* Sync up to trunk head (r60691).
[reactos.git] / dll / directx / wine / d3dx9_36 / math.c
1 /*
2 * Mathematical operations specific to D3DX9.
3 *
4 * Copyright (C) 2008 David Adam
5 * Copyright (C) 2008 Luis Busquets
6 * Copyright (C) 2008 Jérôme Gardou
7 * Copyright (C) 2008 Philip Nilsson
8 * Copyright (C) 2008 Henri Verbeet
9 *
10 * This library is free software; you can redistribute it and/or
11 * modify it under the terms of the GNU Lesser General Public
12 * License as published by the Free Software Foundation; either
13 * version 2.1 of the License, or (at your option) any later version.
14 *
15 * This library is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 * Lesser General Public License for more details.
19 *
20 * You should have received a copy of the GNU Lesser General Public
21 * License along with this library; if not, write to the Free Software
22 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
23 */
24
25 #define NONAMELESSUNION
26
27 #include <config.h>
28 //#include "windef.h"
29 //#include "wingdi.h"
30 #include "d3dx9_36_private.h"
31
32 #include <wine/debug.h>
33
34 WINE_DEFAULT_DEBUG_CHANNEL(d3dx);
35
36 #ifdef _MSC_VER
37 #define copysignf(x, y) ((x) < 0.0f ? -fabsf(y) : fabsf(y))
38 #endif
39
40 struct ID3DXMatrixStackImpl
41 {
42 ID3DXMatrixStack ID3DXMatrixStack_iface;
43 LONG ref;
44
45 unsigned int current;
46 unsigned int stack_size;
47 D3DXMATRIX *stack;
48 };
49
50 static const unsigned int INITIAL_STACK_SIZE = 32;
51
52 /*_________________D3DXColor____________________*/
53
54 D3DXCOLOR* WINAPI D3DXColorAdjustContrast(D3DXCOLOR *pout, const D3DXCOLOR *pc, FLOAT s)
55 {
56 TRACE("pout %p, pc %p, s %f\n", pout, pc, s);
57
58 pout->r = 0.5f + s * (pc->r - 0.5f);
59 pout->g = 0.5f + s * (pc->g - 0.5f);
60 pout->b = 0.5f + s * (pc->b - 0.5f);
61 pout->a = pc->a;
62 return pout;
63 }
64
65 D3DXCOLOR* WINAPI D3DXColorAdjustSaturation(D3DXCOLOR *pout, const D3DXCOLOR *pc, FLOAT s)
66 {
67 FLOAT grey;
68
69 TRACE("pout %p, pc %p, s %f\n", pout, pc, s);
70
71 grey = pc->r * 0.2125f + pc->g * 0.7154f + pc->b * 0.0721f;
72 pout->r = grey + s * (pc->r - grey);
73 pout->g = grey + s * (pc->g - grey);
74 pout->b = grey + s * (pc->b - grey);
75 pout->a = pc->a;
76 return pout;
77 }
78
79 /*_________________Misc__________________________*/
80
81 FLOAT WINAPI D3DXFresnelTerm(FLOAT costheta, FLOAT refractionindex)
82 {
83 FLOAT a, d, g, result;
84
85 TRACE("costheta %f, refractionindex %f\n", costheta, refractionindex);
86
87 g = sqrtf(refractionindex * refractionindex + costheta * costheta - 1.0f);
88 a = g + costheta;
89 d = g - costheta;
90 result = (costheta * a - 1.0f) * (costheta * a - 1.0f) / ((costheta * d + 1.0f) * (costheta * d + 1.0f)) + 1.0f;
91 result *= 0.5f * d * d / (a * a);
92
93 return result;
94 }
95
96 /*_________________D3DXMatrix____________________*/
97
98 D3DXMATRIX * WINAPI D3DXMatrixAffineTransformation(D3DXMATRIX *out, FLOAT scaling, const D3DXVECTOR3 *rotationcenter,
99 const D3DXQUATERNION *rotation, const D3DXVECTOR3 *translation)
100 {
101 TRACE("out %p, scaling %f, rotationcenter %p, rotation %p, translation %p\n",
102 out, scaling, rotationcenter, rotation, translation);
103
104 D3DXMatrixIdentity(out);
105
106 if (rotation)
107 {
108 FLOAT temp00, temp01, temp02, temp10, temp11, temp12, temp20, temp21, temp22;
109
110 temp00 = 1.0f - 2.0f * (rotation->y * rotation->y + rotation->z * rotation->z);
111 temp01 = 2.0f * (rotation->x * rotation->y + rotation->z * rotation->w);
112 temp02 = 2.0f * (rotation->x * rotation->z - rotation->y * rotation->w);
113 temp10 = 2.0f * (rotation->x * rotation->y - rotation->z * rotation->w);
114 temp11 = 1.0f - 2.0f * (rotation->x * rotation->x + rotation->z * rotation->z);
115 temp12 = 2.0f * (rotation->y * rotation->z + rotation->x * rotation->w);
116 temp20 = 2.0f * (rotation->x * rotation->z + rotation->y * rotation->w);
117 temp21 = 2.0f * (rotation->y * rotation->z - rotation->x * rotation->w);
118 temp22 = 1.0f - 2.0f * (rotation->x * rotation->x + rotation->y * rotation->y);
119
120 out->u.m[0][0] = scaling * temp00;
121 out->u.m[0][1] = scaling * temp01;
122 out->u.m[0][2] = scaling * temp02;
123 out->u.m[1][0] = scaling * temp10;
124 out->u.m[1][1] = scaling * temp11;
125 out->u.m[1][2] = scaling * temp12;
126 out->u.m[2][0] = scaling * temp20;
127 out->u.m[2][1] = scaling * temp21;
128 out->u.m[2][2] = scaling * temp22;
129
130 if (rotationcenter)
131 {
132 out->u.m[3][0] = rotationcenter->x * (1.0f - temp00) - rotationcenter->y * temp10
133 - rotationcenter->z * temp20;
134 out->u.m[3][1] = rotationcenter->y * (1.0f - temp11) - rotationcenter->x * temp01
135 - rotationcenter->z * temp21;
136 out->u.m[3][2] = rotationcenter->z * (1.0f - temp22) - rotationcenter->x * temp02
137 - rotationcenter->y * temp12;
138 }
139 }
140 else
141 {
142 out->u.m[0][0] = scaling;
143 out->u.m[1][1] = scaling;
144 out->u.m[2][2] = scaling;
145 }
146
147 if (translation)
148 {
149 out->u.m[3][0] += translation->x;
150 out->u.m[3][1] += translation->y;
151 out->u.m[3][2] += translation->z;
152 }
153
154 return out;
155 }
156
157 D3DXMATRIX * WINAPI D3DXMatrixAffineTransformation2D(D3DXMATRIX *out, FLOAT scaling,
158 const D3DXVECTOR2 *rotationcenter, FLOAT rotation, const D3DXVECTOR2 *translation)
159 {
160 FLOAT tmp1, tmp2, s;
161
162 TRACE("out %p, scaling %f, rotationcenter %p, rotation %f, translation %p\n",
163 out, scaling, rotationcenter, rotation, translation);
164
165 s = sinf(rotation / 2.0f);
166 tmp1 = 1.0f - 2.0f * s * s;
167 tmp2 = 2.0 * s * cosf(rotation / 2.0f);
168
169 D3DXMatrixIdentity(out);
170 out->u.m[0][0] = scaling * tmp1;
171 out->u.m[0][1] = scaling * tmp2;
172 out->u.m[1][0] = -scaling * tmp2;
173 out->u.m[1][1] = scaling * tmp1;
174
175 if (rotationcenter)
176 {
177 FLOAT x, y;
178
179 x = rotationcenter->x;
180 y = rotationcenter->y;
181
182 out->u.m[3][0] = y * tmp2 - x * tmp1 + x;
183 out->u.m[3][1] = -x * tmp2 - y * tmp1 + y;
184 }
185
186 if (translation)
187 {
188 out->u.m[3][0] += translation->x;
189 out->u.m[3][1] += translation->y;
190 }
191
192 return out;
193 }
194
195 HRESULT WINAPI D3DXMatrixDecompose(D3DXVECTOR3 *poutscale, D3DXQUATERNION *poutrotation, D3DXVECTOR3 *pouttranslation, const D3DXMATRIX *pm)
196 {
197 D3DXMATRIX normalized;
198 D3DXVECTOR3 vec;
199
200 TRACE("poutscale %p, poutrotation %p, pouttranslation %p, pm %p\n", poutscale, poutrotation, pouttranslation, pm);
201
202 /*Compute the scaling part.*/
203 vec.x=pm->u.m[0][0];
204 vec.y=pm->u.m[0][1];
205 vec.z=pm->u.m[0][2];
206 poutscale->x=D3DXVec3Length(&vec);
207
208 vec.x=pm->u.m[1][0];
209 vec.y=pm->u.m[1][1];
210 vec.z=pm->u.m[1][2];
211 poutscale->y=D3DXVec3Length(&vec);
212
213 vec.x=pm->u.m[2][0];
214 vec.y=pm->u.m[2][1];
215 vec.z=pm->u.m[2][2];
216 poutscale->z=D3DXVec3Length(&vec);
217
218 /*Compute the translation part.*/
219 pouttranslation->x=pm->u.m[3][0];
220 pouttranslation->y=pm->u.m[3][1];
221 pouttranslation->z=pm->u.m[3][2];
222
223 /*Let's calculate the rotation now*/
224 if ( (poutscale->x == 0.0f) || (poutscale->y == 0.0f) || (poutscale->z == 0.0f) ) return D3DERR_INVALIDCALL;
225
226 normalized.u.m[0][0]=pm->u.m[0][0]/poutscale->x;
227 normalized.u.m[0][1]=pm->u.m[0][1]/poutscale->x;
228 normalized.u.m[0][2]=pm->u.m[0][2]/poutscale->x;
229 normalized.u.m[1][0]=pm->u.m[1][0]/poutscale->y;
230 normalized.u.m[1][1]=pm->u.m[1][1]/poutscale->y;
231 normalized.u.m[1][2]=pm->u.m[1][2]/poutscale->y;
232 normalized.u.m[2][0]=pm->u.m[2][0]/poutscale->z;
233 normalized.u.m[2][1]=pm->u.m[2][1]/poutscale->z;
234 normalized.u.m[2][2]=pm->u.m[2][2]/poutscale->z;
235
236 D3DXQuaternionRotationMatrix(poutrotation,&normalized);
237 return S_OK;
238 }
239
240 FLOAT WINAPI D3DXMatrixDeterminant(const D3DXMATRIX *pm)
241 {
242 FLOAT t[3], v[4];
243
244 TRACE("pm %p\n", pm);
245
246 t[0] = pm->u.m[2][2] * pm->u.m[3][3] - pm->u.m[2][3] * pm->u.m[3][2];
247 t[1] = pm->u.m[1][2] * pm->u.m[3][3] - pm->u.m[1][3] * pm->u.m[3][2];
248 t[2] = pm->u.m[1][2] * pm->u.m[2][3] - pm->u.m[1][3] * pm->u.m[2][2];
249 v[0] = pm->u.m[1][1] * t[0] - pm->u.m[2][1] * t[1] + pm->u.m[3][1] * t[2];
250 v[1] = -pm->u.m[1][0] * t[0] + pm->u.m[2][0] * t[1] - pm->u.m[3][0] * t[2];
251
252 t[0] = pm->u.m[1][0] * pm->u.m[2][1] - pm->u.m[2][0] * pm->u.m[1][1];
253 t[1] = pm->u.m[1][0] * pm->u.m[3][1] - pm->u.m[3][0] * pm->u.m[1][1];
254 t[2] = pm->u.m[2][0] * pm->u.m[3][1] - pm->u.m[3][0] * pm->u.m[2][1];
255 v[2] = pm->u.m[3][3] * t[0] - pm->u.m[2][3] * t[1] + pm->u.m[1][3] * t[2];
256 v[3] = -pm->u.m[3][2] * t[0] + pm->u.m[2][2] * t[1] - pm->u.m[1][2] * t[2];
257
258 return pm->u.m[0][0] * v[0] + pm->u.m[0][1] * v[1] +
259 pm->u.m[0][2] * v[2] + pm->u.m[0][3] * v[3];
260 }
261
262 D3DXMATRIX* WINAPI D3DXMatrixInverse(D3DXMATRIX *pout, FLOAT *pdeterminant, const D3DXMATRIX *pm)
263 {
264 FLOAT det, t[3], v[16];
265 UINT i, j;
266
267 TRACE("pout %p, pdeterminant %p, pm %p\n", pout, pdeterminant, pm);
268
269 t[0] = pm->u.m[2][2] * pm->u.m[3][3] - pm->u.m[2][3] * pm->u.m[3][2];
270 t[1] = pm->u.m[1][2] * pm->u.m[3][3] - pm->u.m[1][3] * pm->u.m[3][2];
271 t[2] = pm->u.m[1][2] * pm->u.m[2][3] - pm->u.m[1][3] * pm->u.m[2][2];
272 v[0] = pm->u.m[1][1] * t[0] - pm->u.m[2][1] * t[1] + pm->u.m[3][1] * t[2];
273 v[4] = -pm->u.m[1][0] * t[0] + pm->u.m[2][0] * t[1] - pm->u.m[3][0] * t[2];
274
275 t[0] = pm->u.m[1][0] * pm->u.m[2][1] - pm->u.m[2][0] * pm->u.m[1][1];
276 t[1] = pm->u.m[1][0] * pm->u.m[3][1] - pm->u.m[3][0] * pm->u.m[1][1];
277 t[2] = pm->u.m[2][0] * pm->u.m[3][1] - pm->u.m[3][0] * pm->u.m[2][1];
278 v[8] = pm->u.m[3][3] * t[0] - pm->u.m[2][3] * t[1] + pm->u.m[1][3] * t[2];
279 v[12] = -pm->u.m[3][2] * t[0] + pm->u.m[2][2] * t[1] - pm->u.m[1][2] * t[2];
280
281 det = pm->u.m[0][0] * v[0] + pm->u.m[0][1] * v[4] +
282 pm->u.m[0][2] * v[8] + pm->u.m[0][3] * v[12];
283 if (det == 0.0f)
284 return NULL;
285 if (pdeterminant)
286 *pdeterminant = det;
287
288 t[0] = pm->u.m[2][2] * pm->u.m[3][3] - pm->u.m[2][3] * pm->u.m[3][2];
289 t[1] = pm->u.m[0][2] * pm->u.m[3][3] - pm->u.m[0][3] * pm->u.m[3][2];
290 t[2] = pm->u.m[0][2] * pm->u.m[2][3] - pm->u.m[0][3] * pm->u.m[2][2];
291 v[1] = -pm->u.m[0][1] * t[0] + pm->u.m[2][1] * t[1] - pm->u.m[3][1] * t[2];
292 v[5] = pm->u.m[0][0] * t[0] - pm->u.m[2][0] * t[1] + pm->u.m[3][0] * t[2];
293
294 t[0] = pm->u.m[0][0] * pm->u.m[2][1] - pm->u.m[2][0] * pm->u.m[0][1];
295 t[1] = pm->u.m[3][0] * pm->u.m[0][1] - pm->u.m[0][0] * pm->u.m[3][1];
296 t[2] = pm->u.m[2][0] * pm->u.m[3][1] - pm->u.m[3][0] * pm->u.m[2][1];
297 v[9] = -pm->u.m[3][3] * t[0] - pm->u.m[2][3] * t[1]- pm->u.m[0][3] * t[2];
298 v[13] = pm->u.m[3][2] * t[0] + pm->u.m[2][2] * t[1] + pm->u.m[0][2] * t[2];
299
300 t[0] = pm->u.m[1][2] * pm->u.m[3][3] - pm->u.m[1][3] * pm->u.m[3][2];
301 t[1] = pm->u.m[0][2] * pm->u.m[3][3] - pm->u.m[0][3] * pm->u.m[3][2];
302 t[2] = pm->u.m[0][2] * pm->u.m[1][3] - pm->u.m[0][3] * pm->u.m[1][2];
303 v[2] = pm->u.m[0][1] * t[0] - pm->u.m[1][1] * t[1] + pm->u.m[3][1] * t[2];
304 v[6] = -pm->u.m[0][0] * t[0] + pm->u.m[1][0] * t[1] - pm->u.m[3][0] * t[2];
305
306 t[0] = pm->u.m[0][0] * pm->u.m[1][1] - pm->u.m[1][0] * pm->u.m[0][1];
307 t[1] = pm->u.m[3][0] * pm->u.m[0][1] - pm->u.m[0][0] * pm->u.m[3][1];
308 t[2] = pm->u.m[1][0] * pm->u.m[3][1] - pm->u.m[3][0] * pm->u.m[1][1];
309 v[10] = pm->u.m[3][3] * t[0] + pm->u.m[1][3] * t[1] + pm->u.m[0][3] * t[2];
310 v[14] = -pm->u.m[3][2] * t[0] - pm->u.m[1][2] * t[1] - pm->u.m[0][2] * t[2];
311
312 t[0] = pm->u.m[1][2] * pm->u.m[2][3] - pm->u.m[1][3] * pm->u.m[2][2];
313 t[1] = pm->u.m[0][2] * pm->u.m[2][3] - pm->u.m[0][3] * pm->u.m[2][2];
314 t[2] = pm->u.m[0][2] * pm->u.m[1][3] - pm->u.m[0][3] * pm->u.m[1][2];
315 v[3] = -pm->u.m[0][1] * t[0] + pm->u.m[1][1] * t[1] - pm->u.m[2][1] * t[2];
316 v[7] = pm->u.m[0][0] * t[0] - pm->u.m[1][0] * t[1] + pm->u.m[2][0] * t[2];
317
318 v[11] = -pm->u.m[0][0] * (pm->u.m[1][1] * pm->u.m[2][3] - pm->u.m[1][3] * pm->u.m[2][1]) +
319 pm->u.m[1][0] * (pm->u.m[0][1] * pm->u.m[2][3] - pm->u.m[0][3] * pm->u.m[2][1]) -
320 pm->u.m[2][0] * (pm->u.m[0][1] * pm->u.m[1][3] - pm->u.m[0][3] * pm->u.m[1][1]);
321
322 v[15] = pm->u.m[0][0] * (pm->u.m[1][1] * pm->u.m[2][2] - pm->u.m[1][2] * pm->u.m[2][1]) -
323 pm->u.m[1][0] * (pm->u.m[0][1] * pm->u.m[2][2] - pm->u.m[0][2] * pm->u.m[2][1]) +
324 pm->u.m[2][0] * (pm->u.m[0][1] * pm->u.m[1][2] - pm->u.m[0][2] * pm->u.m[1][1]);
325
326 det = 1.0f / det;
327
328 for (i = 0; i < 4; i++)
329 for (j = 0; j < 4; j++)
330 pout->u.m[i][j] = v[4 * i + j] * det;
331
332 return pout;
333 }
334
335 D3DXMATRIX * WINAPI D3DXMatrixLookAtLH(D3DXMATRIX *out, const D3DXVECTOR3 *eye, const D3DXVECTOR3 *at,
336 const D3DXVECTOR3 *up)
337 {
338 D3DXVECTOR3 right, upn, vec;
339
340 TRACE("out %p, eye %p, at %p, up %p\n", out, eye, at, up);
341
342 D3DXVec3Subtract(&vec, at, eye);
343 D3DXVec3Normalize(&vec, &vec);
344 D3DXVec3Cross(&right, up, &vec);
345 D3DXVec3Cross(&upn, &vec, &right);
346 D3DXVec3Normalize(&right, &right);
347 D3DXVec3Normalize(&upn, &upn);
348 out->u.m[0][0] = right.x;
349 out->u.m[1][0] = right.y;
350 out->u.m[2][0] = right.z;
351 out->u.m[3][0] = -D3DXVec3Dot(&right, eye);
352 out->u.m[0][1] = upn.x;
353 out->u.m[1][1] = upn.y;
354 out->u.m[2][1] = upn.z;
355 out->u.m[3][1] = -D3DXVec3Dot(&upn, eye);
356 out->u.m[0][2] = vec.x;
357 out->u.m[1][2] = vec.y;
358 out->u.m[2][2] = vec.z;
359 out->u.m[3][2] = -D3DXVec3Dot(&vec, eye);
360 out->u.m[0][3] = 0.0f;
361 out->u.m[1][3] = 0.0f;
362 out->u.m[2][3] = 0.0f;
363 out->u.m[3][3] = 1.0f;
364
365 return out;
366 }
367
368 D3DXMATRIX * WINAPI D3DXMatrixLookAtRH(D3DXMATRIX *out, const D3DXVECTOR3 *eye, const D3DXVECTOR3 *at,
369 const D3DXVECTOR3 *up)
370 {
371 D3DXVECTOR3 right, upn, vec;
372
373 TRACE("out %p, eye %p, at %p, up %p\n", out, eye, at, up);
374
375 D3DXVec3Subtract(&vec, at, eye);
376 D3DXVec3Normalize(&vec, &vec);
377 D3DXVec3Cross(&right, up, &vec);
378 D3DXVec3Cross(&upn, &vec, &right);
379 D3DXVec3Normalize(&right, &right);
380 D3DXVec3Normalize(&upn, &upn);
381 out->u.m[0][0] = -right.x;
382 out->u.m[1][0] = -right.y;
383 out->u.m[2][0] = -right.z;
384 out->u.m[3][0] = D3DXVec3Dot(&right, eye);
385 out->u.m[0][1] = upn.x;
386 out->u.m[1][1] = upn.y;
387 out->u.m[2][1] = upn.z;
388 out->u.m[3][1] = -D3DXVec3Dot(&upn, eye);
389 out->u.m[0][2] = -vec.x;
390 out->u.m[1][2] = -vec.y;
391 out->u.m[2][2] = -vec.z;
392 out->u.m[3][2] = D3DXVec3Dot(&vec, eye);
393 out->u.m[0][3] = 0.0f;
394 out->u.m[1][3] = 0.0f;
395 out->u.m[2][3] = 0.0f;
396 out->u.m[3][3] = 1.0f;
397
398 return out;
399 }
400
401 D3DXMATRIX* WINAPI D3DXMatrixMultiply(D3DXMATRIX *pout, const D3DXMATRIX *pm1, const D3DXMATRIX *pm2)
402 {
403 D3DXMATRIX out;
404 int i,j;
405
406 TRACE("pout %p, pm1 %p, pm2 %p\n", pout, pm1, pm2);
407
408 for (i=0; i<4; i++)
409 {
410 for (j=0; j<4; j++)
411 {
412 out.u.m[i][j] = pm1->u.m[i][0] * pm2->u.m[0][j] + pm1->u.m[i][1] * pm2->u.m[1][j] + pm1->u.m[i][2] * pm2->u.m[2][j] + pm1->u.m[i][3] * pm2->u.m[3][j];
413 }
414 }
415
416 *pout = out;
417 return pout;
418 }
419
420 D3DXMATRIX* WINAPI D3DXMatrixMultiplyTranspose(D3DXMATRIX *pout, const D3DXMATRIX *pm1, const D3DXMATRIX *pm2)
421 {
422 D3DXMATRIX temp;
423 int i, j;
424
425 TRACE("pout %p, pm1 %p, pm2 %p\n", pout, pm1, pm2);
426
427 for (i = 0; i < 4; i++)
428 for (j = 0; j < 4; j++)
429 temp.u.m[j][i] = pm1->u.m[i][0] * pm2->u.m[0][j] + pm1->u.m[i][1] * pm2->u.m[1][j] + pm1->u.m[i][2] * pm2->u.m[2][j] + pm1->u.m[i][3] * pm2->u.m[3][j];
430
431 *pout = temp;
432 return pout;
433 }
434
435 D3DXMATRIX* WINAPI D3DXMatrixOrthoLH(D3DXMATRIX *pout, FLOAT w, FLOAT h, FLOAT zn, FLOAT zf)
436 {
437 TRACE("pout %p, w %f, h %f, zn %f, zf %f\n", pout, w, h, zn, zf);
438
439 D3DXMatrixIdentity(pout);
440 pout->u.m[0][0] = 2.0f / w;
441 pout->u.m[1][1] = 2.0f / h;
442 pout->u.m[2][2] = 1.0f / (zf - zn);
443 pout->u.m[3][2] = zn / (zn - zf);
444 return pout;
445 }
446
447 D3DXMATRIX* WINAPI D3DXMatrixOrthoOffCenterLH(D3DXMATRIX *pout, FLOAT l, FLOAT r, FLOAT b, FLOAT t, FLOAT zn, FLOAT zf)
448 {
449 TRACE("pout %p, l %f, r %f, b %f, t %f, zn %f, zf %f\n", pout, l, r, b, t, zn, zf);
450
451 D3DXMatrixIdentity(pout);
452 pout->u.m[0][0] = 2.0f / (r - l);
453 pout->u.m[1][1] = 2.0f / (t - b);
454 pout->u.m[2][2] = 1.0f / (zf -zn);
455 pout->u.m[3][0] = -1.0f -2.0f *l / (r - l);
456 pout->u.m[3][1] = 1.0f + 2.0f * t / (b - t);
457 pout->u.m[3][2] = zn / (zn -zf);
458 return pout;
459 }
460
461 D3DXMATRIX* WINAPI D3DXMatrixOrthoOffCenterRH(D3DXMATRIX *pout, FLOAT l, FLOAT r, FLOAT b, FLOAT t, FLOAT zn, FLOAT zf)
462 {
463 TRACE("pout %p, l %f, r %f, b %f, t %f, zn %f, zf %f\n", pout, l, r, b, t, zn, zf);
464
465 D3DXMatrixIdentity(pout);
466 pout->u.m[0][0] = 2.0f / (r - l);
467 pout->u.m[1][1] = 2.0f / (t - b);
468 pout->u.m[2][2] = 1.0f / (zn -zf);
469 pout->u.m[3][0] = -1.0f -2.0f *l / (r - l);
470 pout->u.m[3][1] = 1.0f + 2.0f * t / (b - t);
471 pout->u.m[3][2] = zn / (zn -zf);
472 return pout;
473 }
474
475 D3DXMATRIX* WINAPI D3DXMatrixOrthoRH(D3DXMATRIX *pout, FLOAT w, FLOAT h, FLOAT zn, FLOAT zf)
476 {
477 TRACE("pout %p, w %f, h %f, zn %f, zf %f\n", pout, w, h, zn, zf);
478
479 D3DXMatrixIdentity(pout);
480 pout->u.m[0][0] = 2.0f / w;
481 pout->u.m[1][1] = 2.0f / h;
482 pout->u.m[2][2] = 1.0f / (zn - zf);
483 pout->u.m[3][2] = zn / (zn - zf);
484 return pout;
485 }
486
487 D3DXMATRIX* WINAPI D3DXMatrixPerspectiveFovLH(D3DXMATRIX *pout, FLOAT fovy, FLOAT aspect, FLOAT zn, FLOAT zf)
488 {
489 TRACE("pout %p, fovy %f, aspect %f, zn %f, zf %f\n", pout, fovy, aspect, zn, zf);
490
491 D3DXMatrixIdentity(pout);
492 pout->u.m[0][0] = 1.0f / (aspect * tan(fovy/2.0f));
493 pout->u.m[1][1] = 1.0f / tan(fovy/2.0f);
494 pout->u.m[2][2] = zf / (zf - zn);
495 pout->u.m[2][3] = 1.0f;
496 pout->u.m[3][2] = (zf * zn) / (zn - zf);
497 pout->u.m[3][3] = 0.0f;
498 return pout;
499 }
500
501 D3DXMATRIX* WINAPI D3DXMatrixPerspectiveFovRH(D3DXMATRIX *pout, FLOAT fovy, FLOAT aspect, FLOAT zn, FLOAT zf)
502 {
503 TRACE("pout %p, fovy %f, aspect %f, zn %f, zf %f\n", pout, fovy, aspect, zn, zf);
504
505 D3DXMatrixIdentity(pout);
506 pout->u.m[0][0] = 1.0f / (aspect * tan(fovy/2.0f));
507 pout->u.m[1][1] = 1.0f / tan(fovy/2.0f);
508 pout->u.m[2][2] = zf / (zn - zf);
509 pout->u.m[2][3] = -1.0f;
510 pout->u.m[3][2] = (zf * zn) / (zn - zf);
511 pout->u.m[3][3] = 0.0f;
512 return pout;
513 }
514
515 D3DXMATRIX* WINAPI D3DXMatrixPerspectiveLH(D3DXMATRIX *pout, FLOAT w, FLOAT h, FLOAT zn, FLOAT zf)
516 {
517 TRACE("pout %p, w %f, h %f, zn %f, zf %f\n", pout, w, h, zn, zf);
518
519 D3DXMatrixIdentity(pout);
520 pout->u.m[0][0] = 2.0f * zn / w;
521 pout->u.m[1][1] = 2.0f * zn / h;
522 pout->u.m[2][2] = zf / (zf - zn);
523 pout->u.m[3][2] = (zn * zf) / (zn - zf);
524 pout->u.m[2][3] = 1.0f;
525 pout->u.m[3][3] = 0.0f;
526 return pout;
527 }
528
529 D3DXMATRIX* WINAPI D3DXMatrixPerspectiveOffCenterLH(D3DXMATRIX *pout, FLOAT l, FLOAT r, FLOAT b, FLOAT t, FLOAT zn, FLOAT zf)
530 {
531 TRACE("pout %p, l %f, r %f, b %f, t %f, zn %f, zf %f\n", pout, l, r, b, t, zn, zf);
532
533 D3DXMatrixIdentity(pout);
534 pout->u.m[0][0] = 2.0f * zn / (r - l);
535 pout->u.m[1][1] = -2.0f * zn / (b - t);
536 pout->u.m[2][0] = -1.0f - 2.0f * l / (r - l);
537 pout->u.m[2][1] = 1.0f + 2.0f * t / (b - t);
538 pout->u.m[2][2] = - zf / (zn - zf);
539 pout->u.m[3][2] = (zn * zf) / (zn -zf);
540 pout->u.m[2][3] = 1.0f;
541 pout->u.m[3][3] = 0.0f;
542 return pout;
543 }
544
545 D3DXMATRIX* WINAPI D3DXMatrixPerspectiveOffCenterRH(D3DXMATRIX *pout, FLOAT l, FLOAT r, FLOAT b, FLOAT t, FLOAT zn, FLOAT zf)
546 {
547 TRACE("pout %p, l %f, r %f, b %f, t %f, zn %f, zf %f\n", pout, l, r, b, t, zn, zf);
548
549 D3DXMatrixIdentity(pout);
550 pout->u.m[0][0] = 2.0f * zn / (r - l);
551 pout->u.m[1][1] = -2.0f * zn / (b - t);
552 pout->u.m[2][0] = 1.0f + 2.0f * l / (r - l);
553 pout->u.m[2][1] = -1.0f -2.0f * t / (b - t);
554 pout->u.m[2][2] = zf / (zn - zf);
555 pout->u.m[3][2] = (zn * zf) / (zn -zf);
556 pout->u.m[2][3] = -1.0f;
557 pout->u.m[3][3] = 0.0f;
558 return pout;
559 }
560
561 D3DXMATRIX* WINAPI D3DXMatrixPerspectiveRH(D3DXMATRIX *pout, FLOAT w, FLOAT h, FLOAT zn, FLOAT zf)
562 {
563 TRACE("pout %p, w %f, h %f, zn %f, zf %f\n", pout, w, h, zn, zf);
564
565 D3DXMatrixIdentity(pout);
566 pout->u.m[0][0] = 2.0f * zn / w;
567 pout->u.m[1][1] = 2.0f * zn / h;
568 pout->u.m[2][2] = zf / (zn - zf);
569 pout->u.m[3][2] = (zn * zf) / (zn - zf);
570 pout->u.m[2][3] = -1.0f;
571 pout->u.m[3][3] = 0.0f;
572 return pout;
573 }
574
575 D3DXMATRIX* WINAPI D3DXMatrixReflect(D3DXMATRIX *pout, const D3DXPLANE *pplane)
576 {
577 D3DXPLANE Nplane;
578
579 TRACE("pout %p, pplane %p\n", pout, pplane);
580
581 D3DXPlaneNormalize(&Nplane, pplane);
582 D3DXMatrixIdentity(pout);
583 pout->u.m[0][0] = 1.0f - 2.0f * Nplane.a * Nplane.a;
584 pout->u.m[0][1] = -2.0f * Nplane.a * Nplane.b;
585 pout->u.m[0][2] = -2.0f * Nplane.a * Nplane.c;
586 pout->u.m[1][0] = -2.0f * Nplane.a * Nplane.b;
587 pout->u.m[1][1] = 1.0f - 2.0f * Nplane.b * Nplane.b;
588 pout->u.m[1][2] = -2.0f * Nplane.b * Nplane.c;
589 pout->u.m[2][0] = -2.0f * Nplane.c * Nplane.a;
590 pout->u.m[2][1] = -2.0f * Nplane.c * Nplane.b;
591 pout->u.m[2][2] = 1.0f - 2.0f * Nplane.c * Nplane.c;
592 pout->u.m[3][0] = -2.0f * Nplane.d * Nplane.a;
593 pout->u.m[3][1] = -2.0f * Nplane.d * Nplane.b;
594 pout->u.m[3][2] = -2.0f * Nplane.d * Nplane.c;
595 return pout;
596 }
597
598 D3DXMATRIX * WINAPI D3DXMatrixRotationAxis(D3DXMATRIX *out, const D3DXVECTOR3 *v, FLOAT angle)
599 {
600 D3DXVECTOR3 nv;
601 FLOAT sangle, cangle, cdiff;
602
603 TRACE("out %p, v %p, angle %f\n", out, v, angle);
604
605 D3DXVec3Normalize(&nv, v);
606 sangle = sinf(angle);
607 cangle = cosf(angle);
608 cdiff = 1.0f - cangle;
609
610 out->u.m[0][0] = cdiff * nv.x * nv.x + cangle;
611 out->u.m[1][0] = cdiff * nv.x * nv.y - sangle * nv.z;
612 out->u.m[2][0] = cdiff * nv.x * nv.z + sangle * nv.y;
613 out->u.m[3][0] = 0.0f;
614 out->u.m[0][1] = cdiff * nv.y * nv.x + sangle * nv.z;
615 out->u.m[1][1] = cdiff * nv.y * nv.y + cangle;
616 out->u.m[2][1] = cdiff * nv.y * nv.z - sangle * nv.x;
617 out->u.m[3][1] = 0.0f;
618 out->u.m[0][2] = cdiff * nv.z * nv.x - sangle * nv.y;
619 out->u.m[1][2] = cdiff * nv.z * nv.y + sangle * nv.x;
620 out->u.m[2][2] = cdiff * nv.z * nv.z + cangle;
621 out->u.m[3][2] = 0.0f;
622 out->u.m[0][3] = 0.0f;
623 out->u.m[1][3] = 0.0f;
624 out->u.m[2][3] = 0.0f;
625 out->u.m[3][3] = 1.0f;
626
627 return out;
628 }
629
630 D3DXMATRIX* WINAPI D3DXMatrixRotationQuaternion(D3DXMATRIX *pout, const D3DXQUATERNION *pq)
631 {
632 TRACE("pout %p, pq %p\n", pout, pq);
633
634 D3DXMatrixIdentity(pout);
635 pout->u.m[0][0] = 1.0f - 2.0f * (pq->y * pq->y + pq->z * pq->z);
636 pout->u.m[0][1] = 2.0f * (pq->x *pq->y + pq->z * pq->w);
637 pout->u.m[0][2] = 2.0f * (pq->x * pq->z - pq->y * pq->w);
638 pout->u.m[1][0] = 2.0f * (pq->x * pq->y - pq->z * pq->w);
639 pout->u.m[1][1] = 1.0f - 2.0f * (pq->x * pq->x + pq->z * pq->z);
640 pout->u.m[1][2] = 2.0f * (pq->y *pq->z + pq->x *pq->w);
641 pout->u.m[2][0] = 2.0f * (pq->x * pq->z + pq->y * pq->w);
642 pout->u.m[2][1] = 2.0f * (pq->y *pq->z - pq->x *pq->w);
643 pout->u.m[2][2] = 1.0f - 2.0f * (pq->x * pq->x + pq->y * pq->y);
644 return pout;
645 }
646
647 D3DXMATRIX* WINAPI D3DXMatrixRotationX(D3DXMATRIX *pout, FLOAT angle)
648 {
649 TRACE("pout %p, angle %f\n", pout, angle);
650
651 D3DXMatrixIdentity(pout);
652 pout->u.m[1][1] = cos(angle);
653 pout->u.m[2][2] = cos(angle);
654 pout->u.m[1][2] = sin(angle);
655 pout->u.m[2][1] = -sin(angle);
656 return pout;
657 }
658
659 D3DXMATRIX* WINAPI D3DXMatrixRotationY(D3DXMATRIX *pout, FLOAT angle)
660 {
661 TRACE("pout %p, angle %f\n", pout, angle);
662
663 D3DXMatrixIdentity(pout);
664 pout->u.m[0][0] = cos(angle);
665 pout->u.m[2][2] = cos(angle);
666 pout->u.m[0][2] = -sin(angle);
667 pout->u.m[2][0] = sin(angle);
668 return pout;
669 }
670
671 D3DXMATRIX * WINAPI D3DXMatrixRotationYawPitchRoll(D3DXMATRIX *out, FLOAT yaw, FLOAT pitch, FLOAT roll)
672 {
673 FLOAT sroll, croll, spitch, cpitch, syaw, cyaw;
674
675 TRACE("out %p, yaw %f, pitch %f, roll %f\n", out, yaw, pitch, roll);
676
677 sroll = sinf(roll);
678 croll = cosf(roll);
679 spitch = sinf(pitch);
680 cpitch = cosf(pitch);
681 syaw = sinf(yaw);
682 cyaw = cosf(yaw);
683
684 out->u.m[0][0] = sroll * spitch * syaw + croll * cyaw;
685 out->u.m[0][1] = sroll * cpitch;
686 out->u.m[0][2] = sroll * spitch * cyaw - croll * syaw;
687 out->u.m[0][3] = 0.0f;
688 out->u.m[1][0] = croll * spitch * syaw - sroll * cyaw;
689 out->u.m[1][1] = croll * cpitch;
690 out->u.m[1][2] = croll * spitch * cyaw + sroll * syaw;
691 out->u.m[1][3] = 0.0f;
692 out->u.m[2][0] = cpitch * syaw;
693 out->u.m[2][1] = -spitch;
694 out->u.m[2][2] = cpitch * cyaw;
695 out->u.m[2][3] = 0.0f;
696 out->u.m[3][0] = 0.0f;
697 out->u.m[3][1] = 0.0f;
698 out->u.m[3][2] = 0.0f;
699 out->u.m[3][3] = 1.0f;
700
701 return out;
702 }
703
704 D3DXMATRIX* WINAPI D3DXMatrixRotationZ(D3DXMATRIX *pout, FLOAT angle)
705 {
706 TRACE("pout %p, angle %f\n", pout, angle);
707
708 D3DXMatrixIdentity(pout);
709 pout->u.m[0][0] = cos(angle);
710 pout->u.m[1][1] = cos(angle);
711 pout->u.m[0][1] = sin(angle);
712 pout->u.m[1][0] = -sin(angle);
713 return pout;
714 }
715
716 D3DXMATRIX* WINAPI D3DXMatrixScaling(D3DXMATRIX *pout, FLOAT sx, FLOAT sy, FLOAT sz)
717 {
718 TRACE("pout %p, sx %f, sy %f, sz %f\n", pout, sx, sy, sz);
719
720 D3DXMatrixIdentity(pout);
721 pout->u.m[0][0] = sx;
722 pout->u.m[1][1] = sy;
723 pout->u.m[2][2] = sz;
724 return pout;
725 }
726
727 D3DXMATRIX* WINAPI D3DXMatrixShadow(D3DXMATRIX *pout, const D3DXVECTOR4 *plight, const D3DXPLANE *pplane)
728 {
729 D3DXPLANE Nplane;
730 FLOAT dot;
731
732 TRACE("pout %p, plight %p, pplane %p\n", pout, plight, pplane);
733
734 D3DXPlaneNormalize(&Nplane, pplane);
735 dot = D3DXPlaneDot(&Nplane, plight);
736 pout->u.m[0][0] = dot - Nplane.a * plight->x;
737 pout->u.m[0][1] = -Nplane.a * plight->y;
738 pout->u.m[0][2] = -Nplane.a * plight->z;
739 pout->u.m[0][3] = -Nplane.a * plight->w;
740 pout->u.m[1][0] = -Nplane.b * plight->x;
741 pout->u.m[1][1] = dot - Nplane.b * plight->y;
742 pout->u.m[1][2] = -Nplane.b * plight->z;
743 pout->u.m[1][3] = -Nplane.b * plight->w;
744 pout->u.m[2][0] = -Nplane.c * plight->x;
745 pout->u.m[2][1] = -Nplane.c * plight->y;
746 pout->u.m[2][2] = dot - Nplane.c * plight->z;
747 pout->u.m[2][3] = -Nplane.c * plight->w;
748 pout->u.m[3][0] = -Nplane.d * plight->x;
749 pout->u.m[3][1] = -Nplane.d * plight->y;
750 pout->u.m[3][2] = -Nplane.d * plight->z;
751 pout->u.m[3][3] = dot - Nplane.d * plight->w;
752 return pout;
753 }
754
755 D3DXMATRIX* WINAPI D3DXMatrixTransformation(D3DXMATRIX *pout, const D3DXVECTOR3 *pscalingcenter, const D3DXQUATERNION *pscalingrotation, const D3DXVECTOR3 *pscaling, const D3DXVECTOR3 *protationcenter, const D3DXQUATERNION *protation, const D3DXVECTOR3 *ptranslation)
756 {
757 D3DXMATRIX m1, m2, m3, m4, m5, m6, m7;
758 D3DXQUATERNION prc;
759 D3DXVECTOR3 psc, pt;
760
761 TRACE("pout %p, pscalingcenter %p, pscalingrotation %p, pscaling %p, protationcentr %p, protation %p, ptranslation %p\n",
762 pout, pscalingcenter, pscalingrotation, pscaling, protationcenter, protation, ptranslation);
763
764 if ( !pscalingcenter )
765 {
766 psc.x = 0.0f;
767 psc.y = 0.0f;
768 psc.z = 0.0f;
769 }
770 else
771 {
772 psc.x = pscalingcenter->x;
773 psc.y = pscalingcenter->y;
774 psc.z = pscalingcenter->z;
775 }
776
777 if ( !protationcenter )
778 {
779 prc.x = 0.0f;
780 prc.y = 0.0f;
781 prc.z = 0.0f;
782 }
783 else
784 {
785 prc.x = protationcenter->x;
786 prc.y = protationcenter->y;
787 prc.z = protationcenter->z;
788 }
789
790 if ( !ptranslation )
791 {
792 pt.x = 0.0f;
793 pt.y = 0.0f;
794 pt.z = 0.0f;
795 }
796 else
797 {
798 pt.x = ptranslation->x;
799 pt.y = ptranslation->y;
800 pt.z = ptranslation->z;
801 }
802
803 D3DXMatrixTranslation(&m1, -psc.x, -psc.y, -psc.z);
804
805 if ( !pscalingrotation )
806 {
807 D3DXMatrixIdentity(&m2);
808 D3DXMatrixIdentity(&m4);
809 }
810 else
811 {
812 D3DXMatrixRotationQuaternion(&m4, pscalingrotation);
813 D3DXMatrixInverse(&m2, NULL, &m4);
814 }
815
816 if ( !pscaling ) D3DXMatrixIdentity(&m3);
817 else D3DXMatrixScaling(&m3, pscaling->x, pscaling->y, pscaling->z);
818
819 if ( !protation ) D3DXMatrixIdentity(&m6);
820 else D3DXMatrixRotationQuaternion(&m6, protation);
821
822 D3DXMatrixTranslation(&m5, psc.x - prc.x, psc.y - prc.y, psc.z - prc.z);
823 D3DXMatrixTranslation(&m7, prc.x + pt.x, prc.y + pt.y, prc.z + pt.z);
824 D3DXMatrixMultiply(&m1, &m1, &m2);
825 D3DXMatrixMultiply(&m1, &m1, &m3);
826 D3DXMatrixMultiply(&m1, &m1, &m4);
827 D3DXMatrixMultiply(&m1, &m1, &m5);
828 D3DXMatrixMultiply(&m1, &m1, &m6);
829 D3DXMatrixMultiply(pout, &m1, &m7);
830 return pout;
831 }
832
833 D3DXMATRIX* WINAPI D3DXMatrixTransformation2D(D3DXMATRIX *pout, const D3DXVECTOR2 *pscalingcenter, FLOAT scalingrotation, const D3DXVECTOR2 *pscaling, const D3DXVECTOR2 *protationcenter, FLOAT rotation, const D3DXVECTOR2 *ptranslation)
834 {
835 D3DXQUATERNION rot, sca_rot;
836 D3DXVECTOR3 rot_center, sca, sca_center, trans;
837
838 TRACE("pout %p, pscalingcenter %p, scalingrotation %f, pscaling %p, protztioncenter %p, rotation %f, ptranslation %p\n",
839 pout, pscalingcenter, scalingrotation, pscaling, protationcenter, rotation, ptranslation);
840
841 if ( pscalingcenter )
842 {
843 sca_center.x=pscalingcenter->x;
844 sca_center.y=pscalingcenter->y;
845 sca_center.z=0.0f;
846 }
847 else
848 {
849 sca_center.x=0.0f;
850 sca_center.y=0.0f;
851 sca_center.z=0.0f;
852 }
853
854 if ( pscaling )
855 {
856 sca.x=pscaling->x;
857 sca.y=pscaling->y;
858 sca.z=1.0f;
859 }
860 else
861 {
862 sca.x=1.0f;
863 sca.y=1.0f;
864 sca.z=1.0f;
865 }
866
867 if ( protationcenter )
868 {
869 rot_center.x=protationcenter->x;
870 rot_center.y=protationcenter->y;
871 rot_center.z=0.0f;
872 }
873 else
874 {
875 rot_center.x=0.0f;
876 rot_center.y=0.0f;
877 rot_center.z=0.0f;
878 }
879
880 if ( ptranslation )
881 {
882 trans.x=ptranslation->x;
883 trans.y=ptranslation->y;
884 trans.z=0.0f;
885 }
886 else
887 {
888 trans.x=0.0f;
889 trans.y=0.0f;
890 trans.z=0.0f;
891 }
892
893 rot.w=cos(rotation/2.0f);
894 rot.x=0.0f;
895 rot.y=0.0f;
896 rot.z=sin(rotation/2.0f);
897
898 sca_rot.w=cos(scalingrotation/2.0f);
899 sca_rot.x=0.0f;
900 sca_rot.y=0.0f;
901 sca_rot.z=sin(scalingrotation/2.0f);
902
903 D3DXMatrixTransformation(pout, &sca_center, &sca_rot, &sca, &rot_center, &rot, &trans);
904
905 return pout;
906 }
907
908 D3DXMATRIX* WINAPI D3DXMatrixTranslation(D3DXMATRIX *pout, FLOAT x, FLOAT y, FLOAT z)
909 {
910 TRACE("pout %p, x %f, y %f, z %f\n", pout, x, y, z);
911
912 D3DXMatrixIdentity(pout);
913 pout->u.m[3][0] = x;
914 pout->u.m[3][1] = y;
915 pout->u.m[3][2] = z;
916 return pout;
917 }
918
919 D3DXMATRIX* WINAPI D3DXMatrixTranspose(D3DXMATRIX *pout, const D3DXMATRIX *pm)
920 {
921 const D3DXMATRIX m = *pm;
922 int i,j;
923
924 TRACE("pout %p, pm %p\n", pout, pm);
925
926 for (i=0; i<4; i++)
927 for (j=0; j<4; j++) pout->u.m[i][j] = m.u.m[j][i];
928
929 return pout;
930 }
931
932 /*_________________D3DXMatrixStack____________________*/
933
934
935 static inline struct ID3DXMatrixStackImpl *impl_from_ID3DXMatrixStack(ID3DXMatrixStack *iface)
936 {
937 return CONTAINING_RECORD(iface, struct ID3DXMatrixStackImpl, ID3DXMatrixStack_iface);
938 }
939
940 static HRESULT WINAPI ID3DXMatrixStackImpl_QueryInterface(ID3DXMatrixStack *iface, REFIID riid, void **out)
941 {
942 TRACE("iface %p, riid %s, out %p.\n", iface, debugstr_guid(riid), out);
943
944 if (IsEqualGUID(riid, &IID_ID3DXMatrixStack)
945 || IsEqualGUID(riid, &IID_IUnknown))
946 {
947 ID3DXMatrixStack_AddRef(iface);
948 *out = iface;
949 return S_OK;
950 }
951
952 WARN("%s not implemented, returning E_NOINTERFACE.\n", debugstr_guid(riid));
953
954 *out = NULL;
955 return E_NOINTERFACE;
956 }
957
958 static ULONG WINAPI ID3DXMatrixStackImpl_AddRef(ID3DXMatrixStack *iface)
959 {
960 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
961 ULONG ref = InterlockedIncrement(&This->ref);
962 TRACE("(%p) : AddRef from %d\n", This, ref - 1);
963 return ref;
964 }
965
966 static ULONG WINAPI ID3DXMatrixStackImpl_Release(ID3DXMatrixStack *iface)
967 {
968 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
969 ULONG ref = InterlockedDecrement(&This->ref);
970 if (!ref)
971 {
972 HeapFree(GetProcessHeap(), 0, This->stack);
973 HeapFree(GetProcessHeap(), 0, This);
974 }
975 TRACE("(%p) : ReleaseRef to %d\n", This, ref);
976 return ref;
977 }
978
979 static D3DXMATRIX* WINAPI ID3DXMatrixStackImpl_GetTop(ID3DXMatrixStack *iface)
980 {
981 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
982
983 TRACE("iface %p\n", iface);
984
985 return &This->stack[This->current];
986 }
987
988 static HRESULT WINAPI ID3DXMatrixStackImpl_LoadIdentity(ID3DXMatrixStack *iface)
989 {
990 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
991
992 TRACE("iface %p\n", iface);
993
994 D3DXMatrixIdentity(&This->stack[This->current]);
995
996 return D3D_OK;
997 }
998
999 static HRESULT WINAPI ID3DXMatrixStackImpl_LoadMatrix(ID3DXMatrixStack *iface, const D3DXMATRIX *pm)
1000 {
1001 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1002
1003 TRACE("iface %p, pm %p\n", iface, pm);
1004
1005 This->stack[This->current] = *pm;
1006
1007 return D3D_OK;
1008 }
1009
1010 static HRESULT WINAPI ID3DXMatrixStackImpl_MultMatrix(ID3DXMatrixStack *iface, const D3DXMATRIX *pm)
1011 {
1012 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1013
1014 TRACE("iface %p, pm %p\n", iface, pm);
1015
1016 D3DXMatrixMultiply(&This->stack[This->current], &This->stack[This->current], pm);
1017
1018 return D3D_OK;
1019 }
1020
1021 static HRESULT WINAPI ID3DXMatrixStackImpl_MultMatrixLocal(ID3DXMatrixStack *iface, const D3DXMATRIX *pm)
1022 {
1023 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1024
1025 TRACE("iface %p, pm %p\n", iface, pm);
1026
1027 D3DXMatrixMultiply(&This->stack[This->current], pm, &This->stack[This->current]);
1028
1029 return D3D_OK;
1030 }
1031
1032 static HRESULT WINAPI ID3DXMatrixStackImpl_Pop(ID3DXMatrixStack *iface)
1033 {
1034 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1035
1036 TRACE("iface %p\n", iface);
1037
1038 /* Popping the last element on the stack returns D3D_OK, but does nothing. */
1039 if (!This->current) return D3D_OK;
1040
1041 if (This->current <= This->stack_size / 4 && This->stack_size >= INITIAL_STACK_SIZE * 2)
1042 {
1043 unsigned int new_size;
1044 D3DXMATRIX *new_stack;
1045
1046 new_size = This->stack_size / 2;
1047 new_stack = HeapReAlloc(GetProcessHeap(), 0, This->stack, new_size * sizeof(*new_stack));
1048 if (new_stack)
1049 {
1050 This->stack_size = new_size;
1051 This->stack = new_stack;
1052 }
1053 }
1054
1055 --This->current;
1056
1057 return D3D_OK;
1058 }
1059
1060 static HRESULT WINAPI ID3DXMatrixStackImpl_Push(ID3DXMatrixStack *iface)
1061 {
1062 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1063
1064 TRACE("iface %p\n", iface);
1065
1066 if (This->current == This->stack_size - 1)
1067 {
1068 unsigned int new_size;
1069 D3DXMATRIX *new_stack;
1070
1071 if (This->stack_size > UINT_MAX / 2) return E_OUTOFMEMORY;
1072
1073 new_size = This->stack_size * 2;
1074 new_stack = HeapReAlloc(GetProcessHeap(), 0, This->stack, new_size * sizeof(*new_stack));
1075 if (!new_stack) return E_OUTOFMEMORY;
1076
1077 This->stack_size = new_size;
1078 This->stack = new_stack;
1079 }
1080
1081 ++This->current;
1082 This->stack[This->current] = This->stack[This->current - 1];
1083
1084 return D3D_OK;
1085 }
1086
1087 static HRESULT WINAPI ID3DXMatrixStackImpl_RotateAxis(ID3DXMatrixStack *iface, const D3DXVECTOR3 *pv, FLOAT angle)
1088 {
1089 D3DXMATRIX temp;
1090 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1091
1092 TRACE("iface %p, pv %p, angle %f\n", iface, pv, angle);
1093
1094 D3DXMatrixRotationAxis(&temp, pv, angle);
1095 D3DXMatrixMultiply(&This->stack[This->current], &This->stack[This->current], &temp);
1096
1097 return D3D_OK;
1098 }
1099
1100 static HRESULT WINAPI ID3DXMatrixStackImpl_RotateAxisLocal(ID3DXMatrixStack *iface, const D3DXVECTOR3 *pv, FLOAT angle)
1101 {
1102 D3DXMATRIX temp;
1103 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1104
1105 TRACE("iface %p, pv %p, angle %f\n", iface, pv, angle);
1106
1107 D3DXMatrixRotationAxis(&temp, pv, angle);
1108 D3DXMatrixMultiply(&This->stack[This->current], &temp, &This->stack[This->current]);
1109
1110 return D3D_OK;
1111 }
1112
1113 static HRESULT WINAPI ID3DXMatrixStackImpl_RotateYawPitchRoll(ID3DXMatrixStack *iface, FLOAT x, FLOAT y, FLOAT z)
1114 {
1115 D3DXMATRIX temp;
1116 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1117
1118 TRACE("iface %p, x %f, y %f, z %f\n", iface, x, y, z);
1119
1120 D3DXMatrixRotationYawPitchRoll(&temp, x, y, z);
1121 D3DXMatrixMultiply(&This->stack[This->current], &This->stack[This->current], &temp);
1122
1123 return D3D_OK;
1124 }
1125
1126 static HRESULT WINAPI ID3DXMatrixStackImpl_RotateYawPitchRollLocal(ID3DXMatrixStack *iface, FLOAT x, FLOAT y, FLOAT z)
1127 {
1128 D3DXMATRIX temp;
1129 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1130
1131 TRACE("iface %p, x %f, y %f, z %f\n", iface, x, y, z);
1132
1133 D3DXMatrixRotationYawPitchRoll(&temp, x, y, z);
1134 D3DXMatrixMultiply(&This->stack[This->current], &temp, &This->stack[This->current]);
1135
1136 return D3D_OK;
1137 }
1138
1139 static HRESULT WINAPI ID3DXMatrixStackImpl_Scale(ID3DXMatrixStack *iface, FLOAT x, FLOAT y, FLOAT z)
1140 {
1141 D3DXMATRIX temp;
1142 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1143
1144 TRACE("iface %p,x %f, y %f, z %f\n", iface, x, y, z);
1145
1146 D3DXMatrixScaling(&temp, x, y, z);
1147 D3DXMatrixMultiply(&This->stack[This->current], &This->stack[This->current], &temp);
1148
1149 return D3D_OK;
1150 }
1151
1152 static HRESULT WINAPI ID3DXMatrixStackImpl_ScaleLocal(ID3DXMatrixStack *iface, FLOAT x, FLOAT y, FLOAT z)
1153 {
1154 D3DXMATRIX temp;
1155 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1156
1157 TRACE("iface %p, x %f, y %f, z %f\n", iface, x, y, z);
1158
1159 D3DXMatrixScaling(&temp, x, y, z);
1160 D3DXMatrixMultiply(&This->stack[This->current], &temp, &This->stack[This->current]);
1161
1162 return D3D_OK;
1163 }
1164
1165 static HRESULT WINAPI ID3DXMatrixStackImpl_Translate(ID3DXMatrixStack *iface, FLOAT x, FLOAT y, FLOAT z)
1166 {
1167 D3DXMATRIX temp;
1168 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1169
1170 TRACE("iface %p, x %f, y %f, z %f\n", iface, x, y, z);
1171
1172 D3DXMatrixTranslation(&temp, x, y, z);
1173 D3DXMatrixMultiply(&This->stack[This->current], &This->stack[This->current], &temp);
1174
1175 return D3D_OK;
1176 }
1177
1178 static HRESULT WINAPI ID3DXMatrixStackImpl_TranslateLocal(ID3DXMatrixStack *iface, FLOAT x, FLOAT y, FLOAT z)
1179 {
1180 D3DXMATRIX temp;
1181 struct ID3DXMatrixStackImpl *This = impl_from_ID3DXMatrixStack(iface);
1182
1183 TRACE("iface %p, x %f, y %f, z %f\n", iface, x, y, z);
1184
1185 D3DXMatrixTranslation(&temp, x, y, z);
1186 D3DXMatrixMultiply(&This->stack[This->current], &temp,&This->stack[This->current]);
1187
1188 return D3D_OK;
1189 }
1190
1191 static const ID3DXMatrixStackVtbl ID3DXMatrixStack_Vtbl =
1192 {
1193 ID3DXMatrixStackImpl_QueryInterface,
1194 ID3DXMatrixStackImpl_AddRef,
1195 ID3DXMatrixStackImpl_Release,
1196 ID3DXMatrixStackImpl_Pop,
1197 ID3DXMatrixStackImpl_Push,
1198 ID3DXMatrixStackImpl_LoadIdentity,
1199 ID3DXMatrixStackImpl_LoadMatrix,
1200 ID3DXMatrixStackImpl_MultMatrix,
1201 ID3DXMatrixStackImpl_MultMatrixLocal,
1202 ID3DXMatrixStackImpl_RotateAxis,
1203 ID3DXMatrixStackImpl_RotateAxisLocal,
1204 ID3DXMatrixStackImpl_RotateYawPitchRoll,
1205 ID3DXMatrixStackImpl_RotateYawPitchRollLocal,
1206 ID3DXMatrixStackImpl_Scale,
1207 ID3DXMatrixStackImpl_ScaleLocal,
1208 ID3DXMatrixStackImpl_Translate,
1209 ID3DXMatrixStackImpl_TranslateLocal,
1210 ID3DXMatrixStackImpl_GetTop
1211 };
1212
1213 HRESULT WINAPI D3DXCreateMatrixStack(DWORD flags, ID3DXMatrixStack **stack)
1214 {
1215 struct ID3DXMatrixStackImpl *object;
1216
1217 TRACE("flags %#x, stack %p.\n", flags, stack);
1218
1219 if (!(object = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, sizeof(*object))))
1220 {
1221 *stack = NULL;
1222 return E_OUTOFMEMORY;
1223 }
1224 object->ID3DXMatrixStack_iface.lpVtbl = &ID3DXMatrixStack_Vtbl;
1225 object->ref = 1;
1226
1227 if (!(object->stack = HeapAlloc(GetProcessHeap(), 0, INITIAL_STACK_SIZE * sizeof(*object->stack))))
1228 {
1229 HeapFree(GetProcessHeap(), 0, object);
1230 *stack = NULL;
1231 return E_OUTOFMEMORY;
1232 }
1233
1234 object->current = 0;
1235 object->stack_size = INITIAL_STACK_SIZE;
1236 D3DXMatrixIdentity(&object->stack[0]);
1237
1238 TRACE("Created matrix stack %p.\n", object);
1239
1240 *stack = &object->ID3DXMatrixStack_iface;
1241 return D3D_OK;
1242 }
1243
1244 /*_________________D3DXPLANE________________*/
1245
1246 D3DXPLANE* WINAPI D3DXPlaneFromPointNormal(D3DXPLANE *pout, const D3DXVECTOR3 *pvpoint, const D3DXVECTOR3 *pvnormal)
1247 {
1248 TRACE("pout %p, pvpoint %p, pvnormal %p\n", pout, pvpoint, pvnormal);
1249
1250 pout->a = pvnormal->x;
1251 pout->b = pvnormal->y;
1252 pout->c = pvnormal->z;
1253 pout->d = -D3DXVec3Dot(pvpoint, pvnormal);
1254 return pout;
1255 }
1256
1257 D3DXPLANE* WINAPI D3DXPlaneFromPoints(D3DXPLANE *pout, const D3DXVECTOR3 *pv1, const D3DXVECTOR3 *pv2, const D3DXVECTOR3 *pv3)
1258 {
1259 D3DXVECTOR3 edge1, edge2, normal, Nnormal;
1260
1261 TRACE("pout %p, pv1 %p, pv2 %p, pv3 %p\n", pout, pv1, pv2, pv3);
1262
1263 edge1.x = 0.0f; edge1.y = 0.0f; edge1.z = 0.0f;
1264 edge2.x = 0.0f; edge2.y = 0.0f; edge2.z = 0.0f;
1265 D3DXVec3Subtract(&edge1, pv2, pv1);
1266 D3DXVec3Subtract(&edge2, pv3, pv1);
1267 D3DXVec3Cross(&normal, &edge1, &edge2);
1268 D3DXVec3Normalize(&Nnormal, &normal);
1269 D3DXPlaneFromPointNormal(pout, pv1, &Nnormal);
1270 return pout;
1271 }
1272
1273 D3DXVECTOR3* WINAPI D3DXPlaneIntersectLine(D3DXVECTOR3 *pout, const D3DXPLANE *pp, const D3DXVECTOR3 *pv1, const D3DXVECTOR3 *pv2)
1274 {
1275 D3DXVECTOR3 direction, normal;
1276 FLOAT dot, temp;
1277
1278 TRACE("pout %p, pp %p, pv1 %p, pv2 %p\n", pout, pp, pv1, pv2);
1279
1280 normal.x = pp->a;
1281 normal.y = pp->b;
1282 normal.z = pp->c;
1283 direction.x = pv2->x - pv1->x;
1284 direction.y = pv2->y - pv1->y;
1285 direction.z = pv2->z - pv1->z;
1286 dot = D3DXVec3Dot(&normal, &direction);
1287 if ( !dot ) return NULL;
1288 temp = ( pp->d + D3DXVec3Dot(&normal, pv1) ) / dot;
1289 pout->x = pv1->x - temp * direction.x;
1290 pout->y = pv1->y - temp * direction.y;
1291 pout->z = pv1->z - temp * direction.z;
1292 return pout;
1293 }
1294
1295 D3DXPLANE * WINAPI D3DXPlaneNormalize(D3DXPLANE *out, const D3DXPLANE *p)
1296 {
1297 FLOAT norm;
1298
1299 TRACE("out %p, p %p\n", out, p);
1300
1301 norm = sqrtf(p->a * p->a + p->b * p->b + p->c * p->c);
1302 if (norm)
1303 {
1304 out->a = p->a / norm;
1305 out->b = p->b / norm;
1306 out->c = p->c / norm;
1307 out->d = p->d / norm;
1308 }
1309 else
1310 {
1311 out->a = 0.0f;
1312 out->b = 0.0f;
1313 out->c = 0.0f;
1314 out->d = 0.0f;
1315 }
1316
1317 return out;
1318 }
1319
1320 D3DXPLANE* WINAPI D3DXPlaneTransform(D3DXPLANE *pout, const D3DXPLANE *pplane, const D3DXMATRIX *pm)
1321 {
1322 const D3DXPLANE plane = *pplane;
1323
1324 TRACE("pout %p, pplane %p, pm %p\n", pout, pplane, pm);
1325
1326 pout->a = pm->u.m[0][0] * plane.a + pm->u.m[1][0] * plane.b + pm->u.m[2][0] * plane.c + pm->u.m[3][0] * plane.d;
1327 pout->b = pm->u.m[0][1] * plane.a + pm->u.m[1][1] * plane.b + pm->u.m[2][1] * plane.c + pm->u.m[3][1] * plane.d;
1328 pout->c = pm->u.m[0][2] * plane.a + pm->u.m[1][2] * plane.b + pm->u.m[2][2] * plane.c + pm->u.m[3][2] * plane.d;
1329 pout->d = pm->u.m[0][3] * plane.a + pm->u.m[1][3] * plane.b + pm->u.m[2][3] * plane.c + pm->u.m[3][3] * plane.d;
1330 return pout;
1331 }
1332
1333 D3DXPLANE* WINAPI D3DXPlaneTransformArray(D3DXPLANE* out, UINT outstride, const D3DXPLANE* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
1334 {
1335 UINT i;
1336
1337 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1338
1339 for (i = 0; i < elements; ++i) {
1340 D3DXPlaneTransform(
1341 (D3DXPLANE*)((char*)out + outstride * i),
1342 (const D3DXPLANE*)((const char*)in + instride * i),
1343 matrix);
1344 }
1345 return out;
1346 }
1347
1348 /*_________________D3DXQUATERNION________________*/
1349
1350 D3DXQUATERNION* WINAPI D3DXQuaternionBaryCentric(D3DXQUATERNION *pout, const D3DXQUATERNION *pq1, const D3DXQUATERNION *pq2, const D3DXQUATERNION *pq3, FLOAT f, FLOAT g)
1351 {
1352 D3DXQUATERNION temp1, temp2;
1353
1354 TRACE("pout %p, pq1 %p, pq2 %p, pq3 %p, f %f, g %f\n", pout, pq1, pq2, pq3, f, g);
1355
1356 D3DXQuaternionSlerp(pout, D3DXQuaternionSlerp(&temp1, pq1, pq2, f + g), D3DXQuaternionSlerp(&temp2, pq1, pq3, f+g), g / (f + g));
1357 return pout;
1358 }
1359
1360 D3DXQUATERNION * WINAPI D3DXQuaternionExp(D3DXQUATERNION *out, const D3DXQUATERNION *q)
1361 {
1362 FLOAT norm;
1363
1364 TRACE("out %p, q %p\n", out, q);
1365
1366 norm = sqrtf(q->x * q->x + q->y * q->y + q->z * q->z);
1367 if (norm)
1368 {
1369 out->x = sinf(norm) * q->x / norm;
1370 out->y = sinf(norm) * q->y / norm;
1371 out->z = sinf(norm) * q->z / norm;
1372 out->w = cosf(norm);
1373 }
1374 else
1375 {
1376 out->x = 0.0f;
1377 out->y = 0.0f;
1378 out->z = 0.0f;
1379 out->w = 1.0f;
1380 }
1381
1382 return out;
1383 }
1384
1385 D3DXQUATERNION* WINAPI D3DXQuaternionInverse(D3DXQUATERNION *pout, const D3DXQUATERNION *pq)
1386 {
1387 D3DXQUATERNION out;
1388 FLOAT norm;
1389
1390 TRACE("pout %p, pq %p\n", pout, pq);
1391
1392 norm = D3DXQuaternionLengthSq(pq);
1393
1394 out.x = -pq->x / norm;
1395 out.y = -pq->y / norm;
1396 out.z = -pq->z / norm;
1397 out.w = pq->w / norm;
1398
1399 *pout =out;
1400 return pout;
1401 }
1402
1403 D3DXQUATERNION * WINAPI D3DXQuaternionLn(D3DXQUATERNION *out, const D3DXQUATERNION *q)
1404 {
1405 FLOAT t;
1406
1407 TRACE("out %p, q %p\n", out, q);
1408
1409 if ((q->w >= 1.0f) || (q->w == -1.0f))
1410 t = 1.0f;
1411 else
1412 t = acosf(q->w) / sqrtf(1.0f - q->w * q->w);
1413
1414 out->x = t * q->x;
1415 out->y = t * q->y;
1416 out->z = t * q->z;
1417 out->w = 0.0f;
1418
1419 return out;
1420 }
1421
1422 D3DXQUATERNION* WINAPI D3DXQuaternionMultiply(D3DXQUATERNION *pout, const D3DXQUATERNION *pq1, const D3DXQUATERNION *pq2)
1423 {
1424 D3DXQUATERNION out;
1425
1426 TRACE("pout %p, pq1 %p, pq2 %p\n", pout, pq1, pq2);
1427
1428 out.x = pq2->w * pq1->x + pq2->x * pq1->w + pq2->y * pq1->z - pq2->z * pq1->y;
1429 out.y = pq2->w * pq1->y - pq2->x * pq1->z + pq2->y * pq1->w + pq2->z * pq1->x;
1430 out.z = pq2->w * pq1->z + pq2->x * pq1->y - pq2->y * pq1->x + pq2->z * pq1->w;
1431 out.w = pq2->w * pq1->w - pq2->x * pq1->x - pq2->y * pq1->y - pq2->z * pq1->z;
1432 *pout = out;
1433 return pout;
1434 }
1435
1436 D3DXQUATERNION * WINAPI D3DXQuaternionNormalize(D3DXQUATERNION *out, const D3DXQUATERNION *q)
1437 {
1438 FLOAT norm;
1439
1440 TRACE("out %p, q %p\n", out, q);
1441
1442 norm = D3DXQuaternionLength(q);
1443
1444 out->x = q->x / norm;
1445 out->y = q->y / norm;
1446 out->z = q->z / norm;
1447 out->w = q->w / norm;
1448
1449 return out;
1450 }
1451
1452 D3DXQUATERNION * WINAPI D3DXQuaternionRotationAxis(D3DXQUATERNION *out, const D3DXVECTOR3 *v, FLOAT angle)
1453 {
1454 D3DXVECTOR3 temp;
1455
1456 TRACE("out %p, v %p, angle %f\n", out, v, angle);
1457
1458 D3DXVec3Normalize(&temp, v);
1459
1460 out->x = sinf(angle / 2.0f) * temp.x;
1461 out->y = sinf(angle / 2.0f) * temp.y;
1462 out->z = sinf(angle / 2.0f) * temp.z;
1463 out->w = cosf(angle / 2.0f);
1464
1465 return out;
1466 }
1467
1468 D3DXQUATERNION * WINAPI D3DXQuaternionRotationMatrix(D3DXQUATERNION *out, const D3DXMATRIX *m)
1469 {
1470 FLOAT s, trace;
1471
1472 TRACE("out %p, m %p\n", out, m);
1473
1474 trace = m->u.m[0][0] + m->u.m[1][1] + m->u.m[2][2] + 1.0f;
1475 if (trace > 1.0f)
1476 {
1477 s = 2.0f * sqrtf(trace);
1478 out->x = (m->u.m[1][2] - m->u.m[2][1]) / s;
1479 out->y = (m->u.m[2][0] - m->u.m[0][2]) / s;
1480 out->z = (m->u.m[0][1] - m->u.m[1][0]) / s;
1481 out->w = 0.25f * s;
1482 }
1483 else
1484 {
1485 int i, maxi = 0;
1486
1487 for (i = 1; i < 3; i++)
1488 {
1489 if (m->u.m[i][i] > m->u.m[maxi][maxi])
1490 maxi = i;
1491 }
1492
1493 switch (maxi)
1494 {
1495 case 0:
1496 s = 2.0f * sqrtf(1.0f + m->u.m[0][0] - m->u.m[1][1] - m->u.m[2][2]);
1497 out->x = 0.25f * s;
1498 out->y = (m->u.m[0][1] + m->u.m[1][0]) / s;
1499 out->z = (m->u.m[0][2] + m->u.m[2][0]) / s;
1500 out->w = (m->u.m[1][2] - m->u.m[2][1]) / s;
1501 break;
1502
1503 case 1:
1504 s = 2.0f * sqrtf(1.0f + m->u.m[1][1] - m->u.m[0][0] - m->u.m[2][2]);
1505 out->x = (m->u.m[0][1] + m->u.m[1][0]) / s;
1506 out->y = 0.25f * s;
1507 out->z = (m->u.m[1][2] + m->u.m[2][1]) / s;
1508 out->w = (m->u.m[2][0] - m->u.m[0][2]) / s;
1509 break;
1510
1511 case 2:
1512 s = 2.0f * sqrtf(1.0f + m->u.m[2][2] - m->u.m[0][0] - m->u.m[1][1]);
1513 out->x = (m->u.m[0][2] + m->u.m[2][0]) / s;
1514 out->y = (m->u.m[1][2] + m->u.m[2][1]) / s;
1515 out->z = 0.25f * s;
1516 out->w = (m->u.m[0][1] - m->u.m[1][0]) / s;
1517 break;
1518 }
1519 }
1520
1521 return out;
1522 }
1523
1524 D3DXQUATERNION * WINAPI D3DXQuaternionRotationYawPitchRoll(D3DXQUATERNION *out, FLOAT yaw, FLOAT pitch, FLOAT roll)
1525 {
1526 FLOAT syaw, cyaw, spitch, cpitch, sroll, croll;
1527
1528 TRACE("out %p, yaw %f, pitch %f, roll %f\n", out, yaw, pitch, roll);
1529
1530 syaw = sinf(yaw / 2.0f);
1531 cyaw = cosf(yaw / 2.0f);
1532 spitch = sinf(pitch / 2.0f);
1533 cpitch = cosf(pitch / 2.0f);
1534 sroll = sinf(roll / 2.0f);
1535 croll = cosf(roll / 2.0f);
1536
1537 out->x = syaw * cpitch * sroll + cyaw * spitch * croll;
1538 out->y = syaw * cpitch * croll - cyaw * spitch * sroll;
1539 out->z = cyaw * cpitch * sroll - syaw * spitch * croll;
1540 out->w = cyaw * cpitch * croll + syaw * spitch * sroll;
1541
1542 return out;
1543 }
1544
1545 D3DXQUATERNION * WINAPI D3DXQuaternionSlerp(D3DXQUATERNION *out, const D3DXQUATERNION *q1,
1546 const D3DXQUATERNION *q2, FLOAT t)
1547 {
1548 FLOAT dot, temp;
1549
1550 TRACE("out %p, q1 %p, q2 %p, t %f\n", out, q1, q2, t);
1551
1552 temp = 1.0f - t;
1553 dot = D3DXQuaternionDot(q1, q2);
1554 if (dot < 0.0f)
1555 {
1556 t = -t;
1557 dot = -dot;
1558 }
1559
1560 if (1.0f - dot > 0.001f)
1561 {
1562 FLOAT theta = acosf(dot);
1563
1564 temp = sinf(theta * temp) / sinf(theta);
1565 t = sinf(theta * t) / sinf(theta);
1566 }
1567
1568 out->x = temp * q1->x + t * q2->x;
1569 out->y = temp * q1->y + t * q2->y;
1570 out->z = temp * q1->z + t * q2->z;
1571 out->w = temp * q1->w + t * q2->w;
1572
1573 return out;
1574 }
1575
1576 D3DXQUATERNION* WINAPI D3DXQuaternionSquad(D3DXQUATERNION *pout, const D3DXQUATERNION *pq1, const D3DXQUATERNION *pq2, const D3DXQUATERNION *pq3, const D3DXQUATERNION *pq4, FLOAT t)
1577 {
1578 D3DXQUATERNION temp1, temp2;
1579
1580 TRACE("pout %p, pq1 %p, pq2 %p, pq3 %p, pq4 %p, t %f\n", pout, pq1, pq2, pq3, pq4, t);
1581
1582 D3DXQuaternionSlerp(pout, D3DXQuaternionSlerp(&temp1, pq1, pq4, t), D3DXQuaternionSlerp(&temp2, pq2, pq3, t), 2.0f * t * (1.0f - t));
1583 return pout;
1584 }
1585
1586 static D3DXQUATERNION add_diff(const D3DXQUATERNION *q1, const D3DXQUATERNION *q2, const FLOAT add)
1587 {
1588 D3DXQUATERNION temp;
1589
1590 temp.x = q1->x + add * q2->x;
1591 temp.y = q1->y + add * q2->y;
1592 temp.z = q1->z + add * q2->z;
1593 temp.w = q1->w + add * q2->w;
1594
1595 return temp;
1596 }
1597
1598 void WINAPI D3DXQuaternionSquadSetup(D3DXQUATERNION *paout, D3DXQUATERNION *pbout, D3DXQUATERNION *pcout, const D3DXQUATERNION *pq0, const D3DXQUATERNION *pq1, const D3DXQUATERNION *pq2, const D3DXQUATERNION *pq3)
1599 {
1600 D3DXQUATERNION q, temp1, temp2, temp3, zero;
1601
1602 TRACE("paout %p, pbout %p, pcout %p, pq0 %p, pq1 %p, pq2 %p, pq3 %p\n", paout, pbout, pcout, pq0, pq1, pq2, pq3);
1603
1604 zero.x = 0.0f;
1605 zero.y = 0.0f;
1606 zero.z = 0.0f;
1607 zero.w = 0.0f;
1608
1609 if ( D3DXQuaternionDot(pq0, pq1) < 0.0f )
1610 temp2 = add_diff(&zero, pq0, -1.0f);
1611 else
1612 temp2 = *pq0;
1613
1614 if ( D3DXQuaternionDot(pq1, pq2) < 0.0f )
1615 *pcout = add_diff(&zero, pq2, -1.0f);
1616 else
1617 *pcout = *pq2;
1618
1619 if ( D3DXQuaternionDot(pcout, pq3) < 0.0f )
1620 temp3 = add_diff(&zero, pq3, -1.0f);
1621 else
1622 temp3 = *pq3;
1623
1624 D3DXQuaternionInverse(&temp1, pq1);
1625 D3DXQuaternionMultiply(&temp2, &temp1, &temp2);
1626 D3DXQuaternionLn(&temp2, &temp2);
1627 D3DXQuaternionMultiply(&q, &temp1, pcout);
1628 D3DXQuaternionLn(&q, &q);
1629 temp1 = add_diff(&temp2, &q, 1.0f);
1630 temp1.x *= -0.25f;
1631 temp1.y *= -0.25f;
1632 temp1.z *= -0.25f;
1633 temp1.w *= -0.25f;
1634 D3DXQuaternionExp(&temp1, &temp1);
1635 D3DXQuaternionMultiply(paout, pq1, &temp1);
1636
1637 D3DXQuaternionInverse(&temp1, pcout);
1638 D3DXQuaternionMultiply(&temp2, &temp1, pq1);
1639 D3DXQuaternionLn(&temp2, &temp2);
1640 D3DXQuaternionMultiply(&q, &temp1, &temp3);
1641 D3DXQuaternionLn(&q, &q);
1642 temp1 = add_diff(&temp2, &q, 1.0f);
1643 temp1.x *= -0.25f;
1644 temp1.y *= -0.25f;
1645 temp1.z *= -0.25f;
1646 temp1.w *= -0.25f;
1647 D3DXQuaternionExp(&temp1, &temp1);
1648 D3DXQuaternionMultiply(pbout, pcout, &temp1);
1649
1650 return;
1651 }
1652
1653 void WINAPI D3DXQuaternionToAxisAngle(const D3DXQUATERNION *pq, D3DXVECTOR3 *paxis, FLOAT *pangle)
1654 {
1655 TRACE("pq %p, paxis %p, pangle %p\n", pq, paxis, pangle);
1656
1657 paxis->x = pq->x;
1658 paxis->y = pq->y;
1659 paxis->z = pq->z;
1660 *pangle = 2.0f * acos(pq->w);
1661 }
1662
1663 /*_________________D3DXVec2_____________________*/
1664
1665 D3DXVECTOR2* WINAPI D3DXVec2BaryCentric(D3DXVECTOR2 *pout, const D3DXVECTOR2 *pv1, const D3DXVECTOR2 *pv2, const D3DXVECTOR2 *pv3, FLOAT f, FLOAT g)
1666 {
1667 TRACE("pout %p, pv1 %p, pv2 %p, pv3 %p, f %f, g %f\n", pout, pv1, pv2, pv3, f, g);
1668
1669 pout->x = (1.0f-f-g) * (pv1->x) + f * (pv2->x) + g * (pv3->x);
1670 pout->y = (1.0f-f-g) * (pv1->y) + f * (pv2->y) + g * (pv3->y);
1671 return pout;
1672 }
1673
1674 D3DXVECTOR2* WINAPI D3DXVec2CatmullRom(D3DXVECTOR2 *pout, const D3DXVECTOR2 *pv0, const D3DXVECTOR2 *pv1, const D3DXVECTOR2 *pv2, const D3DXVECTOR2 *pv3, FLOAT s)
1675 {
1676 TRACE("pout %p, pv0 %p, pv1 %p, pv2 %p, pv3 %p, s %f\n", pout, pv0, pv1, pv2, pv3, s);
1677
1678 pout->x = 0.5f * (2.0f * pv1->x + (pv2->x - pv0->x) *s + (2.0f *pv0->x - 5.0f * pv1->x + 4.0f * pv2->x - pv3->x) * s * s + (pv3->x -3.0f * pv2->x + 3.0f * pv1->x - pv0->x) * s * s * s);
1679 pout->y = 0.5f * (2.0f * pv1->y + (pv2->y - pv0->y) *s + (2.0f *pv0->y - 5.0f * pv1->y + 4.0f * pv2->y - pv3->y) * s * s + (pv3->y -3.0f * pv2->y + 3.0f * pv1->y - pv0->y) * s * s * s);
1680 return pout;
1681 }
1682
1683 D3DXVECTOR2* WINAPI D3DXVec2Hermite(D3DXVECTOR2 *pout, const D3DXVECTOR2 *pv1, const D3DXVECTOR2 *pt1, const D3DXVECTOR2 *pv2, const D3DXVECTOR2 *pt2, FLOAT s)
1684 {
1685 FLOAT h1, h2, h3, h4;
1686
1687 TRACE("pout %p, pv1 %p, pt1 %p, pv2 %p, pt2 %p, s %f\n", pout, pv1, pt1, pv2, pt2, s);
1688
1689 h1 = 2.0f * s * s * s - 3.0f * s * s + 1.0f;
1690 h2 = s * s * s - 2.0f * s * s + s;
1691 h3 = -2.0f * s * s * s + 3.0f * s * s;
1692 h4 = s * s * s - s * s;
1693
1694 pout->x = h1 * (pv1->x) + h2 * (pt1->x) + h3 * (pv2->x) + h4 * (pt2->x);
1695 pout->y = h1 * (pv1->y) + h2 * (pt1->y) + h3 * (pv2->y) + h4 * (pt2->y);
1696 return pout;
1697 }
1698
1699 D3DXVECTOR2* WINAPI D3DXVec2Normalize(D3DXVECTOR2 *pout, const D3DXVECTOR2 *pv)
1700 {
1701 FLOAT norm;
1702
1703 TRACE("pout %p, pv %p\n", pout, pv);
1704
1705 norm = D3DXVec2Length(pv);
1706 if ( !norm )
1707 {
1708 pout->x = 0.0f;
1709 pout->y = 0.0f;
1710 }
1711 else
1712 {
1713 pout->x = pv->x / norm;
1714 pout->y = pv->y / norm;
1715 }
1716
1717 return pout;
1718 }
1719
1720 D3DXVECTOR4* WINAPI D3DXVec2Transform(D3DXVECTOR4 *pout, const D3DXVECTOR2 *pv, const D3DXMATRIX *pm)
1721 {
1722 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
1723
1724 pout->x = pm->u.m[0][0] * pv->x + pm->u.m[1][0] * pv->y + pm->u.m[3][0];
1725 pout->y = pm->u.m[0][1] * pv->x + pm->u.m[1][1] * pv->y + pm->u.m[3][1];
1726 pout->z = pm->u.m[0][2] * pv->x + pm->u.m[1][2] * pv->y + pm->u.m[3][2];
1727 pout->w = pm->u.m[0][3] * pv->x + pm->u.m[1][3] * pv->y + pm->u.m[3][3];
1728 return pout;
1729 }
1730
1731 D3DXVECTOR4* WINAPI D3DXVec2TransformArray(D3DXVECTOR4* out, UINT outstride, const D3DXVECTOR2* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
1732 {
1733 UINT i;
1734
1735 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1736
1737 for (i = 0; i < elements; ++i) {
1738 D3DXVec2Transform(
1739 (D3DXVECTOR4*)((char*)out + outstride * i),
1740 (const D3DXVECTOR2*)((const char*)in + instride * i),
1741 matrix);
1742 }
1743 return out;
1744 }
1745
1746 D3DXVECTOR2* WINAPI D3DXVec2TransformCoord(D3DXVECTOR2 *pout, const D3DXVECTOR2 *pv, const D3DXMATRIX *pm)
1747 {
1748 D3DXVECTOR2 v;
1749 FLOAT norm;
1750
1751 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
1752
1753 v = *pv;
1754 norm = pm->u.m[0][3] * pv->x + pm->u.m[1][3] * pv->y + pm->u.m[3][3];
1755
1756 pout->x = (pm->u.m[0][0] * v.x + pm->u.m[1][0] * v.y + pm->u.m[3][0]) / norm;
1757 pout->y = (pm->u.m[0][1] * v.x + pm->u.m[1][1] * v.y + pm->u.m[3][1]) / norm;
1758
1759 return pout;
1760 }
1761
1762 D3DXVECTOR2* WINAPI D3DXVec2TransformCoordArray(D3DXVECTOR2* out, UINT outstride, const D3DXVECTOR2* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
1763 {
1764 UINT i;
1765
1766 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1767
1768 for (i = 0; i < elements; ++i) {
1769 D3DXVec2TransformCoord(
1770 (D3DXVECTOR2*)((char*)out + outstride * i),
1771 (const D3DXVECTOR2*)((const char*)in + instride * i),
1772 matrix);
1773 }
1774 return out;
1775 }
1776
1777 D3DXVECTOR2* WINAPI D3DXVec2TransformNormal(D3DXVECTOR2 *pout, const D3DXVECTOR2 *pv, const D3DXMATRIX *pm)
1778 {
1779 const D3DXVECTOR2 v = *pv;
1780
1781 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
1782
1783 pout->x = pm->u.m[0][0] * v.x + pm->u.m[1][0] * v.y;
1784 pout->y = pm->u.m[0][1] * v.x + pm->u.m[1][1] * v.y;
1785 return pout;
1786 }
1787
1788 D3DXVECTOR2* WINAPI D3DXVec2TransformNormalArray(D3DXVECTOR2* out, UINT outstride, const D3DXVECTOR2 *in, UINT instride, const D3DXMATRIX *matrix, UINT elements)
1789 {
1790 UINT i;
1791
1792 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1793
1794 for (i = 0; i < elements; ++i) {
1795 D3DXVec2TransformNormal(
1796 (D3DXVECTOR2*)((char*)out + outstride * i),
1797 (const D3DXVECTOR2*)((const char*)in + instride * i),
1798 matrix);
1799 }
1800 return out;
1801 }
1802
1803 /*_________________D3DXVec3_____________________*/
1804
1805 D3DXVECTOR3* WINAPI D3DXVec3BaryCentric(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv1, const D3DXVECTOR3 *pv2, const D3DXVECTOR3 *pv3, FLOAT f, FLOAT g)
1806 {
1807 TRACE("pout %p, pv1 %p, pv2 %p, pv3 %p, f %f, g %f\n", pout, pv1, pv2, pv3, f, g);
1808
1809 pout->x = (1.0f-f-g) * (pv1->x) + f * (pv2->x) + g * (pv3->x);
1810 pout->y = (1.0f-f-g) * (pv1->y) + f * (pv2->y) + g * (pv3->y);
1811 pout->z = (1.0f-f-g) * (pv1->z) + f * (pv2->z) + g * (pv3->z);
1812 return pout;
1813 }
1814
1815 D3DXVECTOR3* WINAPI D3DXVec3CatmullRom( D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv0, const D3DXVECTOR3 *pv1, const D3DXVECTOR3 *pv2, const D3DXVECTOR3 *pv3, FLOAT s)
1816 {
1817 TRACE("pout %p, pv0 %p, pv1 %p, pv2 %p, pv3 %p, s %f\n", pout, pv0, pv1, pv2, pv3, s);
1818
1819 pout->x = 0.5f * (2.0f * pv1->x + (pv2->x - pv0->x) *s + (2.0f *pv0->x - 5.0f * pv1->x + 4.0f * pv2->x - pv3->x) * s * s + (pv3->x -3.0f * pv2->x + 3.0f * pv1->x - pv0->x) * s * s * s);
1820 pout->y = 0.5f * (2.0f * pv1->y + (pv2->y - pv0->y) *s + (2.0f *pv0->y - 5.0f * pv1->y + 4.0f * pv2->y - pv3->y) * s * s + (pv3->y -3.0f * pv2->y + 3.0f * pv1->y - pv0->y) * s * s * s);
1821 pout->z = 0.5f * (2.0f * pv1->z + (pv2->z - pv0->z) *s + (2.0f *pv0->z - 5.0f * pv1->z + 4.0f * pv2->z - pv3->z) * s * s + (pv3->z -3.0f * pv2->z + 3.0f * pv1->z - pv0->z) * s * s * s);
1822 return pout;
1823 }
1824
1825 D3DXVECTOR3* WINAPI D3DXVec3Hermite(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv1, const D3DXVECTOR3 *pt1, const D3DXVECTOR3 *pv2, const D3DXVECTOR3 *pt2, FLOAT s)
1826 {
1827 FLOAT h1, h2, h3, h4;
1828
1829 TRACE("pout %p, pv1 %p, pt1 %p, pv2 %p, pt2 %p, s %f\n", pout, pv1, pt1, pv2, pt2, s);
1830
1831 h1 = 2.0f * s * s * s - 3.0f * s * s + 1.0f;
1832 h2 = s * s * s - 2.0f * s * s + s;
1833 h3 = -2.0f * s * s * s + 3.0f * s * s;
1834 h4 = s * s * s - s * s;
1835
1836 pout->x = h1 * (pv1->x) + h2 * (pt1->x) + h3 * (pv2->x) + h4 * (pt2->x);
1837 pout->y = h1 * (pv1->y) + h2 * (pt1->y) + h3 * (pv2->y) + h4 * (pt2->y);
1838 pout->z = h1 * (pv1->z) + h2 * (pt1->z) + h3 * (pv2->z) + h4 * (pt2->z);
1839 return pout;
1840 }
1841
1842 D3DXVECTOR3* WINAPI D3DXVec3Normalize(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv)
1843 {
1844 FLOAT norm;
1845
1846 TRACE("pout %p, pv %p\n", pout, pv);
1847
1848 norm = D3DXVec3Length(pv);
1849 if ( !norm )
1850 {
1851 pout->x = 0.0f;
1852 pout->y = 0.0f;
1853 pout->z = 0.0f;
1854 }
1855 else
1856 {
1857 pout->x = pv->x / norm;
1858 pout->y = pv->y / norm;
1859 pout->z = pv->z / norm;
1860 }
1861
1862 return pout;
1863 }
1864
1865 D3DXVECTOR3* WINAPI D3DXVec3Project(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv, const D3DVIEWPORT9 *pviewport, const D3DXMATRIX *pprojection, const D3DXMATRIX *pview, const D3DXMATRIX *pworld)
1866 {
1867 D3DXMATRIX m;
1868
1869 TRACE("pout %p, pv %p, pviewport %p, pprojection %p, pview %p, pworld %p\n", pout, pv, pviewport, pprojection, pview, pworld);
1870
1871 D3DXMatrixIdentity(&m);
1872 if (pworld) D3DXMatrixMultiply(&m, &m, pworld);
1873 if (pview) D3DXMatrixMultiply(&m, &m, pview);
1874 if (pprojection) D3DXMatrixMultiply(&m, &m, pprojection);
1875
1876 D3DXVec3TransformCoord(pout, pv, &m);
1877
1878 if (pviewport)
1879 {
1880 pout->x = pviewport->X + ( 1.0f + pout->x ) * pviewport->Width / 2.0f;
1881 pout->y = pviewport->Y + ( 1.0f - pout->y ) * pviewport->Height / 2.0f;
1882 pout->z = pviewport->MinZ + pout->z * ( pviewport->MaxZ - pviewport->MinZ );
1883 }
1884 return pout;
1885 }
1886
1887 D3DXVECTOR3* WINAPI D3DXVec3ProjectArray(D3DXVECTOR3* out, UINT outstride, const D3DXVECTOR3* in, UINT instride, const D3DVIEWPORT9* viewport, const D3DXMATRIX* projection, const D3DXMATRIX* view, const D3DXMATRIX* world, UINT elements)
1888 {
1889 UINT i;
1890
1891 TRACE("out %p, outstride %u, in %p, instride %u, viewport %p, projection %p, view %p, world %p, elements %u\n",
1892 out, outstride, in, instride, viewport, projection, view, world, elements);
1893
1894 for (i = 0; i < elements; ++i) {
1895 D3DXVec3Project(
1896 (D3DXVECTOR3*)((char*)out + outstride * i),
1897 (const D3DXVECTOR3*)((const char*)in + instride * i),
1898 viewport, projection, view, world);
1899 }
1900 return out;
1901 }
1902
1903 D3DXVECTOR4* WINAPI D3DXVec3Transform(D3DXVECTOR4 *pout, const D3DXVECTOR3 *pv, const D3DXMATRIX *pm)
1904 {
1905 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
1906
1907 pout->x = pm->u.m[0][0] * pv->x + pm->u.m[1][0] * pv->y + pm->u.m[2][0] * pv->z + pm->u.m[3][0];
1908 pout->y = pm->u.m[0][1] * pv->x + pm->u.m[1][1] * pv->y + pm->u.m[2][1] * pv->z + pm->u.m[3][1];
1909 pout->z = pm->u.m[0][2] * pv->x + pm->u.m[1][2] * pv->y + pm->u.m[2][2] * pv->z + pm->u.m[3][2];
1910 pout->w = pm->u.m[0][3] * pv->x + pm->u.m[1][3] * pv->y + pm->u.m[2][3] * pv->z + pm->u.m[3][3];
1911 return pout;
1912 }
1913
1914 D3DXVECTOR4* WINAPI D3DXVec3TransformArray(D3DXVECTOR4* out, UINT outstride, const D3DXVECTOR3* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
1915 {
1916 UINT i;
1917
1918 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1919
1920 for (i = 0; i < elements; ++i) {
1921 D3DXVec3Transform(
1922 (D3DXVECTOR4*)((char*)out + outstride * i),
1923 (const D3DXVECTOR3*)((const char*)in + instride * i),
1924 matrix);
1925 }
1926 return out;
1927 }
1928
1929 D3DXVECTOR3* WINAPI D3DXVec3TransformCoord(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv, const D3DXMATRIX *pm)
1930 {
1931 D3DXVECTOR3 out;
1932 FLOAT norm;
1933
1934 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
1935
1936 norm = pm->u.m[0][3] * pv->x + pm->u.m[1][3] * pv->y + pm->u.m[2][3] *pv->z + pm->u.m[3][3];
1937
1938 out.x = (pm->u.m[0][0] * pv->x + pm->u.m[1][0] * pv->y + pm->u.m[2][0] * pv->z + pm->u.m[3][0]) / norm;
1939 out.y = (pm->u.m[0][1] * pv->x + pm->u.m[1][1] * pv->y + pm->u.m[2][1] * pv->z + pm->u.m[3][1]) / norm;
1940 out.z = (pm->u.m[0][2] * pv->x + pm->u.m[1][2] * pv->y + pm->u.m[2][2] * pv->z + pm->u.m[3][2]) / norm;
1941
1942 *pout = out;
1943
1944 return pout;
1945 }
1946
1947 D3DXVECTOR3* WINAPI D3DXVec3TransformCoordArray(D3DXVECTOR3* out, UINT outstride, const D3DXVECTOR3* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
1948 {
1949 UINT i;
1950
1951 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1952
1953 for (i = 0; i < elements; ++i) {
1954 D3DXVec3TransformCoord(
1955 (D3DXVECTOR3*)((char*)out + outstride * i),
1956 (const D3DXVECTOR3*)((const char*)in + instride * i),
1957 matrix);
1958 }
1959 return out;
1960 }
1961
1962 D3DXVECTOR3* WINAPI D3DXVec3TransformNormal(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv, const D3DXMATRIX *pm)
1963 {
1964 const D3DXVECTOR3 v = *pv;
1965
1966 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
1967
1968 pout->x = pm->u.m[0][0] * v.x + pm->u.m[1][0] * v.y + pm->u.m[2][0] * v.z;
1969 pout->y = pm->u.m[0][1] * v.x + pm->u.m[1][1] * v.y + pm->u.m[2][1] * v.z;
1970 pout->z = pm->u.m[0][2] * v.x + pm->u.m[1][2] * v.y + pm->u.m[2][2] * v.z;
1971 return pout;
1972
1973 }
1974
1975 D3DXVECTOR3* WINAPI D3DXVec3TransformNormalArray(D3DXVECTOR3* out, UINT outstride, const D3DXVECTOR3* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
1976 {
1977 UINT i;
1978
1979 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
1980
1981 for (i = 0; i < elements; ++i) {
1982 D3DXVec3TransformNormal(
1983 (D3DXVECTOR3*)((char*)out + outstride * i),
1984 (const D3DXVECTOR3*)((const char*)in + instride * i),
1985 matrix);
1986 }
1987 return out;
1988 }
1989
1990 D3DXVECTOR3* WINAPI D3DXVec3Unproject(D3DXVECTOR3 *pout, const D3DXVECTOR3 *pv, const D3DVIEWPORT9 *pviewport, const D3DXMATRIX *pprojection, const D3DXMATRIX *pview, const D3DXMATRIX *pworld)
1991 {
1992 D3DXMATRIX m;
1993
1994 TRACE("pout %p, pv %p, pviewport %p, pprojection %p, pview %p, pworlds %p\n", pout, pv, pviewport, pprojection, pview, pworld);
1995
1996 D3DXMatrixIdentity(&m);
1997 if (pworld) D3DXMatrixMultiply(&m, &m, pworld);
1998 if (pview) D3DXMatrixMultiply(&m, &m, pview);
1999 if (pprojection) D3DXMatrixMultiply(&m, &m, pprojection);
2000 D3DXMatrixInverse(&m, NULL, &m);
2001
2002 *pout = *pv;
2003 if (pviewport)
2004 {
2005 pout->x = 2.0f * ( pout->x - pviewport->X ) / pviewport->Width - 1.0f;
2006 pout->y = 1.0f - 2.0f * ( pout->y - pviewport->Y ) / pviewport->Height;
2007 pout->z = ( pout->z - pviewport->MinZ) / ( pviewport->MaxZ - pviewport->MinZ );
2008 }
2009 D3DXVec3TransformCoord(pout, pout, &m);
2010 return pout;
2011 }
2012
2013 D3DXVECTOR3* WINAPI D3DXVec3UnprojectArray(D3DXVECTOR3* out, UINT outstride, const D3DXVECTOR3* in, UINT instride, const D3DVIEWPORT9* viewport, const D3DXMATRIX* projection, const D3DXMATRIX* view, const D3DXMATRIX* world, UINT elements)
2014 {
2015 UINT i;
2016
2017 TRACE("out %p, outstride %u, in %p, instride %u, viewport %p, projection %p, view %p, world %p, elements %u\n",
2018 out, outstride, in, instride, viewport, projection, view, world, elements);
2019
2020 for (i = 0; i < elements; ++i) {
2021 D3DXVec3Unproject(
2022 (D3DXVECTOR3*)((char*)out + outstride * i),
2023 (const D3DXVECTOR3*)((const char*)in + instride * i),
2024 viewport, projection, view, world);
2025 }
2026 return out;
2027 }
2028
2029 /*_________________D3DXVec4_____________________*/
2030
2031 D3DXVECTOR4* WINAPI D3DXVec4BaryCentric(D3DXVECTOR4 *pout, const D3DXVECTOR4 *pv1, const D3DXVECTOR4 *pv2, const D3DXVECTOR4 *pv3, FLOAT f, FLOAT g)
2032 {
2033 TRACE("pout %p, pv1 %p, pv2 %p, pv3 %p, f %f, g %f\n", pout, pv1, pv2, pv3, f, g);
2034
2035 pout->x = (1.0f-f-g) * (pv1->x) + f * (pv2->x) + g * (pv3->x);
2036 pout->y = (1.0f-f-g) * (pv1->y) + f * (pv2->y) + g * (pv3->y);
2037 pout->z = (1.0f-f-g) * (pv1->z) + f * (pv2->z) + g * (pv3->z);
2038 pout->w = (1.0f-f-g) * (pv1->w) + f * (pv2->w) + g * (pv3->w);
2039 return pout;
2040 }
2041
2042 D3DXVECTOR4* WINAPI D3DXVec4CatmullRom(D3DXVECTOR4 *pout, const D3DXVECTOR4 *pv0, const D3DXVECTOR4 *pv1, const D3DXVECTOR4 *pv2, const D3DXVECTOR4 *pv3, FLOAT s)
2043 {
2044 TRACE("pout %p, pv0 %p, pv1 %p, pv2 %p, pv3 %p, s %f\n", pout, pv0, pv1, pv2, pv3, s);
2045
2046 pout->x = 0.5f * (2.0f * pv1->x + (pv2->x - pv0->x) *s + (2.0f *pv0->x - 5.0f * pv1->x + 4.0f * pv2->x - pv3->x) * s * s + (pv3->x -3.0f * pv2->x + 3.0f * pv1->x - pv0->x) * s * s * s);
2047 pout->y = 0.5f * (2.0f * pv1->y + (pv2->y - pv0->y) *s + (2.0f *pv0->y - 5.0f * pv1->y + 4.0f * pv2->y - pv3->y) * s * s + (pv3->y -3.0f * pv2->y + 3.0f * pv1->y - pv0->y) * s * s * s);
2048 pout->z = 0.5f * (2.0f * pv1->z + (pv2->z - pv0->z) *s + (2.0f *pv0->z - 5.0f * pv1->z + 4.0f * pv2->z - pv3->z) * s * s + (pv3->z -3.0f * pv2->z + 3.0f * pv1->z - pv0->z) * s * s * s);
2049 pout->w = 0.5f * (2.0f * pv1->w + (pv2->w - pv0->w) *s + (2.0f *pv0->w - 5.0f * pv1->w + 4.0f * pv2->w - pv3->w) * s * s + (pv3->w -3.0f * pv2->w + 3.0f * pv1->w - pv0->w) * s * s * s);
2050 return pout;
2051 }
2052
2053 D3DXVECTOR4* WINAPI D3DXVec4Cross(D3DXVECTOR4 *pout, const D3DXVECTOR4 *pv1, const D3DXVECTOR4 *pv2, const D3DXVECTOR4 *pv3)
2054 {
2055 D3DXVECTOR4 out;
2056
2057 TRACE("pout %p, pv1 %p, pv2 %p, pv3 %p\n", pout, pv1, pv2, pv3);
2058
2059 out.x = pv1->y * (pv2->z * pv3->w - pv3->z * pv2->w) - pv1->z * (pv2->y * pv3->w - pv3->y * pv2->w) + pv1->w * (pv2->y * pv3->z - pv2->z *pv3->y);
2060 out.y = -(pv1->x * (pv2->z * pv3->w - pv3->z * pv2->w) - pv1->z * (pv2->x * pv3->w - pv3->x * pv2->w) + pv1->w * (pv2->x * pv3->z - pv3->x * pv2->z));
2061 out.z = pv1->x * (pv2->y * pv3->w - pv3->y * pv2->w) - pv1->y * (pv2->x *pv3->w - pv3->x * pv2->w) + pv1->w * (pv2->x * pv3->y - pv3->x * pv2->y);
2062 out.w = -(pv1->x * (pv2->y * pv3->z - pv3->y * pv2->z) - pv1->y * (pv2->x * pv3->z - pv3->x *pv2->z) + pv1->z * (pv2->x * pv3->y - pv3->x * pv2->y));
2063 *pout = out;
2064 return pout;
2065 }
2066
2067 D3DXVECTOR4* WINAPI D3DXVec4Hermite(D3DXVECTOR4 *pout, const D3DXVECTOR4 *pv1, const D3DXVECTOR4 *pt1, const D3DXVECTOR4 *pv2, const D3DXVECTOR4 *pt2, FLOAT s)
2068 {
2069 FLOAT h1, h2, h3, h4;
2070
2071 TRACE("pout %p, pv1 %p, pt1 %p, pv2 %p, pt2 %p, s %f\n", pout, pv1, pt1, pv2, pt2, s);
2072
2073 h1 = 2.0f * s * s * s - 3.0f * s * s + 1.0f;
2074 h2 = s * s * s - 2.0f * s * s + s;
2075 h3 = -2.0f * s * s * s + 3.0f * s * s;
2076 h4 = s * s * s - s * s;
2077
2078 pout->x = h1 * (pv1->x) + h2 * (pt1->x) + h3 * (pv2->x) + h4 * (pt2->x);
2079 pout->y = h1 * (pv1->y) + h2 * (pt1->y) + h3 * (pv2->y) + h4 * (pt2->y);
2080 pout->z = h1 * (pv1->z) + h2 * (pt1->z) + h3 * (pv2->z) + h4 * (pt2->z);
2081 pout->w = h1 * (pv1->w) + h2 * (pt1->w) + h3 * (pv2->w) + h4 * (pt2->w);
2082 return pout;
2083 }
2084
2085 D3DXVECTOR4* WINAPI D3DXVec4Normalize(D3DXVECTOR4 *pout, const D3DXVECTOR4 *pv)
2086 {
2087 FLOAT norm;
2088
2089 TRACE("pout %p, pv %p\n", pout, pv);
2090
2091 norm = D3DXVec4Length(pv);
2092
2093 pout->x = pv->x / norm;
2094 pout->y = pv->y / norm;
2095 pout->z = pv->z / norm;
2096 pout->w = pv->w / norm;
2097
2098 return pout;
2099 }
2100
2101 D3DXVECTOR4* WINAPI D3DXVec4Transform(D3DXVECTOR4 *pout, const D3DXVECTOR4 *pv, const D3DXMATRIX *pm)
2102 {
2103 D3DXVECTOR4 out;
2104
2105 TRACE("pout %p, pv %p, pm %p\n", pout, pv, pm);
2106
2107 out.x = pm->u.m[0][0] * pv->x + pm->u.m[1][0] * pv->y + pm->u.m[2][0] * pv->z + pm->u.m[3][0] * pv->w;
2108 out.y = pm->u.m[0][1] * pv->x + pm->u.m[1][1] * pv->y + pm->u.m[2][1] * pv->z + pm->u.m[3][1] * pv->w;
2109 out.z = pm->u.m[0][2] * pv->x + pm->u.m[1][2] * pv->y + pm->u.m[2][2] * pv->z + pm->u.m[3][2] * pv->w;
2110 out.w = pm->u.m[0][3] * pv->x + pm->u.m[1][3] * pv->y + pm->u.m[2][3] * pv->z + pm->u.m[3][3] * pv->w;
2111 *pout = out;
2112 return pout;
2113 }
2114
2115 D3DXVECTOR4* WINAPI D3DXVec4TransformArray(D3DXVECTOR4* out, UINT outstride, const D3DXVECTOR4* in, UINT instride, const D3DXMATRIX* matrix, UINT elements)
2116 {
2117 UINT i;
2118
2119 TRACE("out %p, outstride %u, in %p, instride %u, matrix %p, elements %u\n", out, outstride, in, instride, matrix, elements);
2120
2121 for (i = 0; i < elements; ++i) {
2122 D3DXVec4Transform(
2123 (D3DXVECTOR4*)((char*)out + outstride * i),
2124 (const D3DXVECTOR4*)((const char*)in + instride * i),
2125 matrix);
2126 }
2127 return out;
2128 }
2129
2130 unsigned short float_32_to_16(const float in)
2131 {
2132 int exp = 0, origexp;
2133 float tmp = fabs(in);
2134 int sign = (copysignf(1, in) < 0);
2135 unsigned int mantissa;
2136 unsigned short ret;
2137
2138 /* Deal with special numbers */
2139 if (isinf(in)) return (sign ? 0xffff : 0x7fff);
2140 if (isnan(in)) return (sign ? 0xffff : 0x7fff);
2141 if (in == 0.0f) return (sign ? 0x8000 : 0x0000);
2142
2143 if (tmp < powf(2, 10))
2144 {
2145 do
2146 {
2147 tmp *= 2.0f;
2148 exp--;
2149 } while (tmp < powf(2, 10));
2150 }
2151 else if (tmp >= powf(2, 11))
2152 {
2153 do
2154 {
2155 tmp /= 2.0f;
2156 exp++;
2157 } while (tmp >= powf(2, 11));
2158 }
2159
2160 exp += 10; /* Normalize the mantissa */
2161 exp += 15; /* Exponent is encoded with excess 15 */
2162
2163 origexp = exp;
2164
2165 mantissa = (unsigned int) tmp;
2166 if ((tmp - mantissa == 0.5f && mantissa % 2 == 1) || /* round half to even */
2167 (tmp - mantissa > 0.5f))
2168 {
2169 mantissa++; /* round to nearest, away from zero */
2170 }
2171 if (mantissa == 2048)
2172 {
2173 mantissa = 1024;
2174 exp++;
2175 }
2176
2177 if (exp > 31)
2178 {
2179 /* too big */
2180 ret = 0x7fff; /* INF */
2181 }
2182 else if (exp <= 0)
2183 {
2184 unsigned int rounding = 0;
2185
2186 /* Denormalized half float */
2187
2188 /* return 0x0000 (=0.0) for numbers too small to represent in half floats */
2189 if (exp < -11)
2190 return (sign ? 0x8000 : 0x0000);
2191
2192 exp = origexp;
2193
2194 /* the 13 extra bits from single precision are used for rounding */
2195 mantissa = (unsigned int)(tmp * powf(2, 13));
2196 mantissa >>= 1 - exp; /* denormalize */
2197
2198 mantissa -= ~(mantissa >> 13) & 1; /* round half to even */
2199 /* remove 13 least significant bits to get half float precision */
2200 mantissa >>= 12;
2201 rounding = mantissa & 1;
2202 mantissa >>= 1;
2203
2204 ret = mantissa + rounding;
2205 }
2206 else
2207 {
2208 ret = (exp << 10) | (mantissa & 0x3ff);
2209 }
2210
2211 ret |= ((sign ? 1 : 0) << 15); /* Add the sign */
2212 return ret;
2213 }
2214
2215 D3DXFLOAT16 *WINAPI D3DXFloat32To16Array(D3DXFLOAT16 *pout, const FLOAT *pin, UINT n)
2216 {
2217 unsigned int i;
2218
2219 TRACE("pout %p, pin %p, n %u\n", pout, pin, n);
2220
2221 for (i = 0; i < n; ++i)
2222 {
2223 pout[i].value = float_32_to_16(pin[i]);
2224 }
2225
2226 return pout;
2227 }
2228
2229 /* Native d3dx9's D3DXFloat16to32Array lacks support for NaN and Inf. Specifically, e = 16 is treated as a
2230 * regular number - e.g., 0x7fff is converted to 131008.0 and 0xffff to -131008.0. */
2231 static inline float float_16_to_32(const unsigned short in)
2232 {
2233 const unsigned short s = (in & 0x8000);
2234 const unsigned short e = (in & 0x7C00) >> 10;
2235 const unsigned short m = in & 0x3FF;
2236 const float sgn = (s ? -1.0f : 1.0f);
2237
2238 if (e == 0)
2239 {
2240 if (m == 0) return sgn * 0.0f; /* +0.0 or -0.0 */
2241 else return sgn * powf(2, -14.0f) * (m / 1024.0f);
2242 }
2243 else
2244 {
2245 return sgn * powf(2, e - 15.0f) * (1.0f + (m / 1024.0f));
2246 }
2247 }
2248
2249 FLOAT *WINAPI D3DXFloat16To32Array(FLOAT *pout, const D3DXFLOAT16 *pin, UINT n)
2250 {
2251 unsigned int i;
2252
2253 TRACE("pout %p, pin %p, n %u\n", pout, pin, n);
2254
2255 for (i = 0; i < n; ++i)
2256 {
2257 pout[i] = float_16_to_32(pin[i].value);
2258 }
2259
2260 return pout;
2261 }
2262
2263 /*_________________D3DXSH________________*/
2264
2265 FLOAT* WINAPI D3DXSHAdd(FLOAT *out, UINT order, const FLOAT *a, const FLOAT *b)
2266 {
2267 UINT i;
2268
2269 TRACE("out %p, order %u, a %p, b %p\n", out, order, a, b);
2270
2271 for (i = 0; i < order * order; i++)
2272 out[i] = a[i] + b[i];
2273
2274 return out;
2275 }
2276
2277 FLOAT WINAPI D3DXSHDot(UINT order, const FLOAT *a, const FLOAT *b)
2278 {
2279 FLOAT s;
2280 UINT i;
2281
2282 TRACE("order %u, a %p, b %p\n", order, a, b);
2283
2284 s = a[0] * b[0];
2285 for (i = 1; i < order * order; i++)
2286 s += a[i] * b[i];
2287
2288 return s;
2289 }
2290
2291 static void weightedcapintegrale(FLOAT *out, FLOAT order, FLOAT angle)
2292 {
2293 FLOAT coeff[3];
2294
2295 coeff[0] = cosf(angle);
2296
2297 out[0] = 2.0f * D3DX_PI * (1.0f - coeff[0]);
2298 out[1] = D3DX_PI * sinf(angle) * sinf(angle);
2299 if (order <= 2)
2300 return;
2301
2302 out[2] = coeff[0] * out[1];
2303 if (order == 3)
2304 return;
2305
2306 coeff[1] = coeff[0] * coeff[0];
2307 coeff[2] = coeff[1] * coeff[1];
2308
2309 out[3] = D3DX_PI * (-1.25f * coeff[2] + 1.5f * coeff[1] - 0.25f);
2310 if (order == 4)
2311 return;
2312
2313 out[4] = -0.25f * D3DX_PI * coeff[0] * (7.0f * coeff[2] - 10.0f * coeff[1] + 3.0f);
2314 if (order == 5)
2315 return;
2316
2317 out[5] = D3DX_PI * (-2.625f * coeff[2] * coeff[1] + 4.375f * coeff[2] - 1.875f * coeff[1] + 0.125f);
2318 }
2319
2320 HRESULT WINAPI D3DXSHEvalConeLight(UINT order, const D3DXVECTOR3 *dir, FLOAT radius,
2321 FLOAT Rintensity, FLOAT Gintensity, FLOAT Bintensity, FLOAT *rout, FLOAT *gout, FLOAT *bout)
2322 {
2323 FLOAT cap[6], clamped_angle, norm, scale, temp;
2324 UINT i, index, j;
2325
2326 TRACE("order %u, dir %p, radius %f, red %f, green %f, blue %f, rout %p, gout %p, bout %p\n",
2327 order, dir, radius, Rintensity, Gintensity, Bintensity, rout, gout, bout);
2328
2329 if (radius <= 0.0f)
2330 return D3DXSHEvalDirectionalLight(order, dir, Rintensity, Gintensity, Bintensity, rout, gout, bout);
2331
2332 clamped_angle = (radius > D3DX_PI / 2.0f) ? (D3DX_PI / 2.0f) : radius;
2333 norm = sinf(clamped_angle) * sinf(clamped_angle);
2334
2335 if (order > D3DXSH_MAXORDER)
2336 {
2337 WARN("Order clamped at D3DXSH_MAXORDER\n");
2338 order = D3DXSH_MAXORDER;
2339 }
2340
2341 weightedcapintegrale(cap, order, radius);
2342 D3DXSHEvalDirection(rout, order, dir);
2343
2344 for (i = 0; i < order; i++)
2345 {
2346 scale = cap[i] / norm;
2347
2348 for (j = 0; j < 2 * i + 1; j++)
2349 {
2350 index = i * i + j;
2351 temp = rout[index] * scale;
2352
2353 rout[index] = temp * Rintensity;
2354 if (gout)
2355 gout[index] = temp * Gintensity;
2356 if (bout)
2357 bout[index] = temp * Bintensity;
2358 }
2359 }
2360
2361 return D3D_OK;
2362 }
2363
2364 FLOAT* WINAPI D3DXSHEvalDirection(FLOAT *out, UINT order, const D3DXVECTOR3 *dir)
2365 {
2366 const FLOAT dirxx = dir->x * dir->x;
2367 const FLOAT dirxy = dir->x * dir->y;
2368 const FLOAT dirxz = dir->x * dir->z;
2369 const FLOAT diryy = dir->y * dir->y;
2370 const FLOAT diryz = dir->y * dir->z;
2371 const FLOAT dirzz = dir->z * dir->z;
2372 const FLOAT dirxxxx = dirxx * dirxx;
2373 const FLOAT diryyyy = diryy * diryy;
2374 const FLOAT dirzzzz = dirzz * dirzz;
2375 const FLOAT dirxyxy = dirxy * dirxy;
2376
2377 TRACE("out %p, order %u, dir %p\n", out, order, dir);
2378
2379 if ((order < D3DXSH_MINORDER) || (order > D3DXSH_MAXORDER))
2380 return out;
2381
2382 out[0] = 0.5f / sqrtf(D3DX_PI);
2383 out[1] = -0.5f / sqrtf(D3DX_PI / 3.0f) * dir->y;
2384 out[2] = 0.5f / sqrtf(D3DX_PI / 3.0f) * dir->z;
2385 out[3] = -0.5f / sqrtf(D3DX_PI / 3.0f) * dir->x;
2386 if (order == 2)
2387 return out;
2388
2389 out[4] = 0.5f / sqrtf(D3DX_PI / 15.0f) * dirxy;
2390 out[5] = -0.5f / sqrtf(D3DX_PI / 15.0f) * diryz;
2391 out[6] = 0.25f / sqrtf(D3DX_PI / 5.0f) * (3.0f * dirzz - 1.0f);
2392 out[7] = -0.5f / sqrtf(D3DX_PI / 15.0f) * dirxz;
2393 out[8] = 0.25f / sqrtf(D3DX_PI / 15.0f) * (dirxx - diryy);
2394 if (order == 3)
2395 return out;
2396
2397 out[9] = -sqrtf(70.0f / D3DX_PI) / 8.0f * dir->y * (3.0f * dirxx - diryy);
2398 out[10] = sqrtf(105.0f / D3DX_PI) / 2.0f * dirxy * dir->z;
2399 out[11] = -sqrtf(42.0 / D3DX_PI) / 8.0f * dir->y * (-1.0f + 5.0f * dirzz);
2400 out[12] = sqrtf(7.0f / D3DX_PI) / 4.0f * dir->z * (5.0f * dirzz - 3.0f);
2401 out[13] = sqrtf(42.0 / D3DX_PI) / 8.0f * dir->x * (1.0f - 5.0f * dirzz);
2402 out[14] = sqrtf(105.0f / D3DX_PI) / 4.0f * dir->z * (dirxx - diryy);
2403 out[15] = -sqrtf(70.0f / D3DX_PI) / 8.0f * dir->x * (dirxx - 3.0f * diryy);
2404 if (order == 4)
2405 return out;
2406
2407 out[16] = 0.75f * sqrtf(35.0f / D3DX_PI) * dirxy * (dirxx - diryy);
2408 out[17] = 3.0f * dir->z * out[9];
2409 out[18] = 0.75f * sqrtf(5.0f / D3DX_PI) * dirxy * (7.0f * dirzz - 1.0f);
2410 out[19] = 0.375f * sqrtf(10.0f / D3DX_PI) * diryz * (3.0f - 7.0f * dirzz);
2411 out[20] = 3.0f / (16.0f * sqrtf(D3DX_PI)) * (35.0f * dirzzzz - 30.f * dirzz + 3.0f);
2412 out[21] = 0.375f * sqrtf(10.0f / D3DX_PI) * dirxz * (3.0f - 7.0f * dirzz);
2413 out[22] = 0.375f * sqrtf(5.0f / D3DX_PI) * (dirxx - diryy) * (7.0f * dirzz - 1.0f);
2414 out[23] = 3.0 * dir->z * out[15];
2415 out[24] = 3.0f / 16.0f * sqrtf(35.0f / D3DX_PI) * (dirxxxx - 6.0f * dirxyxy + diryyyy);
2416 if (order == 5)
2417 return out;
2418
2419 out[25] = -3.0f/ 32.0f * sqrtf(154.0f / D3DX_PI) * dir->y * (5.0f * dirxxxx - 10.0f * dirxyxy + diryyyy);
2420 out[26] = 0.75f * sqrtf(385.0f / D3DX_PI) * dirxy * dir->z * (dirxx - diryy);
2421 out[27] = sqrtf(770.0f / D3DX_PI) / 32.0f * dir->y * (3.0f * dirxx - diryy) * (1.0f - 9.0f * dirzz);
2422 out[28] = sqrtf(1155.0f / D3DX_PI) / 4.0f * dirxy * dir->z * (3.0f * dirzz - 1.0f);
2423 out[29] = sqrtf(165.0f / D3DX_PI) / 16.0f * dir->y * (14.0f * dirzz - 21.0f * dirzzzz - 1.0f);
2424 out[30] = sqrtf(11.0f / D3DX_PI) / 16.0f * dir->z * (63.0f * dirzzzz - 70.0f * dirzz + 15.0f);
2425 out[31] = sqrtf(165.0f / D3DX_PI) / 16.0f * dir->x * (14.0f * dirzz - 21.0f * dirzzzz - 1.0f);
2426 out[32] = sqrtf(1155.0f / D3DX_PI) / 8.0f * dir->z * (dirxx - diryy) * (3.0f * dirzz - 1.0f);
2427 out[33] = sqrtf(770.0f / D3DX_PI) / 32.0f * dir->x * (dirxx - 3.0f * diryy) * (1.0f - 9.0f * dirzz);
2428 out[34] = 3.0f / 16.0f * sqrtf(385.0f / D3DX_PI) * dir->z * (dirxxxx - 6.0 * dirxyxy + diryyyy);
2429 out[35] = -3.0f/ 32.0f * sqrtf(154.0f / D3DX_PI) * dir->x * (dirxxxx - 10.0f * dirxyxy + 5.0f * diryyyy);
2430
2431 return out;
2432 }
2433
2434 HRESULT WINAPI D3DXSHEvalDirectionalLight(UINT order, const D3DXVECTOR3 *dir, FLOAT Rintensity, FLOAT Gintensity, FLOAT Bintensity, FLOAT *Rout, FLOAT *Gout, FLOAT *Bout)
2435 {
2436 FLOAT s, temp;
2437 UINT j;
2438
2439 TRACE("Order %u, Vector %p, Red %f, Green %f, Blue %f, Rout %p, Gout %p, Bout %p\n", order, dir, Rintensity, Gintensity, Bintensity, Rout, Gout, Bout);
2440
2441 s = 0.75f;
2442 if ( order > 2 )
2443 s += 5.0f / 16.0f;
2444 if ( order > 4 )
2445 s -= 3.0f / 32.0f;
2446 s /= D3DX_PI;
2447
2448 D3DXSHEvalDirection(Rout, order, dir);
2449 for (j = 0; j < order * order; j++)
2450 {
2451 temp = Rout[j] / s;
2452
2453 Rout[j] = Rintensity * temp;
2454 if ( Gout )
2455 Gout[j] = Gintensity * temp;
2456 if ( Bout )
2457 Bout[j] = Bintensity * temp;
2458 }
2459
2460 return D3D_OK;
2461 }
2462
2463 HRESULT WINAPI D3DXSHEvalHemisphereLight(UINT order, const D3DXVECTOR3 *dir, D3DXCOLOR top, D3DXCOLOR bottom,
2464 FLOAT *rout, FLOAT *gout, FLOAT *bout)
2465 {
2466 FLOAT a[2], temp[4];
2467 UINT i, j;
2468
2469 TRACE("order %u, dir %p, rout %p, gout %p, bout %p\n", order, dir, rout, gout, bout);
2470
2471 D3DXSHEvalDirection(temp, 2, dir);
2472
2473 a[0] = (top.r + bottom.r) * 3.0f * D3DX_PI;
2474 a[1] = (top.r - bottom.r) * D3DX_PI;
2475 for (i = 0; i < order; i++)
2476 for (j = 0; j < 2 * i + 1; j++)
2477 if (i < 2)
2478 rout[i * i + j] = temp[i * i + j] * a[i];
2479 else
2480 rout[i * i + j] = 0.0f;
2481
2482 if (gout)
2483 {
2484 a[0] = (top.g + bottom.g) * 3.0f * D3DX_PI;
2485 a[1] = (top.g - bottom.g) * D3DX_PI;
2486 for (i = 0; i < order; i++)
2487 for (j = 0; j < 2 * i + 1; j++)
2488 if (i < 2)
2489 gout[i * i + j] = temp[i * i + j] * a[i];
2490 else
2491 gout[i * i + j] = 0.0f;
2492 }
2493
2494 if (bout)
2495 {
2496 a[0] = (top.b + bottom.b) * 3.0f * D3DX_PI;
2497 a[1] = (top.b - bottom.b) * D3DX_PI;
2498 for (i = 0; i < order; i++)
2499 for (j = 0; j < 2 * i + 1; j++)
2500 if (i < 2)
2501 bout[i * i + j] = temp[i * i + j] * a[i];
2502 else
2503 bout[i * i + j] = 0.0f;
2504 }
2505
2506 return D3D_OK;
2507 }
2508
2509 HRESULT WINAPI D3DXSHEvalSphericalLight(UINT order, const D3DXVECTOR3 *dir, FLOAT radius,
2510 FLOAT Rintensity, FLOAT Gintensity, FLOAT Bintensity, FLOAT *rout, FLOAT *gout, FLOAT *bout)
2511 {
2512 D3DXVECTOR3 normal;
2513 FLOAT cap[6], clamped_angle, dist, temp;
2514 UINT i, index, j;
2515
2516 TRACE("order %u, dir %p, radius %f, red %f, green %f, blue %f, rout %p, gout %p, bout %p\n",
2517 order, dir, radius, Rintensity, Gintensity, Bintensity, rout, gout, bout);
2518
2519 if (order > D3DXSH_MAXORDER)
2520 {
2521 WARN("Order clamped at D3DXSH_MAXORDER\n");
2522 order = D3DXSH_MAXORDER;
2523 }
2524
2525 if (radius < 0.0f)
2526 radius = -radius;
2527
2528 dist = D3DXVec3Length(dir);
2529 clamped_angle = (dist <= radius) ? D3DX_PI / 2.0f : asinf(radius / dist);
2530
2531 weightedcapintegrale(cap, order, clamped_angle);
2532 D3DXVec3Normalize(&normal, dir);
2533 D3DXSHEvalDirection(rout, order, &normal);
2534
2535 for (i = 0; i < order; i++)
2536 for (j = 0; j < 2 * i + 1; j++)
2537 {
2538 index = i * i + j;
2539 temp = rout[index] * cap[i];
2540
2541 rout[index] = temp * Rintensity;
2542 if (gout)
2543 gout[index] = temp * Gintensity;
2544 if (bout)
2545 bout[index] = temp * Bintensity;
2546 }
2547
2548 return D3D_OK;
2549 }
2550
2551 FLOAT * WINAPI D3DXSHMultiply2(FLOAT *out, const FLOAT *a, const FLOAT *b)
2552 {
2553 FLOAT ta, tb;
2554
2555 TRACE("out %p, a %p, b %p\n", out, a, b);
2556
2557 ta = 0.28209479f * a[0];
2558 tb = 0.28209479f * b[0];
2559
2560 out[0] = 0.28209479f * D3DXSHDot(2, a, b);
2561 out[1] = ta * b[1] + tb * a[1];
2562 out[2] = ta * b[2] + tb * a[2];
2563 out[3] = ta * b[3] + tb * a[3];
2564
2565 return out;
2566 }
2567
2568 FLOAT * WINAPI D3DXSHMultiply3(FLOAT *out, const FLOAT *a, const FLOAT *b)
2569 {
2570 FLOAT t, ta, tb;
2571
2572 TRACE("out %p, a %p, b %p\n", out, a, b);
2573
2574 out[0] = 0.28209479f * a[0] * b[0];
2575
2576 ta = 0.28209479f * a[0] - 0.12615662f * a[6] - 0.21850968f * a[8];
2577 tb = 0.28209479f * b[0] - 0.12615662f * b[6] - 0.21850968f * b[8];
2578 out[1] = ta * b[1] + tb * a[1];
2579 t = a[1] * b[1];
2580 out[0] += 0.28209479f * t;
2581 out[6] = -0.12615662f * t;
2582 out[8] = -0.21850968f * t;
2583
2584 ta = 0.21850968f * a[5];
2585 tb = 0.21850968f * b[5];
2586 out[1] += ta * b[2] + tb * a[2];
2587 out[2] = ta * b[1] + tb * a[1];
2588 t = a[1] * b[2] +a[2] * b[1];
2589 out[5] = 0.21850968f * t;
2590
2591 ta = 0.21850968f * a[4];
2592 tb = 0.21850968f * b[4];
2593 out[1] += ta * b[3] + tb * a[3];
2594 out[3] = ta * b[1] + tb * a[1];
2595 t = a[1] * b[3] + a[3] * b[1];
2596 out[4] = 0.21850968f * t;
2597
2598 ta = 0.28209480f * a[0] + 0.25231326f * a[6];
2599 tb = 0.28209480f * b[0] + 0.25231326f * b[6];
2600 out[2] += ta * b[2] + tb * a[2];
2601 t = a[2] * b[2];
2602 out[0] += 0.28209480f * t;
2603 out[6] += 0.25231326f * t;
2604
2605 ta = 0.21850969f * a[7];
2606 tb = 0.21850969f * b[7];
2607 out[2] += ta * b[3] + tb * a[3];
2608 out[3] += ta * b[2] + tb * a[2];
2609 t = a[2] * b[3] + a[3] * b[2];
2610 out[7] = 0.21850969f * t;
2611
2612 ta = 0.28209479f * a[0] - 0.12615663f * a[6] + 0.21850969f * a[8];
2613 tb = 0.28209479f * b[0] - 0.12615663f * b[6] + 0.21850969f * b[8];
2614 out[3] += ta * b[3] + tb * a[3];
2615 t = a[3] * b[3];
2616 out[0] += 0.28209479f * t;
2617 out[6] -= 0.12615663f * t;
2618 out[8] += 0.21850969f * t;
2619
2620 ta = 0.28209479f * a[0] - 0.18022375f * a[6];
2621 tb = 0.28209479f * b[0] - 0.18022375f * b[6];
2622 out[4] += ta * b[4] + tb * a[4];
2623 t = a[4] * b[4];
2624 out[0] += 0.28209479f * t;
2625 out[6] -= 0.18022375f * t;
2626
2627 ta = 0.15607835f * a[7];
2628 tb = 0.15607835f * b[7];
2629 out[4] += ta * b[5] + tb * a[5];
2630 out[5] += ta * b[4] + tb * a[4];
2631 t = a[4] * b[5] + a[5] * b[4];
2632 out[7] += 0.15607834f * t;
2633
2634 ta = 0.28209479f * a[0] + 0.09011186 * a[6] - 0.15607835f * a[8];
2635 tb = 0.28209479f * b[0] + 0.09011186 * b[6] - 0.15607835f * b[8];
2636 out[5] += ta * b[5] + tb * a[5];
2637 t = a[5] * b[5];
2638 out[0] += 0.28209479f * t;
2639 out[6] += 0.09011186f * t;
2640 out[8] -= 0.15607835f * t;
2641
2642 ta = 0.28209480f * a[0];
2643 tb = 0.28209480f * b[0];
2644 out[6] += ta * b[6] + tb * a[6];
2645 t = a[6] * b[6];
2646 out[0] += 0.28209480f * t;
2647 out[6] += 0.18022376f * t;
2648
2649 ta = 0.28209479f * a[0] + 0.09011186 * a[6] + 0.15607835f * a[8];
2650 tb = 0.28209479f * b[0] + 0.09011186 * b[6] + 0.15607835f * b[8];
2651 out[7] += ta * b[7] + tb * a[7];
2652 t = a[7] * b[7];
2653 out[0] += 0.28209479f * t;
2654 out[6] += 0.09011186f * t;
2655 out[8] += 0.15607835f * t;
2656
2657 ta = 0.28209479f * a[0] - 0.18022375f * a[6];
2658 tb = 0.28209479f * b[0] - 0.18022375f * b[6];
2659 out[8] += ta * b[8] + tb * a[8];
2660 t = a[8] * b[8];
2661 out[0] += 0.28209479f * t;
2662 out[6] -= 0.18022375f * t;
2663
2664 return out;
2665 }
2666
2667 FLOAT * WINAPI D3DXSHMultiply4(FLOAT *out, const FLOAT *a, const FLOAT *b)
2668 {
2669 FLOAT ta, tb, t;
2670
2671 TRACE("out %p, a %p, b %p\n", out, a, b);
2672
2673 out[0] = 0.28209479f * a[0] * b[0];
2674
2675 ta = 0.28209479f * a[0] - 0.12615663f * a[6] - 0.21850969f * a[8];
2676 tb = 0.28209479f * b[0] - 0.12615663f * b[6] - 0.21850969f * b[8];
2677 out[1] = ta * b[1] + tb * a[1];
2678 t = a[1] * b[1];
2679 out[0] += 0.28209479f * t;
2680 out[6] = -0.12615663f * t;
2681 out[8] = -0.21850969f * t;
2682
2683 ta = 0.21850969f * a[3] - 0.05839917f * a[13] - 0.22617901f * a[15];
2684 tb = 0.21850969f * b[3] - 0.05839917f * b[13] - 0.22617901f * b[15];
2685 out[1] += ta * b[4] + tb * a[4];
2686 out[4] = ta * b[1] + tb * a[1];
2687 t = a[1] * b[4] + a[4] * b[1];
2688 out[3] = 0.21850969f * t;
2689 out[13] = -0.05839917f * t;
2690 out[15] = -0.22617901f * t;
2691
2692 ta = 0.21850969f * a[2] - 0.14304817f * a[12] - 0.18467439f * a[14];
2693 tb = 0.21850969f * b[2] - 0.14304817f * b[12] - 0.18467439f * b[14];
2694 out[1] += ta * b[5] + tb * a[5];
2695 out[5] = ta * b[1] + tb * a[1];
2696 t = a[1] * b[5] + a[5] * b[1];
2697 out[2] = 0.21850969f * t;
2698 out[12] = -0.14304817f * t;
2699 out[14] = -0.18467439f * t;
2700
2701 ta = 0.20230066f * a[11];
2702 tb = 0.20230066f * b[11];
2703 out[1] += ta * b[6] + tb * a[6];
2704 out[6] += ta * b[1] + tb * a[1];
2705 t = a[1] * b[6] + a[6] * b[1];
2706 out[11] = 0.20230066f * t;
2707
2708 ta = 0.22617901f * a[9] + 0.05839917f * a[11];
2709 tb = 0.22617901f * b[9] + 0.05839917f * b[11];
2710 out[1] += ta * b[8] + tb * a[8];
2711 out[8] += ta * b[1] + tb * a[1];
2712 t = a[1] * b[8] + a[8] * b[1];
2713 out[9] = 0.22617901f * t;
2714 out[11] += 0.05839917f * t;
2715
2716 ta = 0.28209480f * a[0] + 0.25231326f * a[6];
2717 tb = 0.28209480f * b[0] + 0.25231326f * b[6];
2718 out[2] += ta * b[2] + tb * a[2];
2719 t = a[2] * b[2];
2720 out[0] += 0.28209480f * t;
2721 out[6] += 0.25231326f * t;
2722
2723 ta = 0.24776671f * a[12];
2724 tb = 0.24776671f * b[12];
2725 out[2] += ta * b[6] + tb * a[6];
2726 out[6] += ta * b[2] + tb * a[2];
2727 t = a[2] * b[6] + a[6] * b[2];
2728 out[12] += 0.24776671f * t;
2729
2730 ta = 0.28209480f * a[0] - 0.12615663f * a[6] + 0.21850969f * a[8];
2731 tb = 0.28209480f * b[0] - 0.12615663f * b[6] + 0.21850969f * b[8];
2732 out[3] += ta * b[3] + tb * a[3];
2733 t = a[3] * b[3];
2734 out[0] += 0.28209480f * t;
2735 out[6] -= 0.12615663f * t;
2736 out[8] += 0.21850969f * t;
2737
2738 ta = 0.20230066f * a[13];
2739 tb = 0.20230066f * b[13];
2740 out[3] += ta * b[6] + tb * a[6];
2741 out[6] += ta * b[3] + tb * a[3];
2742 t = a[3] * b[6] + a[6] * b[3];
2743 out[13] += 0.20230066f * t;
2744
2745 ta = 0.21850969f * a[2] - 0.14304817f * a[12] + 0.18467439f * a[14];
2746 tb = 0.21850969f * b[2] - 0.14304817f * b[12] + 0.18467439f * b[14];
2747 out[3] += ta * b[7] + tb * a[7];
2748 out[7] = ta * b[3] + tb * a[3];
2749 t = a[3] * b[7] + a[7] * b[3];
2750 out[2] += 0.21850969f * t;
2751 out[12] -= 0.14304817f * t;
2752 out[14] += 0.18467439f * t;
2753
2754 ta = -0.05839917f * a[13] + 0.22617901f * a[15];
2755 tb = -0.05839917f * b[13] + 0.22617901f * b[15];
2756 out[3] += ta * b[8] + tb * a[8];
2757 out[8] += ta * b[3] + tb * a[3];
2758 t = a[3] * b[8] + a[8] * b[3];
2759 out[13] -= 0.05839917f * t;
2760 out[15] += 0.22617901f * t;
2761
2762 ta = 0.28209479f * a[0] - 0.18022375f * a[6];
2763 tb = 0.28209479f * b[0] - 0.18022375f * b[6];
2764 out[4] += ta * b[4] + tb * a[4];
2765 t = a[4] * b[4];
2766 out[0] += 0.28209479f * t;
2767 out[6] -= 0.18022375f * t;
2768
2769 ta = 0.15607835f * a[7];
2770 tb = 0.15607835f * b[7];
2771 out[4] += ta * b[5] + tb * a[5];
2772 out[5] += ta * b[4] + tb * a[4];
2773 t = a[4] * b[5] + a[5] * b[4];
2774 out[7] += 0.15607835f * t;
2775
2776 ta = 0.22617901f * a[3] - 0.09403160f * a[13];
2777 tb = 0.22617901f * b[3] - 0.09403160f * b[13];
2778 out[4] += ta * b[9] + tb * a[9];
2779 out[9] += ta * b[4] + tb * a[4];
2780 t = a[4] * b[9] + a[9] * b[4];
2781 out[3] += 0.22617901f * t;
2782 out[13] -= 0.09403160f * t;
2783
2784 ta = 0.18467439f * a[2] - 0.18806319f * a[12];
2785 tb = 0.18467439f * b[2] - 0.18806319f * b[12];
2786 out[4] += ta * b[10] + tb * a [10];
2787 out[10] = ta * b[4] + tb * a[4];
2788 t = a[4] * b[10] + a[10] * b[4];
2789 out[2] += 0.18467439f * t;
2790 out[12] -= 0.18806319f * t;
2791
2792 ta = -0.05839917f * a[3] + 0.14567312f * a[13] + 0.09403160f * a[15];
2793 tb = -0.05839917f * b[3] + 0.14567312f * b[13] + 0.09403160f * b[15];
2794 out[4] += ta * b[11] + tb * a[11];
2795 out[11] += ta * b[4] + tb * a[4];
2796 t = a[4] * b[11] + a[11] * b[4];
2797 out[3] -= 0.05839917f * t;
2798 out[13] += 0.14567312f * t;
2799 out[15] += 0.09403160f * t;
2800
2801 ta = 0.28209479f * a[0] + 0.09011186f * a[6] - 0.15607835f * a[8];
2802 tb = 0.28209479f * b[0] + 0.09011186f * b[6] - 0.15607835f * b[8];
2803 out[5] += ta * b[5] + tb * a[5];
2804 t = a[5] * b[5];
2805 out[0] += 0.28209479f * t;
2806 out[6] += 0.09011186f * t;
2807 out[8] -= 0.15607835f * t;
2808
2809 ta = 0.14867701f * a[14];
2810 tb = 0.14867701f * b[14];
2811 out[5] += ta * b[9] + tb * a[9];
2812 out[9] += ta * b[5] + tb * a[5];
2813 t = a[5] * b[9] + a[9] * b[5];
2814 out[14] += 0.14867701f * t;
2815
2816 ta = 0.18467439f * a[3] + 0.11516472f * a[13] - 0.14867701f * a[15];
2817 tb = 0.18467439f * b[3] + 0.11516472f * b[13] - 0.14867701f * b[15];
2818 out[5] += ta * b[10] + tb * a[10];
2819 out[10] += ta * b[5] + tb * a[5];
2820 t = a[5] * b[10] + a[10] * b[5];
2821 out[3] += 0.18467439f * t;
2822 out[13] += 0.11516472f * t;
2823 out[15] -= 0.14867701f * t;
2824
2825 ta = 0.23359668f * a[2] + 0.05947080f * a[12] - 0.11516472f * a[14];
2826 tb = 0.23359668f * b[2] + 0.05947080f * b[12] - 0.11516472f * b[14];
2827 out[5] += ta * b[11] + tb * a[11];
2828 out[11] += ta * b[5] + tb * a[5];
2829 t = a[5] * b[11] + a[11] * b[5];
2830 out[2] += 0.23359668f * t;
2831 out[12] += 0.05947080f * t;
2832 out[14] -= 0.11516472f * t;
2833
2834 ta = 0.28209479f * a[0];
2835 tb = 0.28209479f * b[0];
2836 out[6] += ta * b[6] + tb * a[6];
2837 t = a[6] * b[6];
2838 out[0] += 0.28209479f * t;
2839 out[6] += 0.18022376f * t;
2840
2841 ta = 0.09011186f * a[6] + 0.28209479f * a[0] + 0.15607835f * a[8];
2842 tb = 0.09011186f * b[6] + 0.28209479f * b[0] + 0.15607835f * b[8];
2843 out[7] += ta * b[7] + tb * a[7];
2844 t = a[7] * b[7];
2845 out[6] += 0.09011186f * t;
2846 out[0] += 0.28209479f * t;
2847 out[8] += 0.15607835f * t;
2848
2849 ta = 0.14867701f * a[9] + 0.18467439f * a[1] + 0.11516472f * a[11];
2850 tb = 0.14867701f * b[9] + 0.18467439f * b[1] + 0.11516472f * b[11];
2851 out[7] += ta * b[10] + tb * a[10];
2852 out[10] += ta * b[7] + tb * a[7];
2853 t = a[7] * b[10] + a[10] * b[7];
2854 out[9] += 0.14867701f * t;
2855 out[1] += 0.18467439f * t;
2856 out[11] += 0.11516472f * t;
2857
2858 ta = 0.05947080f * a[12] + 0.23359668f * a[2] + 0.11516472f * a[14];
2859 tb = 0.05947080f * b[12] + 0.23359668f * b[2] + 0.11516472f * b[14];
2860 out[7] += ta * b[13] + tb * a[13];
2861 out[13] += ta * b[7]+ tb * a[7];
2862 t = a[7] * b[13] + a[13] * b[7];
2863 out[12] += 0.05947080f * t;
2864 out[2] += 0.23359668f * t;
2865 out[14] += 0.11516472f * t;
2866
2867 ta = 0.14867701f * a[15];
2868 tb = 0.14867701f * b[15];
2869 out[7] += ta * b[14] + tb * a[14];
2870 out[14] += ta * b[7] + tb * a[7];
2871 t = a[7] * b[14] + a[14] * b[7];
2872 out[15] += 0.14867701f * t;
2873
2874 ta = 0.28209479f * a[0] - 0.18022375f * a[6];
2875 tb = 0.28209479f * b[0] - 0.18022375f * b[6];
2876 out[8] += ta * b[8] + tb * a[8];
2877 t = a[8] * b[8];
2878 out[0] += 0.28209479f * t;
2879 out[6] -= 0.18022375f * t;
2880
2881 ta = -0.09403160f * a[11];
2882 tb = -0.09403160f * b[11];
2883 out[8] += ta * b[9] + tb * a[9];
2884 out[9] += ta * b[8] + tb * a[8];
2885 t = a[8] * b[9] + a[9] * b[8];
2886 out[11] -= 0.09403160f * t;
2887
2888 ta = -0.09403160f * a[15];
2889 tb = -0.09403160f * b[15];
2890 out[8] += ta * b[13] + tb * a[13];
2891 out[13] += ta * b[8] + tb * a[8];
2892 t = a[8] * b[13] + a[13] * b[8];
2893 out[15] -= 0.09403160f * t;
2894
2895 ta = 0.18467439f * a[2] - 0.18806319f * a[12];
2896 tb = 0.18467439f * b[2] - 0.18806319f * b[12];
2897 out[8] += ta * b[14] + tb * a[14];
2898 out[14] += ta * b[8] + tb * a[8];
2899 t = a[8] * b[14] + a[14] * b[8];
2900 out[2] += 0.18467439f * t;
2901 out[12] -= 0.18806319f * t;
2902
2903 ta = -0.21026104f * a[6] + 0.28209479f * a[0];
2904 tb = -0.21026104f * b[6] + 0.28209479f * b[0];
2905 out[9] += ta * b[9] + tb * a[9];
2906 t = a[9] * b[9];
2907 out[6] -= 0.21026104f * t;
2908 out[0] += 0.28209479f * t;
2909
2910 ta = 0.28209479f * a[0];
2911 tb = 0.28209479f * b[0];
2912 out[10] += ta * b[10] + tb * a[10];
2913 t = a[10] * b[10];
2914 out[0] += 0.28209479f * t;
2915
2916 ta = 0.28209479f * a[0] + 0.12615663f * a[6] - 0.14567312f * a[8];
2917 tb = 0.28209479f * b[0] + 0.12615663f * b[6] - 0.14567312f * b[8];
2918 out[11] += ta * b[11] + tb * a[11];
2919 t = a[11] * b[11];
2920 out[0] += 0.28209479f * t;
2921 out[6] += 0.12615663f * t;
2922 out[8] -= 0.14567312f * t;
2923
2924 ta = 0.28209479f * a[0] + 0.16820885f * a[6];
2925 tb = 0.28209479f * b[0] + 0.16820885f * b[6];
2926 out[12] += ta * b[12] + tb * a[12];
2927 t = a[12] * b[12];
2928 out[0] += 0.28209479f * t;
2929 out[6] += 0.16820885f * t;
2930
2931 ta =0.28209479f * a[0] + 0.14567312f * a[8] + 0.12615663f * a[6];
2932 tb =0.28209479f * b[0] + 0.14567312f * b[8] + 0.12615663f * b[6];
2933 out[13] += ta * b[13] + tb * a[13];
2934 t = a[13] * b[13];
2935 out[0] += 0.28209479f * t;
2936 out[8] += 0.14567312f * t;
2937 out[6] += 0.12615663f * t;
2938
2939 ta = 0.28209479f * a[0];
2940 tb = 0.28209479f * b[0];
2941 out[14] += ta * b[14] + tb * a[14];
2942 t = a[14] * b[14];
2943 out[0] += 0.28209479f * t;
2944
2945 ta = 0.28209479f * a[0] - 0.21026104f * a[6];
2946 tb = 0.28209479f * b[0] - 0.21026104f * b[6];
2947 out[15] += ta * b[15] + tb * a[15];
2948 t = a[15] * b[15];
2949 out[0] += 0.28209479f * t;
2950 out[6] -= 0.21026104f * t;
2951
2952 return out;
2953 }
2954
2955 static void rotate_X(FLOAT *out, UINT order, FLOAT a, FLOAT *in)
2956 {
2957 out[0] = in[0];
2958
2959 out[1] = a * in[2];
2960 out[2] = -a * in[1];
2961 out[3] = in[3];
2962
2963 out[4] = a * in[7];
2964 out[5] = -in[5];
2965 out[6] = -0.5f * in[6] - 0.8660253882f * in[8];
2966 out[7] = -a * in[4];
2967 out[8] = -0.8660253882f * in[6] + 0.5f * in[8];
2968 out[9] = -a * 0.7905694842f * in[12] + a * 0.6123724580f * in[14];
2969
2970 out[10] = -in[10];
2971 out[11] = -a * 0.6123724580f * in[12] - a * 0.7905694842f * in[14];
2972 out[12] = a * 0.7905694842f * in[9] + a * 0.6123724580f * in[11];
2973 out[13] = -0.25f * in[13] - 0.9682458639f * in[15];
2974 out[14] = -a * 0.6123724580f * in[9] + a * 0.7905694842f * in[11];
2975 out[15] = -0.9682458639f * in[13] + 0.25f * in[15];
2976 if (order == 4)
2977 return;
2978
2979 out[16] = -a * 0.9354143739f * in[21] + a * 0.3535533845f * in[23];
2980 out[17] = -0.75f * in[17] + 0.6614378095f * in[19];
2981 out[18] = -a * 0.3535533845f * in[21] - a * 0.9354143739f * in[23];
2982 out[19] = 0.6614378095f * in[17] + 0.75f * in[19];
2983 out[20] = 0.375f * in[20] + 0.5590170026f * in[22] + 0.7395099998f * in[24];
2984 out[21] = a * 0.9354143739f * in[16] + a * 0.3535533845f * in[18];
2985 out[22] = 0.5590170026f * in[20] + 0.5f * in[22] - 0.6614378691f * in[24];
2986 out[23] = -a * 0.3535533845f * in[16] + a * 0.9354143739f * in[18];
2987 out[24] = 0.7395099998f * in[20] - 0.6614378691f * in[22] + 0.125f * in[24];
2988 if (order == 5)
2989 return;
2990
2991 out[25] = a * 0.7015607357f * in[30] - a * 0.6846531630f * in[32] + a * 0.1976423711f * in[34];
2992 out[26] = -0.5f * in[26] + 0.8660253882f * in[28];
2993 out[27] = a * 0.5229125023f * in[30] + a * 0.3061861992f * in[32] - a * 0.7954951525 * in[34];
2994 out[28] = 0.8660253882f * in[26] + 0.5f * in[28];
2995 out[29] = a * 0.4841229022f * in[30] + a * 0.6614378691f * in[32] + a * 0.5728219748f * in[34];
2996 out[30] = -a * 0.7015607357f * in[25] - a * 0.5229125023f * in[27] - a * 0.4841229022f * in[29];
2997 out[31] = 0.125f * in[31] + 0.4050463140f * in[33] + 0.9057110548f * in[35];
2998 out[32] = a * 0.6846531630f * in[25] - a * 0.3061861992f * in[27] - a * 0.6614378691f * in[29];
2999 out[33] = 0.4050463140f * in[31] + 0.8125f * in[33] - 0.4192627370f * in[35];
3000 out[34] = -a * 0.1976423711f * in[25] + a * 0.7954951525f * in[27] - a * 0.5728219748f * in[29];
3001 out[35] = 0.9057110548f * in[31] - 0.4192627370f * in[33] + 0.0624999329f * in[35];
3002 }
3003
3004 FLOAT* WINAPI D3DXSHRotate(FLOAT *out, UINT order, const D3DXMATRIX *matrix, const FLOAT *in)
3005 {
3006 FLOAT alpha, beta, gamma, sinb, temp[36], temp1[36];
3007
3008 TRACE("out %p, order %u, matrix %p, in %p\n", out, order, matrix, in);
3009
3010 out[0] = in[0];
3011
3012 if ((order > D3DXSH_MAXORDER) || (order < D3DXSH_MINORDER))
3013 return out;
3014
3015 if (order <= 3)
3016 {
3017 out[1] = matrix->u.m[1][1] * in[1] - matrix->u.m[2][1] * in[2] + matrix->u.m[0][1] * in[3];
3018 out[2] = -matrix->u.m[1][2] * in[1] + matrix->u.m[2][2] * in[2] - matrix->u.m[0][2] * in[3];
3019 out[3] = matrix->u.m[1][0] * in[1] - matrix->u.m[2][0] * in[2] + matrix->u.m[0][0] * in[3];
3020
3021 if (order == 3)
3022 {
3023 FLOAT coeff[]={
3024 matrix->u.m[1][0] * matrix->u.m[0][0], matrix->u.m[1][1] * matrix->u.m[0][1],
3025 matrix->u.m[1][1] * matrix->u.m[2][1], matrix->u.m[1][0] * matrix->u.m[2][0],
3026 matrix->u.m[2][0] * matrix->u.m[2][0], matrix->u.m[2][1] * matrix->u.m[2][1],
3027 matrix->u.m[0][0] * matrix->u.m[2][0], matrix->u.m[0][1] * matrix->u.m[2][1],
3028 matrix->u.m[0][1] * matrix->u.m[0][1], matrix->u.m[1][0] * matrix->u.m[1][0],
3029 matrix->u.m[1][1] * matrix->u.m[1][1], matrix->u.m[0][0] * matrix->u.m[0][0], };
3030
3031 out[4] = (matrix->u.m[1][1] * matrix->u.m[0][0] + matrix->u.m[0][1] * matrix->u.m[1][0]) * in[4];
3032 out[4] -= (matrix->u.m[1][0] * matrix->u.m[2][1] + matrix->u.m[1][1] * matrix->u.m[2][0]) * in[5];
3033 out[4] += 1.7320508076f * matrix->u.m[2][0] * matrix->u.m[2][1] * in[6];
3034 out[4] -= (matrix->u.m[0][1] * matrix->u.m[2][0] + matrix->u.m[0][0] * matrix->u.m[2][1]) * in[7];
3035 out[4] += (matrix->u.m[0][0] * matrix->u.m[0][1] - matrix->u.m[1][0] * matrix->u.m[1][1]) * in[8];
3036
3037 out[5] = (matrix->u.m[1][1] * matrix->u.m[2][2] + matrix->u.m[1][2] * matrix->u.m[2][1]) * in[5];
3038 out[5] -= (matrix->u.m[1][1] * matrix->u.m[0][2] + matrix->u.m[1][2] * matrix->u.m[0][1]) * in[4];
3039 out[5] -= 1.7320508076f * matrix->u.m[2][2] * matrix->u.m[2][1] * in[6];
3040 out[5] += (matrix->u.m[0][2] * matrix->u.m[2][1] + matrix->u.m[0][1] * matrix->u.m[2][2]) * in[7];
3041 out[5] -= (matrix->u.m[0][1] * matrix->u.m[0][2] - matrix->u.m[1][1] * matrix->u.m[1][2]) * in[8];
3042
3043 out[6] = (matrix->u.m[2][2] * matrix->u.m[2][2] - 0.5f * (coeff[4] + coeff[5])) * in[6];
3044 out[6] -= (0.5773502692f * (coeff[0] + coeff[1]) - 1.1547005384f * matrix->u.m[1][2] * matrix->u.m[0][2]) * in[4];
3045 out[6] += (0.5773502692f * (coeff[2] + coeff[3]) - 1.1547005384f * matrix->u.m[1][2] * matrix->u.m[2][2]) * in[5];
3046 out[6] += (0.5773502692f * (coeff[6] + coeff[7]) - 1.1547005384f * matrix->u.m[0][2] * matrix->u.m[2][2]) * in[7];
3047 out[6] += (0.2886751347f * (coeff[9] - coeff[8] + coeff[10] - coeff[11]) - 0.5773502692f *
3048 (matrix->u.m[1][2] * matrix->u.m[1][2] - matrix->u.m[0][2] * matrix->u.m[0][2])) * in[8];
3049
3050 out[7] = (matrix->u.m[0][0] * matrix->u.m[2][2] + matrix->u.m[0][2] * matrix->u.m[2][0]) * in[7];
3051 out[7] -= (matrix->u.m[1][0] * matrix->u.m[0][2] + matrix->u.m[1][2] * matrix->u.m[0][0]) * in[4];
3052 out[7] += (matrix->u.m[1][0] * matrix->u.m[2][2] + matrix->u.m[1][2] * matrix->u.m[2][0]) * in[5];
3053 out[7] -= 1.7320508076f * matrix->u.m[2][2] * matrix->u.m[2][0] * in[6];
3054 out[7] -= (matrix->u.m[0][0] * matrix->u.m[0][2] - matrix->u.m[1][0] * matrix->u.m[1][2]) * in[8];
3055
3056 out[8] = 0.5f * (coeff[11] - coeff[8] - coeff[9] + coeff[10]) * in[8];
3057 out[8] += (coeff[0] - coeff[1]) * in[4];
3058 out[8] += (coeff[2] - coeff[3]) * in[5];
3059 out[8] += 0.86602540f * (coeff[4] - coeff[5]) * in[6];
3060 out[8] += (coeff[7] - coeff[6]) * in[7];
3061 }
3062
3063 return out;
3064 }
3065
3066 if (fabsf(matrix->u.m[2][2]) != 1.0f)
3067 {
3068 sinb = sqrtf(1.0f - matrix->u.m[2][2] * matrix->u.m[2][2]);
3069 alpha = atan2f(matrix->u.m[2][1] / sinb, matrix->u.m[2][0] / sinb);
3070 beta = atan2f(sinb, matrix->u.m[2][2]);
3071 gamma = atan2f(matrix->u.m[1][2] / sinb, -matrix->u.m[0][2] / sinb);
3072 }
3073 else
3074 {
3075 alpha = atan2f(matrix->u.m[0][1], matrix->u.m[0][0]);
3076 beta = 0.0f;
3077 gamma = 0.0f;
3078 }
3079
3080 D3DXSHRotateZ(temp, order, gamma, in);
3081 rotate_X(temp1, order, 1.0f, temp);
3082 D3DXSHRotateZ(temp, order, beta, temp1);
3083 rotate_X(temp1, order, -1.0f, temp);
3084 D3DXSHRotateZ(out, order, alpha, temp1);
3085
3086 return out;
3087 }
3088
3089 FLOAT * WINAPI D3DXSHRotateZ(FLOAT *out, UINT order, FLOAT angle, const FLOAT *in)
3090 {
3091 UINT i, sum = 0;
3092 FLOAT c[5], s[5];
3093
3094 TRACE("out %p, order %u, angle %f, in %p\n", out, order, angle, in);
3095
3096 order = min(max(order, D3DXSH_MINORDER), D3DXSH_MAXORDER);
3097
3098 out[0] = in[0];
3099
3100 for (i = 1; i < order; i++)
3101 {
3102 UINT j;
3103
3104 c[i - 1] = cosf(i * angle);
3105 s[i - 1] = sinf(i * angle);
3106 sum += i * 2;
3107
3108 out[sum - i] = c[i - 1] * in[sum - i];
3109 out[sum - i] += s[i - 1] * in[sum + i];
3110 for (j = i - 1; j > 0; j--)
3111 {
3112 out[sum - j] = 0.0f;
3113 out[sum - j] = c[j - 1] * in[sum - j];
3114 out[sum - j] += s[j - 1] * in[sum + j];
3115 }
3116
3117 if (in == out)
3118 out[sum] = 0.0f;
3119 else
3120 out[sum] = in[sum];
3121
3122 for (j = 1; j < i; j++)
3123 {
3124 out[sum + j] = 0.0f;
3125 out[sum + j] = -s[j - 1] * in[sum - j];
3126 out[sum + j] += c[j - 1] * in[sum + j];
3127 }
3128 out[sum + i] = -s[i - 1] * in[sum - i];
3129 out[sum + i] += c[i - 1] * in[sum + i];
3130 }
3131
3132 return out;
3133 }
3134
3135 FLOAT* WINAPI D3DXSHScale(FLOAT *out, UINT order, const FLOAT *a, const FLOAT scale)
3136 {
3137 UINT i;
3138
3139 TRACE("out %p, order %u, a %p, scale %f\n", out, order, a, scale);
3140
3141 for (i = 0; i < order * order; i++)
3142 out[i] = a[i] * scale;
3143
3144 return out;
3145 }