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 >= 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
else
 
        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 8014	Rev 8232
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	if 0	-
15	make_vector:	-
16	fninit	-
17	fild word[edi+x3d] ;edi+x3d	-
18	fisub word[esi+x3d] ;esi+x3d	-
19	fstp dword[ebx+vec_x]	-
20		-
21	fild word[edi+y3d]	-
22	fisub word[esi+y3d]	-
23	fstp dword[ebx+vec_y]	-
24		-
25	fild word[edi+z3d]	-
26	fisub word[esi+z3d]	-
27	fstp dword[ebx+vec_z]	-
28		-
29	ret	-
30	end if	-
31	reverse_mx_3x3:	14	reverse_mx_3x3:
32	; esi - source matrix	15	; esi - source matrix
33	; edi - desired reversed matrix	16	; edi - desired reversed matrix
34		17
35	push ebp	18	push ebp
36	mov ebp,esp	19	mov ebp,esp
37	sub esp,4	20	sub esp,4
38	.det equ ebp-4	21	.det equ ebp-4
39		22
40	fninit	23	fninit
41	fld dword[esi]	24	fld dword[esi]
42	fmul dword[esi+16]	25	fmul dword[esi+16]
43	fmul dword[esi+32]	26	fmul dword[esi+32]
44	fld dword[esi+12]	27	fld dword[esi+12]
45	fmul dword[esi+28]	28	fmul dword[esi+28]
46	fmul dword[esi+8]	29	fmul dword[esi+8]
47	faddp	30	faddp
48	fld dword[esi+24]	31	fld dword[esi+24]
49	fmul dword[esi+4]	32	fmul dword[esi+4]
50	fmul dword[esi+20]	33	fmul dword[esi+20]
51	faddp	34	faddp
52	fld dword[esi]	35	fld dword[esi]
53	fmul dword[esi+28]	36	fmul dword[esi+28]
54	fmul dword[esi+20]	37	fmul dword[esi+20]
55	fchs	38	fchs
56	faddp	39	faddp
57	fld dword[esi+24]	40	fld dword[esi+24]
58	fmul dword[esi+16]	41	fmul dword[esi+16]
59	fmul dword[esi+8]	42	fmul dword[esi+8]
60	fchs	43	fchs
61	faddp	44	faddp
62	fld dword[esi+12]	45	fld dword[esi+12]
63	fmul dword[esi+4]	46	fmul dword[esi+4]
64	fmul dword[esi+32]	47	fmul dword[esi+32]
65	fchs	48	fchs
66	faddp	49	faddp
67	fstp dword[.det]	50	fstp dword[.det]
68	cmp dword[.det],0	51	cmp dword[.det],0
69	jne @f	52	jne @f
70	int3	53	int3
71	@@:	54	@@:
72	; fld1	55	; fld1
73	; fdiv dword[.det]	56	; fdiv dword[.det]
74	; fstp dword[.det]	57	; fstp dword[.det]
75		58
76	fld dword[esi+16]	59	fld dword[esi+16]
77	fmul dword[esi+32]	60	fmul dword[esi+32]
78	fld dword[esi+20]	61	fld dword[esi+20]
79	fmul dword[esi+28]	62	fmul dword[esi+28]
80	fchs	63	fchs
81	faddp	64	faddp
82	fdiv dword[.det]	65	fdiv dword[.det]
83	fstp dword[edi]	66	fstp dword[edi]
84		67
85	fld dword[esi+8]	68	fld dword[esi+8]
86	fmul dword[esi+28]	69	fmul dword[esi+28]
87	fld dword[esi+4]	70	fld dword[esi+4]
88	fmul dword[esi+32]	71	fmul dword[esi+32]
89	fchs	72	fchs
90	faddp	73	faddp
91	fdiv dword[.det]	74	fdiv dword[.det]
92	fstp dword[edi+4]	75	fstp dword[edi+4]
93		76
94	fld dword[esi+4]	77	fld dword[esi+4]
95	fmul dword[esi+20]	78	fmul dword[esi+20]
96	fld dword[esi+8]	79	fld dword[esi+8]
97	fmul dword[esi+16]	80	fmul dword[esi+16]
98	fchs	81	fchs
99	faddp	82	faddp
100	fdiv dword[.det]	83	fdiv dword[.det]
101	fstp dword[edi+8]	84	fstp dword[edi+8]
102		85
103	fld dword[esi+20]	86	fld dword[esi+20]
104	fmul dword[esi+24]	87	fmul dword[esi+24]
105	fld dword[esi+12]	88	fld dword[esi+12]
106	fmul dword[esi+32]	89	fmul dword[esi+32]
107	fchs	90	fchs
108	faddp	91	faddp
109	fdiv dword[.det]	92	fdiv dword[.det]
110	fstp dword[edi+12]	93	fstp dword[edi+12]
111		94
112	fld dword[esi]	95	fld dword[esi]
113	fmul dword[esi+32]	96	fmul dword[esi+32]
114	fld dword[esi+8]	97	fld dword[esi+8]
115	fmul dword[esi+24]	98	fmul dword[esi+24]
116	fchs	99	fchs
117	faddp	100	faddp
118	fdiv dword[.det]	101	fdiv dword[.det]
119	fstp dword[edi+16]	102	fstp dword[edi+16]
120		103
121	fld dword[esi+8]	104	fld dword[esi+8]
122	fmul dword[esi+12]	105	fmul dword[esi+12]
123	fld dword[esi]	106	fld dword[esi]
124	fmul dword[esi+20]	107	fmul dword[esi+20]
125	fchs	108	fchs
126	faddp	109	faddp
127	fdiv dword[.det]	110	fdiv dword[.det]
128	fstp dword[edi+20]	111	fstp dword[edi+20]
129		112
130	fld dword[esi+12]	113	fld dword[esi+12]
131	fmul dword[esi+28]	114	fmul dword[esi+28]
132	fld dword[esi+16]	115	fld dword[esi+16]
133	fmul dword[esi+24]	116	fmul dword[esi+24]
134	fchs	117	fchs
135	faddp	118	faddp
136	fdiv dword[.det]	119	fdiv dword[.det]
137	fstp dword[edi+24]	120	fstp dword[edi+24]
138		121
139	fld dword[esi+4]	122	fld dword[esi+4]
140	fmul dword[esi+24]	123	fmul dword[esi+24]
141	fld dword[esi]	124	fld dword[esi]
142	fmul dword[esi+28]	125	fmul dword[esi+28]
143	fchs	126	fchs
144	faddp	127	faddp
145	fdiv dword[.det]	128	fdiv dword[.det]
146	fstp dword[edi+28]	129	fstp dword[edi+28]
147		130
148	fld dword[esi]	131	fld dword[esi]
149	fmul dword[esi+16]	132	fmul dword[esi+16]
150	fld dword[esi+4]	133	fld dword[esi+4]
151	fmul dword[esi+12]	134	fmul dword[esi+12]
152	fchs	135	fchs
153	faddp	136	faddp
154	fdiv dword[.det]	137	fdiv dword[.det]
155	fstp dword[edi+32]	138	fstp dword[edi+32]
156		139
157		140
158	mov esp,ebp	141	mov esp,ebp
159	pop ebp	142	pop ebp
160	ret	143	ret
161		144
162	make_vector_r:	145	make_vector_r:
163	if Ext < SSE2	146	if Ext < SSE2
164	fninit	147	fninit
165	fld dword[edi] ;edi+x3d	148	fld dword[edi] ;edi+x3d
166	fsub dword[esi] ;esi+x3d	149	fsub dword[esi] ;esi+x3d
167	fstp dword[ebx+vec_x]	150	fstp dword[ebx+vec_x]
168		151
169	fld dword[edi+4]	152	fld dword[edi+4]
170	fsub dword[esi+4]	153	fsub dword[esi+4]
171	fstp dword[ebx+vec_y]	154	fstp dword[ebx+vec_y]
172		155
173	fld dword[edi+8]	156	fld dword[edi+8]
174	fsub dword[esi+8]	157	fsub dword[esi+8]
175	fstp dword[ebx+vec_z]	158	fstp dword[ebx+vec_z]
176	else	159	else
177	movups xmm0,[esi]	160	movups xmm0,[esi]
178	movups xmm1,[edi]	161	movups xmm1,[edi]
179	subps xmm1,xmm0	162	subps xmm1,xmm0
180	movlps [ebx],xmm1	163	movlps [ebx],xmm1
181	movhlps xmm1,xmm1	164	movhlps xmm1,xmm1
182	movss [ebx+8],xmm1	165	movss [ebx+8],xmm1
183	end if	166	end if
184		167
185	ret	168	ret
186	;---------------------- in: -------------------------------	169	;---------------------- in: -------------------------------
187	;--------------------------- esi - pointer to 1st vector --	170	;--------------------------- esi - pointer to 1st vector --
188	;--------------------------- edi - pointer to 2nd vector --	171	;--------------------------- edi - pointer to 2nd vector --
189	;--------------------------- ebx - pointer to result vector	172	;--------------------------- ebx - pointer to result vector
190	;---------------------- out : none	173	;---------------------- out : none
191	cross_product:	174	cross_product:
192	fninit	175	fninit
193	fld dword [esi+vec_y]	176	fld dword [esi+vec_y]
194	fmul dword [edi+vec_z]	177	fmul dword [edi+vec_z]
195	fld dword [esi+vec_z]	178	fld dword [esi+vec_z]
196	fmul dword [edi+vec_y]	179	fmul dword [edi+vec_y]
197	fsubp ;st1 ,st	180	fsubp ;st1 ,st
198	fstp dword [ebx+vec_x]	181	fstp dword [ebx+vec_x]
199		182
200	fld dword [esi+vec_z]	183	fld dword [esi+vec_z]
201	fmul dword [edi+vec_x]	184	fmul dword [edi+vec_x]
202	fld dword [esi+vec_x]	185	fld dword [esi+vec_x]
203	fmul dword [edi+vec_z]	186	fmul dword [edi+vec_z]
204	fsubp ;st1 ,st	187	fsubp ;st1 ,st
205	fstp dword [ebx+vec_y]	188	fstp dword [ebx+vec_y]
206		189
207	fld dword [esi+vec_x]	190	fld dword [esi+vec_x]
208	fmul dword [edi+vec_y]	191	fmul dword [edi+vec_y]
209	fld dword [esi+vec_y]	192	fld dword [esi+vec_y]
210	fmul dword [edi+vec_x]	193	fmul dword [edi+vec_x]
211	fsubp ;st1 ,st	194	fsubp ;st1 ,st
212	fstp dword [ebx+vec_z]	195	fstp dword [ebx+vec_z]
213	ret	196	ret
214	;----------------------- in: ------------------------------	197	;----------------------- in: ------------------------------
215	;---------------------------- edi - pointer to vector -----	198	;---------------------------- edi - pointer to vector -----
216	;----------------------- out : none	199	;----------------------- out : none
217	normalize_vector:	200	normalize_vector:
-		201	if Ext >= SSE3
-		202	movups xmm0,[edi]
-		203	andps xmm0,[zero_hgst_dd]
-		204	movups xmm1,xmm0
-		205	mulps xmm0,xmm0
-		206	haddps xmm0,xmm0
-		207	haddps xmm0,xmm0
-		208	rsqrtps xmm0,xmm0
-		209	mulps xmm0,xmm1
-		210	movlps [edi],xmm0
-		211	movhlps xmm0,xmm0
-		212	movss [edi+8],xmm0
-		213	else
-		214
218	fninit	215	fninit
219	fld dword [edi+vec_x]	216	fld dword [edi+vec_x]
220	fmul st, st	217	fmul st, st
221	fld dword [edi+vec_y]	218	fld dword [edi+vec_y]
222	fmul st, st	219	fmul st, st
223	fld dword [edi+vec_z]	220	fld dword [edi+vec_z]
224	fmul st, st	221	fmul st, st
225	faddp st1, st	222	faddp st1, st
226	faddp st1, st	223	faddp st1, st
227	fsqrt	224	fsqrt
228		225
229	ftst	226	ftst
230	fstsw ax	227	fstsw ax
231	sahf	228	sahf
232	jnz @f	229	jnz @f
233		230
234	fst dword [edi+vec_x]	231	fst dword [edi+vec_x]
235	fst dword [edi+vec_y]	232	fst dword [edi+vec_y]
236	fstp dword [edi+vec_z]	233	fstp dword [edi+vec_z]
237	ret	234	ret
238	@@:	235	@@:
239	fld st	236	fld st
240	fld st	237	fld st
241	fdivr dword [edi+vec_x]	238	fdivr dword [edi+vec_x]
242	fstp dword [edi+vec_x]	239	fstp dword [edi+vec_x]
243	fdivr dword [edi+vec_y]	240	fdivr dword [edi+vec_y]
244	fstp dword [edi+vec_y]	241	fstp dword [edi+vec_y]
245	fdivr dword [edi+vec_z]	242	fdivr dword [edi+vec_z]
246	fstp dword [edi+vec_z]	243	fstp dword [edi+vec_z]
-		244	end if
247	ret	245	ret
248	;------------------in: -------------------------	246	;------------------in: -------------------------
249	;------------------ esi - pointer to 1st vector	247	;------------------ esi - pointer to 1st vector
250	;------------------ edi - pointer to 2nd vector	248	;------------------ edi - pointer to 2nd vector
251	;------------------out: ------------------------	249	;------------------out: ------------------------
252	;------------------ st0 - dot-product	250	;------------------ st0 - dot-product
253	dot_product:	251	dot_product:
254	fninit	252	fninit
-		253	;if Ext >=SSE3
-		254	; movups xmm0,[esi]
-		255	; movups xmm1,[edi]
-		256	; andps xmm0,[zero_hgst_dd]
-		257	; mulps xmm0,xmm1
-		258	; haddps xmm0,xmm0
-		259	; haddps xmm0,xmm0
-		260	; movss [esp-4],xmm0
-		261	; fld dword[esp-4]
-		262	;else
255	fld dword [esi+vec_x]	263	fld dword [esi+vec_x]
256	fmul dword [edi+vec_x]	264	fmul dword [edi+vec_x]
257	fld dword [esi+vec_y]	265	fld dword [esi+vec_y]
258	fmul dword [edi+vec_y]	266	fmul dword [edi+vec_y]
259	fld dword [esi+vec_z]	267	fld dword [esi+vec_z]
260	fmul dword [edi+vec_z]	268	fmul dword [edi+vec_z]
261	faddp	269	faddp
262	faddp	270	faddp
-		271	;end if
263	ret	272	ret
264		273
265	; DOS version Coded by Mikolaj Felix aka Majuma	274	; DOS version Coded by Mikolaj Felix aka Majuma
266	; mfelix@polbox.com	275	; mfelix@polbox.com
267	; www.majuma.xt.pl	276	; www.majuma.xt.pl
268	; into FASM translation by Macgub	277	; into FASM translation by Macgub
269	init_sincos_tab:	278	init_sincos_tab:
270	.counter equ dword [ebp-4] ; cur angle	279	.counter equ dword [ebp-4] ; cur angle
271		280
272	push ebp	281	push ebp
273	mov ebp,esp	282	mov ebp,esp
274		283
275	xor eax,eax	284	xor eax,eax
276	push eax ; init .counter	285	push eax ; init .counter
277	mov edi,cos_tab	286	mov edi,cos_tab
278	mov esi,sin_tab	287	mov esi,sin_tab
279	mov ecx,256	288	mov ecx,256
280	fninit	289	fninit
281		290
282	fld .counter	291	fld .counter
283	@@:	292	@@:
284	fld st	293	fld st
285	fsincos	294	fsincos
286	fstp dword [edi]	295	fstp dword [edi]
287	fstp dword [esi]	296	fstp dword [esi]
288	; fadd [piD180]	297	; fadd [piD180]
289	fadd [piD128]	298	fadd [piD128]
290	add esi,4	299	add esi,4
291	add edi,4	300	add edi,4
292	loop @b	301	loop @b
293	ffree st	302	ffree st
294		303
295	mov esp,ebp	304	mov esp,ebp
296	pop ebp	305	pop ebp
297	ret	306	ret
298	;------	307	;------
299	; esi - offset (pointer) to angles, edi offset to 3x3 matrix	308	; esi - offset (pointer) to angles, edi offset to 3x3 matrix
300	make_rotation_matrix:	309	make_rotation_matrix:
301	.sinx equ dword[ebp-4]	310	.sinx equ dword[ebp-4]
302	.cosx equ dword[ebp-8]	311	.cosx equ dword[ebp-8]
303	.siny equ dword[ebp-12]	312	.siny equ dword[ebp-12]
304	.cosy equ dword[ebp-16]	313	.cosy equ dword[ebp-16]
305	.sinz equ dword[ebp-20]	314	.sinz equ dword[ebp-20]
306	.cosz equ dword[ebp-24]	315	.cosz equ dword[ebp-24]
307	push ebp	316	push ebp
308	mov ebp,esp	317	mov ebp,esp
309	sub esp,24	318	sub esp,24
310		319
311	movzx ebx,word[esi]	320	movzx ebx,word[esi]
312	shl ebx,2	321	shl ebx,2
313	mov eax,dword[sin_tab+ebx]	322	mov eax,dword[sin_tab+ebx]
314	mov .sinx,eax	323	mov .sinx,eax
315	mov edx,dword[cos_tab+ebx]	324	mov edx,dword[cos_tab+ebx]
316	mov .cosx,edx	325	mov .cosx,edx
317		326
318	movzx ebx,word[esi+2]	327	movzx ebx,word[esi+2]
319	shl ebx,2	328	shl ebx,2
320	mov eax,dword[sin_tab+ebx]	329	mov eax,dword[sin_tab+ebx]
321	mov .siny,eax	330	mov .siny,eax
322	mov edx,dword[cos_tab+ebx]	331	mov edx,dword[cos_tab+ebx]
323	mov .cosy,edx	332	mov .cosy,edx
324		333
325	movzx ebx,word[esi+4]	334	movzx ebx,word[esi+4]
326	shl ebx,2	335	shl ebx,2
327	mov eax,dword[sin_tab+ebx]	336	mov eax,dword[sin_tab+ebx]
328	mov .sinz,eax	337	mov .sinz,eax
329	mov edx,dword[cos_tab+ebx]	338	mov edx,dword[cos_tab+ebx]
330	mov .cosz,edx	339	mov .cosz,edx
331		340
332	fninit	341	fninit
333	fld .cosy	342	fld .cosy
334	fmul .cosz	343	fmul .cosz
335	fstp dword[edi]	344	fstp dword[edi]
336		345
337	fld .sinx	346	fld .sinx
338	fmul .siny	347	fmul .siny
339	fmul .cosz	348	fmul .cosz
340	fld .cosx	349	fld .cosx
341	fmul .sinz	350	fmul .sinz
342	fchs	351	fchs
343	faddp	352	faddp
344	fstp dword[edi+12]	353	fstp dword[edi+12]
345		354
346	fld .cosx	355	fld .cosx
347	fmul .siny	356	fmul .siny
348	fmul .cosz	357	fmul .cosz
349	fld .sinx	358	fld .sinx
350	fmul .sinz	359	fmul .sinz
351	faddp	360	faddp
352	fstp dword[edi+24]	361	fstp dword[edi+24]
353		362
354	fld .cosy	363	fld .cosy
355	fmul .sinz	364	fmul .sinz
356	fstp dword[edi+4]	365	fstp dword[edi+4]
357		366
358	fld .sinx	367	fld .sinx
359	fmul .siny	368	fmul .siny
360	fmul .sinz	369	fmul .sinz
361	fld .cosx	370	fld .cosx
362	fmul .cosz	371	fmul .cosz
363	faddp	372	faddp
364	fstp dword[edi+16]	373	fstp dword[edi+16]
365		374
366	fld .cosx	375	fld .cosx
367	fmul .siny	376	fmul .siny
368	fmul .sinz	377	fmul .sinz
369	fld .sinx	378	fld .sinx
370	fchs	379	fchs
371	fmul .cosz	380	fmul .cosz
372	faddp	381	faddp
373	fstp dword[edi+28]	382	fstp dword[edi+28]
374		383
375	fld .siny	384	fld .siny
376	fchs	385	fchs
377	fstp dword[edi+8]	386	fstp dword[edi+8]
378		387
379	fld .cosy	388	fld .cosy
380	fmul .sinx	389	fmul .sinx
381	fstp dword[edi+20]	390	fstp dword[edi+20]
382		391
383	fld .cosx	392	fld .cosx
384	fmul .cosy	393	fmul .cosy
385	fstp dword[edi+32]	394	fstp dword[edi+32]
386		395
387	mov esp,ebp	396	mov esp,ebp
388	pop ebp	397	pop ebp
389	ret	398	ret
390	;---------------------	399	;---------------------
391	; in: esi - ptr to points(normals], each point(normal) coeficient as dword	400	; in: esi - ptr to points(normals], each point(normal) coeficient as dword
392	; edi - ptr to rotated points(normals)	401	; edi - ptr to rotated points(normals)
393	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix	402	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
394	; ecx - number of points(normals)	403	; ecx - number of points(normals)
395	rotary:	404	rotary:
396	if Ext	405	if Ext
397	fninit	406	fninit
398	.again:	407	.again:
399		408
400	fld dword[esi]	409	fld dword[esi]
401	fmul dword[ebx]	410	fmul dword[ebx]
402	fld dword[esi+4]	411	fld dword[esi+4]
403	fmul dword[ebx+12]	412	fmul dword[ebx+12]
404	faddp	413	faddp
405	fld dword[esi+8]	414	fld dword[esi+8]
406	fmul dword[ebx+24]	415	fmul dword[ebx+24]
407	faddp	416	faddp
408	fstp dword[edi]	417	fstp dword[edi]
409		418
410		419
411	fld dword[esi+4]	420	fld dword[esi+4]
412	fmul dword[ebx+16]	421	fmul dword[ebx+16]
413	fld dword[esi]	422	fld dword[esi]
414	fmul dword[ebx+4]	423	fmul dword[ebx+4]
415	faddp	424	faddp
416	fld dword[esi+8]	425	fld dword[esi+8]
417	fmul dword[ebx+28]	426	fmul dword[ebx+28]
418	faddp	427	faddp
419	fstp dword[edi+4]	428	fstp dword[edi+4]
420		429
421		430
422	fld dword[esi+8]	431	fld dword[esi+8]
423	fmul dword[ebx+32]	432	fmul dword[ebx+32]
424	fld dword[esi]	433	fld dword[esi]
425	fmul dword[ebx+8]	434	fmul dword[ebx+8]
426	fld dword[esi+4]	435	fld dword[esi+4]
427	fmul dword[ebx+20]	436	fmul dword[ebx+20]
428	faddp	437	faddp
429	faddp	438	faddp
430	fstp dword[edi+8]	439	fstp dword[edi+8]
431		440
432		441
433	add esi,12	442	add esi,12
434	add edi,12	443	add edi,12
435	loop .again	444	loop .again
436	mov [edi],dword -1	445	mov [edi],dword -1
437	else	446	else
438	; Copyright (C) 1999-2001 Brian Paul	447	; Copyright (C) 1999-2001 Brian Paul
439	; Copyright (C) Maciej Guba	448	; Copyright (C) Maciej Guba
440	;---------------------	449	;---------------------
441	; in: esi - ptr to points(normals], each point(normal) coeficient as dword	450	; in: esi - ptr to points(normals], each point(normal) coeficient as dword
442	; edi - ptr to rotated points(normals)	451	; edi - ptr to rotated points(normals)
443	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix	452	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
444	; ecx - number of points(normals)	453	; ecx - number of points(normals)
445	;align 32	454	;align 32
446	movups xmm4,[ebx]	455	movups xmm4,[ebx]
447	; lddqu xmm4,[ebx] ; I tried sse3 :D	456	; lddqu xmm4,[ebx] ; I tried sse3 :D
448	movups xmm5,[ebx+12]	457	movups xmm5,[ebx+12]
449	movups xmm6,[ebx+24]	458	movups xmm6,[ebx+24]
450	;align 32	459	;align 32
451	.again:	460	.again:
452	movss xmm0,dword[esi]	461	movss xmm0,dword[esi]
453	shufps xmm0,xmm0,0	462	shufps xmm0,xmm0,0
454	mulps xmm0,xmm4	463	mulps xmm0,xmm4
455		464
456	movss xmm1,dword[esi+4]	465	movss xmm1,dword[esi+4]
457	shufps xmm1,xmm1,0	466	shufps xmm1,xmm1,0
458	mulps xmm1,xmm5	467	mulps xmm1,xmm5
459		468
460	movss xmm2,dword[esi+8]	469	movss xmm2,dword[esi+8]
461	shufps xmm2,xmm2,0	470	shufps xmm2,xmm2,0
462	mulps xmm2,xmm6	471	mulps xmm2,xmm6
463		472
464	addps xmm0,xmm1	473	addps xmm0,xmm1
465	addps xmm0,xmm2	474	addps xmm0,xmm2
466		475
467	movups [edi],xmm0	476	movups [edi],xmm0
468		477
469	add esi,12	478	add esi,12
470	add edi,12	479	add edi,12
471	dec ecx	480	dec ecx
472	jne .again	481	jne .again
473	mov [edi],dword -1	482	mov [edi],dword -1
474	end if	483	end if

Subversion Repositories Kolibri OS

(root)/programs/demos/view3ds/3dmath.inc @ 8666 – Rev 8014 → 8232