WebSVN – Kolibri OS – Diff – /programs/demos/view3ds/3dmath.inc


x3d equ 0
y3d equ 2
z3d equ 4
vec_x equ 0
vec_y equ 4
vec_z equ 8
; 3d point - triple integer word coordinate
; vector   - triple float dword coordinate
;----------------------in: --------------------------------
;------------------------ esi - pointer to 1st 3d point ---
;------------------------ edi - pointer to 2nd 3d point ---
;------------------------ ebx - pointer to result vector --
;---------------------- out : none ------------------------
reverse_mx_3x3:
; esi - source matrix
; edi - desired reversed matrix
 
  push  ebp
  mov   ebp,esp
  sub   esp,4
  .det  equ  ebp-4
 
  fninit
  fld  dword[esi]
  fmul dword[esi+16]
  fmul dword[esi+32]
  fld  dword[esi+12]
  fmul dword[esi+28]
  fmul dword[esi+8]
  faddp
  fld  dword[esi+24]
  fmul dword[esi+4]
  fmul dword[esi+20]
  faddp
  fld  dword[esi]
  fmul dword[esi+28]
  fmul dword[esi+20]
  fchs
  faddp
  fld  dword[esi+24]
  fmul dword[esi+16]
  fmul dword[esi+8]
  fchs
  faddp
  fld  dword[esi+12]
  fmul dword[esi+4]
  fmul dword[esi+32]
  fchs
  faddp
  fstp dword[.det]
  cmp  dword[.det],0
  jne  @f
  int3
 @@:
 ; fld1
 ; fdiv dword[.det]
 ; fstp dword[.det]
 
  fld  dword[esi+16]
  fmul dword[esi+32]
  fld  dword[esi+20]
  fmul dword[esi+28]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi]
 
  fld  dword[esi+8]
  fmul dword[esi+28]
  fld  dword[esi+4]
  fmul dword[esi+32]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+4]
 
  fld  dword[esi+4]
  fmul dword[esi+20]
  fld  dword[esi+8]
  fmul dword[esi+16]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+8]
 
  fld  dword[esi+20]
  fmul dword[esi+24]
  fld  dword[esi+12]
  fmul dword[esi+32]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+12]
 
  fld  dword[esi]
  fmul dword[esi+32]
  fld  dword[esi+8]
  fmul dword[esi+24]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+16]
 
  fld  dword[esi+8]
  fmul dword[esi+12]
  fld  dword[esi]
  fmul dword[esi+20]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+20]
 
  fld  dword[esi+12]
  fmul dword[esi+28]
  fld  dword[esi+16]
  fmul dword[esi+24]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+24]
 
  fld  dword[esi+4]
  fmul dword[esi+24]
  fld  dword[esi]
  fmul dword[esi+28]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+28]
 
  fld  dword[esi]
  fmul dword[esi+16]
  fld  dword[esi+4]
  fmul dword[esi+12]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+32]
 
 
  mov  esp,ebp
  pop  ebp
ret
 
make_vector_r:
   if Ext < SSE2
        fninit
        fld     dword[edi]                ;edi+x3d
        fsub    dword[esi]                ;esi+x3d
        fstp    dword[ebx+vec_x]
 
        fld     dword[edi+4]
        fsub    dword[esi+4]
        fstp    dword[ebx+vec_y]
 
        fld     dword[edi+8]
        fsub    dword[esi+8]
        fstp    dword[ebx+vec_z]
    else
        movups  xmm0,[esi]
        movups  xmm1,[edi]
        subps   xmm1,xmm0
        movlps  [ebx],xmm1
        movhlps  xmm1,xmm1
        movss   [ebx+8],xmm1
     end if
 
ret
;---------------------- in: -------------------------------
;--------------------------- esi - pointer to 1st vector --
;--------------------------- edi - pointer to 2nd vector --
;--------------------------- ebx - pointer to result vector
;---------------------- out : none
cross_product:
        fninit
        fld     dword [esi+vec_y]
        fmul    dword [edi+vec_z]
        fld     dword [esi+vec_z]
        fmul    dword [edi+vec_y]
        fsubp   ;st1 ,st
        fstp    dword [ebx+vec_x]
 
        fld     dword [esi+vec_z]
        fmul    dword [edi+vec_x]
        fld     dword [esi+vec_x]
        fmul    dword [edi+vec_z]
        fsubp   ;st1 ,st
        fstp    dword [ebx+vec_y]
 
        fld     dword [esi+vec_x]
        fmul    dword [edi+vec_y]
        fld     dword [esi+vec_y]
        fmul    dword [edi+vec_x]
        fsubp   ;st1 ,st
        fstp    dword [ebx+vec_z]
ret
;----------------------- in: ------------------------------
;---------------------------- edi - pointer to vector -----
;----------------------- out : none
normalize_vector:
if Ext = SSE2 | Ext = SSE | Ext = SSE3
        movups  xmm0,[edi]
        andps   xmm0,[zero_hgst_dd]
        movups  xmm1,xmm0
        mulps   xmm0,xmm0
        movhlps xmm2,xmm0
        addps   xmm0,xmm2
        movaps  xmm2,xmm0
        shufps  xmm2,xmm2,11100001b
        addps   xmm0,xmm2
        shufps  xmm0,xmm0,0
        rsqrtps xmm0,xmm0
        mulps   xmm0,xmm1
        movlps  [edi],xmm0
        movhlps xmm0,xmm0
        movss   [edi+8],xmm0
end if
if 0 ; Ext >= SSE3
        movups  xmm0,[edi]
        andps   xmm0,[zero_hgst_dd]
        movups  xmm1,xmm0
        mulps   xmm0,xmm0
        haddps  xmm0,xmm0
        haddps  xmm0,xmm0
        rsqrtps xmm0,xmm0
        mulps   xmm0,xmm1
        movlps  [edi],xmm0
        movhlps xmm0,xmm0
        movss   [edi+8],xmm0
end if
if Ext < SSE
 
        fninit
        fld     dword [edi+vec_x]
        fmul    st, st
        fld     dword [edi+vec_y]
        fmul    st, st
        fld     dword [edi+vec_z]
        fmul    st, st
        faddp   st1, st
        faddp   st1, st
        fsqrt
 
        ftst
        fstsw ax
        sahf
        jnz     @f
 
        fst     dword [edi+vec_x]
        fst     dword [edi+vec_y]
        fstp    dword [edi+vec_z]
        ret
      @@:
        fld st
        fld st
        fdivr dword [edi+vec_x]
        fstp  dword [edi+vec_x]
        fdivr dword [edi+vec_y]
        fstp  dword [edi+vec_y]
        fdivr dword [edi+vec_z]
        fstp  dword [edi+vec_z]
end if
ret
;------------------in: -------------------------
;------------------ esi - pointer to 1st vector
;------------------ edi - pointer to 2nd vector
;------------------out: ------------------------
;------------------ st0 - dot-product
dot_product:
        fninit
;if Ext >=SSE3
;        movups  xmm0,[esi]
;        movups  xmm1,[edi]
;        andps   xmm0,[zero_hgst_dd]
;        mulps   xmm0,xmm1
;        haddps  xmm0,xmm0
;        haddps  xmm0,xmm0
;        movss   [esp-4],xmm0
;        fld     dword[esp-4]
;else
        fld     dword [esi+vec_x]
        fmul    dword [edi+vec_x]
        fld     dword [esi+vec_y]
        fmul    dword [edi+vec_y]
        fld     dword [esi+vec_z]
        fmul    dword [edi+vec_z]
        faddp
        faddp
;end if
ret
 
; DOS version Coded by Mikolaj Felix aka Majuma
; mfelix@polbox.com
; www.majuma.xt.pl
; into FASM translation by Macgub
init_sincos_tab:
.counter   equ  dword [ebp-4]  ; cur angle
 
     push       ebp
     mov        ebp,esp
 
     xor        eax,eax
     push       eax            ; init .counter
     mov        edi,cos_tab
     mov        esi,sin_tab
     mov        ecx,256
     fninit
 
     fld        .counter
  @@:
     fld        st
     fsincos
     fstp       dword [edi]
     fstp       dword [esi]
;     fadd       [piD180]
     fadd       [piD128]
     add        esi,4
     add        edi,4
     loop       @b
     ffree      st
 
     mov        esp,ebp
     pop        ebp
ret
;------
; esi - offset (pointer) to angles, edi offset to 3x3 matrix
make_rotation_matrix:
   .sinx   equ dword[ebp-4]
   .cosx   equ dword[ebp-8]
   .siny   equ dword[ebp-12]
   .cosy   equ dword[ebp-16]
   .sinz   equ dword[ebp-20]
   .cosz   equ dword[ebp-24]
     push      ebp
     mov       ebp,esp
     sub       esp,24
 
     movzx     ebx,word[esi]
     shl       ebx,2
     mov       eax,dword[sin_tab+ebx]
     mov       .sinx,eax
     mov       edx,dword[cos_tab+ebx]
     mov       .cosx,edx
 
     movzx     ebx,word[esi+2]
     shl       ebx,2
     mov       eax,dword[sin_tab+ebx]
     mov       .siny,eax
     mov       edx,dword[cos_tab+ebx]
     mov       .cosy,edx
 
     movzx     ebx,word[esi+4]
     shl       ebx,2
     mov       eax,dword[sin_tab+ebx]
     mov       .sinz,eax
     mov       edx,dword[cos_tab+ebx]
     mov       .cosz,edx
 
     fninit
     fld       .cosy
     fmul      .cosz
     fstp      dword[edi]
 
     fld       .sinx
     fmul      .siny
     fmul      .cosz
     fld       .cosx
     fmul      .sinz
     fchs
     faddp
     fstp      dword[edi+12]
 
     fld       .cosx
     fmul      .siny
     fmul      .cosz
     fld       .sinx
     fmul      .sinz
     faddp
     fstp      dword[edi+24]
 
     fld       .cosy
     fmul      .sinz
     fstp      dword[edi+4]
 
     fld       .sinx
     fmul      .siny
     fmul      .sinz
     fld       .cosx
     fmul      .cosz
     faddp
     fstp      dword[edi+16]
 
     fld       .cosx
     fmul      .siny
     fmul      .sinz
     fld       .sinx
     fchs
     fmul      .cosz
     faddp
     fstp      dword[edi+28]
 
     fld       .siny
     fchs
     fstp      dword[edi+8]
 
     fld       .cosy
     fmul      .sinx
     fstp      dword[edi+20]
 
     fld       .cosx
     fmul      .cosy
     fstp      dword[edi+32]
 
     mov       esp,ebp
     pop       ebp
ret
;---------------------
;  in:  esi - ptr to points(normals], each point(normal) coeficient as dword
;       edi - ptr to rotated points(normals)
;       ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
;       ecx - number of points(normals)
rotary:
if Ext
    fninit
 .again:
 
    fld     dword[esi]
    fmul    dword[ebx]
    fld     dword[esi+4]
    fmul    dword[ebx+12]
    faddp
    fld     dword[esi+8]
    fmul    dword[ebx+24]
    faddp
    fstp    dword[edi]
 
 
    fld     dword[esi+4]
    fmul    dword[ebx+16]
    fld     dword[esi]
    fmul    dword[ebx+4]
    faddp
    fld     dword[esi+8]
    fmul    dword[ebx+28]
    faddp
    fstp    dword[edi+4]
 
 
    fld     dword[esi+8]
    fmul    dword[ebx+32]
    fld     dword[esi]
    fmul    dword[ebx+8]
    fld     dword[esi+4]
    fmul    dword[ebx+20]
    faddp
    faddp
    fstp    dword[edi+8]
 
 
    add     esi,12
    add     edi,12
    loop    .again
    mov     [edi],dword -1
else
;   Copyright (C) 1999-2001  Brian Paul
;   Copyright (C)            Maciej Guba
;---------------------
;  in:  esi - ptr to points(normals], each point(normal) coeficient as dword
;       edi - ptr to rotated points(normals)
;       ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
;       ecx - number of points(normals)
;align 32
    movups   xmm4,[ebx]
 ;   lddqu    xmm4,[ebx]   ; I tried sse3 :D
    movups   xmm5,[ebx+12]
    movups   xmm6,[ebx+24]
;align 32
  .again:
    movss    xmm0,dword[esi]
    shufps   xmm0,xmm0,0
    mulps    xmm0,xmm4
 
    movss    xmm1,dword[esi+4]
    shufps   xmm1,xmm1,0
    mulps    xmm1,xmm5
 
    movss    xmm2,dword[esi+8]
    shufps   xmm2,xmm2,0
    mulps    xmm2,xmm6
 
    addps    xmm0,xmm1
    addps    xmm0,xmm2
 
    movups   [edi],xmm0
 
    add      esi,12
    add      edi,12
    dec      ecx
    jne      .again
    mov      [edi],dword -1
end if
ret
;----------------------------------------------
;  esi - pointer to 3x3 matrix
add_scale_to_matrix:
     fninit
     fld     [rsscale]
     fld     dword[esi]            ;-----
     fmul    st,st1
     fstp    dword[esi]
     fld     dword[esi+12]           ; x scale
     fmul    st,st1
     fstp    dword[esi+12]
     fld     dword[esi+24]
     fmul    st,st1
     fstp    dword[esi+24]         ;------
 
     fld     dword[esi+4]          ;-----
     fmul    st,st1
     fstp    dword[esi+4]
     fld     dword[esi+16]            ; y scale
     fmul    st,st1
     fstp    dword[esi+16]
     fld     dword[esi+28]
     fmul    st,st1
     fstp    dword[esi+28]         ;------
 
 
     fld     dword[esi+8]          ;-----
     fmul    st,st1
     fstp    dword[esi+8]
     fld     dword[esi+20]              ; z scale
     fmul    st,st1
     fstp    dword[esi+20]
     fld     dword[esi+32]
     fmulp    st1,st
     fstp    dword[esi+32]         ;------
 
ret
 
;in   esi - offset to 3d points  (point as 3 dwords float)
;     edi - offset to 2d points  ( as 3 words integer)
;     ecx - number of points
translate_points:  ; just convert into integer; z coord still needed
    fninit
  .again:
  if 0
    fld    dword[esi+8]
 ;   fmul   [rsscale]
    fist   word[edi+4]
 
    fisub  [zobs]
    fchs
 
    fld    dword[esi]
;    fmul   [rsscale]
    fisub  [xobs]
    fimul  [zobs]
    fdiv   st0,st1
 
    fiadd  [xobs]
    fiadd  [vect_x]
    fistp  word[edi]
 
    fld    dword[esi+4]
;    fmul   [rsscale]
    fisub  [yobs]
    fimul  [zobs]
    fdivrp  ;   st0,st1
 
    fiadd  [yobs]
    fiadd  [vect_y]
    fistp  word[edi+2]
   end if
   ; movups   xmm0,[esi]
   ; cvtps2dq xmm0,xmm0
   ; packsdw xmm0,xmm0
   ; movq     [edi]
    fld    dword[esi]
    fiadd  [vect_x]
    fistp  word[edi]
    fld    dword[esi+4]
    fiadd  [vect_y]
    fistp  word[edi+2]
    fld    dword[esi+8]
    fistp  word[edi+4]
    add    esi,12
    add    edi,6
    dec    ecx
    jnz    .again
 
ret
>

Rev 8232	Rev 8666
1	x3d equ 0	1	x3d equ 0
2	y3d equ 2	2	y3d equ 2
3	z3d equ 4	3	z3d equ 4
4	vec_x equ 0	4	vec_x equ 0
5	vec_y equ 4	5	vec_y equ 4
6	vec_z equ 8	6	vec_z equ 8
7	; 3d point - triple integer word coordinate	7	; 3d point - triple integer word coordinate
8	; vector - triple float dword coordinate	8	; vector - triple float dword coordinate
9	;----------------------in: --------------------------------	9	;----------------------in: --------------------------------
10	;------------------------ esi - pointer to 1st 3d point ---	10	;------------------------ esi - pointer to 1st 3d point ---
11	;------------------------ edi - pointer to 2nd 3d point ---	11	;------------------------ edi - pointer to 2nd 3d point ---
12	;------------------------ ebx - pointer to result vector --	12	;------------------------ ebx - pointer to result vector --
13	;---------------------- out : none ------------------------	13	;---------------------- out : none ------------------------
14	reverse_mx_3x3:	14	reverse_mx_3x3:
15	; esi - source matrix	15	; esi - source matrix
16	; edi - desired reversed matrix	16	; edi - desired reversed matrix
17		17
18	push ebp	18	push ebp
19	mov ebp,esp	19	mov ebp,esp
20	sub esp,4	20	sub esp,4
21	.det equ ebp-4	21	.det equ ebp-4
22		22
23	fninit	23	fninit
24	fld dword[esi]	24	fld dword[esi]
25	fmul dword[esi+16]	25	fmul dword[esi+16]
26	fmul dword[esi+32]	26	fmul dword[esi+32]
27	fld dword[esi+12]	27	fld dword[esi+12]
28	fmul dword[esi+28]	28	fmul dword[esi+28]
29	fmul dword[esi+8]	29	fmul dword[esi+8]
30	faddp	30	faddp
31	fld dword[esi+24]	31	fld dword[esi+24]
32	fmul dword[esi+4]	32	fmul dword[esi+4]
33	fmul dword[esi+20]	33	fmul dword[esi+20]
34	faddp	34	faddp
35	fld dword[esi]	35	fld dword[esi]
36	fmul dword[esi+28]	36	fmul dword[esi+28]
37	fmul dword[esi+20]	37	fmul dword[esi+20]
38	fchs	38	fchs
39	faddp	39	faddp
40	fld dword[esi+24]	40	fld dword[esi+24]
41	fmul dword[esi+16]	41	fmul dword[esi+16]
42	fmul dword[esi+8]	42	fmul dword[esi+8]
43	fchs	43	fchs
44	faddp	44	faddp
45	fld dword[esi+12]	45	fld dword[esi+12]
46	fmul dword[esi+4]	46	fmul dword[esi+4]
47	fmul dword[esi+32]	47	fmul dword[esi+32]
48	fchs	48	fchs
49	faddp	49	faddp
50	fstp dword[.det]	50	fstp dword[.det]
51	cmp dword[.det],0	51	cmp dword[.det],0
52	jne @f	52	jne @f
53	int3	53	int3
54	@@:	54	@@:
55	; fld1	55	; fld1
56	; fdiv dword[.det]	56	; fdiv dword[.det]
57	; fstp dword[.det]	57	; fstp dword[.det]
58		58
59	fld dword[esi+16]	59	fld dword[esi+16]
60	fmul dword[esi+32]	60	fmul dword[esi+32]
61	fld dword[esi+20]	61	fld dword[esi+20]
62	fmul dword[esi+28]	62	fmul dword[esi+28]
63	fchs	63	fchs
64	faddp	64	faddp
65	fdiv dword[.det]	65	fdiv dword[.det]
66	fstp dword[edi]	66	fstp dword[edi]
67		67
68	fld dword[esi+8]	68	fld dword[esi+8]
69	fmul dword[esi+28]	69	fmul dword[esi+28]
70	fld dword[esi+4]	70	fld dword[esi+4]
71	fmul dword[esi+32]	71	fmul dword[esi+32]
72	fchs	72	fchs
73	faddp	73	faddp
74	fdiv dword[.det]	74	fdiv dword[.det]
75	fstp dword[edi+4]	75	fstp dword[edi+4]
76		76
77	fld dword[esi+4]	77	fld dword[esi+4]
78	fmul dword[esi+20]	78	fmul dword[esi+20]
79	fld dword[esi+8]	79	fld dword[esi+8]
80	fmul dword[esi+16]	80	fmul dword[esi+16]
81	fchs	81	fchs
82	faddp	82	faddp
83	fdiv dword[.det]	83	fdiv dword[.det]
84	fstp dword[edi+8]	84	fstp dword[edi+8]
85		85
86	fld dword[esi+20]	86	fld dword[esi+20]
87	fmul dword[esi+24]	87	fmul dword[esi+24]
88	fld dword[esi+12]	88	fld dword[esi+12]
89	fmul dword[esi+32]	89	fmul dword[esi+32]
90	fchs	90	fchs
91	faddp	91	faddp
92	fdiv dword[.det]	92	fdiv dword[.det]
93	fstp dword[edi+12]	93	fstp dword[edi+12]
94		94
95	fld dword[esi]	95	fld dword[esi]
96	fmul dword[esi+32]	96	fmul dword[esi+32]
97	fld dword[esi+8]	97	fld dword[esi+8]
98	fmul dword[esi+24]	98	fmul dword[esi+24]
99	fchs	99	fchs
100	faddp	100	faddp
101	fdiv dword[.det]	101	fdiv dword[.det]
102	fstp dword[edi+16]	102	fstp dword[edi+16]
103		103
104	fld dword[esi+8]	104	fld dword[esi+8]
105	fmul dword[esi+12]	105	fmul dword[esi+12]
106	fld dword[esi]	106	fld dword[esi]
107	fmul dword[esi+20]	107	fmul dword[esi+20]
108	fchs	108	fchs
109	faddp	109	faddp
110	fdiv dword[.det]	110	fdiv dword[.det]
111	fstp dword[edi+20]	111	fstp dword[edi+20]
112		112
113	fld dword[esi+12]	113	fld dword[esi+12]
114	fmul dword[esi+28]	114	fmul dword[esi+28]
115	fld dword[esi+16]	115	fld dword[esi+16]
116	fmul dword[esi+24]	116	fmul dword[esi+24]
117	fchs	117	fchs
118	faddp	118	faddp
119	fdiv dword[.det]	119	fdiv dword[.det]
120	fstp dword[edi+24]	120	fstp dword[edi+24]
121		121
122	fld dword[esi+4]	122	fld dword[esi+4]
123	fmul dword[esi+24]	123	fmul dword[esi+24]
124	fld dword[esi]	124	fld dword[esi]
125	fmul dword[esi+28]	125	fmul dword[esi+28]
126	fchs	126	fchs
127	faddp	127	faddp
128	fdiv dword[.det]	128	fdiv dword[.det]
129	fstp dword[edi+28]	129	fstp dword[edi+28]
130		130
131	fld dword[esi]	131	fld dword[esi]
132	fmul dword[esi+16]	132	fmul dword[esi+16]
133	fld dword[esi+4]	133	fld dword[esi+4]
134	fmul dword[esi+12]	134	fmul dword[esi+12]
135	fchs	135	fchs
136	faddp	136	faddp
137	fdiv dword[.det]	137	fdiv dword[.det]
138	fstp dword[edi+32]	138	fstp dword[edi+32]
139		139
140		140
141	mov esp,ebp	141	mov esp,ebp
142	pop ebp	142	pop ebp
143	ret	143	ret
144		144
145	make_vector_r:	145	make_vector_r:
146	if Ext < SSE2	146	if Ext < SSE2
147	fninit	147	fninit
148	fld dword[edi] ;edi+x3d	148	fld dword[edi] ;edi+x3d
149	fsub dword[esi] ;esi+x3d	149	fsub dword[esi] ;esi+x3d
150	fstp dword[ebx+vec_x]	150	fstp dword[ebx+vec_x]
151		151
152	fld dword[edi+4]	152	fld dword[edi+4]
153	fsub dword[esi+4]	153	fsub dword[esi+4]
154	fstp dword[ebx+vec_y]	154	fstp dword[ebx+vec_y]
155		155
156	fld dword[edi+8]	156	fld dword[edi+8]
157	fsub dword[esi+8]	157	fsub dword[esi+8]
158	fstp dword[ebx+vec_z]	158	fstp dword[ebx+vec_z]
159	else	159	else
160	movups xmm0,[esi]	160	movups xmm0,[esi]
161	movups xmm1,[edi]	161	movups xmm1,[edi]
162	subps xmm1,xmm0	162	subps xmm1,xmm0
163	movlps [ebx],xmm1	163	movlps [ebx],xmm1
164	movhlps xmm1,xmm1	164	movhlps xmm1,xmm1
165	movss [ebx+8],xmm1	165	movss [ebx+8],xmm1
166	end if	166	end if
167		167
168	ret	168	ret
169	;---------------------- in: -------------------------------	169	;---------------------- in: -------------------------------
170	;--------------------------- esi - pointer to 1st vector --	170	;--------------------------- esi - pointer to 1st vector --
171	;--------------------------- edi - pointer to 2nd vector --	171	;--------------------------- edi - pointer to 2nd vector --
172	;--------------------------- ebx - pointer to result vector	172	;--------------------------- ebx - pointer to result vector
173	;---------------------- out : none	173	;---------------------- out : none
174	cross_product:	174	cross_product:
175	fninit	175	fninit
176	fld dword [esi+vec_y]	176	fld dword [esi+vec_y]
177	fmul dword [edi+vec_z]	177	fmul dword [edi+vec_z]
178	fld dword [esi+vec_z]	178	fld dword [esi+vec_z]
179	fmul dword [edi+vec_y]	179	fmul dword [edi+vec_y]
180	fsubp ;st1 ,st	180	fsubp ;st1 ,st
181	fstp dword [ebx+vec_x]	181	fstp dword [ebx+vec_x]
182		182
183	fld dword [esi+vec_z]	183	fld dword [esi+vec_z]
184	fmul dword [edi+vec_x]	184	fmul dword [edi+vec_x]
185	fld dword [esi+vec_x]	185	fld dword [esi+vec_x]
186	fmul dword [edi+vec_z]	186	fmul dword [edi+vec_z]
187	fsubp ;st1 ,st	187	fsubp ;st1 ,st
188	fstp dword [ebx+vec_y]	188	fstp dword [ebx+vec_y]
189		189
190	fld dword [esi+vec_x]	190	fld dword [esi+vec_x]
191	fmul dword [edi+vec_y]	191	fmul dword [edi+vec_y]
192	fld dword [esi+vec_y]	192	fld dword [esi+vec_y]
193	fmul dword [edi+vec_x]	193	fmul dword [edi+vec_x]
194	fsubp ;st1 ,st	194	fsubp ;st1 ,st
195	fstp dword [ebx+vec_z]	195	fstp dword [ebx+vec_z]
196	ret	196	ret
197	;----------------------- in: ------------------------------	197	;----------------------- in: ------------------------------
198	;---------------------------- edi - pointer to vector -----	198	;---------------------------- edi - pointer to vector -----
199	;----------------------- out : none	199	;----------------------- out : none
200	normalize_vector:	200	normalize_vector:
-		201	if Ext = SSE2 \| Ext = SSE \| Ext = SSE3
-		202	movups xmm0,[edi]
-		203	andps xmm0,[zero_hgst_dd]
-		204	movups xmm1,xmm0
-		205	mulps xmm0,xmm0
-		206	movhlps xmm2,xmm0
-		207	addps xmm0,xmm2
-		208	movaps xmm2,xmm0
-		209	shufps xmm2,xmm2,11100001b
-		210	addps xmm0,xmm2
-		211	shufps xmm0,xmm0,0
-		212	rsqrtps xmm0,xmm0
-		213	mulps xmm0,xmm1
-		214	movlps [edi],xmm0
-		215	movhlps xmm0,xmm0
-		216	movss [edi+8],xmm0
-		217	end if
201	if Ext >= SSE3	218	if 0 ; Ext >= SSE3
202	movups xmm0,[edi]	219	movups xmm0,[edi]
203	andps xmm0,[zero_hgst_dd]	220	andps xmm0,[zero_hgst_dd]
204	movups xmm1,xmm0	221	movups xmm1,xmm0
205	mulps xmm0,xmm0	222	mulps xmm0,xmm0
206	haddps xmm0,xmm0	223	haddps xmm0,xmm0
207	haddps xmm0,xmm0	224	haddps xmm0,xmm0
208	rsqrtps xmm0,xmm0	225	rsqrtps xmm0,xmm0
209	mulps xmm0,xmm1	226	mulps xmm0,xmm1
210	movlps [edi],xmm0	227	movlps [edi],xmm0
211	movhlps xmm0,xmm0	228	movhlps xmm0,xmm0
212	movss [edi+8],xmm0	229	movss [edi+8],xmm0
213	else	230	end if
-		231	if Ext < SSE
214		232
215	fninit	233	fninit
216	fld dword [edi+vec_x]	234	fld dword [edi+vec_x]
217	fmul st, st	235	fmul st, st
218	fld dword [edi+vec_y]	236	fld dword [edi+vec_y]
219	fmul st, st	237	fmul st, st
220	fld dword [edi+vec_z]	238	fld dword [edi+vec_z]
221	fmul st, st	239	fmul st, st
222	faddp st1, st	240	faddp st1, st
223	faddp st1, st	241	faddp st1, st
224	fsqrt	242	fsqrt
225		243
226	ftst	244	ftst
227	fstsw ax	245	fstsw ax
228	sahf	246	sahf
229	jnz @f	247	jnz @f
230		248
231	fst dword [edi+vec_x]	249	fst dword [edi+vec_x]
232	fst dword [edi+vec_y]	250	fst dword [edi+vec_y]
233	fstp dword [edi+vec_z]	251	fstp dword [edi+vec_z]
234	ret	252	ret
235	@@:	253	@@:
236	fld st	254	fld st
237	fld st	255	fld st
238	fdivr dword [edi+vec_x]	256	fdivr dword [edi+vec_x]
239	fstp dword [edi+vec_x]	257	fstp dword [edi+vec_x]
240	fdivr dword [edi+vec_y]	258	fdivr dword [edi+vec_y]
241	fstp dword [edi+vec_y]	259	fstp dword [edi+vec_y]
242	fdivr dword [edi+vec_z]	260	fdivr dword [edi+vec_z]
243	fstp dword [edi+vec_z]	261	fstp dword [edi+vec_z]
244	end if	262	end if
245	ret	263	ret
246	;------------------in: -------------------------	264	;------------------in: -------------------------
247	;------------------ esi - pointer to 1st vector	265	;------------------ esi - pointer to 1st vector
248	;------------------ edi - pointer to 2nd vector	266	;------------------ edi - pointer to 2nd vector
249	;------------------out: ------------------------	267	;------------------out: ------------------------
250	;------------------ st0 - dot-product	268	;------------------ st0 - dot-product
251	dot_product:	269	dot_product:
252	fninit	270	fninit
253	;if Ext >=SSE3	271	;if Ext >=SSE3
254	; movups xmm0,[esi]	272	; movups xmm0,[esi]
255	; movups xmm1,[edi]	273	; movups xmm1,[edi]
256	; andps xmm0,[zero_hgst_dd]	274	; andps xmm0,[zero_hgst_dd]
257	; mulps xmm0,xmm1	275	; mulps xmm0,xmm1
258	; haddps xmm0,xmm0	276	; haddps xmm0,xmm0
259	; haddps xmm0,xmm0	277	; haddps xmm0,xmm0
260	; movss [esp-4],xmm0	278	; movss [esp-4],xmm0
261	; fld dword[esp-4]	279	; fld dword[esp-4]
262	;else	280	;else
263	fld dword [esi+vec_x]	281	fld dword [esi+vec_x]
264	fmul dword [edi+vec_x]	282	fmul dword [edi+vec_x]
265	fld dword [esi+vec_y]	283	fld dword [esi+vec_y]
266	fmul dword [edi+vec_y]	284	fmul dword [edi+vec_y]
267	fld dword [esi+vec_z]	285	fld dword [esi+vec_z]
268	fmul dword [edi+vec_z]	286	fmul dword [edi+vec_z]
269	faddp	287	faddp
270	faddp	288	faddp
271	;end if	289	;end if
272	ret	290	ret
273		291
274	; DOS version Coded by Mikolaj Felix aka Majuma	292	; DOS version Coded by Mikolaj Felix aka Majuma
275	; mfelix@polbox.com	293	; mfelix@polbox.com
276	; www.majuma.xt.pl	294	; www.majuma.xt.pl
277	; into FASM translation by Macgub	295	; into FASM translation by Macgub
278	init_sincos_tab:	296	init_sincos_tab:
279	.counter equ dword [ebp-4] ; cur angle	297	.counter equ dword [ebp-4] ; cur angle
280		298
281	push ebp	299	push ebp
282	mov ebp,esp	300	mov ebp,esp
283		301
284	xor eax,eax	302	xor eax,eax
285	push eax ; init .counter	303	push eax ; init .counter
286	mov edi,cos_tab	304	mov edi,cos_tab
287	mov esi,sin_tab	305	mov esi,sin_tab
288	mov ecx,256	306	mov ecx,256
289	fninit	307	fninit
290		308
291	fld .counter	309	fld .counter
292	@@:	310	@@:
293	fld st	311	fld st
294	fsincos	312	fsincos
295	fstp dword [edi]	313	fstp dword [edi]
296	fstp dword [esi]	314	fstp dword [esi]
297	; fadd [piD180]	315	; fadd [piD180]
298	fadd [piD128]	316	fadd [piD128]
299	add esi,4	317	add esi,4
300	add edi,4	318	add edi,4
301	loop @b	319	loop @b
302	ffree st	320	ffree st
303		321
304	mov esp,ebp	322	mov esp,ebp
305	pop ebp	323	pop ebp
306	ret	324	ret
307	;------	325	;------
308	; esi - offset (pointer) to angles, edi offset to 3x3 matrix	326	; esi - offset (pointer) to angles, edi offset to 3x3 matrix
309	make_rotation_matrix:	327	make_rotation_matrix:
310	.sinx equ dword[ebp-4]	328	.sinx equ dword[ebp-4]
311	.cosx equ dword[ebp-8]	329	.cosx equ dword[ebp-8]
312	.siny equ dword[ebp-12]	330	.siny equ dword[ebp-12]
313	.cosy equ dword[ebp-16]	331	.cosy equ dword[ebp-16]
314	.sinz equ dword[ebp-20]	332	.sinz equ dword[ebp-20]
315	.cosz equ dword[ebp-24]	333	.cosz equ dword[ebp-24]
316	push ebp	334	push ebp
317	mov ebp,esp	335	mov ebp,esp
318	sub esp,24	336	sub esp,24
319		337
320	movzx ebx,word[esi]	338	movzx ebx,word[esi]
321	shl ebx,2	339	shl ebx,2
322	mov eax,dword[sin_tab+ebx]	340	mov eax,dword[sin_tab+ebx]
323	mov .sinx,eax	341	mov .sinx,eax
324	mov edx,dword[cos_tab+ebx]	342	mov edx,dword[cos_tab+ebx]
325	mov .cosx,edx	343	mov .cosx,edx
326		344
327	movzx ebx,word[esi+2]	345	movzx ebx,word[esi+2]
328	shl ebx,2	346	shl ebx,2
329	mov eax,dword[sin_tab+ebx]	347	mov eax,dword[sin_tab+ebx]
330	mov .siny,eax	348	mov .siny,eax
331	mov edx,dword[cos_tab+ebx]	349	mov edx,dword[cos_tab+ebx]
332	mov .cosy,edx	350	mov .cosy,edx
333		351
334	movzx ebx,word[esi+4]	352	movzx ebx,word[esi+4]
335	shl ebx,2	353	shl ebx,2
336	mov eax,dword[sin_tab+ebx]	354	mov eax,dword[sin_tab+ebx]
337	mov .sinz,eax	355	mov .sinz,eax
338	mov edx,dword[cos_tab+ebx]	356	mov edx,dword[cos_tab+ebx]
339	mov .cosz,edx	357	mov .cosz,edx
340		358
341	fninit	359	fninit
342	fld .cosy	360	fld .cosy
343	fmul .cosz	361	fmul .cosz
344	fstp dword[edi]	362	fstp dword[edi]
345		363
346	fld .sinx	364	fld .sinx
347	fmul .siny	365	fmul .siny
348	fmul .cosz	366	fmul .cosz
349	fld .cosx	367	fld .cosx
350	fmul .sinz	368	fmul .sinz
351	fchs	369	fchs
352	faddp	370	faddp
353	fstp dword[edi+12]	371	fstp dword[edi+12]
354		372
355	fld .cosx	373	fld .cosx
356	fmul .siny	374	fmul .siny
357	fmul .cosz	375	fmul .cosz
358	fld .sinx	376	fld .sinx
359	fmul .sinz	377	fmul .sinz
360	faddp	378	faddp
361	fstp dword[edi+24]	379	fstp dword[edi+24]
362		380
363	fld .cosy	381	fld .cosy
364	fmul .sinz	382	fmul .sinz
365	fstp dword[edi+4]	383	fstp dword[edi+4]
366		384
367	fld .sinx	385	fld .sinx
368	fmul .siny	386	fmul .siny
369	fmul .sinz	387	fmul .sinz
370	fld .cosx	388	fld .cosx
371	fmul .cosz	389	fmul .cosz
372	faddp	390	faddp
373	fstp dword[edi+16]	391	fstp dword[edi+16]
374		392
375	fld .cosx	393	fld .cosx
376	fmul .siny	394	fmul .siny
377	fmul .sinz	395	fmul .sinz
378	fld .sinx	396	fld .sinx
379	fchs	397	fchs
380	fmul .cosz	398	fmul .cosz
381	faddp	399	faddp
382	fstp dword[edi+28]	400	fstp dword[edi+28]
383		401
384	fld .siny	402	fld .siny
385	fchs	403	fchs
386	fstp dword[edi+8]	404	fstp dword[edi+8]
387		405
388	fld .cosy	406	fld .cosy
389	fmul .sinx	407	fmul .sinx
390	fstp dword[edi+20]	408	fstp dword[edi+20]
391		409
392	fld .cosx	410	fld .cosx
393	fmul .cosy	411	fmul .cosy
394	fstp dword[edi+32]	412	fstp dword[edi+32]
395		413
396	mov esp,ebp	414	mov esp,ebp
397	pop ebp	415	pop ebp
398	ret	416	ret
399	;---------------------	417	;---------------------
400	; in: esi - ptr to points(normals], each point(normal) coeficient as dword	418	; in: esi - ptr to points(normals], each point(normal) coeficient as dword
401	; edi - ptr to rotated points(normals)	419	; edi - ptr to rotated points(normals)
402	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix	420	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
403	; ecx - number of points(normals)	421	; ecx - number of points(normals)
404	rotary:	422	rotary:
405	if Ext	423	if Ext
406	fninit	424	fninit
407	.again:	425	.again:
408		426
409	fld dword[esi]	427	fld dword[esi]
410	fmul dword[ebx]	428	fmul dword[ebx]
411	fld dword[esi+4]	429	fld dword[esi+4]
412	fmul dword[ebx+12]	430	fmul dword[ebx+12]
413	faddp	431	faddp
414	fld dword[esi+8]	432	fld dword[esi+8]
415	fmul dword[ebx+24]	433	fmul dword[ebx+24]
416	faddp	434	faddp
417	fstp dword[edi]	435	fstp dword[edi]
418		436
419		437
420	fld dword[esi+4]	438	fld dword[esi+4]
421	fmul dword[ebx+16]	439	fmul dword[ebx+16]
422	fld dword[esi]	440	fld dword[esi]
423	fmul dword[ebx+4]	441	fmul dword[ebx+4]
424	faddp	442	faddp
425	fld dword[esi+8]	443	fld dword[esi+8]
426	fmul dword[ebx+28]	444	fmul dword[ebx+28]
427	faddp	445	faddp
428	fstp dword[edi+4]	446	fstp dword[edi+4]
429		447
430		448
431	fld dword[esi+8]	449	fld dword[esi+8]
432	fmul dword[ebx+32]	450	fmul dword[ebx+32]
433	fld dword[esi]	451	fld dword[esi]
434	fmul dword[ebx+8]	452	fmul dword[ebx+8]
435	fld dword[esi+4]	453	fld dword[esi+4]
436	fmul dword[ebx+20]	454	fmul dword[ebx+20]
437	faddp	455	faddp
438	faddp	456	faddp
439	fstp dword[edi+8]	457	fstp dword[edi+8]
440		458
441		459
442	add esi,12	460	add esi,12
443	add edi,12	461	add edi,12
444	loop .again	462	loop .again
445	mov [edi],dword -1	463	mov [edi],dword -1
446	else	464	else
447	; Copyright (C) 1999-2001 Brian Paul	465	; Copyright (C) 1999-2001 Brian Paul
448	; Copyright (C) Maciej Guba	466	; Copyright (C) Maciej Guba
449	;---------------------	467	;---------------------
450	; in: esi - ptr to points(normals], each point(normal) coeficient as dword	468	; in: esi - ptr to points(normals], each point(normal) coeficient as dword
451	; edi - ptr to rotated points(normals)	469	; edi - ptr to rotated points(normals)
452	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix	470	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
453	; ecx - number of points(normals)	471	; ecx - number of points(normals)
454	;align 32	472	;align 32
455	movups xmm4,[ebx]	473	movups xmm4,[ebx]
456	; lddqu xmm4,[ebx] ; I tried sse3 :D	474	; lddqu xmm4,[ebx] ; I tried sse3 :D
457	movups xmm5,[ebx+12]	475	movups xmm5,[ebx+12]
458	movups xmm6,[ebx+24]	476	movups xmm6,[ebx+24]
459	;align 32	477	;align 32
460	.again:	478	.again:
461	movss xmm0,dword[esi]	479	movss xmm0,dword[esi]
462	shufps xmm0,xmm0,0	480	shufps xmm0,xmm0,0
463	mulps xmm0,xmm4	481	mulps xmm0,xmm4
464		482
465	movss xmm1,dword[esi+4]	483	movss xmm1,dword[esi+4]
466	shufps xmm1,xmm1,0	484	shufps xmm1,xmm1,0
467	mulps xmm1,xmm5	485	mulps xmm1,xmm5
468		486
469	movss xmm2,dword[esi+8]	487	movss xmm2,dword[esi+8]
470	shufps xmm2,xmm2,0	488	shufps xmm2,xmm2,0
471	mulps xmm2,xmm6	489	mulps xmm2,xmm6
472		490
473	addps xmm0,xmm1	491	addps xmm0,xmm1
474	addps xmm0,xmm2	492	addps xmm0,xmm2
475		493
476	movups [edi],xmm0	494	movups [edi],xmm0
477		495
478	add esi,12	496	add esi,12
479	add edi,12	497	add edi,12
480	dec ecx	498	dec ecx
481	jne .again	499	jne .again
482	mov [edi],dword -1	500	mov [edi],dword -1

Subversion Repositories Kolibri OS

(root)/programs/demos/view3ds/3dmath.inc – Rev 8232 → 8666