       jnz	.bt_dx12_make
       mov	ecx,6
       xor	edx,edx
     @@:
       push	edx   ;dword 0
       loop	@b
       jmp	.bt_dx12_done
 .bt_dx12_make:
       mov	ax,.x2
       sub	ax,.x1
       cwde
       movsx	ebx,bx
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dx12,eax
       push	 eax
 
       mov     ax,.z2
       sub     ax,.z1
       cwde
       shl     eax,CATMULL_SHIFT
       cdq
       idiv    ebx
       push    eax
if Ext>=SSE
 
       sub	 esp,16
       cvtsi2ss  xmm3,ebx	     ;rcps
   ;    mov       eax,255
       cvtsi2ss  xmm4,[i255d]
       divss	 xmm3,xmm4
       rcpss	 xmm3,xmm3
    ;   mulss     xmm3,xmm4
       shufps	 xmm3,xmm3,0
 
       movd	 mm0,[.b_x1]
       movd	 mm1,[.b_x2]
       movd	 mm2,[.e_x1]
       movd	 mm3,[.e_x2]
       pxor	 mm4,mm4
       punpcklwd mm0,mm4
       punpcklwd mm1,mm4
       punpcklwd mm2,mm4
       punpcklwd mm3,mm4
 
       psubd	  mm1,mm0
       psubd	  mm3,mm2
 
 ;       cvtpi2ps  xmm0,mm0
 ;       movlhps   xmm0,xmm0
 ;       cvtpi2ps  xmm0,mm2
       cvtpi2ps  xmm1,mm1
       movlhps	 xmm1,xmm1
       cvtpi2ps  xmm1,mm3
 ;       subps     xmm1,xmm0
 
 
;       cvtpi2ps  xmm0,mm3
  ;     divps     xmm1,xmm3
       mulps	 xmm1,xmm3
       shufps	 xmm1,xmm1,10110001b
       cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
       movhlps	 xmm1,xmm1
       cvtps2pi  mm1,xmm1
       movq	 .dey12,mm0
       movq	 .dby12,mm1
 
 
else
 
       mov	ax,word[.b_x2]
       sub	ax,word[.b_x1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dbx12,eax
       push	 eax
 
       mov	ax,word[.b_y2]
       sub	ax,word[.b_y1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dby12,eax
       push	 eax
 
       mov	ax,word[.e_x2]
       sub	ax,word[.e_x1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dex12,eax
       push	 eax
 
       mov	ax,word[.e_y2]
       sub	ax,word[.e_y1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dey12,eax
       push	 eax
 
end if
 
   .bt_dx12_done:
 
       mov	bx,.y3	     ; calc delta13
       sub	bx,.y1
       jnz	.bt_dx13_make
       mov	ecx,6
       xor	edx,edx
     @@:
       push	edx   ;dword 0
       loop	@b
       jmp	.bt_dx13_done
 .bt_dx13_make:
       mov	ax,.x3
       sub	ax,.x1
       cwde
       movsx	ebx,bx
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dx13,eax
       push	 eax
 
       mov     ax,.z3
       sub     ax,.z1
       cwde
       shl     eax,CATMULL_SHIFT
       cdq
       idiv    ebx
  ;    mov    .dz13,eax
       push    eax
 
if Ext>=SSE
 
       sub	 esp,16
       cvtsi2ss  xmm3,ebx	     ;rcps
  ;     mov       eax,255
       cvtsi2ss  xmm4,[i255d]
       divss	 xmm3,xmm4
       rcpss	 xmm3,xmm3
    ;   mulss     xmm3,xmm4
       shufps	 xmm3,xmm3,0
 
       movd	 mm0,[.b_x1]
       movd	 mm1,[.b_x3]
       movd	 mm2,[.e_x1]
       movd	 mm3,[.e_x3]
       pxor	 mm4,mm4
       punpcklwd mm0,mm4
       punpcklwd mm1,mm4
       punpcklwd mm2,mm4
       punpcklwd mm3,mm4
 
       psubd	  mm1,mm0
       psubd	  mm3,mm2
 
     ;  cvtpi2ps  xmm0,mm0
     ;  movlhps   xmm0,xmm0
     ;  cvtpi2ps  xmm0,mm2
       cvtpi2ps  xmm1,mm1
       movlhps	 xmm1,xmm1
       cvtpi2ps  xmm1,mm3
    ;   subps     xmm1,xmm0
 
 
;       cvtpi2ps  xmm0,mm3
  ;     divps     xmm1,xmm3
       mulps	 xmm1,xmm3
       shufps	 xmm1,xmm1,10110001b
       cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
       movhlps	 xmm1,xmm1
       cvtps2pi  mm1,xmm1
       movq	 .dey13,mm0
       movq	 .dby13,mm1
 
 
else
 
       mov	ax,word[.b_x3]
       sub	ax,word[.b_x1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dbx13,eax
       push	 eax
 
       mov	ax,word[.b_y3]
       sub	ax,word[.b_y1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dby13,eax
       push	 eax
 
       mov	ax,word[.e_x3]
       sub	ax,word[.e_x1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dex13,eax
       push	 eax
 
       mov	ax,word[.e_y3]
       sub	ax,word[.e_y1]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dey13,eax
       push	 eax
end if
 
   .bt_dx13_done:
 
       mov	bx,.y3	     ; calc delta23
       sub	bx,.y2
       jnz	.bt_dx23_make
       mov	ecx,6
       xor	edx,edx
     @@:
       push	edx   ;dword 0
       loop	@b
       jmp	.bt_dx23_done
 .bt_dx23_make:
       mov	ax,.x3
       sub	ax,.x2
       cwde
       movsx	ebx,bx
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dx23,eax
       push	 eax
 
       mov     ax,.z3
       sub     ax,.z2
       cwde
       shl     eax,CATMULL_SHIFT
       cdq
       idiv    ebx
      ; mov     .dz23,eax
       push    eax
      ;  sub     esp,40
if Ext>=SSE
 
       sub	 esp,16
       cvtsi2ss  xmm3,ebx	     ;rcps
   ;    mov       eax,255
       cvtsi2ss  xmm4,[i255d]
       divss	 xmm3,xmm4
       rcpss	 xmm3,xmm3
    ;   mulss     xmm3,xmm4
       shufps	 xmm3,xmm3,0
 
       movd	 mm0,[.b_x2]
       movd	 mm1,[.b_x3]
       movd	 mm2,[.e_x2]
       movd	 mm3,[.e_x3]
       pxor	 mm4,mm4
       punpcklwd mm0,mm4
       punpcklwd mm1,mm4
       punpcklwd mm2,mm4
       punpcklwd mm3,mm4
 
       psubd	  mm1,mm0
       psubd	  mm3,mm2
 
;       cvtpi2ps  xmm0,mm0
;       movlhps   xmm0,xmm0
;       cvtpi2ps  xmm0,mm2
       cvtpi2ps  xmm1,mm1
       movlhps	 xmm1,xmm1
       cvtpi2ps  xmm1,mm3
;       subps     xmm1,xmm0
 
 
;       cvtpi2ps  xmm0,mm3
  ;     divps     xmm1,xmm3
       mulps	 xmm1,xmm3
       shufps	 xmm1,xmm1,10110001b
       cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
       movhlps	 xmm1,xmm1
       cvtps2pi  mm1,xmm1
       movq	 .dey23,mm0
       movq	 .dby23,mm1
 
else
 
       mov	ax,word[.b_x3]
       sub	ax,word[.b_x2]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dbx23,eax
       push	 eax
 
       mov	ax,word[.b_y3]
       sub	ax,word[.b_y2]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dby23,eax
       push	 eax
 
       mov	ax,word[.e_x3]
       sub	ax,word[.e_x2]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dex23,eax
       push	 eax
 
       mov	ax,word[.e_y3]
       sub	ax,word[.e_y2]
       cwde
       shl	eax,ROUND
       cdq
       idiv	ebx
 ;      mov      .dey23,eax
       push	 eax
 
end if
 
   .bt_dx23_done:
       sub	esp,48
 
       movsx	eax,.x1
       shl	eax,ROUND
       mov	.cx1,eax
       mov	.cx2,eax
     ;  push     eax
     ;  push     eax
 
       movsx	eax,word[.b_x1]
       shl	eax,ROUND
       mov	.cbx1,eax
       mov	.cbx2,eax
    ;   push     eax
    ;   push     eax
 
       movsx	eax,word[.b_y1]
       shl	eax,ROUND
       mov	.cby1,eax
       mov	.cby2,eax
    ;   push     eax
    ;   push     eax
 
       movsx	eax,word[.e_x1]
       shl	eax,ROUND
       mov	.cex1,eax
       mov	.cex2,eax
    ;   push     eax
    ;   push     eax
 
       movsx	eax,word[.e_y1]
       shl	eax,ROUND
       mov	.cey1,eax
       mov	.cey2,eax
    ;   push     eax
    ;   push     eax
 
       movsx	eax,.z1
       shl	eax,CATMULL_SHIFT
       mov	.cz1,eax
       mov	.cz2,eax
     ;  push     eax
     ;  push     eax
 
       movsx	ecx,.y1
       cmp	cx,.y2
       jge	.loop12_done
  .loop12:
       call	.call_bump_line
 
if Ext >= MMX
       movq	mm0,.cby2
       movq	mm1,.cby1
       movq	mm2,.cey2
       movq	mm3,.cey1
       movq	mm4,.cz1
       movq	mm5,.cz2
       paddd	mm0,.dby12
       paddd	mm1,.dby13
       paddd	mm2,.dey12
       paddd	mm3,.dey13
       paddd	mm4,.dz13
       paddd	mm5,.dz12
       movq	.cby2,mm0
       movq	.cby1,mm1
       movq	.cey1,mm3
       movq	.cey2,mm2
       movq	.cz1,mm4
       movq	.cz2,mm5
else
       mov	edx,.dbx13
       add	.cbx1,edx
       mov	eax,.dbx12
       add	.cbx2,eax
       mov	ebx,.dby13
       add	.cby1,ebx
       mov	edx,.dby12
       add	.cby2,edx
 
       mov	eax,.dex13
       add	.cex1,eax
       mov	ebx,.dex12
       add	.cex2,ebx
       mov	edx,.dey13
       add	.cey1,edx
       mov	eax,.dey12
       add	.cey2,eax
 
       mov	eax,.dx13
       add	.cx1,eax
       mov	ebx,.dx12
       add	.cx2,ebx
 
       mov	ebx,.dz13
       add	.cz1,ebx
       mov	edx,.dz12
       add	.cz2,edx
end if
       inc	ecx
       cmp	cx,.y2
       jl	.loop12
    .loop12_done:
 
       movsx	ecx,.y2
       cmp	cx,.y3
       jge	.loop23_done
 
       movsx	eax,.z2
       shl	eax,CATMULL_SHIFT
       mov	.cz2,eax
 
       movsx	eax,.x2
       shl	eax,ROUND
       mov	.cx2,eax
 
       movzx	eax,word[.b_x2]
       shl	eax,ROUND
       mov	.cbx2,eax
 
       movzx	eax,word[.b_y2]
       shl	eax,ROUND
       mov	.cby2,eax
 
       movzx	eax,word[.e_x2]
       shl	eax,ROUND
       mov	.cex2,eax
 
       movzx	eax,word[.e_y2]
       shl	eax,ROUND
       mov	.cey2,eax
 
     .loop23:
       call	.call_bump_line
 
if Ext >= MMX
       movq	mm0,.cby2
       movq	mm1,.cby1
       movq	mm2,.cey2
       movq	mm3,.cey1
       movq	mm4,.cz1
       movq	mm5,.cz2
       paddd	mm0,.dby23
       paddd	mm1,.dby13
       paddd	mm2,.dey23
       paddd	mm3,.dey13
       paddd	mm4,.dz13
       paddd	mm5,.dz23
       movq	.cby2,mm0
       movq	.cby1,mm1
       movq	.cey1,mm3
       movq	.cey2,mm2
       movq	.cz1,mm4
       movq	.cz2,mm5
else
       mov	eax,.dx13
       add	.cx1,eax
       mov	ebx,.dx23
       add	.cx2,ebx
 
       mov	edx,.dbx13
       add	.cbx1,edx
       mov	eax,.dbx23
       add	.cbx2,eax
       mov	ebx,.dby13
       add	.cby1,ebx
       mov	edx,.dby23
       add	.cby2,edx
 
       mov	eax,.dex13
       add	.cex1,eax
       mov	ebx,.dex23
       add	.cex2,ebx
       mov	edx,.dey13
       add	.cey1,edx
       mov	eax,.dey23
       add	.cey2,eax
 
       mov	ebx,.dz13
       add	.cz1,ebx
       mov	edx,.dz23
       add	.cz2,edx
end if
       inc	ecx
       cmp	cx,.y3
       jl	.loop23
    .loop23_done:
 
       mov	esp,ebp
ret   34
 
.call_bump_line:
 
      ; push     ebp
      ; push     ecx
       pushad
 
       push	dword .cz1
       push	dword .cz2
       push	.z_buff
       push	.t_emap
       push	.t_bmap
       push	dword .cey2
       push	.cex2
 
 
       push	dword .cby2
       push	.cbx2
       push	dword .cey1
       push	.cex1
       push	dword .cby1
       push	.cbx1
       push	ecx
 
       mov	eax,.cx1
       sar	eax,ROUND
       mov	ebx,.cx2
       sar	ebx,ROUND
 
       call	bump_line_z
 
       popad
ret
bump_line_z:
;--------------in: eax - x1
;--------------    ebx - x2
;--------------    edi - pointer to screen buffer
;stack - another parameters :
.y	equ dword [ebp+4]
.bx1	equ	  [ebp+8]   ;   ---
.by1	equ dword [ebp+12]  ;       |
.ex1	equ	  [ebp+16]  ;       |
.ey1	equ dword [ebp+20]  ;       |>   bump and env coords
.bx2	equ	  [ebp+24]  ;       |>   shifted shl ROUND
.by2	equ dword [ebp+28]  ;       |
.ex2	equ	  [ebp+32]  ;       |
.ey2	equ dword [ebp+36]  ;   ---
.bmap	equ dword [ebp+40]
.emap	equ dword [ebp+44]
.z_buff equ dword [ebp+48]
.z2	equ dword [ebp+52]  ;   -- |>   z coords shifted
.z1	equ dword [ebp+56]  ;   --       shl  CATMULL_SHIFT
 
.x1	equ dword [ebp-4]
.x2	equ dword [ebp-8]
.dbx	equ dword [ebp-12]
.dby	equ	  [ebp-16]
.dex	equ dword [ebp-20]
.dey	equ	  [ebp-24]
.dz	equ dword [ebp-28]
.cbx	equ dword [ebp-32]
.cby	equ	  [ebp-36]
.cex	equ dword [ebp-40]
.cey	equ	  [ebp-44]
.cz	equ dword [ebp-48]
.czbuff equ dword [ebp-52]
.temp1	equ	   ebp-60
.temp2	equ	   ebp-68
.temp3	equ	   ebp-76
.temp4	equ	   ebp-84
.temp5	equ	   ebp-92
 
	mov	ebp,esp
 
	mov	ecx,.y
	or	ecx,ecx
	jl	.bl_end
	cmp	ecx,SIZE_Y
	jge	.bl_end
 
	cmp	eax,ebx
	jl	.bl_ok
	je	.bl_end
 
	xchg	eax,ebx
if Ext=NON
	mov	edx,.bx1
	xchg	edx,.bx2
	mov	.bx1,edx
	mov	edx,.by1
	xchg	edx,.by2
	mov	.by1,edx
 
	mov	edx,.ex1
	xchg	edx,.ex2
	mov	.ex1,edx
	mov	edx,.ey1
	xchg	edx,.ey2
	mov	.ey1,edx
end if
if Ext = MMX
	movq	mm0,.bx1
	movq	mm1,.ex1
	movq	mm2,.bx2
	movq	mm3,.ex2
	movq	.bx2,mm0
	movq	.ex2,mm1
	movq	.bx1,mm2
	movq	.ex1,mm3
end if
if Ext >= SSE
	movups	xmm0,.bx1
	movups	xmm1,.bx2
	movups	.bx2,xmm0
	movups	.bx1,xmm1
end if
 
	mov	edx,.z1
	xchg	edx,.z2
	mov	.z1,edx
 
  .bl_ok:
 
	push	eax
	push	ebx	      ;store x1, x2
 
	cmp	.x1,SIZE_X
	jge	.bl_end
	cmp	.x2,0
	jle	.bl_end
 
	mov	ebx,.x2
	sub	ebx,.x1
 
if Ext >= SSE
 
       sub	 esp,16
       cvtsi2ss  xmm3,ebx	     ;rcps
       shufps	 xmm3,xmm3,0
 
       cvtpi2ps  xmm0,.bx1 ;mm0
       movlhps	 xmm0,xmm0
       cvtpi2ps  xmm0,.ex1 ;mm2
       cvtpi2ps  xmm1,.bx2 ;mm1
       movlhps	 xmm1,xmm1
       cvtpi2ps  xmm1,.ex2 ;mm3
       subps	 xmm1,xmm0
 
       divps	 xmm1,xmm3
 
       shufps	 xmm1,xmm1,10110001b
       cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
       movhlps	 xmm1,xmm1
       cvtps2pi  mm1,xmm1
       movq	 .dey,mm0
       movq	 .dby,mm1
 
else
 
	mov	eax,.bx2       ; calc .dbx
	sub	eax,.bx1
	cdq
	idiv	ebx
	push	eax
 
	mov	eax,.by2       ; calc .dby
	sub	eax,.by1
	cdq
	idiv	ebx
	push	eax
 
	mov	eax,.ex2       ; calc .dex
	sub	eax,.ex1
	cdq
	idiv	ebx
	push	eax
 
	mov	eax,.ey2       ; calc .dey
	sub	eax,.ey1
	cdq
	idiv	ebx
	push	eax
 
end if
 
	mov	eax,.z2        ; calc .dz
	sub	eax,.z1
	cdq
	idiv	ebx
	push	eax
 
	cmp	.x1,0	      ; set correctly begin variable
	jge	@f	      ; CLIPPING ON FUNCTION
			      ; cutting triangle exceedes screen
	mov	ebx,.x1
	neg	ebx
	imul	ebx	      ; eax = .dz * abs(.x1)
	add	.z1,eax
	mov	.x1,0
 
	mov	eax,.dbx
	imul	ebx
	add    .bx1,eax
 
	mov	eax,.dby
	imul	ebx
	add	.by1,eax
 
	mov	eax,.dex
	imul	ebx
	add	.ex1,eax
 
	mov	eax,.dey
	imul	ebx
	add	.ey1,eax
      @@:
	cmp	.x2,SIZE_X
	jl	@f
	mov	.x2,SIZE_X
      @@:
	mov	eax,SIZE_X	 ;calc memory begin in buffers
	mov	ebx,.y
	mul	ebx
	mov	ebx,.x1
	add	eax,ebx
	mov	ebx,eax
	lea	eax,[eax*3]
	add	edi,eax
	mov	esi,.z_buff	  ; z-buffer filled with dd variables
	shl	ebx,2
	add	esi,ebx
 
	mov	ecx,.x2
	sub	ecx,.x1
	; init current variables
	push	dword .bx1
	push	.by1
	push	dword .ex1
	push	.ey1
 
	push	.z1		 ; current z shl CATMULL_SHIFT
	push	esi
; It's my first attempt at MMX :), have mercy - Macgub
 
;;     if Ext = MMX
;      mov     dword[.temp1],esi
;      mov     dword[.temp1+4],esi
;;        movq    mm0,.cbyq          ; mm0 - current bump coords
;;        movq    mm1,.ceyq          ; mm1 - current env  coords
;;        movq    mm2,.dbyq          ; mm2 - delta bump
;;        movq    mm3,.deyq          ; mm3 - delta env
;     movd    mm6,.z1            ; mm6 - cur z
;     movq    mm7,qword.[temp1]  ; mm7 = lo = hi dword = current z buff
;;        mov     dword [.temp2],1
;;        mov     dword [.temp2+4],-1
;;        mov     dword [.temp3],TEXTURE_SIZE
;;        mov     dword [.temp3+4],TEXTURE_SIZE
;;        mov     esi,.bmap
;;        mov     dword [.temp4],esi
;;        mov     dword [.temp4+4],esi
;;        mov     dword [.temp5],TEX_X
;;        mov     dword [.temp5+4],- TEX_X
;     mov     dword [.temp1],TEX_SHIFT
;     mov     dword [.temp1+4],0
;;     end if
 
     .draw:
    ; if TEX = SHIFTING   ;bump drawing only in shifting mode
	mov	esi,.czbuff	 ; .czbuff current address in buffer
	mov	ebx,.cz 	 ; .cz - cur z position
	cmp	ebx,dword[esi]
	jge	.skip
;;      if Ext=NON
	mov	eax,.cby
	sar	eax,ROUND
	mov	esi,.cbx
	sar	esi,ROUND
;;      else
;;        movq    mm4,mm0          ; mm4 - copies of cur bump coords
;;        psrad   mm4,ROUND        ; mm4 =  lo dword = y b coord, hi dword = x b coord
;;        movd    eax,mm4    ; -
;;        psrlq   mm4,32     ; -
;;        movd    esi,mm4    ; -
;;
;;;       punpckldq  mm5,mm4     ;
;;;       psllq   mm5,TEX_SHIFT
;;;       paddq   mm4,mm5   ; mm4 - lo dword index to b. map
;;
;; ;       packqd  mm4,mm5
 ;       movq    mm5,mm4       ; mm5 ~~ current bump map index?
 
;;      end if
	shl	eax,TEX_SHIFT	;-
	add	esi,eax 	;-  ; esi - current bump map index
;;     if  Ext = NON
	mov	ebx,esi
	dec	ebx
	and	ebx,TEXTURE_SIZE
	add	ebx,.bmap
	movzx	eax,byte [ebx]
 
	mov	ebx,esi
	inc	ebx
	and	ebx,TEXTURE_SIZE
	add	ebx,.bmap
	movzx	ebx,byte [ebx]
;;      else                   ;-------------------------------------------
;;        mov     dword [.temp1],esi       ;-
;;        mov     dword [.temp1+4],esi     ;-
;;        movq    mm5, qword[.temp1]       ;-
;;        paddd   mm5, qword[.temp2]  ; .temp2 == low dword = 1, high dword = -1
;;        pand    mm5, qword[.temp3]  ; .temp3 == low = high dword = TEX_SIZE
;;        paddd   mm5, qword[.temp4]  ; .temp4 == low = high dword = .bmap
;;        movd    ebx,mm5
;;        psrlq   mm5,32
;;        movd    eax,mm5
;;        movzx   ebx,byte[ebx]
;;        movzx   eax,byte[eax]
;;     end if
	sub	eax,ebx
;;     if Ext=NON
	mov	ebx,esi
	sub	ebx,TEX_X
	and	ebx,TEXTURE_SIZE
	add	ebx,.bmap
	movzx	edx,byte [ebx]
 
	mov	ebx,esi
	add	ebx,TEX_X
	and	ebx,TEXTURE_SIZE
	add	ebx,.bmap
	movzx	ebx,byte [ebx]
;;     else
;;        movq    mm5, qword[.temp1]  ;-
;;        paddd   mm5, qword[.temp5]  ; .temp5 == low dword = TEX_X, high dword = -TEX_X
;;        pand    mm5, qword[.temp3]  ; .temp3 == low = high dword = TEX_SIZE
;;        paddd   mm5, qword[.temp4]  ; .temp4 == low = high dword = offset .bmap
;;        movd    ebx,mm5
;;        psrlq   mm5,32
;;        movd    edx,mm5
;;        movzx   ebx,byte[ebx]
;;        movzx   edx,byte[edx]
;;     end if
	sub	edx,ebx
     ;  eax - horizontal sub
     ;  edx - vertical   sub
;;    if Ext=NON
	mov	ebx,.cex       ;.cex - current env map X
	sar	ebx,ROUND
	add	eax,ebx        ; eax - modified x coord
 
	mov	ebx,.cey       ;.cey - current  env map y
	sar	ebx,ROUND
	add	edx,ebx        ; edx - modified y coord
;;    else
;;        movq    mm5,mm1        ; mm5 - copy of cur env coords
;;        psrad   mm5,ROUND
;;        movq    qword[.temp1],mm5
;;        add     eax,dword [.temp1]
;;        add     edx,dword [.temp1+4]
;;      ;  movd    ebx,mm5
;;      ;  add     eax,ebx
;;      ;  psrlq   mm5,32
;;      ;  movd    ebx,mm5
      ;  add     edx,ebx
;;    end if
 
	or	eax,eax
	jl	.black
	cmp	eax,TEX_X
	jg	.black
	or	edx,edx
	jl	.black
	cmp	edx,TEX_Y
	jg	.black
 
	shl	edx,TEX_SHIFT
	add	edx,eax
	lea	edx,[edx*3]
	add	edx,.emap
	mov	eax,dword[edx]
	jmp	.put_pixel
     .black:
	xor	eax,eax
     .put_pixel:
	stosd
	dec	edi
	mov	ebx,.cz
	mov	esi,.czbuff
	mov	dword[esi],ebx
	jmp	.no_skip
     .skip:
	add	edi,3
     .no_skip:
	add	.czbuff,4
;;   if Ext = NON
if Ext >= MMX
	movq	mm0,.cby
	movq	mm1,.cey
	paddd	mm0,.dby
	paddd	mm1,.dey
	movq	.cby,mm0
	movq	.cey,mm1
else
	mov	eax,.dbx
	add	.cbx,eax
	mov	eax,.dby
	add	.cby,eax
	mov	eax,.dex
	add	.cex,eax
	mov	eax,.dey
	add	.cey,eax
end if
;;    else
;;        paddd   mm0,mm2
;;        paddd   mm1,mm3
;;    end if
	mov	eax,.dz
	add	.cz,eax
 
	dec	ecx
	jnz	.draw
   ;   end if
  .bl_end:
	mov	esp,ebp
ret 56

Rev 1776	Rev 1931
1	;SIZE_X equ 350	1	;SIZE_X equ 350
2	;SIZE_Y equ 350	2	;SIZE_Y equ 350
3	ROUND equ 8	3	ROUND equ 8
4	;TEX_X equ 512	4	;TEX_X equ 512
5	;TEX_Y equ 512	5	;TEX_Y equ 512
6	;TEXTURE_SIZE EQU (512*512)-1	6	;TEXTURE_SIZE EQU (512*512)-1
7	;TEX_SHIFT EQU 9	7	;TEX_SHIFT EQU 9
8	CATMULL_SHIFT equ 8	8	CATMULL_SHIFT equ 8
9	;TEXTURE_SIZE EQU (TEX_X * TEX_Y)-1	9	;TEXTURE_SIZE EQU (TEX_X * TEX_Y)-1
10	;Ext = NON	10	;Ext = NON
11	;MMX = 1	11	;MMX = 1
12	;NON = 0	12	;NON = 0
13	;------- Big thanks to Majuma (www.majuma.xt.pl) for absolutely great---	13	;------- Big thanks to Majuma (www.majuma.xt.pl) for absolutely great---
14	;------- DOS 13h mode demos --------------------------------------------	14	;------- DOS 13h mode demos --------------------------------------------
15	;------- Procedure draws bump triangle using Catmull Z-buffer algorithm-	15	;------- Procedure draws bump triangle using Catmull Z-buffer algorithm-
16	;------- (Z coordinate interpolation)-----------------------------------	16	;------- (Z coordinate interpolation)-----------------------------------
17	bump_triangle_z:	17	bump_triangle_z:
18	;------------------in - eax - x1 shl 16 + y1 -----------	18	;------------------in - eax - x1 shl 16 + y1 -----------
19	;---------------------- ebx - x2 shl 16 + y2 -----------	19	;---------------------- ebx - x2 shl 16 + y2 -----------
20	;---------------------- ecx - x3 shl 16 + y3 -----------	20	;---------------------- ecx - x3 shl 16 + y3 -----------
21	;---------------------- edx - pointer to bump map ------	21	;---------------------- edx - pointer to bump map ------
22	;---------------------- esi - pointer to environment map	22	;---------------------- esi - pointer to environment map
23	;---------------------- edi - pointer to screen buffer--	23	;---------------------- edi - pointer to screen buffer--
24	;---------------------- stack : bump coordinates--------	24	;---------------------- stack : bump coordinates--------
25	;---------------------- environment coordinates-	25	;---------------------- environment coordinates-
26	;---------------------- Z position coordinates--	26	;---------------------- Z position coordinates--
27	;---------------------- pointer io Z buffer-----	27	;---------------------- pointer io Z buffer-----
28	;-- Z-buffer - filled with coordinates as dword --------	28	;-- Z-buffer - filled with coordinates as dword --------
29	;-- (Z coor. as word) shl CATMULL_SHIFT ----------------	29	;-- (Z coor. as word) shl CATMULL_SHIFT ----------------
30	.b_x1 equ ebp+4 ; procedure don't save registers !!!	30	.b_x1 equ ebp+4 ; procedure don't save registers !!!
31	.b_y1 equ ebp+6 ; each coordinate as word	31	.b_y1 equ ebp+6 ; each coordinate as word
32	.b_x2 equ ebp+8	32	.b_x2 equ ebp+8
33	.b_y2 equ ebp+10	33	.b_y2 equ ebp+10
34	.b_x3 equ ebp+12	34	.b_x3 equ ebp+12
35	.b_y3 equ ebp+14	35	.b_y3 equ ebp+14
36	.e_x1 equ ebp+16	36	.e_x1 equ ebp+16
37	.e_y1 equ ebp+18	37	.e_y1 equ ebp+18
38	.e_x2 equ ebp+20	38	.e_x2 equ ebp+20
39	.e_y2 equ ebp+22	39	.e_y2 equ ebp+22
40	.e_x3 equ ebp+24	40	.e_x3 equ ebp+24
41	.e_y3 equ ebp+26	41	.e_y3 equ ebp+26
42	.z1 equ word[ebp+28]	42	.z1 equ word[ebp+28]
43	.z2 equ word[ebp+30]	43	.z2 equ word[ebp+30]
44	.z3 equ word[ebp+32]	44	.z3 equ word[ebp+32]
45	.z_buff equ dword[ebp+34] ; pointer to Z-buffer	45	.z_buff equ dword[ebp+34] ; pointer to Z-buffer
46		46
47		47
48	.t_bmap equ dword[ebp-4] ; pointer to bump map	48	.t_bmap equ dword[ebp-4] ; pointer to bump map
49	.t_emap equ dword[ebp-8] ; pointer to e. map	49	.t_emap equ dword[ebp-8] ; pointer to e. map
50	.x1 equ word[ebp-10]	50	.x1 equ word[ebp-10]
51	.y1 equ word[ebp-12]	51	.y1 equ word[ebp-12]
52	.x2 equ word[ebp-14]	52	.x2 equ word[ebp-14]
53	.y2 equ word[ebp-16]	53	.y2 equ word[ebp-16]
54	.x3 equ word[ebp-18]	54	.x3 equ word[ebp-18]
55	.y3 equ word[ebp-20]	55	.y3 equ word[ebp-20]
56		56
57	.dx12 equ dword[ebp-24]	57	.dx12 equ dword[ebp-24]
58	.dz12 equ [ebp-28]	58	.dz12 equ [ebp-28]
59	.dbx12 equ dword[ebp-32]	59	.dbx12 equ dword[ebp-32]
60	.dby12 equ [ebp-36]	60	.dby12 equ [ebp-36]
61	.dex12 equ dword[ebp-40]	61	.dex12 equ dword[ebp-40]
62	.dey12 equ [ebp-44]	62	.dey12 equ [ebp-44]
63		63
64	.dx13 equ dword[ebp-48]	64	.dx13 equ dword[ebp-48]
65	.dz13 equ [ebp-52]	65	.dz13 equ [ebp-52]
66	.dbx13 equ dword[ebp-56]	66	.dbx13 equ dword[ebp-56]
67	.dby13 equ [ebp-60]	67	.dby13 equ [ebp-60]
68	.dex13 equ dword[ebp-64]	68	.dex13 equ dword[ebp-64]
69	.dey13 equ [ebp-68]	69	.dey13 equ [ebp-68]
70		70
71	.dx23 equ dword[ebp-72]	71	.dx23 equ dword[ebp-72]
72	.dz23 equ [ebp-76]	72	.dz23 equ [ebp-76]
73	.dbx23 equ dword[ebp-80]	73	.dbx23 equ dword[ebp-80]
74	.dby23 equ [ebp-84]	74	.dby23 equ [ebp-84]
75	.dex23 equ dword[ebp-88]	75	.dex23 equ dword[ebp-88]
76	.dey23 equ [ebp-92]	76	.dey23 equ [ebp-92]
77		77
78	.cx1 equ dword[ebp-96] ; current variables	78	.cx1 equ dword[ebp-96] ; current variables
79	.cz1 equ [ebp-100]	79	.cz1 equ [ebp-100]
80	.cx2 equ dword[ebp-104]	80	.cx2 equ dword[ebp-104]
81	.cz2 equ [ebp-108]	81	.cz2 equ [ebp-108]
82	.cbx1 equ dword[ebp-112]	82	.cbx1 equ dword[ebp-112]
83	.cby1 equ [ebp-116]	83	.cby1 equ [ebp-116]
84	.cbx2 equ dword[ebp-120]	84	.cbx2 equ dword[ebp-120]
85	.cby2 equ [ebp-124]	85	.cby2 equ [ebp-124]
86	.cex1 equ dword[ebp-128]	86	.cex1 equ dword[ebp-128]
87	.cey1 equ [ebp-132]	87	.cey1 equ [ebp-132]
88	.cex2 equ dword[ebp-136]	88	.cex2 equ dword[ebp-136]
89	.cey2 equ [ebp-140]	89	.cey2 equ [ebp-140]
90		90
91	mov ebp,esp	91	mov ebp,esp
92	push edx ; store bump map	92	push edx ; store bump map
93	push esi ; store e. map	93	push esi ; store e. map
94	; sub esp,120	94	; sub esp,120
95	.sort3: ; sort triangle coordinates...	95	.sort3: ; sort triangle coordinates...
96	cmp ax,bx	96	cmp ax,bx
97	jle .sort1	97	jle .sort1
98	xchg eax,ebx	98	xchg eax,ebx
99	mov edx,dword[.b_x1]	99	mov edx,dword[.b_x1]
100	xchg edx,dword[.b_x2]	100	xchg edx,dword[.b_x2]
101	mov dword[.b_x1],edx	101	mov dword[.b_x1],edx
102	mov edx,dword[.e_x1]	102	mov edx,dword[.e_x1]
103	xchg edx,dword[.e_x2]	103	xchg edx,dword[.e_x2]
104	mov dword[.e_x1],edx	104	mov dword[.e_x1],edx
105	mov dx,.z1	105	mov dx,.z1
106	xchg dx,.z2	106	xchg dx,.z2
107	mov .z1,dx	107	mov .z1,dx
108	.sort1:	108	.sort1:
109	cmp bx,cx	109	cmp bx,cx
110	jle .sort2	110	jle .sort2
111	xchg ebx,ecx	111	xchg ebx,ecx
112	mov edx,dword[.b_x2]	112	mov edx,dword[.b_x2]
113	xchg edx,dword[.b_x3]	113	xchg edx,dword[.b_x3]
114	mov dword[.b_x2],edx	114	mov dword[.b_x2],edx
115	mov edx,dword[.e_x2]	115	mov edx,dword[.e_x2]
116	xchg edx,dword[.e_x3]	116	xchg edx,dword[.e_x3]
117	mov dword[.e_x2],edx	117	mov dword[.e_x2],edx
118	mov dx,.z2	118	mov dx,.z2
119	xchg dx,.z3	119	xchg dx,.z3
120	mov .z2,dx	120	mov .z2,dx
121	jmp .sort3	121	jmp .sort3
122	.sort2:	122	.sort2:
123	push eax ; store triangle coords in variables	123	push eax ; store triangle coords in variables
124	push ebx	124	push ebx
125	push ecx	125	push ecx
126		126
127	mov edx,80008000h ; eax,ebx,ecx are ANDd together into edx which means that	127	mov edx,80008000h ; eax,ebx,ecx are ANDd together into edx which means that
128	and edx,ebx ; if all of them are negative a sign flag is raised	128	and edx,ebx ; if all of them are negative a sign flag is raised
129	and edx,ecx	129	and edx,ecx
130	and edx,eax	130	and edx,eax
131	test edx,80008000h ; Check both X&Y at once	131	test edx,80008000h ; Check both X&Y at once
132	jne .loop23_done	132	jne .loop23_done
133	; mov edx,eax ; eax,ebx,ecx are ORd together into edx which means that	133	; mov edx,eax ; eax,ebx,ecx are ORd together into edx which means that
134	; or edx,ebx ; if any one of them is negative a sign flag is raised	134	; or edx,ebx ; if any one of them is negative a sign flag is raised
135	; or edx,ecx	135	; or edx,ecx
136	; test edx,80000000h ; Check only X	136	; test edx,80000000h ; Check only X
137	; jne .loop23_done	137	; jne .loop23_done
138		138
139	; cmp .x1,SIZE_X ; {	139	; cmp .x1,SIZE_X ; {
140	; jg .loop23_done	140	; jg .loop23_done
141	; cmp .x2,SIZE_X ; This can be optimized with effort	141	; cmp .x2,SIZE_X ; This can be optimized with effort
142	; jg .loop23_done	142	; jg .loop23_done
143	; cmp .x3,SIZE_X	143	; cmp .x3,SIZE_X
144	; jg .loop23_done ; {	144	; jg .loop23_done ; {
145		145
146		146
147	mov bx,.y2 ; calc delta 12	147	mov bx,.y2 ; calc delta 12
148	sub bx,.y1	148	sub bx,.y1
149	jnz .bt_dx12_make	149	jnz .bt_dx12_make
150	mov ecx,6	150	mov ecx,6
151	xor edx,edx	151	xor edx,edx
152	@@:	152	@@:
153	push edx ;dword 0	153	push edx ;dword 0
154	loop @b	154	loop @b
155	jmp .bt_dx12_done	155	jmp .bt_dx12_done
156	.bt_dx12_make:	156	.bt_dx12_make:
157	mov ax,.x2	157	mov ax,.x2
158	sub ax,.x1	158	sub ax,.x1
159	cwde	159	cwde
160	movsx ebx,bx	160	movsx ebx,bx
161	shl eax,ROUND	161	shl eax,ROUND
162	cdq	162	cdq
163	idiv ebx	163	idiv ebx
164	; mov .dx12,eax	164	; mov .dx12,eax
165	push eax	165	push eax
166		166
167	mov ax,.z2	167	mov ax,.z2
168	sub ax,.z1	168	sub ax,.z1
169	cwde	169	cwde
170	shl eax,CATMULL_SHIFT	170	shl eax,CATMULL_SHIFT
171	cdq	171	cdq
172	idiv ebx	172	idiv ebx
173	push eax	173	push eax
174	if Ext>=SSE	174	if Ext>=SSE
175		175
176	sub esp,16	176	sub esp,16
177	cvtsi2ss xmm3,ebx ;rcps	177	cvtsi2ss xmm3,ebx ;rcps
178	; mov eax,255	178	; mov eax,255
179	cvtsi2ss xmm4,[i255d]	179	cvtsi2ss xmm4,[i255d]
180	divss xmm3,xmm4	180	divss xmm3,xmm4
181	rcpss xmm3,xmm3	181	rcpss xmm3,xmm3
182	; mulss xmm3,xmm4	182	; mulss xmm3,xmm4
183	shufps xmm3,xmm3,0	183	shufps xmm3,xmm3,0
184		184
185	movd mm0,[.b_x1]	185	movd mm0,[.b_x1]
186	movd mm1,[.b_x2]	186	movd mm1,[.b_x2]
187	movd mm2,[.e_x1]	187	movd mm2,[.e_x1]
188	movd mm3,[.e_x2]	188	movd mm3,[.e_x2]
189	pxor mm4,mm4	189	pxor mm4,mm4
190	punpcklwd mm0,mm4	190	punpcklwd mm0,mm4
191	punpcklwd mm1,mm4	191	punpcklwd mm1,mm4
192	punpcklwd mm2,mm4	192	punpcklwd mm2,mm4
193	punpcklwd mm3,mm4	193	punpcklwd mm3,mm4
194		194
195	psubd mm1,mm0	195	psubd mm1,mm0
196	psubd mm3,mm2	196	psubd mm3,mm2
197		197
198	; cvtpi2ps xmm0,mm0	198	; cvtpi2ps xmm0,mm0
199	; movlhps xmm0,xmm0	199	; movlhps xmm0,xmm0
200	; cvtpi2ps xmm0,mm2	200	; cvtpi2ps xmm0,mm2
201	cvtpi2ps xmm1,mm1	201	cvtpi2ps xmm1,mm1
202	movlhps xmm1,xmm1	202	movlhps xmm1,xmm1
203	cvtpi2ps xmm1,mm3	203	cvtpi2ps xmm1,mm3
204	; subps xmm1,xmm0	204	; subps xmm1,xmm0
205		205
206		206
207	; cvtpi2ps xmm0,mm3	207	; cvtpi2ps xmm0,mm3
208	; divps xmm1,xmm3	208	; divps xmm1,xmm3
209	mulps xmm1,xmm3	209	mulps xmm1,xmm3
210	shufps xmm1,xmm1,10110001b	210	shufps xmm1,xmm1,10110001b
211	cvtps2pi mm0,xmm1 ; mm0 -> 2 delta dwords	211	cvtps2pi mm0,xmm1 ; mm0 -> 2 delta dwords
212	movhlps xmm1,xmm1	212	movhlps xmm1,xmm1
213	cvtps2pi mm1,xmm1	213	cvtps2pi mm1,xmm1
214	movq .dey12,mm0	214	movq .dey12,mm0
215	movq .dby12,mm1	215	movq .dby12,mm1
216		216
217		217
218	else	218	else
219		219
220	mov ax,word[.b_x2]	220	mov ax,word[.b_x2]
221	sub ax,word[.b_x1]	221	sub ax,word[.b_x1]
222	cwde	222	cwde
223	shl eax,ROUND	223	shl eax,ROUND
224	cdq	224	cdq
225	idiv ebx	225	idiv ebx
226	; mov .dbx12,eax	226	; mov .dbx12,eax
227	push eax	227	push eax
228		228
229	mov ax,word[.b_y2]	229	mov ax,word[.b_y2]
230	sub ax,word[.b_y1]	230	sub ax,word[.b_y1]
231	cwde	231	cwde
232	shl eax,ROUND	232	shl eax,ROUND
233	cdq	233	cdq
234	idiv ebx	234	idiv ebx
235	; mov .dby12,eax	235	; mov .dby12,eax
236	push eax	236	push eax
237		237
238	mov ax,word[.e_x2]	238	mov ax,word[.e_x2]
239	sub ax,word[.e_x1]	239	sub ax,word[.e_x1]
240	cwde	240	cwde
241	shl eax,ROUND	241	shl eax,ROUND
242	cdq	242	cdq
243	idiv ebx	243	idiv ebx
244	; mov .dex12,eax	244	; mov .dex12,eax
245	push eax	245	push eax
246		246
247	mov ax,word[.e_y2]	247	mov ax,word[.e_y2]
248	sub ax,word[.e_y1]	248	sub ax,word[.e_y1]
249	cwde	249	cwde
250	shl eax,ROUND	250	shl eax,ROUND
251	cdq	251	cdq
252	idiv ebx	252	idiv ebx
253	; mov .dey12,eax	253	; mov .dey12,eax
254	push eax	254	push eax
255		255
256	end if	256	end if
257		257
258	.bt_dx12_done:	258	.bt_dx12_done:
259		259
260	mov bx,.y3 ; calc delta13	260	mov bx,.y3 ; calc delta13
261	sub bx,.y1	261	sub bx,.y1
262	jnz .bt_dx13_make	262	jnz .bt_dx13_make
263	mov ecx,6	263	mov ecx,6
264	xor edx,edx	264	xor edx,edx
265	@@:	265	@@:
266	push edx ;dword 0	266	push edx ;dword 0
267	loop @b	267	loop @b
268	jmp .bt_dx13_done	268	jmp .bt_dx13_done
269	.bt_dx13_make:	269	.bt_dx13_make:
270	mov ax,.x3	270	mov ax,.x3
271	sub ax,.x1	271	sub ax,.x1
272	cwde	272	cwde
273	movsx ebx,bx	273	movsx ebx,bx
274	shl eax,ROUND	274	shl eax,ROUND
275	cdq	275	cdq
276	idiv ebx	276	idiv ebx
277	; mov .dx13,eax	277	; mov .dx13,eax
278	push eax	278	push eax
279		279
280	mov ax,.z3	280	mov ax,.z3
281	sub ax,.z1	281	sub ax,.z1
282	cwde	282	cwde
283	shl eax,CATMULL_SHIFT	283	shl eax,CATMULL_SHIFT
284	cdq	284	cdq
285	idiv ebx	285	idiv ebx
286	; mov .dz13,eax	286	; mov .dz13,eax
287	push eax	287	push eax
288		288
289	if Ext>=SSE	289	if Ext>=SSE
290		290
291	sub esp,16	291	sub esp,16
292	cvtsi2ss xmm3,ebx ;rcps	292	cvtsi2ss xmm3,ebx ;rcps
293	; mov eax,255	293	; mov eax,255
294	cvtsi2ss xmm4,[i255d]	294	cvtsi2ss xmm4,[i255d]
295	divss xmm3,xmm4	295	divss xmm3,xmm4
296	rcpss xmm3,xmm3	296	rcpss xmm3,xmm3
297	; mulss xmm3,xmm4	297	; mulss xmm3,xmm4
298	shufps xmm3,xmm3,0	298	shufps xmm3,xmm3,0
299		299
300	movd mm0,[.b_x1]	300	movd mm0,[.b_x1]
301	movd mm1,[.b_x3]	301	movd mm1,[.b_x3]
302	movd mm2,[.e_x1]	302	movd mm2,[.e_x1]
303	movd mm3,[.e_x3]	303	movd mm3,[.e_x3]
304	pxor mm4,mm4	304	pxor mm4,mm4
305	punpcklwd mm0,mm4	305	punpcklwd mm0,mm4
306	punpcklwd mm1,mm4	306	punpcklwd mm1,mm4
307	punpcklwd mm2,mm4	307	punpcklwd mm2,mm4
308	punpcklwd mm3,mm4	308	punpcklwd mm3,mm4
309		309
310	psubd mm1,mm0	310	psubd mm1,mm0
311	psubd mm3,mm2	311	psubd mm3,mm2
312		312
313	; cvtpi2ps xmm0,mm0	313	; cvtpi2ps xmm0,mm0
314	; movlhps xmm0,xmm0	314	; movlhps xmm0,xmm0
315	; cvtpi2ps xmm0,mm2	315	; cvtpi2ps xmm0,mm2
316	cvtpi2ps xmm1,mm1	316	cvtpi2ps xmm1,mm1
317	movlhps xmm1,xmm1	317	movlhps xmm1,xmm1
318	cvtpi2ps xmm1,mm3	318	cvtpi2ps xmm1,mm3
319	; subps xmm1,xmm0	319	; subps xmm1,xmm0
320		320
321		321
322	; cvtpi2ps xmm0,mm3	322	; cvtpi2ps xmm0,mm3
323	; divps xmm1,xmm3	323	; divps xmm1,xmm3
324	mulps xmm1,xmm3	324	mulps xmm1,xmm3
325	shufps xmm1,xmm1,10110001b	325	shufps xmm1,xmm1,10110001b
326	cvtps2pi mm0,xmm1 ; mm0 -> 2 delta dwords	326	cvtps2pi mm0,xmm1 ; mm0 -> 2 delta dwords
327	movhlps xmm1,xmm1	327	movhlps xmm1,xmm1
328	cvtps2pi mm1,xmm1	328	cvtps2pi mm1,xmm1
329	movq .dey13,mm0	329	movq .dey13,mm0
330	movq .dby13,mm1	330	movq .dby13,mm1
331		331
332		332
333	else	333	else
334		334
335	mov ax,word[.b_x3]	335	mov ax,word[.b_x3]
336	sub ax,word[.b_x1]	336	sub ax,word[.b_x1]
337	cwde	337	cwde
338	shl eax,ROUND	338	shl eax,ROUND
339	cdq	339	cdq
340	idiv ebx	340	idiv ebx
341	; mov .dbx13,eax	341	; mov .dbx13,eax
342	push eax	342	push eax
343		343
344	mov ax,word[.b_y3]	344	mov ax,word[.b_y3]
345	sub ax,word[.b_y1]	345	sub ax,word[.b_y1]
346	cwde	346	cwde
347	shl eax,ROUND	347	shl eax,ROUND
348	cdq	348	cdq
349	idiv ebx	349	idiv ebx
350	; mov .dby13,eax	350	; mov .dby13,eax
351	push eax	351	push eax
352		352
353	mov ax,word[.e_x3]	353	mov ax,word[.e_x3]
354	sub ax,word[.e_x1]	354	sub ax,word[.e_x1]
355	cwde	355	cwde
356	shl eax,ROUND	356	shl eax,ROUND
357	cdq	357	cdq
358	idiv ebx	358	idiv ebx
359	; mov .dex13,eax	359	; mov .dex13,eax
360	push eax	360	push eax
361		361
362	mov ax,word[.e_y3]	362	mov ax,word[.e_y3]
363	sub ax,word[.e_y1]	363	sub ax,word[.e_y1]
364	cwde	364	cwde
365	shl eax,ROUND	365	shl eax,ROUND
366	cdq	366	cdq
367	idiv ebx	367	idiv ebx
368	; mov .dey13,eax	368	; mov .dey13,eax
369	push eax	369	push eax
370	end if	370	end if
371		371
372	.bt_dx13_done:	372	.bt_dx13_done:
373		373
374	mov bx,.y3 ; calc delta23	374	mov bx,.y3 ; calc delta23
375	sub bx,.y2	375	sub bx,.y2
376	jnz .bt_dx23_make	376	jnz .bt_dx23_make
377	mov ecx,6	377	mov ecx,6
378	xor edx,edx	378	xor edx,edx
379	@@:	379	@@:
380	push edx ;dword 0	380	push edx ;dword 0
381	loop @b	381	loop @b
382	jmp .bt_dx23_done	382	jmp .bt_dx23_done
383	.bt_dx23_make:	383	.bt_dx23_make:
384	mov ax,.x3	384	mov ax,.x3
385	sub ax,.x2	385	sub ax,.x2
386	cwde	386	cwde
387	movsx ebx,bx	387	movsx ebx,bx
388	shl eax,ROUND	388	shl eax,ROUND
389	cdq	389	cdq
390	idiv ebx	390	idiv ebx
391	; mov .dx23,eax	391	; mov .dx23,eax
392	push eax	392	push eax
393		393
394	mov ax,.z3	394	mov ax,.z3
395	sub ax,.z2	395	sub ax,.z2
396	cwde	396	cwde
397	shl eax,CATMULL_SHIFT	397	shl eax,CATMULL_SHIFT
398	cdq	398	cdq
399	idiv ebx	399	idiv ebx
400	; mov .dz23,eax	400	; mov .dz23,eax
401	push eax	401	push eax
402	; sub esp,40	402	; sub esp,40
403	if Ext>=SSE	403	if Ext>=SSE
404		404
405	sub esp,16	405	sub esp,16
406	cvtsi2ss xmm3,ebx ;rcps	406	cvtsi2ss xmm3,ebx ;rcps
407	; mov eax,255	407	; mov eax,255
408	cvtsi2ss xmm4,[i255d]	408	cvtsi2ss xmm4,[i255d]
409	divss xmm3,xmm4	409	divss xmm3,xmm4
410	rcpss xmm3,xmm3	410	rcpss xmm3,xmm3
411	; mulss xmm3,xmm4	411	; mulss xmm3,xmm4
412	shufps xmm3,xmm3,0	412	shufps xmm3,xmm3,0
413		413
414	movd mm0,[.b_x2]	414	movd mm0,[.b_x2]
415	movd mm1,[.b_x3]	415	movd mm1,[.b_x3]
416	movd mm2,[.e_x2]	416	movd mm2,[.e_x2]
417	movd mm3,[.e_x3]	417	movd mm3,[.e_x3]
418	pxor mm4,mm4	418	pxor mm4,mm4
419	punpcklwd mm0,mm4	419	punpcklwd mm0,mm4
420	punpcklwd mm1,mm4	420	punpcklwd mm1,mm4
421	punpcklwd mm2,mm4	421	punpcklwd mm2,mm4
422	punpcklwd mm3,mm4	422	punpcklwd mm3,mm4
423		423
424	psubd mm1,mm0	424	psubd mm1,mm0
425	psubd mm3,mm2	425	psubd mm3,mm2
426		426
427	; cvtpi2ps xmm0,mm0	427	; cvtpi2ps xmm0,mm0
428	; movlhps xmm0,xmm0	428	; movlhps xmm0,xmm0
429	; cvtpi2ps xmm0,mm2	429	; cvtpi2ps xmm0,mm2
430	cvtpi2ps xmm1,mm1	430	cvtpi2ps xmm1,mm1
431	movlhps xmm1,xmm1	431	movlhps xmm1,xmm1
432	cvtpi2ps xmm1,mm3	432	cvtpi2ps xmm1,mm3
433	; subps xmm1,xmm0	433	; subps xmm1,xmm0
434		434
435		435
436	; cvtpi2ps xmm0,mm3	436	; cvtpi2ps xmm0,mm3
437	; divps xmm1,xmm3	437	; divps xmm1,xmm3
438	mulps xmm1,xmm3	438	mulps xmm1,xmm3
439	shufps xmm1,xmm1,10110001b	439	shufps xmm1,xmm1,10110001b
440	cvtps2pi mm0,xmm1 ; mm0 -> 2 delta dwords	440	cvtps2pi mm0,xmm1 ; mm0 -> 2 delta dwords
441	movhlps xmm1,xmm1	441	movhlps xmm1,xmm1
442	cvtps2pi mm1,xmm1	442	cvtps2pi mm1,xmm1
443	movq .dey23,mm0	443	movq .dey23,mm0
444	movq .dby23,mm1	444	movq .dby23,mm1
445		445
446	else	446	else
447		447
448	mov ax,word[.b_x3]	448	mov ax,word[.b_x3]
449	sub ax,word[.b_x2]	449	sub ax,word[.b_x2]
450	cwde	450	cwde
451	shl eax,ROUND	451	shl eax,ROUND
452	cdq	452	cdq
453	idiv ebx	453	idiv ebx
454	; mov .dbx23,eax	454	; mov .dbx23,eax
455	push eax	455	push eax
456		456
457	mov ax,word[.b_y3]	457	mov ax,word[.b_y3]
458	sub ax,word[.b_y2]	458	sub ax,word[.b_y2]
459	cwde	459	cwde
460	shl eax,ROUND	460	shl eax,ROUND
461	cdq	461	cdq
462	idiv ebx	462	idiv ebx
463	; mov .dby23,eax	463	; mov .dby23,eax
464	push eax	464	push eax
465		465
466	mov ax,word[.e_x3]	466	mov ax,word[.e_x3]
467	sub ax,word[.e_x2]	467	sub ax,word[.e_x2]
468	cwde	468	cwde
469	shl eax,ROUND	469	shl eax,ROUND
470	cdq	470	cdq
471	idiv ebx	471	idiv ebx
472	; mov .dex23,eax	472	; mov .dex23,eax
473	push eax	473	push eax
474		474
475	mov ax,word[.e_y3]	475	mov ax,word[.e_y3]
476	sub ax,word[.e_y2]	476	sub ax,word[.e_y2]
477	cwde	477	cwde
478	shl eax,ROUND	478	shl eax,ROUND
479	cdq	479	cdq
480	idiv ebx	480	idiv ebx
481	; mov .dey23,eax	481	; mov .dey23,eax
482	push eax	482	push eax
483		483
484	end if	484	end if
485		485
486	.bt_dx23_done:	486	.bt_dx23_done:
487	sub esp,48	487	sub esp,48
488		488
489	movsx eax,.x1	489	movsx eax,.x1
490	shl eax,ROUND	490	shl eax,ROUND
491	mov .cx1,eax	491	mov .cx1,eax
492	mov .cx2,eax	492	mov .cx2,eax
493	; push eax	493	; push eax
494	; push eax	494	; push eax
495		495
496	movsx eax,word[.b_x1]	496	movsx eax,word[.b_x1]
497	shl eax,ROUND	497	shl eax,ROUND
498	mov .cbx1,eax	498	mov .cbx1,eax
499	mov .cbx2,eax	499	mov .cbx2,eax
500	; push eax	500	; push eax

Subversion Repositories Kolibri OS

(root)/programs/demos/3DS/BUMP_CAT.INC @ 1776 – Rev 1776 → 1931