Updated 'View3ds' sources till ver. 0.055. Added revision history file. Added some more 3DS example objects.

git-svn-id: svn://kolibrios.org@1819 a494cfbc-eb01-0410-851d-a64ba20cac60

programs/demos/3DS/BUMP_TEX.INC

@@ -88,6 +88,8 @@ bump_tex_triangle_z:
 .dtx23 equ dword[ebp-(52+4*15)]
 .dty23 equ	[ebp-(52+4*16)]
 
+if Ext < SSE
+
 .cx1   equ dword[ebp-(52+4*17)] 	; current variables
 .cz1   equ	[ebp-(52+4*18)]
 .cx2   equ dword[ebp-(52+4*19)]
@@ -106,6 +108,27 @@ bump_tex_triangle_z:
 .ctx2  equ dword[ebp-(52+4*31)]
 .cty2  equ	[ebp-(52+4*32)]
 
+else
+
+.cx1   equ dword[ebp-(52+4*17)] 	; current variables
+.cz1   equ	[ebp-(52+4*18)]
+.cbx1  equ dword[ebp-(52+4*19)]
+.cby1  equ	[ebp-(52+4*20)]
+.cex1  equ dword[ebp-(52+4*21)]
+.cey1  equ	[ebp-(52+4*22)]
+.ctx1  equ dword[ebp-(52+4*23)]
+.cty1  equ	[ebp-(52+4*24)]
+
+.cx2   equ dword[ebp-(52+4*25)]
+.cz2   equ	[ebp-(52+4*26)]
+.cbx2  equ dword[ebp-(52+4*27)]
+.cby2  equ	[ebp-(52+4*28)]
+.cex2  equ dword[ebp-(52+4*29)]
+.cey2  equ	[ebp-(52+4*30)]
+.ctx2  equ dword[ebp-(52+4*31)]
+.cty2  equ	[ebp-(52+4*32)]
+
+end if
        cld
        mov     ebp,esp
        push    edx	  ; store bump map
@@ -180,18 +203,9 @@ bump_tex_triangle_z:
  .bt_dx12_make:
        movsx	ebx,bx
 
-       mov	ax,.x2
-       sub	ax,.x1
-       cwde
-       shl	eax,ROUND
-       cdq
-       idiv	ebx
- ;     mov      .dx12,eax
-       push	 eax
 
 if Ext>=SSE
-
+       sub	 esp,32
-       sub	 esp,28
    ;    mov       eax,256
        cvtsi2ss  xmm4,[i255d]
        cvtsi2ss  xmm3,ebx	     ;rcps
@@ -220,6 +234,7 @@ if Ext>=SSE
 
        shufps	 xmm1,xmm1,10110001b
 			     ;xmm1--> | dbx | dby | dex | dey |
+;1       movups    .dey12,xmm1
        cvtps2pi  mm0,xmm1 ;mm0,xmm1          ; mm0 -> 2 delta dwords
        movhlps	 xmm1,xmm1
        cvtps2pi  mm1,xmm1 ;mm1,xmm1
@@ -235,6 +250,11 @@ if Ext>=SSE
        mov	 ax,.z2
        sub	 ax,.z1
        cwde
+
+       mov	dx,.x2
+       sub	dx,.x1
+       movsx	edx,dx
+
    ;/    movd      mm1,eax
 
    ;/    punpcklwd  mm0,mm4
@@ -256,17 +276,40 @@ if Ext>=SSE
    ;/  cvtpi2ps  xmm1,mm1
        cvtsi2ss  xmm1,eax
        movlhps	 xmm1,xmm1
+       cvtsi2ss  xmm1,edx
+   ;    movss     xmm1,xmm4
+       shufps	 xmm1,xmm1,00101111b
        cvtpi2ps  xmm1,mm3
 
-       divps	 xmm1,xmm3   ; xmm1--> | xxx | dz | dty | dtx |
+       divps	 xmm1,xmm3   ; xmm1--> | dx | dz | dty | dtx |
 
        shufps	 xmm1,xmm1,11100001b
-			     ; xmm1--> | xxx | dz | dtx | dty |
+			     ; xmm1--> | dx | dz | dtx | dty |
+;1       movlps    .dty12,xmm1
+;1       movhps    .dz12,xmm1
        cvtps2pi  mm0,xmm1    ; mm0 -> 2 delta dwords  | dtx | dty |
        movhlps	 xmm1,xmm1
-       cvtss2si  eax,xmm1    ; eax --> 1 delta dword | dz |
+       cvtps2pi  mm1,xmm1
        movq	 .dty12,mm0
-       mov	 .dz12,eax
+       movq	 .dz12,mm1
+;temporaly conversion to keep upside down
+;     fninit
+;     fld      .dx12
+;     fistp    .dx12
+;     fld      dword .dz12
+;     fistp    dword .dz12
+;     fld      .dbx12
+;     fistp    .dbx12
+;     fld      dword .dby12
+;     fistp    dword .dby12
+;     fld      .dex12
+;     fistp    .dex12
+;     fld      dword .dey12
+;     fistp    dword .dey12
+;     fld      .dtx12
+;     fistp    .dtx12
+;     fld      dword .dty12
+;     fistp    dword .dty12
 ;----
 ;       mov       ax,.z2
 ;       sub       ax,.z1
@@ -317,6 +360,15 @@ if Ext>=SSE
 ;       movq      .dty12,mm0
 ;       movq      .dz12,mm1
 else
+       mov	ax,.x2
+       sub	ax,.x1
+       cwde
+       shl	eax,ROUND
+       cdq
+       idiv	ebx
+ ;     mov      .dx12,eax
+       push	 eax
+
        mov     ax,.z2
        sub     ax,.z1
        cwde
@@ -361,8 +413,6 @@ else
  ;     mov      .dey12,eax
        push	 eax
 
-;end if
-
        mov	ax,word[.t_x2]
        sub	ax,word[.t_x1]
        cwde
@@ -393,20 +443,12 @@ end if
        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
 
 if Ext>=SSE
 
-       sub	 esp,28
+       sub	 esp,32
-   ;    mov       eax,255
+   ;    mov       eax,256
        cvtsi2ss  xmm4,[i255d]
        cvtsi2ss  xmm3,ebx	     ;rcps
        divss	 xmm3,xmm4
@@ -434,16 +476,22 @@ if Ext>=SSE
 
        shufps	 xmm1,xmm1,10110001b
 			     ;xmm1--> | dbx | dby | dex | dey |
-       cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
+;1       movups    .dey13,xmm1
+
+       cvtps2pi  mm0,xmm1 ;mm0,xmm1          ; mm0 -> 2 delta dwords
        movhlps	 xmm1,xmm1
-       cvtps2pi  mm1,xmm1
+       cvtps2pi  mm1,xmm1 ;mm1,xmm1
        movq	 .dey13,mm0
        movq	 .dby13,mm1
-;----------
+
        mov	 ax,.z3
        sub	 ax,.z1
        cwde
 
+       mov	dx,.x3
+       sub	dx,.x1
+       movsx	edx,dx
+
        movd	  mm2,[.t_x1]
        movd	  mm3,[.t_x3]
 
@@ -453,18 +501,53 @@ if Ext>=SSE
 
        cvtsi2ss  xmm1,eax
        movlhps	 xmm1,xmm1
+       cvtsi2ss  xmm1,edx
+       shufps	 xmm1,xmm1,00101111b
        cvtpi2ps  xmm1,mm3
 
-       divps	 xmm1,xmm3   ; xmm1--> | xxx | dz | dty | dtx |
+       divps	 xmm1,xmm3   ; xmm1--> | dx | dz | dty | dtx |
 
        shufps	 xmm1,xmm1,11100001b
-			     ; xmm1--> | xxx | dz | dtx | dty |
+			     ; xmm1--> | dx | dz | dtx | dty |
+;1       movlps    .dty13,xmm1
+;1       movhps    .dz13,xmm1
+
        cvtps2pi  mm0,xmm1    ; mm0 -> 2 delta dwords  | dtx | dty |
        movhlps	 xmm1,xmm1
-       cvtss2si  eax,xmm1    ; eax --> 1 delta dword | dz |
+       cvtps2pi  mm1,xmm1
        movq	 .dty13,mm0
-       mov	 .dz13,eax
+       movq	 .dz13,mm1
+
+;temporaly conversion to keep upside down
+;     fninit
+;     fld      .dx13
+;     fistp    .dx13
+;     fld      dword .dz13
+;     fistp    dword .dz13
+;     fld      .dbx13
+;     fistp    .dbx13
+;     fld      dword .dby13
+;     fistp    dword .dby13
+;     fld      .dex13
+;     fistp    .dex13
+;     fld      dword .dey13
+;     fistp    dword .dey13
+;     fld      .dtx13
+;     fistp    .dtx13
+;     fld      dword .dty13
+;     fistp    dword .dty13
+
 else
+
+       mov	ax,.x3
+       sub	ax,.x1
+       cwde
+       shl	eax,ROUND
+       cdq
+       idiv	ebx
+ ;     mov      .dx13,eax
+       push	 eax
+
        mov     ax,.z3
        sub     ax,.z1
        cwde
@@ -541,20 +624,12 @@ end if
        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
 
 if Ext>=SSE
 
-       sub	 esp,28
+       sub	 esp,32
-    ;   mov       eax,255
+   ;    mov       eax,256
        cvtsi2ss  xmm4,[i255d]
        cvtsi2ss  xmm3,ebx	     ;rcps
        divss	 xmm3,xmm4
@@ -582,16 +657,22 @@ if Ext>=SSE
 
        shufps	 xmm1,xmm1,10110001b
 			     ;xmm1--> | dbx | dby | dex | dey |
-       cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
+;1       movups    .dey23,xmm1
+
+       cvtps2pi  mm0,xmm1 ;mm0,xmm1          ; mm0 -> 2 delta dwords
        movhlps	 xmm1,xmm1
-       cvtps2pi  mm1,xmm1
+       cvtps2pi  mm1,xmm1 ;mm1,xmm1
        movq	 .dey23,mm0
        movq	 .dby23,mm1
-;----------
+
        mov	 ax,.z3
        sub	 ax,.z2
        cwde
 
+       mov	dx,.x3
+       sub	dx,.x2
+       movsx	edx,dx
+
        movd	  mm2,[.t_x2]
        movd	  mm3,[.t_x3]
 
@@ -601,18 +682,50 @@ if Ext>=SSE
 
        cvtsi2ss  xmm1,eax
        movlhps	 xmm1,xmm1
+       cvtsi2ss  xmm1,edx
+       shufps	 xmm1,xmm1,00101111b
        cvtpi2ps  xmm1,mm3
 
-       divps	 xmm1,xmm3   ; xmm1--> | xxx | dz | dty | dtx |
+       divps	 xmm1,xmm3   ; xmm1--> | dx | dz | dty | dtx |
 
        shufps	 xmm1,xmm1,11100001b
-			     ; xmm1--> | xxx | dz | dtx | dty |
+			    ; xmm1--> | dx | dz | dtx | dty |
+;       movlps    .dty23,xmm1
+;       movhps    .dz23,xmm1
        cvtps2pi  mm0,xmm1    ; mm0 -> 2 delta dwords  | dtx | dty |
        movhlps	 xmm1,xmm1
-       cvtss2si  eax,xmm1    ; eax --> 1 delta dword | dz |
+       cvtps2pi  mm1,xmm1    ; mm1 --> 2 delta dwords | dx  |  dz |
        movq	 .dty23,mm0
-       mov	 .dz23,eax
+       movq	 .dz23,mm1
+
+
+;temporaly conversion to keep upside down
+;     fninit
+;     fld      .dx23
+;     fistp    .dx23
+;     fld      dword .dz23
+;     fistp    dword .dz23
+;     fld      .dbx23
+;     fistp    .dbx23
+;     fld      dword .dby23
+;     fistp    dword .dby23
+;     fld      .dex23
+;     fistp    .dex23
+;     fld      dword .dey23
+;     fistp    dword .dey23
+;     fld      .dtx23
+;     fistp    .dtx23
+;     fld      dword .dty23
+;     fistp    dword .dty23
 else
+       mov	ax,.x3
+       sub	ax,.x2
+       cwde
+       shl	eax,ROUND
+       cdq
+       idiv	ebx
+ ;     mov      .dx23,eax
+       push	 eax
 
        mov     ax,.z3
        sub     ax,.z2
@@ -681,6 +794,58 @@ end if
       ;  sub     esp,40
    .bt_dx23_done:
        sub	 esp,64
+;if Ext>=SSE
+;       movsx    eax,.x1
+;       shl      eax,ROUND
+;       cvtsi2ss xmm0,eax
+;       movss    .cx1,xmm0
+;       movss    .cx2,xmm0
+;
+;       movsx    ebx,word[.b_x1]
+;       shl      ebx,ROUND
+;       cvtsi2ss xmm0,ebx
+;       movss    .cbx1,xmm0
+;       movss    .cbx2,xmm0
+;
+;       movsx    ecx,word[.b_y1]
+;       shl      ecx,ROUND
+;       cvtsi2ss xmm0,ecx
+;       movss    .cby1,xmm0
+;       movss    .cby2,xmm0
+
+;       movsx    edx,word[.e_x1]
+;       shl      edx,ROUND
+;       cvtsi2ss xmm0,edx
+;       movss    .cex1,xmm0
+;       movss    .cex2,xmm0
+;
+;       movsx    eax,word[.e_y1]
+;       shl      eax,ROUND
+;       cvtsi2ss xmm0,eax
+;       movss    .cey1,xmm0
+;       movss    .cey2,xmm0
+;
+
+;       movsx    ebx,.z1
+;       shl      ebx,CATMULL_SHIFT
+;       cvtsi2ss xmm0,ebx
+;       movss    .cz1,xmm0
+;       movss    .cz2,xmm0
+;
+;       movsx    ecx,word[.t_x1]
+;       shl      ecx,ROUND
+;       cvtsi2ss xmm0,ecx
+;       movss    .ctx1,xmm0
+;       movss    .ctx2,xmm0
+
+;       movsx    edx,word[.t_y1]
+;       shl      edx,ROUND
+;       cvtsi2ss xmm0,edx
+;       movss    .cty1,xmm0
+;       movss    .cty2,xmm0
+
+;else
+
        movsx	eax,.x1
        shl	eax,ROUND
        mov	.cx1,eax
@@ -737,13 +902,36 @@ end if
        mov	.cty2,edx
       ; push     edx
       ; push     edx
-
+;end if
 
        movsx	ecx,.y1
        cmp	cx,.y2
        jge	.loop12_done
   .loop12:
        call	.call_line
+;if Ext >= SSE
+;       movups  xmm0,.cby1
+;       movups  xmm1,.cty1
+;       movups  xmm2,.cby2
+;       movups  xmm3,.cty2
+;       movups  xmm4,.dby13
+;       movups  xmm5,.dty13
+;       movups  xmm6,.dby12
+;       movups  xmm7,.dty12
+;       addps   xmm0,xmm4
+;       addps   xmm1,xmm5
+;       addps   xmm2,xmm6
+;       addps   xmm3,xmm7
+;
+;;       addps   xmm0,.dby12
+;;       addps   xmm1,.dty12
+;;       addps   xmm2,.dby13
+;;       addps   xmm3,.dty13
+;       movups  .cby1,xmm0
+;       movups  .cty1,xmm1
+;       movups  .cby2,xmm2
+;       movups  .cty2,xmm3
+;end if
 
 if Ext >= MMX
        movq	mm0,.cby2
@@ -770,7 +958,8 @@ if Ext >= MMX
        movq	.cty2,mm5
        movq	.cz1,mm6
        movq	.cz2,mm7
-else
+end if
+if Ext = NON
        mov	edx,.dbx13
        add	.cbx1,edx
        mov	eax,.dbx12
@@ -816,6 +1005,8 @@ end if
        cmp	cx,.y3
        jge	.loop23_done
 
+;if Ext < SSE
+
        movsx	eax,.z2
        shl	eax,CATMULL_SHIFT
        mov	.cz2,eax
@@ -847,10 +1038,76 @@ end if
        movzx	ebx,word[.t_y2]
        shl	ebx,ROUND
        mov	.cty2,ebx
+;else
+;       movsx    eax,.z2
+;       shl      eax,CATMULL_SHIFT
+;       cvtsi2ss xmm0,eax
+;       movss    .cz2,xmm0
+;
+;       movsx    ebx,.x2
+;       shl      ebx,ROUND
+;       cvtsi2ss xmm1,ebx
+;       movss    .cx2,xmm1
+;
+;       movzx    edx,word[.b_x2]
+;       shl      edx,ROUND
+;       cvtsi2ss xmm2,edx
+;       movss    .cbx2,xmm2
+;
+;       movzx    eax,word[.b_y2]
+;       shl      eax,ROUND
+;       cvtsi2ss xmm0,eax
+;       movss    .cby2,xmm0
+;
+;       movzx    ebx,word[.e_x2]
+;       shl      ebx,ROUND
+;       cvtsi2ss xmm1,ebx
+;       movss    .cex2,xmm1
+;
+;       movzx    edx,word[.e_y2]
+;       shl      edx,ROUND
+;       cvtsi2ss xmm2,edx
+;       movss    .cey2,xmm2
+;
+;       movzx    eax,word[.t_x2]
+;       shl      eax,ROUND
+;       cvtsi2ss xmm0,eax
+;       movss    .ctx2,xmm0
+;
+;       movzx    ebx,word[.t_y2]
+;       shl      ebx,ROUND
+;       cvtsi2ss xmm1,ebx
+;       movss    .cty2,xmm1
+
+;end if
 
      .loop23:
        call	.call_line
 
+;if Ext >= SSE
+
+;       movups  xmm0,.cby1
+;       movups  xmm1,.cty1
+;       movups  xmm2,.cby2
+;       movups  xmm3,.cty2
+;       movups  xmm4,.dby13
+;       movups  xmm5,.dty13
+;       movups  xmm6,.dby23
+;       movups  xmm7,.dty23
+;       addps   xmm0,xmm4
+;       addps   xmm1,xmm5
+;       addps   xmm2,xmm6
+;       addps   xmm3,xmm7
+; ;      addps   xmm0,.dby13
+; ;      addps   xmm1,.dty13
+; ;      addps   xmm2,.dby23
+; ;      addps   xmm3,.dty23
+;       movups  .cby1,xmm0
+;       movups  .cty1,xmm1
+;       movups  .cby2,xmm2
+;       movups  .cty2,xmm3
+;
+;end if
 if Ext >= MMX
        movq	mm0,.cby2
        movq	mm1,.cby1
@@ -876,7 +1133,8 @@ if Ext >= MMX
        movq	.cty2,mm5
        movq	.cz1,mm6
        movq	.cz2,mm7
-else
+end if
+If Ext = NON
        mov	edx,.dbx13
        add	.cbx1,edx
        mov	eax,.dbx23
@@ -912,7 +1170,6 @@ else
        add	.cty1,edx
        mov	eax,.dty23
        add	.cty2,eax
-
 end if
        inc	ecx
        cmp	cx,.y3
@@ -923,24 +1180,107 @@ end if
 ret   50
 
 .call_line:
+;if Ext >= SSE
+;       pushad
+;       push     .tex_ptr
+;       fninit
+;       fld      dword .cty1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     dword .cty1
+;       fld      .ctx1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     .ctx1
+;       fld      dword .cz1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;    ;   push     dwod .cz1
+;       fld      dword .cty2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;    ;   push     .cty2
+;       fld      .ctx2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;  ;     push     dword .ctx2
+;       fld      dword .cz2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     dword .cz2
+;
+;       push     .z_buff
+;       push     .t_emap
+;       push     .t_bmap
+;--------------------------------------
+;       fld      dword .cey2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     dword .cey2
+;
+;       fld      .cex2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     .cex2
+;
+;       fld      dword .cby2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+; ;      push     dword .cby2
 
+;       fld      .cbx2
+;       fistp    dword [esp-4]
+;       sub      esp,4
+ ;      push     .cbx2
+;------------------------------------
+;       fld      dword .cey1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     dword .cey1
+;       fld      .cex1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;    ;   push     .cex1
+;       fld      dword .cby1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;       push     dword .cby1
+;       fld      .cbx1
+;       fistp    dword [esp-4]
+;       sub      esp,4
+;;      push     .cbx1
+;       push     ecx
+
+;       fld      .cx1
+;       fistp    dword [esp-4]
+;       mov      eax,[esp-4]
+;       sar      eax,ROUND
+;       fld      .cx2
+;       fistp    dword [esp-4]
+;       mov      ebx,[esp-4]
+;       sar      ebx,ROUND
+
+;       call     bump_tex_line_z
+;
+;       popad
+;else
        pushad
        push	.tex_ptr
-       push	dword .cty2
-       push	.ctx2
        push	dword .cty1
        push	.ctx1
        push	dword .cz1
+       push	dword .cty2
+       push	.ctx2
        push	dword .cz2
        push	.z_buff
        push	.t_emap
        push	.t_bmap
        push	dword .cey2
        push	.cex2
-       push	dword .cey1
-       push	.cex1
        push	dword .cby2
        push	.cbx2
+       push	dword .cey1
+       push	.cex1
        push	dword .cby1
        push	.cbx1
        push	ecx
@@ -953,6 +1293,7 @@ ret   50
        call	bump_tex_line_z
 
        popad
+;end if
 ret
 bump_tex_line_z:
 ;--------------in: eax - x1
@@ -960,62 +1301,46 @@ bump_tex_line_z:
 ;--------------    edi - pointer to screen buffer
 ;stack - another parameters :
 .y	equ dword [ebp+4]
-.bx1q	equ	  [ebp+8]
-.bx2q	equ	  [ebp+16]
-.ex1q	equ	  [ebp+24]
-.ex2q	equ	  [ebp+32]
-.tx1q	equ	  [ebp+60]
-.tx2q	equ	  [ebp+68]
-;.bx1q   equ       [ebp+8]
-;.bx2q   equ       [ebp+16]
-;.ex1q   equ       [ebp+24]
-;.exyq   equ       [ebp+32]
-.bx1	equ dword [ebp+8]   ;   ---
-.by1	equ dword [ebp+12]  ;       |
-.bx2	equ dword [ebp+16]  ;       |
-.by2	equ dword [ebp+20]  ;       |>   b. map and e. map coords
-.ex1	equ dword [ebp+24]  ;       |>   shifted shl ROUND
-.ey1	equ dword [ebp+28]  ;       |
-.ex2	equ dword [ebp+32]  ;       |
-.ey2	equ dword [ebp+36]  ;   ---
-.bmap	equ dword [ebp+40]  ; bump map offset
-.emap	equ dword [ebp+44]  ; env map offset
-.z_buff equ dword [ebp+48]
-.z2	equ dword [ebp+52]  ;   -- |>   z coords shifted
-.z1	equ dword [ebp+56]  ;   --       shl  CATMULL_SHIFT
 
-.tx1	equ dword [ebp+60]  ;  -----
+.bx1	equ  [ebp+8]   ;   ---
-.ty1	equ dword [ebp+64]  ;       |>   shifted shl ROUND
+.by1	equ  [ebp+12]  ;       |
-.tx2	equ dword [ebp+68]  ;       |
+.ex1	equ  [ebp+16]  ;       |
-.ty2	equ dword [ebp+72]  ;   ---
+.ey1	equ  [ebp+20]  ;       |
+.bx2	equ  [ebp+24]  ;       |
+.by2	equ  [ebp+28]  ;       |>   b. map and e. map coords
+.ex2	equ  [ebp+32]  ;       |>   shifted shl ROUND
+.ey2	equ  [ebp+36]  ;   ---
+.bmap	equ  [ebp+40]  ; bump map offset
+.emap	equ  [ebp+44]  ; env map offset
+.z_buff equ  [ebp+48]
+.z2	equ  [ebp+52]
+.tx2	equ  [ebp+56]
+.ty2	equ  [ebp+60]
+.z1	equ  [ebp+64]
+.tx1	equ  [ebp+68]
+.ty1	equ  [ebp+72]
 .tex_map equ dword [ebp+76]  ; texture offset  ( pointer )
 
 
-.x1	equ dword [ebp-4]
+.x1	equ [ebp-4]
-.x2	equ dword [ebp-8]
+.x2	equ [ebp-8]
-.dbx	equ dword [ebp-12]
+.dbx	equ [ebp-12]
-.dby	equ dword [ebp-16]
+.dby	equ [ebp-16]
-.dbyq	equ qword [ebp-16]  ; - new
+.dex	equ [ebp-20]
-.dex	equ dword [ebp-20]
+.dey	equ [ebp-24]
-.dey	equ dword [ebp-24]
+.dz	equ [ebp-28]
-.deyq	equ qword [ebp-24]  ; - new
+.dtx	equ [ebp-32]
-.dz	equ dword [ebp-28]
+.dty	equ [ebp-36]
-.dtx	equ dword [ebp-32]
-.dty	equ dword [ebp-36]
-.dtyq	equ qword [ebp-36]
 
-.cbx	equ dword [ebp-40]
+.cbx	equ [ebp-40]
-.cby	equ dword [ebp-44]
+.cby	equ [ebp-44]
-.cbyq	equ qword [ebp-44]  ; - new
+.cex	equ [ebp-48]
-.cex	equ dword [ebp-48]
+.cey	equ [ebp-52]
-.cey	equ dword [ebp-52]
+.cz	equ [ebp-56]
-.ceyq	equ qword [ebp-52]  ; - new
+.czbuff equ [ebp-60]
-.cz	equ dword [ebp-56]
+.ctx	equ [ebp-64]
-.czbuff equ dword [ebp-60]
+.cty	equ [ebp-68]
-.ctx	equ dword [ebp-64]
+.c_scr	equ [ebp-72]
-.cty	equ dword [ebp-68]
-.ctyq	equ qword [ebp-68]
-.c_scr	equ dword [ebp-72]
 
 .temp1	equ	   ebp-80
 .temp2	equ	   ebp-88
@@ -1025,8 +1350,6 @@ bump_tex_line_z:
 
 	mov	ebp,esp
 
-
-
 	mov	ecx,.y
 	or	ecx,ecx
 	jl	.bl_end
@@ -1059,19 +1382,30 @@ if Ext=NON
 	mov	edx,.ty1
 	xchg	edx,.ty2
 	mov	.ty1,edx
-else
+end if
-	movq	mm0,.bx1q
+if Ext = MMX
-	movq	mm1,.bx2q
+	movq	mm0,.bx1
-	movq	mm2,.ex1q
+	movq	mm1,.bx2
-	movq	mm3,.ex2q
+	movq	mm2,.ex1
-	movq	mm4,.tx1q
+	movq	mm3,.ex2
-	movq	mm5,.tx2q
+	movq	mm4,.tx1
-	movq	.bx2q,mm0
+	movq	mm5,.tx2
-	movq	.bx1q,mm1
+	movq	.bx2,mm0
-	movq	.ex1q,mm3
+	movq	.bx1,mm1
-	movq	.ex2q,mm2
+	movq	.ex1,mm3
-	movq	.tx1q,mm5
+	movq	.ex2,mm2
-	movq	.tx2q,mm4
+	movq	.tx1,mm5
+	movq	.tx2,mm4
+end if
+if Ext>=SSE
+	movups xmm0,.bx1
+	movups xmm1,.bx2
+	movups .bx1,xmm1
+	movups .bx2,xmm0
+	movq	mm0,.tx1
+	movq	mm1,.tx2
+	movq	.tx1,mm1
+	movq	.tx2,mm0
 end if
 
 	mov	edx,.z1
@@ -1080,9 +1414,9 @@ end if
   .bl_ok:
 	push	eax
 	push	ebx	      ;store x1, x2
-	cmp	.x1,SIZE_X
+	cmp	dword .x1,SIZE_X
 	jge	.bl_end
-	cmp	.x2,0
+	cmp	dword .x2,0
 	jle	.bl_end
 
 	mov	ebx,.x2
@@ -1093,43 +1427,51 @@ if Ext>=SSE
        sub	 esp,28
        cvtsi2ss  xmm3,ebx	     ;rcps
        shufps	 xmm3,xmm3,0
+; float using SSE variant  ::-->
+;       movups    xmm0,.bx1  ; new
+;       movups    xmm1,.bx2  ; new
 
-       cvtpi2ps  xmm0,.bx1q ;mm0
+       cvtpi2ps  xmm0,.bx1 ;mm0    ; variant fixed point
        movlhps	 xmm0,xmm0
-       cvtpi2ps  xmm0,.ex1q ;mm2
+       cvtpi2ps  xmm0,.ex1 ;mm2
-       cvtpi2ps  xmm1,.bx2q ;mm1
+       cvtpi2ps  xmm1,.bx2 ;mm1
        movlhps	 xmm1,xmm1
-       cvtpi2ps  xmm1,.ex2q ;mm3
+       cvtpi2ps  xmm1,.ex2 ;mm3
        subps	 xmm1,xmm0
 
        divps	 xmm1,xmm3
 
        shufps	 xmm1,xmm1,10110001b
+;       movups    .dey,xmm1  ; new
        cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
        movhlps	 xmm1,xmm1
        cvtps2pi  mm1,xmm1
-       movq	 .deyq,mm0
+       movq	 .dey,mm0
-       movq	 .dbyq,mm1
+       movq	 .dby,mm1
 
        movd	 mm2,.z1
        movd	 mm3,.z2
 
-       cvtpi2ps  xmm0,.tx1q ;mm0
+       cvtpi2ps  xmm0,.tx1 ;mm0
        movlhps	 xmm0,xmm0
        cvtpi2ps  xmm0,mm2
-       cvtpi2ps  xmm1,.tx2q ;mm1
+       cvtpi2ps  xmm1,.tx2 ;mm1
        movlhps	 xmm1,xmm1
        cvtpi2ps  xmm1,mm3
+;       movups    xmm0,,z1  ; new
+;       movups    xmm1,.z2  ; new
        subps	 xmm1,xmm0
 
        divps	 xmm1,xmm3
 
+;       movups    .dz,xmm1  ;new
+
        shufps	 xmm1,xmm1,10110100b
        cvtps2pi  mm0,xmm1	   ; mm0 -> 2 delta dwords
        movhlps	 xmm1,xmm1
        cvtps2pi  mm1,xmm1
        movd	 .dz,mm0
-       movq	 .dtyq,mm1
+       movq	 .dty,mm1
 
 else
 
@@ -1177,14 +1519,27 @@ else
 	push	eax
 
 end if
-	cmp	.x1,0	      ; set correctly begin variable
+	cmp	dword .x1,0	    ; set correctly begin variable
 	jge	@f	      ; CLIPPING ON FUNCTION
 			      ; cutting triangle exceedes screen
 	mov	ebx,.x1
 	neg	ebx
+
+;if Ext >= SSE
+
+;        cvtsi2ss xmm0,ebx
+;        shufps   xmm0,xmm0,0
+;        movups   xmm1,.dey
+;        mulps    xmm1,xmm0
+;        shufps   xmm1,xmm1,00011011b
+;        movups   xmm2,.bx1
+;        addps    xmm2,xmm1
+;        movups   .bx1,xmm2
+
+	mov	eax,.dz
 	imul	ebx	      ; eax = .dz * abs(.x1)
 	add	.z1,eax
-	mov	.x1,0
+	mov	dword .x1,0
 
 	mov	eax,.dbx
 	imul	ebx
@@ -1211,9 +1566,9 @@ end if
 	add	.ty1,eax
 
       @@:
-	cmp	.x2,SIZE_X
+	cmp	dword .x2,SIZE_X
 	jl	@f
-	mov	.x2,SIZE_X
+	mov	dword .x2,SIZE_X
       @@:
 	mov	eax,SIZE_X	 ;calc memory begin in buffers
 	mul	.y
@@ -1227,21 +1582,21 @@ end if
 	mov	ecx,.x2
 	sub	ecx,.x1
 	; init current variables
-	push	.bx1   ; current b, e and t shifted shl ROUND   .cbx
+	push	dword .bx1   ; current b, e and t shifted shl ROUND   .cbx
-	push	.by1					     ;  .cby
+	push	dword .by1					   ;  .cby
-	push	.ex1					     ;  .cex
+	push	dword .ex1					   ;  .cex
-	push	.ey1					     ;  .cey
+	push	dword .ey1					   ;  .cey
 
-	push	.z1    ; current z shl CATMULL_SHIFT         ; .cz
+	push	dword .z1    ; current z shl CATMULL_SHIFT         ; .cz
 	push	esi					     ; .czbuff
 
-	push	.tx1	  ;         .ctx
+	push	dword .tx1	;         .ctx
-	push	.ty1	  ;         .cty
+	push	dword .ty1	;         .cty
 	push	edi	  ;         .c_scr
 if Ext>=MMX
-	movq	mm7,.ctyq
+	movq	mm7,.cty
-	movq	mm6,.cbyq
+	movq	mm6,.cby
-	movq	mm5,.ceyq
+	movq	mm5,.cey
 ;        movq    mm4,.dtyq
 ;        movq    mm3,.dbyq
 end if
@@ -1266,7 +1621,6 @@ else
 end if
 
 
-
 	shl	eax,TEX_SHIFT
 	add	esi,eax 	;-  ; esi - current bump map index
 
@@ -1296,6 +1650,7 @@ end if
 	movzx	ebx,byte [ebx]
 	sub	edx,ebx
 
+
      ;  eax - horizontal sub    modificated x coord
      ;  edx - vertical   sub    modificated y coord
 if Ext=NON
@@ -1400,8 +1755,8 @@ end if
 	stosd
 
       .skip:
-	add	.czbuff,4
+	add	dword .czbuff,4
-	add	.c_scr,3
+	add	dword .c_scr,3
 
 if Ext=NON
 	mov	eax,.dbx
@@ -1420,9 +1775,9 @@ if Ext=NON
 	add	.cty,edx
 
 else
-	paddd	mm7,.dtyq
+	paddd	mm7,.dty
-	paddd	mm6,.dbyq
+	paddd	mm6,.dby
-	paddd	mm5,.deyq
+	paddd	mm5,.dey
 end if
 	mov	eax,.dz
 	add	.cz,eax

programs/demos/3DS/History.txt

@@ -0,0 +1,96 @@
+View3ds 0.055 - January 2011.
+1. SSE optimizations (file BUMP_TEX.INC).
+2. Compiles correctly in newest FASM.
+-----------------------------------------------------------------------------------
+
+View3ds 0.054 - December 2009 (Updated January 2010).
+1. Skinned window by Leency.
+2. Optimizations.
+3. Re map texture, bumps option - allow spherical mapping around each axle (X,Y,Z).
+4. Problem with too small memory to generate object fixed. (Problem ocurred with
+   house.3ds object and others objects contains less than 1000 faces and points).
+-----------------------------------------------------------------------------------
+
+View3ds 0.053 - (?) 2009
+1. Optimizations.
+-----------------------------------------------------------------------------------
+
+View3ds 0.052 - November 2007.
+1. Memory for file is allocated dynamically.
+2. Optimizations.
+Note: compiling correct only for KolibriOS.
+-----------------------------------------------------------------------------------
+
+View3ds 0.051 - October 2007.
+1. More smooth texture mapping in tex and txgr shading model and others based on
+   tex3.inc, tex_cat.inc shading models.
+2. Predefined three, I hope, nice lights.
+-----------------------------------------------------------------------------------
+
+View3ds 0.05 - (?) 2007
+1. New shading model: cubic environment mapping with 1.5 kb light buffer.
+2. Bumps optionally according to texture. ( It gives so called texture with
+   shifts in bumps + texture shading model. )
+3. Bumps depth button.
+-----------------------------------------------------------------------------------
+
+View3ds 0.04 - March 2007.
+1. New shading model: bump + texture.  Only version with z coordinate
+   interpolation.
+2. SSE optimizations ( files: two_tex.inc and bump_tex.inc )
+-----------------------------------------------------------------------------------
+
+View3ds 0.03 - March 2007.
+1. Two shading models: smooth + texture, spherical environment mapping + texture.
+   Version only with z coordinate interpolation.
+2. Bit changed rotary. I took attempt in SSE instuctions. ( matrix multiplication )
+3. Color position depend drawing model  instead spot light ( I hope,spot light come
+   back in future ).
+-----------------------------------------------------------------------------------
+
+View3ds 0.02 - December 2006.
+1. New shading models - "spot" - spot light (only one, it's very CPU hungry model,
+   and there is no easy way to implement buffer) with light attenuation (In my
+   implmentation works not perfect); "dots" - app draws only points (with no culling).
+2. Some loseless operations "mirror": according to each axis, rotary 90 degrees.
+3. Postprocesing efects: Emboss (use blur to make edges more visable), fire ( be
+   carefull during `fire + embos`parallel using).
+4. "Move" -  Changes meaning of 'add vector' buttons:
+   used "obj"  move  object, "camr" move camera (in this option use culling = 'off'
+   I must make culling procedure bit advanced).
+5. Generate button. Few objects generating (with bad normals)
+6. Some optimizations on MMX.
+7. I took an attempt to memory managing functions, but it decreased speed (no aligned
+   memory in Menuet functions?)
+
+There is really much work to do and I see disorder in app code.
+Greetings for all, especially Madis Kalme for .3ds files support and good vibrations.  ;)
+Merry Christmas and happy new year.
+-----------------------------------------------------------------------------------
+
+App View3ds ver 0.01 - November 2006.
+1. Object generating (for now only one). When no parameter specified or error occured
+   during reading file from disk, app generate object. No teapot.3ds needed on ramdrive,
+   but, if this file exist it would be read as default.
+2. Bug in read_from_file procedure fixed. (In  May, when I made shading light vector
+   depend, I demaged bit Madis' procedure.)
+3. Backface culling on/off option. (Some objects need culling off - they have mismatched
+   normal vectors. Example: iron.3ds, sink.3ds, the generated one.)
+4. Random lights procedure.
+5. Spherical (instead old - planar ) bump, texture mapping.
+6. Bit improved menu.
+7. Blur.
+-----------------------------------------------------------------------------------
+
+App "View3ds" - it's very early, still unfinished version but I want show you it.
+Based on earlier demos.
+1. 3ds object import as a parameter ( for now only short names supported )
+   I updated Kfar app - it can run view3ds with parameter. You must copy it on
+   ramdrive. You could do such funny thing: Run updated Kfar in KolbrInWin,
+   before them set in KolInWin.ini file path to ramdrive, where yo have "view3ds",
+   choose 3ds file ... Emulator will run View3ds with selected file.
+   If no parameter specified, app try open tpot.3ds from ramdrive. If tpot.3ds
+   don't exist on rd, app hang.
+2. Many lights. Every light - unlinear model ( Phong illumination ). For now
+   no light setting option.
+3. Light buffer for grd, flat model.

programs/demos/3DS/VIEW3DS.ASM

@@ -1,10 +1,10 @@
 
-; application : View3ds ver. 0.054 - tiny .3ds files viewer.
+; application : View3ds ver. 0.055 - tiny .3ds files viewer.
-; compiler    : FASM  1.67.11
+; compiler    : FASM
-; system      : KolibriOS/MenuetOS
+; system      : KolibriOS
 ; author      : Macgub aka Maciej Guba
 ; email       : macgub3@wp.pl
-; web         : www.menuet.xt.pl
+; web         : www.macgub.hekko.pl
 ; Fell free to use this intro in your own distribution of KolibriOS/MenuetOS.
 ; Special greetings to all MenuetOS maniax in the world.
 ; I hope because my intros Christian Belive will be near to each of You.
@@ -34,7 +34,7 @@ LIGHT_SIZE equ 22			       ;      |  |
 NON   =   0				       ;     -/  \-
 MMX   =   1
 SSE   =   2
-Ext   =   MMX			;Ext={ NON | MMX | SSE}
+Ext   =   MMX		     ;Ext={ NON | MMX | SSE}
 
 ; 0 for short names (Menuet-compatible), 1 for long names (Kolibri features)
 USE_LFN = 1

programs/demos/3DS/data.inc

@@ -268,7 +268,7 @@ base_vector:
       if Ext=SSE
 	db   ' (SSE)'
       end if
-	db   ' 0.054'
+	db   ' 0.055'
     labellen:
 	STRdata db '-1        '
 
@@ -396,7 +396,7 @@ align 8
 	rand_seed	    dw ?
 align 8
 	buffer	dq	?
-	err_	dd	?
+	errr	dd	?
 	drr	dd	?
 	xx	dd	?
 	yy	dd	?

programs/demos/3DS/readme.txt

@@ -1,11 +1,8 @@
-View3ds 0.054 - tiny viewer to .3ds files.
+View3ds 0.055 - tiny viewer to .3ds files.
 
 What's new?
-1. Skinned window by Leency.
+1. SSE optimizations (file BUMP_TEX.INC).
-2. Optimizations.
+3. Compiles correctly in newest FASM.
-3. Re map texture, bumps option - allow spherical mapping around each axle (X,Y,Z).
-4. Problem with too small memory to generate object fixed. (Problem ocurred with
-   house.3ds object and others objects contains less than 1000 faces and points).
 
 Buttons description:
 1. rotary: choosing rotary axle: x, y, x+y.
@@ -33,5 +30,4 @@ Buttons description:
 18. re-map tex -> re-map texture and bump map coordinates, to change spherical mapping 
     around axle use 'xchg' and 'mirror' buttons, then press 're-map tex' button.
 
-			 Macgub 		december 2009
+			 Macgub 		 Jan 2011
-                                                                                  Last edited Jan 2010