WebSVN – Kolibri OS – Path Comparison – / – /programs/ Rev 2191 and /programs/ Rev 2192

Ignore whitespace Rev 2191 → Rev 2192

 /programs/demos/3DS/BUMP_CAT.INC
 ,10 → 81,10
 .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]
+.cex1  equ dword[ebp-120]
+.cey1  equ      [ebp-124]
+.cbx2  equ dword[ebp-128]
+.cby2  equ      [ebp-132]
 .cex2  equ dword[ebp-136]
 .cey2  equ      [ebp-140]
 ,8 → 533,25
        jge      .loop12_done
   .loop12:
        call     .call_bump_line
+if Ext >= SSE2
+       movups   xmm0,.cey2
+       movups   xmm1,.cey1
+       movups   xmm2,.dey12
+       movups   xmm3,.dey13
+       paddd    xmm0,xmm2
+       paddd    xmm1,xmm3
+       movups   .cey2,xmm0
+       movups   .cey1,xmm1
+       movq     mm4,.cz1
+       movq     mm5,.cz2
+       paddd    mm4,.dz13
+       paddd    mm5,.dz12
+       movq     .cz1,mm4
+       movq     .cz2,mm5
+end if
-if Ext >= MMX
+if (Ext = MMX) | (Ext = SSE)
        movq     mm0,.cby2
        movq     mm1,.cby1
        movq     mm2,.cey2
 ,7 → 570,7
        movq     .cey2,mm2
        movq     .cz1,mm4
        movq     .cz2,mm5
-else
+else if Ext = NON
        mov      edx,.dbx13
        add      .cbx1,edx
        mov      eax,.dbx12
 ,8 → 634,23
      .loop23:
        call     .call_bump_line
-if Ext >= MMX
+if Ext >=  SSE2
+       movups   xmm0,.cey2
+       movups   xmm1,.cey1
+       movups   xmm2,.dey23
+       movups   xmm3,.dey13
+       paddd    xmm0,xmm2
+       paddd    xmm1,xmm3
+       movups   .cey2,xmm0
+       movups   .cey1,xmm1
+       movq     mm4,.cz1
+       movq     mm5,.cz2
+       paddd    mm4,.dz13
+       paddd    mm5,.dz23
+       movq     .cz1,mm4
+       movq     .cz2,mm5
+end if
+if (Ext = MMX) | (Ext = SSE)
        movq     mm0,.cby2
        movq     mm1,.cby1
        movq     mm2,.cey2
 ,7 → 669,7
        movq     .cey2,mm2
        movq     .cz1,mm4
        movq     .cz2,mm5
-else
+else if Ext = NON
        mov      eax,.dx13
        add      .cx1,eax
        mov      ebx,.dx23

 /programs/demos/3DS/BUMP_TEX.INC
 ,24 → 292,6
        cvtps2pi  mm1,xmm1
        movq      .dty12,mm0
        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
 ,25 → 500,6
        movq      .dty13,mm0
        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
 ,24 → 662,6
        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
 ,58 → 739,7
       ;  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
 ,14 → 796,8
        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
+;if Ext >= SSE2
 ;       movups  xmm0,.cby1
 ;       movups  xmm1,.cty1
 ;       movups  xmm2,.cby2
 ,22 → 806,36
 ;       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
+       movsx    ecx,.y1
+       cmp      cx,.y2
+       jge      .loop12_done
+  .loop12:
+;if Ext >= SSE2
+;       fxsave  [sse_repository]
+;end if
+       call     .call_line
+if Ext >= SSE2
+;       fxrstor [sse_repository]
+       movups  xmm0,.cby1
+       movups  xmm1,.cty1
+       movups  xmm2,.cby2
+       movups  xmm3,.cty2
+       movups  xmm4,.dby13
+       movups  xmm5,.dty13
+       movups  xmm6,.dby12
+       movups  xmm7,.dty12
+       paddd   xmm0,xmm4
+       paddd   xmm1,xmm5
+       paddd   xmm2,xmm6
+       paddd   xmm3,xmm7
+       movups  .cby1,xmm0
+       movups  .cty1,xmm1
+       movups  .cby2,xmm2
+       movups  .cty2,xmm3
+end if
-if Ext >= MMX
+if (Ext = MMX) | (Ext = SSE)
        movq     mm0,.cby2
        movq     mm1,.cby1
        movq     mm2,.cey2
 ,7 → 907,6
        cmp      cx,.y3
        jge      .loop23_done
-;if Ext < SSE
        movsx    eax,.z2
        shl      eax,CATMULL_SHIFT
 ,54 → 939,7
        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
+;if Ext >= SSE2
 ;       movups  xmm0,.cby1
 ;       movups  xmm1,.cty1
 ;       movups  xmm2,.cby2
 ,21 → 948,34
 ;       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
+     .loop23:
+;if Ext >= SSE2
+;       fxsave  [sse_repository]
+;end if
+       call     .call_line
+if Ext >= SSE2
+;       fxrstor [sse_repository]
+       movups  xmm0,.cby1
+       movups  xmm1,.cty1
+       movups  xmm2,.cby2
+       movups  xmm3,.cty2
+       movups  xmm4,.dby13
+       movups  xmm5,.dty13
+       movups  xmm6,.dby23
+       movups  xmm7,.dty23
+       paddd   xmm0,xmm4
+       paddd   xmm1,xmm5
+       paddd   xmm2,xmm6
+       paddd   xmm3,xmm7
+       movups  .cby1,xmm0
+       movups  .cty1,xmm1
+       movups  .cby2,xmm2
+       movups  .cty2,xmm3
 ;
-;end if
-if Ext >= MMX
+end if
+if (Ext = MMX) | (Ext = SSE)
        movq     mm0,.cby2
        movq     mm1,.cby1
        movq     mm2,.cey2
 ,90 → 1047,7
 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 .cty1
 ,8 → 1065,16
        push     .cbx2
        push     dword .cey1
        push     .cex1
+;if Ext >= SSE2
+;       sub      esp,8
+;       shufps   xmm0,xmm0,10110100b
+;       movhps   [esp],xmm0 ;================================
+;else
        push     dword .cby1
        push     .cbx1
+;end if
        push     ecx
        mov      eax,.cx1

 /programs/demos/3DS/DATA.INC
 ,7 → 271,7
       if Ext=SSE2
         db   ' (SSE2)'
       end if
-        db   ' 0.059',0
+        db   ' 0.060',0
     labellen:
         STRdata db '-1        '

 /programs/demos/3DS/GRD_TEX.INC
 ,7 → 486,7
       ; pop      ebp ebx eax
        popad
-if Ext >= MMX
+if (Ext = MMX)|(Ext=SSE)
        movq     mm0,.cur1b
        movq     mm1,.cur1r
        movq     mm2,.scan_y1
 ,7 → 505,25
        movq     .cur2b,mm3
        movq     .cur2r,mm4
        movq     .scan_y2,mm5
-else
+end if
+if Ext >= SSE2
+       movups   xmm0,.cur1b
+       movups   xmm1,.dc13b
+       movups   xmm2,.cur2b
+       movups   xmm3,.dc12b
+       movq     mm2,.scan_y1
+       movq     mm5,.scan_y2
+       paddd    xmm0,xmm1
+       paddd    xmm2,xmm3
+       paddd    mm2,.tex_dy13
+       paddd    mm5,.tex_dy12
+       movq     .scan_y1,mm2
+       movq     .scan_y2,mm5
+       movups   .cur1b,xmm0
+       movups   .cur2b,xmm2
+end if
+if Ext = NON
        mov      edx,.dc13b
        add      .cur1b,edx
        mov      esi,.dc13g
 ,7 → 620,7
        popad
-if Ext >= MMX
+if (Ext = MMX)|(Ext=SSE)
        movq     mm0,.cur1b
        movq     mm1,.cur1r
        movq     mm2,.scan_y1
 ,7 → 639,24
        movq     .cur2b,mm3
        movq     .cur2r,mm4
        movq     .scan_y2,mm5
-else
+end if
+if Ext >= SSE2
+       movups   xmm0,.cur1b
+       movups   xmm1,.dc13b
+       movups   xmm2,.cur2b
+       movups   xmm3,.dc23b
+       movq     mm2,.scan_y1
+       movq     mm5,.scan_y2
+       paddd    xmm0,xmm1
+       paddd    xmm2,xmm3
+       paddd    mm2,.tex_dy13
+       paddd    mm5,.tex_dy23
+       movq     .scan_y1,mm2
+       movq     .scan_y2,mm5
+       movups   .cur1b,xmm0
+       movups   .cur2b,xmm2
+end if
+if Ext = NON
        mov      edx,.dc13b
        add      .cur1b,edx
        mov      esi,.dc13g
 ,8 → 775,8
         mov     ecx,dword .z1
         xchg    ecx, .z2
         mov     dword .z1, ecx
-else
+end if
+if (Ext=MMX)
         movq    mm0,.b1        ; b, g
         movq    mm1,.b2
         movq    .b1, mm1
 ,6 → 791,16
         movq    .tex_x2,mm4
 end if
+if Ext>=SSE
+       movups  xmm0,.b1
+       movups  xmm1,.b2
+       movups  .b1,xmm1
+       movups  .b2,xmm0
+       movq    mm4,.tex_x1    ; x, z
+       movq    mm5,.tex_x2
+       movq    .tex_x1,mm5
+       movq    .tex_x2,mm4
+end if
     @@:
         or      bx,bx

 /programs/demos/3DS/TWO_TEX.INC
 ,72 → 32,68
 ;----------------------         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 - first texture
-.b_x3   equ ebp+12
-.b_y3   equ ebp+14       ; e - second texture
-.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
+.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 - first texture
+.b_x3   equ ebp+12
+.b_y3   equ ebp+14       ; e - second texture
+.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 b. texture
-.t_emap equ dword[ebp-8]        ; pointer to e. texture
-.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]
+.t_bmap equ dword[ebp-4]        ; pointer to b. texture
+.t_emap equ dword[ebp-8]        ; pointer to e. texture
+.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]
 .dbx12 equ dword[ebp-28]
 .dby12 equ dword[ebp-32]
-.dby12q equ     [ebp-32]
+.dby12q equ     [ebp-32]
 .dex12 equ dword[ebp-36]
 .dey12 equ dword[ebp-40]
-.dey12q equ     [ebp-40]
+.dey12q equ     [ebp-40]
 .dz12  equ dword[ebp-44]
 .dx13  equ dword[ebp-48]
 .dbx13 equ dword[ebp-52]
 .dby13 equ dword[ebp-56]
-.dby13q equ     [ebp-56]
+.dby13q equ     [ebp-56]
 .dex13 equ dword[ebp-60]
 .dey13 equ dword[ebp-64]
-.dey13q equ     [ebp-64]
+.dey13q equ     [ebp-64]
 .dz13  equ dword[ebp-68]
 .dx23  equ dword[ebp-72]
 .dbx23 equ dword[ebp-76]
 .dby23 equ dword[ebp-80]
-.dby23q equ     [ebp-80]
+.dby23q equ     [ebp-80]
 .dex23 equ dword[ebp-84]
 .dey23 equ dword[ebp-88]
-.dey23q equ     [ebp-88]
+.dey23q equ     [ebp-88]
 .dz23  equ dword[ebp-92]
 .cx1   equ dword[ebp-96]   ; current variables
 .cx2   equ dword[ebp-100]
-;.cbx1q equ      [ebp-104]
 .cbx1  equ dword[ebp-104]
 .cby1  equ [ebp-108]
-;.cbx2q          [ebp-112]
-.cbx2  equ dword[ebp-112]
-.cby2  equ [ebp-116]
-;.cex1q equ      [ebp-120]
-.cex1  equ dword[ebp-120]
-.cey1  equ [ebp-124]
-;.cex2q equ      [ebp-128]
+.cex1  equ dword[ebp-112]
+.cey1  equ [ebp-116]
+.cbx2  equ dword[ebp-120]
+.cby2  equ [ebp-124]
 .cex2  equ dword[ebp-128]
 .cey2  equ [ebp-132]
 ,10 → 106,10
        cld
     end if
        mov     ebp,esp
-       push    edx esi       ; store bump map
+       push    edx esi       ; store bump map
 ;       push    esi        ; store e. map
      ; sub     esp,120
- .sort3:                  ; sort triangle coordinates...
+ .sort3:                  ; sort triangle coordinates...
        cmp     ax,bx
        jle     .sort1
        xchg    eax,ebx
 ,30 → 123,30
        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
+       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
+       jmp      .sort3
  .sort2:
-       push     eax ebx ecx    ; store triangle coords in variables
+       push     eax ebx ecx    ; 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,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
 ,44 → 161,44
     ;   jg      .loop23_done    ; {
-       mov      bx,.y2       ; calc delta 12
-       sub      bx,.y1
-       jnz      .bt_dx12_make
-       mov      ecx,6
-       xor      edx,edx
+       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
+       push     edx   ;dword 0
+       loop     @b
+       jmp      .bt_dx12_done
  .bt_dx12_make:
-       mov      ax,.x2
-       sub      ax,.x1
+       mov      ax,.x2
+       sub      ax,.x1
        cwde
-       movsx    ebx,bx
-       shl      eax,ROUND
+       movsx    ebx,bx
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
 ;      mov      .dx12,eax
-       push      eax
+       push      eax
 if Ext=SSE
-       sub       esp,16
-       cvtsi2ss  xmm3,ebx            ;rcps
+       sub       esp,16
+       cvtsi2ss  xmm3,ebx            ;rcps
     ;   mov       eax,255
        cvtsi2ss  xmm4,[i255d] ;eax
-       divss     xmm3,xmm4
-       rcpss     xmm3,xmm3
+       divss     xmm3,xmm4
+       rcpss     xmm3,xmm3
     ;   mulss     xmm3,xmm4
-       shufps    xmm3,xmm3,0
+       shufps    xmm3,xmm3,0
-       movd      mm0,[.b_x1]
-       movd      mm1,[.b_x2]
-       movd      mm2,[.e_x1]
-       movd      mm3,[.e_x2]
+       movd      mm0,[.b_x1]
+       movd      mm1,[.b_x2]
+       movd      mm2,[.e_x1]
+       movd      mm3,[.e_x2]
      ;  psubsw    mm3,mm2
      ;  psubsw    mm1,mm0
-       pxor      mm4,mm4
+       pxor      mm4,mm4
        punpcklwd mm0,mm4
        punpcklwd mm1,mm4
        punpcklwd mm2,mm4
 ,12 → 208,12
    ;    pslld     mm2,ROUND
    ;    pslld     mm3,ROUND
        cvtpi2ps  xmm0,mm0
-       movlhps   xmm0,xmm0
+       movlhps   xmm0,xmm0
        cvtpi2ps  xmm0,mm2
        cvtpi2ps  xmm1,mm1
-       movlhps   xmm1,xmm1
+       movlhps   xmm1,xmm1
        cvtpi2ps  xmm1,mm3
-       subps     xmm1,xmm0
+       subps     xmm1,xmm0
      ;  pxor      mm4,mm4
      ;  movq      mm5,mm1
 ,13 → 241,13
 ;       movlhps   xmm0,xmm0
 ;       cvtpi2ps  xmm0,mm3
   ;     divps     xmm1,xmm3
-       mulps     xmm1,xmm3
-       shufps    xmm1,xmm1,10110001b
-       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
-       movhlps   xmm1,xmm1
+       mulps     xmm1,xmm3
+       shufps    xmm1,xmm1,10110001b
+       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
+       movhlps   xmm1,xmm1
        cvtps2pi  mm1,xmm1
-       movq      .dey12q,mm0
-       movq      .dby12q,mm1
+       movq      .dey12q,mm0
+       movq      .dby12q,mm1
 ;       movd      .dex12,mm0
 ;       psrlq     mm0,32
 ,75 → 259,75
 ;       movd      .dby12,mm0
 else
-       mov      ax,word[.b_x2]
-       sub      ax,word[.b_x1]
+       mov      ax,word[.b_x2]
+       sub      ax,word[.b_x1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;     mov      .dbx12,eax
-       push      eax
+       push      eax
-       mov      ax,word[.b_y2]
-       sub      ax,word[.b_y1]
+       mov      ax,word[.b_y2]
+       sub      ax,word[.b_y1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;     mov      .dby12,eax
-       push      eax
+       push      eax
  ;     mov       eax,.dbx12
  ;     mov       ebx,.dby12
  ;     int3
-       mov      ax,word[.e_x2]
-       sub      ax,word[.e_x1]
+       mov      ax,word[.e_x2]
+       sub      ax,word[.e_x1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dex12,eax
-       push      eax
+       push      eax
-       mov      ax,word[.e_y2]
-       sub      ax,word[.e_y1]
+       mov      ax,word[.e_y2]
+       sub      ax,word[.e_y1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dey12,eax
-       push      eax
+       push      eax
 end if
-        mov     ax,.z2
-        sub     ax,.z1
-        cwde
-        shl     eax,CATMULL_SHIFT
-        cdq
-        idiv    ebx
-        push    eax
+        mov     ax,.z2
+        sub     ax,.z1
+        cwde
+        shl     eax,CATMULL_SHIFT
+        cdq
+        idiv    ebx
+        push    eax
    .bt_dx12_done:
-       mov      bx,.y3       ; calc delta13
-       sub      bx,.y1
-       jnz      .bt_dx13_make
-       mov      ecx,6
-       xor      edx,edx
+       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
+       push     edx   ;dword 0
+       loop     @b
+       jmp      .bt_dx13_done
  .bt_dx13_make:
-       mov      ax,.x3
-       sub      ax,.x1
+       mov      ax,.x3
+       sub      ax,.x1
        cwde
-       movsx    ebx,bx
-       shl      eax,ROUND
+       movsx    ebx,bx
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dx13,eax
-       push      eax
+       push      eax
 if Ext=SSE
 ,18 → 334,18
        cvtsi2ss  xmm3,ebx
   ;     mov       eax,255
        cvtsi2ss  xmm4,[i255d]
-       divss     xmm3,xmm4
-       rcpss     xmm3,xmm3
+       divss     xmm3,xmm4
+       rcpss     xmm3,xmm3
 ;       mulss     xmm3,xmm4
-       shufps    xmm3,xmm3,0
-       sub       esp,16
+       shufps    xmm3,xmm3,0
+       sub       esp,16
-       movd      mm0,[.b_x1]
-       movd      mm1,[.b_x3]
-       movd      mm2,[.e_x1]
-       movd      mm3,[.e_x3]
+       movd      mm0,[.b_x1]
+       movd      mm1,[.b_x3]
+       movd      mm2,[.e_x1]
+       movd      mm3,[.e_x3]
-       pxor      mm4,mm4
+       pxor      mm4,mm4
        punpcklwd mm0,mm4
        punpcklwd mm1,mm4
        punpcklwd mm2,mm4
 ,59 → 352,59
        punpcklwd mm3,mm4
        cvtpi2ps  xmm0,mm0
-       movlhps   xmm0,xmm0
+       movlhps   xmm0,xmm0
        cvtpi2ps  xmm0,mm2
        cvtpi2ps  xmm1,mm1
-       movlhps   xmm1,xmm1
+       movlhps   xmm1,xmm1
        cvtpi2ps  xmm1,mm3
-       subps     xmm1,xmm0
+       subps     xmm1,xmm0
    ;    divps     xmm1,xmm3
-       mulps     xmm1,xmm3
-       shufps    xmm1,xmm1,10110001b
-       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
-       movhlps   xmm1,xmm1
+       mulps     xmm1,xmm3
+       shufps    xmm1,xmm1,10110001b
+       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
+       movhlps   xmm1,xmm1
        cvtps2pi  mm1,xmm1
-       movq      .dey13q,mm0
-       movq      .dby13q,mm1
+       movq      .dey13q,mm0
+       movq      .dby13q,mm1
 else
-       mov      ax,word[.b_x3]
-       sub      ax,word[.b_x1]
+       mov      ax,word[.b_x3]
+       sub      ax,word[.b_x1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dbx13,eax
-       push      eax
+       push      eax
-       mov      ax,word[.b_y3]
-       sub      ax,word[.b_y1]
+       mov      ax,word[.b_y3]
+       sub      ax,word[.b_y1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dby13,eax
-       push      eax
+       push      eax
-       mov      ax,word[.e_x3]
-       sub      ax,word[.e_x1]
+       mov      ax,word[.e_x3]
+       sub      ax,word[.e_x1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dex13,eax
-       push      eax
+       push      eax
-       mov      ax,word[.e_y3]
-       sub      ax,word[.e_y1]
+       mov      ax,word[.e_y3]
+       sub      ax,word[.e_y1]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dey13,eax
-       push      eax
+       push      eax
 end if
 ,25 → 418,25
        push    eax
    .bt_dx13_done:
-       mov      bx,.y3       ; calc delta23
-       sub      bx,.y2
-       jnz      .bt_dx23_make
-       mov      ecx,6
-       xor      edx,edx
+       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
+       push     edx   ;dword 0
+       loop     @b
+       jmp      .bt_dx23_done
  .bt_dx23_make:
-       mov      ax,.x3
-       sub      ax,.x2
+       mov      ax,.x3
+       sub      ax,.x2
        cwde
-       movsx    ebx,bx
-       shl      eax,ROUND
+       movsx    ebx,bx
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dx23,eax
-       push      eax
+       push      eax
 if Ext=SSE
 ,16 → 443,16
        cvtsi2ss  xmm3,ebx
      ;  mov       eax,255
        cvtsi2ss  xmm4,[i255d] ;eax
-       divss     xmm3,xmm4
-       shufps    xmm3,xmm3,0
-       sub       esp,16
+       divss     xmm3,xmm4
+       shufps    xmm3,xmm3,0
+       sub       esp,16
-       movd      mm0,[.b_x2]
-       movd      mm1,[.b_x3]
-       movd      mm2,[.e_x2]
-       movd      mm3,[.e_x3]
+       movd      mm0,[.b_x2]
+       movd      mm1,[.b_x3]
+       movd      mm2,[.e_x2]
+       movd      mm3,[.e_x3]
-       pxor      mm4,mm4
+       pxor      mm4,mm4
        punpcklwd mm0,mm4
        punpcklwd mm1,mm4
        punpcklwd mm2,mm4
 ,58 → 459,58
        punpcklwd mm3,mm4
        cvtpi2ps  xmm0,mm0
-       movlhps   xmm0,xmm0
+       movlhps   xmm0,xmm0
        cvtpi2ps  xmm0,mm2
        cvtpi2ps  xmm1,mm1
-       movlhps   xmm1,xmm1
+       movlhps   xmm1,xmm1
        cvtpi2ps  xmm1,mm3
-       subps     xmm1,xmm0
+       subps     xmm1,xmm0
-       divps     xmm1,xmm3
-       shufps    xmm1,xmm1,10110001b
-       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
-       movhlps   xmm1,xmm1
+       divps     xmm1,xmm3
+       shufps    xmm1,xmm1,10110001b
+       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
+       movhlps   xmm1,xmm1
        cvtps2pi  mm1,xmm1
-       movq      .dey23q,mm0
-       movq      .dby23q,mm1
+       movq      .dey23q,mm0
+       movq      .dby23q,mm1
 else
-       mov      ax,word[.b_x3]
-       sub      ax,word[.b_x2]
+       mov      ax,word[.b_x3]
+       sub      ax,word[.b_x2]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dbx23,eax
-       push      eax
+       push      eax
-       mov      ax,word[.b_y3]
-       sub      ax,word[.b_y2]
+       mov      ax,word[.b_y3]
+       sub      ax,word[.b_y2]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dby23,eax
-       push      eax
+       push      eax
-       mov      ax,word[.e_x3]
-       sub      ax,word[.e_x2]
+       mov      ax,word[.e_x3]
+       sub      ax,word[.e_x2]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dex23,eax
-       push      eax
+       push      eax
-       mov      ax,word[.e_y3]
-       sub      ax,word[.e_y2]
+       mov      ax,word[.e_y3]
+       sub      ax,word[.e_y2]
        cwde
-       shl      eax,ROUND
+       shl      eax,ROUND
        cdq
-       idiv     ebx
+       idiv     ebx
  ;      mov      .dey23,eax
-       push      eax
+       push      eax
 end if
        mov     ax,.z3
        sub     ax,.z2
 ,187 → 522,203
        push    eax
       ;  sub     esp,40
    .bt_dx23_done:
-       movsx    eax,.x1
-       shl      eax,ROUND
+       movsx    eax,.x1
+       shl      eax,ROUND
      ;  mov      .cx1,eax
      ;  mov      .cx2,eax
-       push     eax eax
+       push     eax eax
      ; push     eax
-       movsx    eax,word[.b_x1]
-       shl      eax,ROUND
-       mov      .cbx1,eax
-       mov      .cbx2,eax
+       movsx    eax,word[.b_x1]
+       shl      eax,ROUND
+       mov      .cbx1,eax
+       mov      .cbx2,eax
      ;  push     eax eax
      ; push     eax
-       movsx    eax,word[.b_y1]
-       shl      eax,ROUND
-       mov      .cby1,eax
-       mov      .cby2,eax
+       movsx    eax,word[.b_y1]
+       shl      eax,ROUND
+       mov      .cby1,eax
+       mov      .cby2,eax
      ;  push     eax eax
      ; push     eax
-       movsx    eax,word[.e_x1]
-       shl      eax,ROUND
-       mov      .cex1,eax
-       mov      .cex2,eax
+       movsx    eax,word[.e_x1]
+       shl      eax,ROUND
+       mov      .cex1,eax
+       mov      .cex2,eax
       ; push     eax eax
       ;push     eax
-       movsx    eax,word[.e_y1]
-       shl      eax,ROUND
-       mov      .cey1,eax
-       mov      .cey2,eax
-       sub      esp,32
+       movsx    eax,word[.e_y1]
+       shl      eax,ROUND
+       mov      .cey1,eax
+       mov      .cey2,eax
+       sub      esp,32
       ; push     eax eax
       ;push     eax
-       movsx    eax,.z1
-       shl      eax,CATMULL_SHIFT
+       movsx    eax,.z1
+       shl      eax,CATMULL_SHIFT
       ; mov      .cz1,eax
       ; mov      .cz2,eax
       push     eax eax
       ;push     eax
-       movsx    ecx,.y1
-       cmp      cx,.y2
-       jge      .loop12_done
+       movsx    ecx,.y1
+       cmp      cx,.y2
+       jge      .loop12_done
   .loop12:
-       call     .call_line
+       call     .call_line
-       mov      eax,.dx13
-       add      .cx1,eax
-       mov      ebx,.dx12
-       add      .cx2,ebx
-if Ext >= MMX
-       movq     mm0,.cby2       ; with this optimization object
-       movq     mm1,.cby1       ; looks bit annoying
-       movq     mm2,.cey2
-       movq     mm3,.cey1
-       paddd    mm0,.dby12q
-       paddd    mm1,.dby13q
-       paddd    mm2,.dey12q
-       paddd    mm3,.dey13q
-       movq     .cby2,mm0
-       movq     .cby1,mm1
-       movq     .cey1,mm3
-       movq     .cey2,mm2
+       mov      eax,.dx13
+       add      .cx1,eax
+       mov      ebx,.dx12
+       add      .cx2,ebx
+if Ext>= SSE2
+       movups   xmm0,.cey1
+       movups   xmm1,.cey2
+       movups   xmm2,.dey12q
+       movups   xmm3,.dey13q
+       paddd    xmm0,xmm3
+       paddd    xmm1,xmm2
+       movups   .cey1,xmm0
+       movups   .cey2,xmm1
+else if (Ext = MMX) | (Ext=SSE)
+       movq     mm0,.cby2       ; with this optimization object
+       movq     mm1,.cby1       ; looks bit annoying
+       movq     mm2,.cey2
+       movq     mm3,.cey1
+       paddd    mm0,.dby12q
+       paddd    mm1,.dby13q
+       paddd    mm2,.dey12q
+       paddd    mm3,.dey13q
+       movq     .cby2,mm0
+       movq     .cby1,mm1
+       movq     .cey1,mm3
+       movq     .cey2,mm2
 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      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,.dex13
+       add      .cex1,eax
+       mov      ebx,.dex12
+       add      .cex2,ebx
+       mov      edx,.dey13
+       add      .cey1,edx
+       mov      eax,.dey12
+       add      .cey2,eax
 end if
-       mov      ebx,.dz13
-       add      .cz1,ebx
-       mov      edx,.dz12
-       add      .cz2,edx
+       mov      ebx,.dz13
+       add      .cz1,ebx
+       mov      edx,.dz12
+       add      .cz2,edx
-       inc      ecx
-       cmp      cx,.y2
-       jl       .loop12
+       inc      ecx
+       cmp      cx,.y2
+       jl       .loop12
     .loop12_done:
-       movsx    ecx,.y2
-       cmp      cx,.y3
-       jge      .loop23_done
+       movsx    ecx,.y2
+       cmp      cx,.y3
+       jge      .loop23_done
-       movsx    eax,.z2
-       shl      eax,CATMULL_SHIFT
-       mov      .cz2,eax
+       movsx    eax,.z2
+       shl      eax,CATMULL_SHIFT
+       mov      .cz2,eax
-       movsx    eax,.x2
-       shl      eax,ROUND
-       mov      .cx2,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_x2]
+       shl      eax,ROUND
+       mov      .cbx2,eax
-       movzx    eax,word[.b_y2]
-       shl      eax,ROUND
-       mov      .cby2,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_x2]
+       shl      eax,ROUND
+       mov      .cex2,eax
-       movzx    eax,word[.e_y2]
-       shl      eax,ROUND
-       mov      .cey2,eax
+       movzx    eax,word[.e_y2]
+       shl      eax,ROUND
+       mov      .cey2,eax
      .loop23:
-       call     .call_line
+       call     .call_line
 ;if Ext = NON
-       mov      eax,.dx13
-       add      .cx1,eax
-       mov      ebx,.dx23
-       add      .cx2,ebx
+       mov      eax,.dx13
+       add      .cx1,eax
+       mov      ebx,.dx23
+       add      .cx2,ebx
+if Ext>= SSE2
+       movups   xmm0,.cey1
+       movups   xmm1,.cey2
+       movups   xmm2,.dey23q
+       movups   xmm3,.dey13q
+       paddd    xmm0,xmm3
+       paddd    xmm1,xmm2
+       movups   .cey1,xmm0
+       movups   .cey2,xmm1
+else if (Ext = MMX) | ( Ext = SSE)
+       movq     mm0,.cby2        ;    with this mmx optimization object looks bit
+       movq     mm1,.cby1        ;    annoying
+       movq     mm2,.cey2
+       movq     mm3,.cey1
+       paddd    mm0,.dby23q
+       paddd    mm1,.dby13q
+       paddd    mm2,.dey23q
+       paddd    mm3,.dey13q
+       movq     .cby2,mm0
+       movq     .cby1,mm1
+       movq     .cey2,mm2
+       movq     .cey1,mm3
-if Ext >= MMX
-       movq     mm0,.cby2        ;    with this mmx optimization object looks bit
-       movq     mm1,.cby1        ;    annoying
-       movq     mm2,.cey2
-       movq     mm3,.cey1
-       paddd    mm0,.dby23q
-       paddd    mm1,.dby13q
-       paddd    mm2,.dey23q
-       paddd    mm3,.dey13q
-       movq     .cby2,mm0
-       movq     .cby1,mm1
-       movq     .cey2,mm2
-       movq     .cey1,mm3
 else
-       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      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      eax,.dex13
+       add      .cex1,eax
+       mov      ebx,.dex23
+       add      .cex2,ebx
+       mov      edx,.dey13
+       add      .cey1,edx
+       mov      eax,.dey23
+       add      .cey2,eax
 end if
-       mov      ebx,.dz13
-       add      .cz1,ebx
-       mov      edx,.dz23
-       add      .cz2,edx
+       mov      ebx,.dz13
+       add      .cz1,ebx
+       mov      edx,.dz23
+       add      .cz2,edx
 ;else
 ;       movq     mm0,.db13q
 ;       movq     mm1,.cbx1q
-       inc      ecx
-       cmp      cx,.y3
-       jl       .loop23
+       inc      ecx
+       cmp      cx,.y3
+       jl       .loop23
     .loop23_done:
-       mov      esp,ebp
+       mov      esp,ebp
 ret   34
 .call_line:
 ,27 → 725,27
        pushad
-       push     .cz1
-       push     .cz2
-       push     .z_buff
-       push     .t_bmap
-       push     .t_emap
-       push     dword .cey2
-       push     .cex2
-       push     dword .cey1
-       push     .cex1
-       push     dword .cby2
-       push     .cbx2
-       push     dword .cby1
-       push     .cbx1
-       push     ecx
+       push     .cz1
+       push     .cz2
+       push     .z_buff
+       push     .t_bmap
+       push     .t_emap
+       push     dword .cey2
+       push     .cex2
+       push     dword .cey1
+       push     .cex1
+       push     dword .cby2
+       push     .cbx2
+       push     dword .cby1
+       push     .cbx1
+       push     ecx
-       mov      eax,.cx1
-       sar      eax,ROUND
-       mov      ebx,.cx2
-       sar      ebx,ROUND
+       mov      eax,.cx1
+       sar      eax,ROUND
+       mov      ebx,.cx2
+       sar      ebx,ROUND
-       call     two_tex_line_z
+       call     two_tex_line_z
        popad
 ret
 ,92 → 754,92
 ;--------------    ebx - x2
 ;--------------    edi - pointer to screen buffer
 ;stack - another parameters :
-.y      equ dword [ebp+4]
-.bx1    equ  [ebp+8]   ;   ---
-.by1    equ  [ebp+12]  ;       |
-.bx2    equ  [ebp+16]  ;       |
-.by2    equ  [ebp+20]  ;       |>   b. texture and e. texture coords
-.ex1    equ  [ebp+24]  ;       |>   shifted shl ROUND
-.ey1    equ  [ebp+28]  ;       |
-.ex2    equ  [ebp+32]  ;       |
-.ey2    equ  [ebp+36]  ;   ---
-.emap   equ  [ebp+40]  ; b texture offset
-.bmap   equ  [ebp+44]  ; e texture offset
+.y      equ dword [ebp+4]
+.bx1    equ  [ebp+8]   ;   ---
+.by1    equ  [ebp+12]  ;       |
+.bx2    equ  [ebp+16]  ;       |
+.by2    equ  [ebp+20]  ;       |>   b. texture and e. texture coords
+.ex1    equ  [ebp+24]  ;       |>   shifted shl ROUND
+.ey1    equ  [ebp+28]  ;       |
+.ex2    equ  [ebp+32]  ;       |
+.ey2    equ  [ebp+36]  ;   ---
+.emap   equ  [ebp+40]  ; b texture offset
+.bmap   equ  [ebp+44]  ; e texture offset
 .z_buff equ dword [ebp+48]
-.z2     equ dword [ebp+52]  ;   -- |>   z coords shifted
-.z1     equ dword [ebp+56]  ;   --       shl  CATMULL_SHIFT
+.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 [ebp-12]
-.dex    equ [ebp-16]
-.dby    equ [ebp-20]
-.dey    equ [ebp-24]
-.dz     equ dword [ebp-28]
-.cbx    equ [ebp-32]
-.cex    equ [ebp-36]
-.cby    equ [ebp-40]
-.cey    equ [ebp-44]
-.cz     equ dword [ebp-48]
+.x1     equ dword [ebp-4]
+.x2     equ dword [ebp-8]
+.dbx    equ [ebp-12]
+.dex    equ [ebp-16]
+.dby    equ [ebp-20]
+.dey    equ [ebp-24]
+.dz     equ dword [ebp-28]
+.cbx    equ [ebp-32]
+.cex    equ [ebp-36]
+.cby    equ [ebp-40]
+.cey    equ [ebp-44]
+.cz     equ dword [ebp-48]
 .czbuff equ dword [ebp-52]
-        mov     ebp,esp
+        mov     ebp,esp
-        mov     ecx,.y
-        or      ecx,ecx
-        jl      .bl_end
-        cmp     ecx,SIZE_Y
-        jge     .bl_end
+        mov     ecx,.y
+        or      ecx,ecx
+        jl      .bl_end
+        cmp     ecx,SIZE_Y
+        jge     .bl_end
-        cmp     eax,ebx
-        jl      @f
-        je      .bl_end
+        cmp     eax,ebx
+        jl      @f
+        je      .bl_end
-        xchg    eax,ebx
+        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,.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
+        mov     edx,.ex1
+        xchg    edx,.ex2
+        mov     .ex1,edx
+        mov     edx,.ey1
+        xchg    edx,.ey2
+        mov     .ey1,edx
 else
-        movq    mm0,.bx1
-        movq    mm1,.ex1
-        movq    mm2,.bx2
-        movq    mm3,.ex2
-        movq    .bx2,mm0
-        movq    .ex2,mm1
-        movq    .bx1,mm2
-        movq    .ex1,mm3
+        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
-        mov     edx,.z1
-        xchg    edx,.z2
-        mov     .z1,edx
+        mov     edx,.z1
+        xchg    edx,.z2
+        mov     .z1,edx
     @@:
-        push    eax ebx
+        push    eax ebx
 ;        push    ebx           ;store x1, x2
-        cmp     .x1,SIZE_X
-        jge     .bl_end
-        cmp     .x2,0
-        jle     .bl_end
+        cmp     .x1,SIZE_X
+        jge     .bl_end
+        cmp     .x2,0
+        jle     .bl_end
-        mov     ebx,.x2
-        sub     ebx,.x1
+        mov     ebx,.x2
+        sub     ebx,.x1
 if Ext>=SSE
-       sub       esp,16
-       cvtsi2ss  xmm3,ebx            ;rcps
-       shufps    xmm3,xmm3,0
+       sub       esp,16
+       cvtsi2ss  xmm3,ebx            ;rcps
+       shufps    xmm3,xmm3,0
   ;     movq      mm0,.bx1q
   ;     movq      mm1,.bx2q
 ,107 → 852,107
   ;     cvtpi2ps  xmm1,mm3
        cvtpi2ps  xmm0,.bx1 ;mm0     ; bx1; by1
-       movlhps   xmm0,xmm0
+       movlhps   xmm0,xmm0
        cvtpi2ps  xmm0,.ex1 ;mm2     ; ex1; ey1
        cvtpi2ps  xmm1,.bx2 ;mm1     ; bx2; by2
-       movlhps   xmm1,xmm1
+       movlhps   xmm1,xmm1
        cvtpi2ps  xmm1,.ex2 ;mm3     ; ex2; ey2
-       subps     xmm1,xmm0
-                                    ; hi             lo
-       divps     xmm1,xmm3 ; xmm1 -> dby; dbx; dey; dex
+       subps     xmm1,xmm0
+                                    ; hi             lo
+       divps     xmm1,xmm3 ; xmm1 -> dby; dbx; dey; dex
-       shufps    xmm1,xmm1,11011000b
-       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
-       movhlps   xmm1,xmm1
+       shufps    xmm1,xmm1,11011000b
+       cvtps2pi  mm0,xmm1          ; mm0 -> 2 delta dwords
+       movhlps   xmm1,xmm1
        cvtps2pi  mm1,xmm1
-       movq      .dex,mm0 ; hi - lo  ->  dbx, dex
-       movq      .dey,mm1 ; hi - lo  ->  dby, dey
+       movq      .dex,mm0 ; hi - lo  ->  dbx, dex
+       movq      .dey,mm1 ; hi - lo  ->  dby, dey
 else
-        mov     eax,.bx2       ; calc .dbx
-        sub     eax,.bx1
-        cdq
-        idiv    ebx
-        push    eax
+        mov     eax,.bx2       ; calc .dbx
+        sub     eax,.bx1
+        cdq
+        idiv    ebx
+        push    eax
-        mov     eax,.ex2       ; calc .dby
-        sub     eax,.ex1
-        cdq
-        idiv    ebx
-        push    eax
+        mov     eax,.ex2       ; calc .dby
+        sub     eax,.ex1
+        cdq
+        idiv    ebx
+        push    eax
-        mov     eax,.by2       ; calc .dex
-        sub     eax,.by1
-        cdq
-        idiv    ebx
-        push    eax
+        mov     eax,.by2       ; calc .dex
+        sub     eax,.by1
+        cdq
+        idiv    ebx
+        push    eax
-        mov     eax,.ey2       ; calc .dey
-        sub     eax,.ey1
-        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
+        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
+        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,.dbx
+        imul    ebx
+        add    .bx1,eax
-        mov     eax,.dby
-        imul    ebx
-        add     .by1,eax
+        mov     eax,.dby
+        imul    ebx
+        add     .by1,eax
-        mov     eax,.dex
-        imul    ebx
-        add     .ex1,eax
+        mov     eax,.dex
+        imul    ebx
+        add     .ex1,eax
-        mov     eax,.dey
-        imul    ebx
-        add     .ey1,eax
+        mov     eax,.dey
+        imul    ebx
+        add     .ey1,eax
       @@:
-        cmp     .x2,SIZE_X
-        jl      @f
-        mov     .x2,SIZE_X
+        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           ; edi - screen
-        mov     esi,.z_buff       ; z-buffer filled with dd variables
-        shl     ebx,2
-        add     esi,ebx           ; esi - Z buffer
+        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           ; edi - screen
+        mov     esi,.z_buff       ; z-buffer filled with dd variables
+        shl     ebx,2
+        add     esi,ebx           ; esi - Z buffer
-        mov     ecx,.x2
-        sub     ecx,.x1
-        ; init current variables
-        push    dword .bx1 ;.by1 .ex1 .ey1 .z1 esi
-        push    dword .ex1
-        push    dword .by1
-        push    dword .ey1
+        mov     ecx,.x2
+        sub     ecx,.x1
+        ; init current variables
+        push    dword .bx1 ;.by1 .ex1 .ey1 .z1 esi
+        push    dword .ex1
+        push    dword .by1
+        push    dword .ey1
-        push    .z1              ; current z shl CATMULL_SHIFT
-        push    esi
+        push    .z1              ; current z shl CATMULL_SHIFT
+        push    esi
 if Ext >= MMX
      pxor   mm0,mm0
 ,104 → 971,104
      .draw:
     ; if TEX = SHIFTING   ;bump drawing only in shifting mode
 if Ext=NON
-        mov     esi,.czbuff      ; .czbuff current address in buffer
-        mov     ebx,.cz          ; .cz - cur z position
-        cmp     ebx,dword[esi]
+        mov     esi,.czbuff      ; .czbuff current address in buffer
+        mov     ebx,.cz          ; .cz - cur z position
+        cmp     ebx,dword[esi]
 else
-        mov     ebx,.cz
-        cmp     ebx,dword[edx]
+        mov     ebx,.cz
+        cmp     ebx,dword[edx]
 end if
-        jge     .skip
+        jge     .skip
 if Ext=NON
-        mov     eax,.cby
-        mov     esi,.cbx
-        sar     eax,ROUND
-        sar     esi,ROUND
-        shl     eax,TEX_SHIFT   ;-
-        add     esi,eax
-        lea     esi,[esi*3]         ;-  ; esi - current b. texture addres
-        add     esi,.bmap
+        mov     eax,.cby
+        mov     esi,.cbx
+        sar     eax,ROUND
+        sar     esi,ROUND
+        shl     eax,TEX_SHIFT   ;-
+        add     esi,eax
+        lea     esi,[esi*3]         ;-  ; esi - current b. texture addres
+        add     esi,.bmap
-        mov     ebx,.cex       ;.cex - current env map X
-        mov     eax,.cey       ;.cey - current  env map y
-        sar     ebx,ROUND
-        sar     eax,ROUND
+        mov     ebx,.cex       ;.cex - current env map X
+        mov     eax,.cey       ;.cey - current  env map y
+        sar     ebx,ROUND
+        sar     eax,ROUND
-        shl     eax,TEX_SHIFT
-        add     ebx,eax
-        lea     ebx,[ebx*3]
-        add     ebx,.emap
+        shl     eax,TEX_SHIFT
+        add     ebx,eax
+        lea     ebx,[ebx*3]
+        add     ebx,.emap
 else
-        movq    mm5,mm4 ;.cey
-        psrad   mm5,ROUND
-        pslld   mm5,TEX_SHIFT
-        movq    mm6,mm3 ;.cex
-        psrad   mm6,ROUND
-        paddd   mm5,mm6
-        movq    mm6,mm5
-        paddd   mm5,mm5
-        paddd   mm5,mm6
-        paddd   mm5,.emap
-        movd    esi,mm5
-        psrlq   mm5,32
-        movd    ebx,mm5
+        movq    mm5,mm4 ;.cey
+        psrad   mm5,ROUND
+        pslld   mm5,TEX_SHIFT
+        movq    mm6,mm3 ;.cex
+        psrad   mm6,ROUND
+        paddd   mm5,mm6
+        movq    mm6,mm5
+        paddd   mm5,mm5
+        paddd   mm5,mm6
+        paddd   mm5,.emap
+        movd    esi,mm5
+        psrlq   mm5,32
+        movd    ebx,mm5
 end if
 if Ext>=MMX
-        movd      mm1,[esi]
-        movd      mm2,[ebx]
-        punpcklbw mm1,mm0
-        punpcklbw mm2,mm0
-        pmullw    mm1,mm2
-        psrlw     mm1,8
-        packuswb  mm1,mm0
-        movd      [edi],mm1
-        mov       ebx,.cz
-        mov       dword[edx],ebx
+        movd      mm1,[esi]
+        movd      mm2,[ebx]
+        punpcklbw mm1,mm0
+        punpcklbw mm2,mm0
+        pmullw    mm1,mm2
+        psrlw     mm1,8
+        packuswb  mm1,mm0
+        movd      [edi],mm1
+        mov       ebx,.cz
+        mov       dword[edx],ebx
 else
-        cld                     ; esi - tex e.
-        lodsb                   ; ebx - tex b.
-        mov     dl,[ebx]
-        mul     dl
-        shr     ax,8
-        stosb
-        inc     ebx
-        lodsb
-        mov     dl,[ebx]
-        mul     dl
-        shr     ax,8
-        stosb
-        inc     ebx
-        lodsb
-        mov     dl,[ebx]
-        mul     dl
-        shr     ax,8
-        stosb
-        mov     ebx,.cz
-        mov     esi,.czbuff
-        mov     dword[esi],ebx
-        jmp     .no_skip
+        cld                     ; esi - tex e.
+        lodsb                   ; ebx - tex b.
+        mov     dl,[ebx]
+        mul     dl
+        shr     ax,8
+        stosb
+        inc     ebx
+        lodsb
+        mov     dl,[ebx]
+        mul     dl
+        shr     ax,8
+        stosb
+        inc     ebx
+        lodsb
+        mov     dl,[ebx]
+        mul     dl
+        shr     ax,8
+        stosb
+        mov     ebx,.cz
+        mov     esi,.czbuff
+        mov     dword[esi],ebx
+        jmp     .no_skip
 end if
      .skip:
-        add     edi,3
+        add     edi,3
    if Ext = NON
      .no_skip:
-        add     .czbuff,4
-        mov     eax,.dbx
-        add     .cbx,eax
-        mov     eax,.dby
-        add     .cby,eax
-        mov     eax,.dex
-        add     .cex,eax
-        mov     eax,.dey
-        add     .cey,eax
+        add     .czbuff,4
+        mov     eax,.dbx
+        add     .cbx,eax
+        mov     eax,.dby
+        add     .cby,eax
+        mov     eax,.dex
+        add     .cex,eax
+        mov     eax,.dey
+        add     .cey,eax
     else
-        add     edx,4
-        paddd   mm3,.dex
-        paddd   mm4,.dey
+        add     edx,4
+        paddd   mm3,.dex
+        paddd   mm4,.dey
   ;      movq    mm5,mm3
   ;      movq    mm6,mm4
   ;      psrad   mm5,ROUND
 ,16 → 1076,16
      ;   movq    .cex,mm3
      ;   movq    .cey,mm4
     end if
-        mov     eax,.dz
-        add     .cz,eax
+        mov     eax,.dz
+        add     .cz,eax
     if Ext = NON
-        dec     ecx
-        jnz     .draw
+        dec     ecx
+        jnz     .draw
     else
-        loop    .draw
+        loop    .draw
     end if
   .bl_end:
-        mov     esp,ebp
+        mov     esp,ebp
 ret 56

 /programs/demos/3DS/VIEW3DS.ASM
 ,5 → 1,5
-; application : View3ds ver. 0.059 - tiny .3ds files viewer.
+; application : View3ds ver. 0.060 - tiny .3ds files viewer.
 ; compiler    : FASM
 ; system      : KolibriOS
 ; author      : Macgub aka Maciej Guba
 ,7 → 35,7
 MMX   =   1
 SSE   =   2
 SSE2  =   3
-Ext   =   MMX               ;Ext={ NON | MMX | SSE | SSE2 }
+Ext   =   MMX              ;Ext={ NON | MMX | SSE | SSE2 }
 ; 0 for short names (Menuet-compatible), 1 for long names (Kolibri features)
 USE_LFN = 1

 /programs/demos/3DS/readme.txt
 ,11 → 1,8
-View3ds 0.059 - tiny viewer to .3ds files.
+View3ds 0.060 - tiny viewer to .3ds files.
 What's new?
-. Bump and pararell two texture  mapping functions optimizations.
-   (files bump_cat.inc & two_tex.inc)
-   On my P4 changes are rather non visable, but on dual core in KlbrInWin
-   optimizations runs preety nice.
+. Header fix by Leency.
+. SSE2 optimizations by me. (Most visable in BUMP_TEX mode.)
 Buttons description:
 . rotary: choosing rotary axle: x, y, x+y.
 ,4 → 30,4
 . 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                  Jun 2011
+                         Macgub                  Aug 2011

Subversion Repositories Kolibri OS

Compare Revisions

Ignore whitespace Rev 2191 → Rev 2192