1 | unit BGRAMatrix3D;
|
---|
2 |
|
---|
3 | {$mode objfpc}{$H+}
|
---|
4 |
|
---|
5 | {$i bgrasse.inc}
|
---|
6 |
|
---|
7 | {$ifdef CPUI386}
|
---|
8 | {$asmmode intel}
|
---|
9 | {$ENDIF}
|
---|
10 | {$ifdef cpux86_64}
|
---|
11 | {$asmmode intel}
|
---|
12 | {$ENDIF}
|
---|
13 |
|
---|
14 | interface
|
---|
15 |
|
---|
16 | uses
|
---|
17 | BGRABitmapTypes, BGRASSE,
|
---|
18 | BGRATransform;
|
---|
19 |
|
---|
20 | type
|
---|
21 | TMatrix3D = packed array[1..3,1..4] of single;
|
---|
22 | TMatrix4D = packed array[1..4,1..4] of single;
|
---|
23 | TProjection3D = packed record
|
---|
24 | Zoom, Center: TPointF;
|
---|
25 | end;
|
---|
26 | TComputeProjectionFunc = function(AViewCoord: TPoint3D_128): TPointF of object;
|
---|
27 |
|
---|
28 | operator*(const A: TMatrix3D; const M: TPoint3D): TPoint3D;
|
---|
29 | operator*(constref A: TMatrix3D; var M: TPoint3D_128): TPoint3D_128;
|
---|
30 | function MultiplyVect3DWithoutTranslation(constref A: TMatrix3D; constref M: TPoint3D_128): TPoint3D_128;
|
---|
31 | operator*(A,B: TMatrix3D): TMatrix3D;
|
---|
32 |
|
---|
33 | function Matrix3D(m11,m12,m13,m14, m21,m22,m23,m24, m31,m32,m33,m34: single): TMatrix3D; overload;
|
---|
34 | function Matrix3D(vx,vy,vz,ofs: TPoint3D): TMatrix3D; overload;
|
---|
35 | function Matrix3D(vx,vy,vz,ofs: TPoint3D_128): TMatrix3D; overload;
|
---|
36 | function MatrixIdentity3D: TMatrix3D;
|
---|
37 | function MatrixInverse3D(A: TMatrix3D): TMatrix3D;
|
---|
38 | function MatrixTranslation3D(ofs: TPoint3D): TMatrix3D;
|
---|
39 | function MatrixScale3D(size: TPoint3D): TMatrix3D;
|
---|
40 | function MatrixRotateX(angle: single): TMatrix3D;
|
---|
41 | function MatrixRotateY(angle: single): TMatrix3D;
|
---|
42 | function MatrixRotateZ(angle: single): TMatrix3D;
|
---|
43 |
|
---|
44 | operator *(const A, B: TMatrix4D): TMatrix4D;
|
---|
45 | function MatrixIdentity4D: TMatrix4D;
|
---|
46 | function AffineMatrixToMatrix4D(AValue: TAffineMatrix): TMatrix4D;
|
---|
47 |
|
---|
48 | {$IFDEF BGRASSE_AVAILABLE}
|
---|
49 | procedure Matrix3D_SSE_Load(const A: TMatrix3D);
|
---|
50 | procedure MatrixMultiplyVect3D_SSE_Aligned(var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
51 | procedure MatrixMultiplyVect3D_SSE3_Aligned(var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
52 | procedure MatrixMultiplyVect3DWithoutTranslation_SSE_Aligned(var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
53 | procedure MatrixMultiplyVect3DWithoutTranslation_SSE3_Aligned(var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
54 | {$ENDIF}
|
---|
55 |
|
---|
56 | implementation
|
---|
57 |
|
---|
58 | procedure multiplyVect3(const A : TMatrix3D; const vx,vy,vz,vt: single; out outx,outy,outz: single);
|
---|
59 | begin
|
---|
60 | outx := vx * A[1,1] + vy * A[1,2] + vz * A[1,3] + vt * A[1,4];
|
---|
61 | outy := vx * A[2,1] + vy * A[2,2] + vz * A[2,3] + vt * A[2,4];
|
---|
62 | outz := vx * A[3,1] + vy * A[3,2] + vz * A[3,3] + vt * A[3,4];
|
---|
63 | end;
|
---|
64 |
|
---|
65 | procedure multiplyVect4(const A : TMatrix4D; const vx,vy,vz,vt: single; out outx,outy,outz,outt: single);
|
---|
66 | begin
|
---|
67 | outx := vx * A[1,1] + vy * A[1,2] + vz * A[1,3] + vt * A[1,4];
|
---|
68 | outy := vx * A[2,1] + vy * A[2,2] + vz * A[2,3] + vt * A[2,4];
|
---|
69 | outz := vx * A[3,1] + vy * A[3,2] + vz * A[3,3] + vt * A[3,4];
|
---|
70 | outt := vx * A[4,1] + vy * A[4,2] + vz * A[4,3] + vt * A[4,4];
|
---|
71 | end;
|
---|
72 |
|
---|
73 | operator*(const A: TMatrix3D; const M: TPoint3D): TPoint3D;
|
---|
74 | begin
|
---|
75 | result.x := M.x * A[1,1] + M.y * A[1,2] + M.z * A[1,3] + A[1,4];
|
---|
76 | result.y := M.x * A[2,1] + M.y * A[2,2] + M.z * A[2,3] + A[2,4];
|
---|
77 | result.z := M.x * A[3,1] + M.y * A[3,2] + M.z * A[3,3] + A[3,4];
|
---|
78 | end;
|
---|
79 |
|
---|
80 | operator*(const A, B: TMatrix4D): TMatrix4D;
|
---|
81 | begin
|
---|
82 | multiplyVect4(A, B[1,1],B[2,1],B[3,1],B[4,1], result[1,1],result[2,1],result[3,1],result[4,1]);
|
---|
83 | multiplyVect4(A, B[1,2],B[2,2],B[3,2],B[4,2], result[1,2],result[2,2],result[3,2],result[4,2]);
|
---|
84 | multiplyVect4(A, B[1,3],B[2,3],B[3,3],B[4,3], result[1,3],result[2,3],result[3,3],result[4,3]);
|
---|
85 | multiplyVect4(A, B[1,4],B[2,4],B[3,4],B[4,4], result[1,4],result[2,4],result[3,4],result[4,4]);
|
---|
86 | end;
|
---|
87 |
|
---|
88 | function MatrixIdentity4D: TMatrix4D;
|
---|
89 | begin
|
---|
90 | result[1,1] := 1; result[2,1] := 0; result[3,1] := 0; result[4,1] := 0;
|
---|
91 | result[1,2] := 0; result[2,2] := 1; result[3,2] := 0; result[4,2] := 0;
|
---|
92 | result[1,3] := 0; result[2,3] := 0; result[3,3] := 1; result[4,3] := 0;
|
---|
93 | result[1,4] := 0; result[2,4] := 0; result[3,4] := 0; result[4,4] := 1;
|
---|
94 | end;
|
---|
95 |
|
---|
96 | function AffineMatrixToMatrix4D(AValue: TAffineMatrix): TMatrix4D;
|
---|
97 | begin
|
---|
98 | result[1,1] := AValue[1,1]; result[2,1] := AValue[1,2]; result[3,1] := 0; result[4,1] := AValue[1,3];
|
---|
99 | result[1,2] := AValue[2,1]; result[2,2] := AValue[2,2]; result[3,2] := 0; result[4,2] := AValue[2,3];
|
---|
100 | result[1,3] := 0; result[2,3] := 0; result[3,3] := 1; result[4,3] := 0;
|
---|
101 | result[1,4] := 0; result[2,4] := 0; result[3,4] := 0; result[4,4] := 1;
|
---|
102 | end;
|
---|
103 |
|
---|
104 | {$IFDEF BGRASSE_AVAILABLE}
|
---|
105 | var SingleConst1 : single = 1;
|
---|
106 |
|
---|
107 | procedure Matrix3D_SSE_Load(const A: TMatrix3D);
|
---|
108 | begin
|
---|
109 | {$IFDEF cpux86_64}
|
---|
110 | asm
|
---|
111 | mov rax, A
|
---|
112 | movups xmm5, [rax]
|
---|
113 | movups xmm6, [rax+16]
|
---|
114 | movups xmm7, [rax+32]
|
---|
115 | end;
|
---|
116 | {$ELSE}
|
---|
117 | asm
|
---|
118 | mov eax, A
|
---|
119 | movups xmm5, [eax]
|
---|
120 | movups xmm6, [eax+16]
|
---|
121 | movups xmm7, [eax+32]
|
---|
122 | end;
|
---|
123 | {$ENDIF}
|
---|
124 | end;
|
---|
125 |
|
---|
126 | procedure MatrixMultiplyVect3D_SSE_Aligned(var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
127 | var oldMt: single;
|
---|
128 | begin
|
---|
129 | oldMt := M.t;
|
---|
130 | M.t := SingleConst1;
|
---|
131 | {$IFDEF cpux86_64}
|
---|
132 | asm
|
---|
133 | mov rax, M
|
---|
134 | movaps xmm0, [rax]
|
---|
135 |
|
---|
136 | mov rax, N
|
---|
137 |
|
---|
138 | movaps xmm2,xmm0
|
---|
139 | mulps xmm2,xmm5
|
---|
140 | //mix1
|
---|
141 | movaps xmm3, xmm2
|
---|
142 | shufps xmm3, xmm3, $4e
|
---|
143 | addps xmm2, xmm3
|
---|
144 | //mix2
|
---|
145 | movaps xmm3, xmm2
|
---|
146 | shufps xmm3, xmm3, $11
|
---|
147 | addps xmm2, xmm3
|
---|
148 |
|
---|
149 | movss [rax], xmm2
|
---|
150 |
|
---|
151 | movaps xmm2,xmm0
|
---|
152 | mulps xmm2,xmm6
|
---|
153 | //mix1
|
---|
154 | movaps xmm3, xmm2
|
---|
155 | shufps xmm3, xmm3, $4e
|
---|
156 | addps xmm2, xmm3
|
---|
157 | //mix2
|
---|
158 | movaps xmm3, xmm2
|
---|
159 | shufps xmm3, xmm3, $11
|
---|
160 | addps xmm2, xmm3
|
---|
161 |
|
---|
162 | movss [rax+4], xmm2
|
---|
163 |
|
---|
164 | mulps xmm0,xmm7
|
---|
165 | //mix1
|
---|
166 | movaps xmm3, xmm0
|
---|
167 | shufps xmm3, xmm3, $4e
|
---|
168 | addps xmm0, xmm3
|
---|
169 | //mix2
|
---|
170 | movaps xmm3, xmm0
|
---|
171 | shufps xmm3, xmm3, $11
|
---|
172 | addps xmm0, xmm3
|
---|
173 |
|
---|
174 | movss [rax+8], xmm0
|
---|
175 | end;
|
---|
176 | {$ELSE}
|
---|
177 | asm
|
---|
178 | mov eax, M
|
---|
179 | movaps xmm0, [eax]
|
---|
180 |
|
---|
181 | mov eax, N
|
---|
182 |
|
---|
183 | movaps xmm2,xmm0
|
---|
184 | mulps xmm2,xmm5
|
---|
185 | //mix1
|
---|
186 | movaps xmm3, xmm2
|
---|
187 | shufps xmm3, xmm3, $4e
|
---|
188 | addps xmm2, xmm3
|
---|
189 | //mix2
|
---|
190 | movaps xmm3, xmm2
|
---|
191 | shufps xmm3, xmm3, $11
|
---|
192 | addps xmm2, xmm3
|
---|
193 |
|
---|
194 | movss [eax], xmm2
|
---|
195 |
|
---|
196 | movaps xmm2,xmm0
|
---|
197 | mulps xmm2,xmm6
|
---|
198 | //mix1
|
---|
199 | movaps xmm3, xmm2
|
---|
200 | shufps xmm3, xmm3, $4e
|
---|
201 | addps xmm2, xmm3
|
---|
202 | //mix2
|
---|
203 | movaps xmm3, xmm2
|
---|
204 | shufps xmm3, xmm3, $11
|
---|
205 | addps xmm2, xmm3
|
---|
206 |
|
---|
207 | movss [eax+4], xmm2
|
---|
208 |
|
---|
209 | mulps xmm0,xmm7
|
---|
210 | //mix1
|
---|
211 | movaps xmm3, xmm0
|
---|
212 | shufps xmm3, xmm3, $4e
|
---|
213 | addps xmm0, xmm3
|
---|
214 | //mix2
|
---|
215 | movaps xmm3, xmm0
|
---|
216 | shufps xmm3, xmm3, $11
|
---|
217 | addps xmm0, xmm3
|
---|
218 |
|
---|
219 | movss [eax+8], xmm0
|
---|
220 | end;
|
---|
221 | {$ENDIF}
|
---|
222 | M.t := oldMt;
|
---|
223 | N.t := 0;
|
---|
224 | end;
|
---|
225 |
|
---|
226 | procedure MatrixMultiplyVect3D_SSE3_Aligned(var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
227 | var oldMt: single;
|
---|
228 | begin
|
---|
229 | oldMt := M.t;
|
---|
230 | M.t := SingleConst1;
|
---|
231 | {$IFDEF cpux86_64}
|
---|
232 | asm
|
---|
233 | mov rax, M
|
---|
234 | movaps xmm0, [rax]
|
---|
235 |
|
---|
236 | mov rax, N
|
---|
237 |
|
---|
238 | movaps xmm2,xmm0
|
---|
239 | mulps xmm2,xmm5
|
---|
240 | haddps xmm2,xmm2
|
---|
241 | haddps xmm2,xmm2
|
---|
242 | movss [rax], xmm2
|
---|
243 |
|
---|
244 | movaps xmm2,xmm0
|
---|
245 | mulps xmm2,xmm6
|
---|
246 | haddps xmm2,xmm2
|
---|
247 | haddps xmm2,xmm2
|
---|
248 | movss [rax+4], xmm2
|
---|
249 |
|
---|
250 | mulps xmm0,xmm7
|
---|
251 | haddps xmm0,xmm0
|
---|
252 | haddps xmm0,xmm0
|
---|
253 | movss [rax+8], xmm0
|
---|
254 | end;
|
---|
255 | {$ELSE}
|
---|
256 | asm
|
---|
257 | mov eax, M
|
---|
258 | movaps xmm0, [eax]
|
---|
259 |
|
---|
260 | mov eax, N
|
---|
261 |
|
---|
262 | movaps xmm2,xmm0
|
---|
263 | mulps xmm2,xmm5
|
---|
264 | haddps xmm2,xmm2
|
---|
265 | haddps xmm2,xmm2
|
---|
266 | movss [eax], xmm2
|
---|
267 |
|
---|
268 | movaps xmm2,xmm0
|
---|
269 | mulps xmm2,xmm6
|
---|
270 | haddps xmm2,xmm2
|
---|
271 | haddps xmm2,xmm2
|
---|
272 | movss [eax+4], xmm2
|
---|
273 |
|
---|
274 | mulps xmm0,xmm7
|
---|
275 | haddps xmm0,xmm0
|
---|
276 | haddps xmm0,xmm0
|
---|
277 | movss [eax+8], xmm0
|
---|
278 | end;
|
---|
279 | {$ENDIF}
|
---|
280 | M.t := oldMt;
|
---|
281 | end;
|
---|
282 |
|
---|
283 | procedure MatrixMultiplyVect3DWithoutTranslation_SSE_Aligned(
|
---|
284 | var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
285 | begin
|
---|
286 | {$IFDEF cpux86_64}
|
---|
287 | asm
|
---|
288 | mov rax, M
|
---|
289 | movaps xmm0, [rax]
|
---|
290 |
|
---|
291 | mov rax, N
|
---|
292 |
|
---|
293 | movaps xmm2,xmm0
|
---|
294 | mulps xmm2,xmm5
|
---|
295 | //mix1
|
---|
296 | movaps xmm3, xmm2
|
---|
297 | shufps xmm3, xmm3, $4e
|
---|
298 | addps xmm2, xmm3
|
---|
299 | //mix2
|
---|
300 | movaps xmm3, xmm2
|
---|
301 | shufps xmm3, xmm3, $11
|
---|
302 | addps xmm2, xmm3
|
---|
303 |
|
---|
304 | movss [rax], xmm2
|
---|
305 |
|
---|
306 | movaps xmm2,xmm0
|
---|
307 | mulps xmm2,xmm6
|
---|
308 | //mix1
|
---|
309 | movaps xmm3, xmm2
|
---|
310 | shufps xmm3, xmm3, $4e
|
---|
311 | addps xmm2, xmm3
|
---|
312 | //mix2
|
---|
313 | movaps xmm3, xmm2
|
---|
314 | shufps xmm3, xmm3, $11
|
---|
315 | addps xmm2, xmm3
|
---|
316 |
|
---|
317 | movss [rax+4], xmm2
|
---|
318 |
|
---|
319 | mulps xmm0,xmm7
|
---|
320 | //mix1
|
---|
321 | movaps xmm3, xmm0
|
---|
322 | shufps xmm3, xmm3, $4e
|
---|
323 | addps xmm0, xmm3
|
---|
324 | //mix2
|
---|
325 | movaps xmm3, xmm0
|
---|
326 | shufps xmm3, xmm3, $11
|
---|
327 | addps xmm0, xmm3
|
---|
328 |
|
---|
329 | movss [rax+8], xmm0
|
---|
330 | end;
|
---|
331 | {$ELSE}
|
---|
332 | asm
|
---|
333 | mov eax, M
|
---|
334 | movaps xmm0, [eax]
|
---|
335 |
|
---|
336 | mov eax, N
|
---|
337 |
|
---|
338 | movaps xmm2,xmm0
|
---|
339 | mulps xmm2,xmm5
|
---|
340 | //mix1
|
---|
341 | movaps xmm3, xmm2
|
---|
342 | shufps xmm3, xmm3, $4e
|
---|
343 | addps xmm2, xmm3
|
---|
344 | //mix2
|
---|
345 | movaps xmm3, xmm2
|
---|
346 | shufps xmm3, xmm3, $11
|
---|
347 | addps xmm2, xmm3
|
---|
348 |
|
---|
349 | movss [eax], xmm2
|
---|
350 |
|
---|
351 | movaps xmm2,xmm0
|
---|
352 | mulps xmm2,xmm6
|
---|
353 | //mix1
|
---|
354 | movaps xmm3, xmm2
|
---|
355 | shufps xmm3, xmm3, $4e
|
---|
356 | addps xmm2, xmm3
|
---|
357 | //mix2
|
---|
358 | movaps xmm3, xmm2
|
---|
359 | shufps xmm3, xmm3, $11
|
---|
360 | addps xmm2, xmm3
|
---|
361 |
|
---|
362 | movss [eax+4], xmm2
|
---|
363 |
|
---|
364 | mulps xmm0,xmm7
|
---|
365 | //mix1
|
---|
366 | movaps xmm3, xmm0
|
---|
367 | shufps xmm3, xmm3, $4e
|
---|
368 | addps xmm0, xmm3
|
---|
369 | //mix2
|
---|
370 | movaps xmm3, xmm0
|
---|
371 | shufps xmm3, xmm3, $11
|
---|
372 | addps xmm0, xmm3
|
---|
373 |
|
---|
374 | movss [eax+8], xmm0
|
---|
375 | end;
|
---|
376 | {$ENDIF}
|
---|
377 | end;
|
---|
378 |
|
---|
379 | procedure MatrixMultiplyVect3DWithoutTranslation_SSE3_Aligned(
|
---|
380 | var M: TPoint3D_128; out N: TPoint3D_128);
|
---|
381 | begin
|
---|
382 | {$IFDEF cpux86_64}
|
---|
383 | asm
|
---|
384 | mov rax, M
|
---|
385 | movaps xmm0, [rax]
|
---|
386 |
|
---|
387 | mov rax, N
|
---|
388 |
|
---|
389 | movaps xmm2,xmm0
|
---|
390 | mulps xmm2,xmm5
|
---|
391 | haddps xmm2,xmm2
|
---|
392 | haddps xmm2,xmm2
|
---|
393 | movss [rax], xmm2
|
---|
394 |
|
---|
395 | movaps xmm2,xmm0
|
---|
396 | mulps xmm2,xmm6
|
---|
397 | haddps xmm2,xmm2
|
---|
398 | haddps xmm2,xmm2
|
---|
399 | movss [rax+4], xmm2
|
---|
400 |
|
---|
401 | mulps xmm0,xmm7
|
---|
402 | haddps xmm0,xmm0
|
---|
403 | haddps xmm0,xmm0
|
---|
404 | movss [rax+8], xmm0
|
---|
405 | end;
|
---|
406 | {$ELSE}
|
---|
407 | asm
|
---|
408 | mov eax, M
|
---|
409 | movaps xmm0, [eax]
|
---|
410 |
|
---|
411 | mov eax, N
|
---|
412 |
|
---|
413 | movaps xmm2,xmm0
|
---|
414 | mulps xmm2,xmm5
|
---|
415 | haddps xmm2,xmm2
|
---|
416 | haddps xmm2,xmm2
|
---|
417 | movss [eax], xmm2
|
---|
418 |
|
---|
419 | movaps xmm2,xmm0
|
---|
420 | mulps xmm2,xmm6
|
---|
421 | haddps xmm2,xmm2
|
---|
422 | haddps xmm2,xmm2
|
---|
423 | movss [eax+4], xmm2
|
---|
424 |
|
---|
425 | mulps xmm0,xmm7
|
---|
426 | haddps xmm0,xmm0
|
---|
427 | haddps xmm0,xmm0
|
---|
428 | movss [eax+8], xmm0
|
---|
429 | end;
|
---|
430 | {$ENDIF}
|
---|
431 | end;
|
---|
432 |
|
---|
433 | {$ENDIF}
|
---|
434 |
|
---|
435 | operator*(constref A: TMatrix3D; var M: TPoint3D_128): TPoint3D_128;
|
---|
436 | {$IFDEF BGRASSE_AVAILABLE}var oldMt: single; resultAddr: pointer;{$ENDIF}
|
---|
437 | begin
|
---|
438 | {$IFDEF BGRASSE_AVAILABLE}
|
---|
439 | if UseSSE then
|
---|
440 | begin
|
---|
441 | oldMt := M.t;
|
---|
442 | M.t := SingleConst1;
|
---|
443 | resultAddr := @result;
|
---|
444 | {$IFDEF cpux86_64}
|
---|
445 | if UseSSE3 then
|
---|
446 | asm
|
---|
447 | mov rax, A
|
---|
448 | movups xmm5, [rax]
|
---|
449 | movups xmm6, [rax+16]
|
---|
450 | movups xmm7, [rax+32]
|
---|
451 |
|
---|
452 | mov rax, M
|
---|
453 | movups xmm0, [rax]
|
---|
454 |
|
---|
455 | mov rax, resultAddr
|
---|
456 |
|
---|
457 | movaps xmm4,xmm0
|
---|
458 | mulps xmm4,xmm5
|
---|
459 | haddps xmm4,xmm4
|
---|
460 | haddps xmm4,xmm4
|
---|
461 | movss [rax], xmm4
|
---|
462 |
|
---|
463 | movaps xmm4,xmm0
|
---|
464 | mulps xmm4,xmm6
|
---|
465 | haddps xmm4,xmm4
|
---|
466 | haddps xmm4,xmm4
|
---|
467 | movss [rax+4], xmm4
|
---|
468 |
|
---|
469 | mulps xmm0,xmm7
|
---|
470 | haddps xmm0,xmm0
|
---|
471 | haddps xmm0,xmm0
|
---|
472 | movss [rax+8], xmm0
|
---|
473 | end else
|
---|
474 | asm
|
---|
475 | mov rax, A
|
---|
476 | movups xmm5, [rax]
|
---|
477 | movups xmm6, [rax+16]
|
---|
478 | movups xmm7, [rax+32]
|
---|
479 |
|
---|
480 | mov rax, M
|
---|
481 | movups xmm0, [rax]
|
---|
482 |
|
---|
483 | mov rax, resultAddr
|
---|
484 |
|
---|
485 | movaps xmm4,xmm0
|
---|
486 | mulps xmm4,xmm5
|
---|
487 | //mix1
|
---|
488 | movaps xmm3, xmm4
|
---|
489 | shufps xmm3, xmm3, $4e
|
---|
490 | addps xmm4, xmm3
|
---|
491 | //mix2
|
---|
492 | movaps xmm3, xmm4
|
---|
493 | shufps xmm3, xmm3, $11
|
---|
494 | addps xmm4, xmm3
|
---|
495 |
|
---|
496 | movss [rax], xmm4
|
---|
497 |
|
---|
498 | movaps xmm4,xmm0
|
---|
499 | mulps xmm4,xmm6
|
---|
500 | //mix1
|
---|
501 | movaps xmm3, xmm4
|
---|
502 | shufps xmm3, xmm3, $4e
|
---|
503 | addps xmm4, xmm3
|
---|
504 | //mix2
|
---|
505 | movaps xmm3, xmm4
|
---|
506 | shufps xmm3, xmm3, $11
|
---|
507 | addps xmm4, xmm3
|
---|
508 |
|
---|
509 | movss [rax+4], xmm4
|
---|
510 |
|
---|
511 | mulps xmm0,xmm7
|
---|
512 | //mix1
|
---|
513 | movaps xmm3, xmm0
|
---|
514 | shufps xmm3, xmm3, $4e
|
---|
515 | addps xmm0, xmm3
|
---|
516 | //mix2
|
---|
517 | movaps xmm3, xmm0
|
---|
518 | shufps xmm3, xmm3, $11
|
---|
519 | addps xmm0, xmm3
|
---|
520 |
|
---|
521 | movss [rax+8], xmm0
|
---|
522 | end;
|
---|
523 | {$ELSE}
|
---|
524 | if UseSSE3 then
|
---|
525 | asm
|
---|
526 | mov eax, A
|
---|
527 | movups xmm5, [eax]
|
---|
528 | movups xmm6, [eax+16]
|
---|
529 | movups xmm7, [eax+32]
|
---|
530 |
|
---|
531 | mov eax, M
|
---|
532 | movups xmm0, [eax]
|
---|
533 |
|
---|
534 | mov eax, resultAddr
|
---|
535 |
|
---|
536 | movaps xmm4,xmm0
|
---|
537 | mulps xmm4,xmm5
|
---|
538 | haddps xmm4,xmm4
|
---|
539 | haddps xmm4,xmm4
|
---|
540 | movss [eax], xmm4
|
---|
541 |
|
---|
542 | movaps xmm4,xmm0
|
---|
543 | mulps xmm4,xmm6
|
---|
544 | haddps xmm4,xmm4
|
---|
545 | haddps xmm4,xmm4
|
---|
546 | movss [eax+4], xmm4
|
---|
547 |
|
---|
548 | mulps xmm0,xmm7
|
---|
549 | haddps xmm0,xmm0
|
---|
550 | haddps xmm0,xmm0
|
---|
551 | movss [eax+8], xmm0
|
---|
552 | end else
|
---|
553 | asm
|
---|
554 | mov eax, A
|
---|
555 | movups xmm5, [eax]
|
---|
556 | movups xmm6, [eax+16]
|
---|
557 | movups xmm7, [eax+32]
|
---|
558 |
|
---|
559 | mov eax, M
|
---|
560 | movups xmm0, [eax]
|
---|
561 |
|
---|
562 | mov eax, resultAddr
|
---|
563 |
|
---|
564 | movaps xmm4,xmm0
|
---|
565 | mulps xmm4,xmm5
|
---|
566 | //mix1
|
---|
567 | movaps xmm3, xmm4
|
---|
568 | shufps xmm3, xmm3, $4e
|
---|
569 | addps xmm4, xmm3
|
---|
570 | //mix2
|
---|
571 | movaps xmm3, xmm4
|
---|
572 | shufps xmm3, xmm3, $11
|
---|
573 | addps xmm4, xmm3
|
---|
574 |
|
---|
575 | movss [eax], xmm4
|
---|
576 |
|
---|
577 | movaps xmm4,xmm0
|
---|
578 | mulps xmm4,xmm6
|
---|
579 | //mix1
|
---|
580 | movaps xmm3, xmm4
|
---|
581 | shufps xmm3, xmm3, $4e
|
---|
582 | addps xmm4, xmm3
|
---|
583 | //mix2
|
---|
584 | movaps xmm3, xmm4
|
---|
585 | shufps xmm3, xmm3, $11
|
---|
586 | addps xmm4, xmm3
|
---|
587 |
|
---|
588 | movss [eax+4], xmm4
|
---|
589 |
|
---|
590 | mulps xmm0,xmm7
|
---|
591 | //mix1
|
---|
592 | movaps xmm3, xmm0
|
---|
593 | shufps xmm3, xmm3, $4e
|
---|
594 | addps xmm0, xmm3
|
---|
595 | //mix2
|
---|
596 | movaps xmm3, xmm0
|
---|
597 | shufps xmm3, xmm3, $11
|
---|
598 | addps xmm0, xmm3
|
---|
599 |
|
---|
600 | movss [eax+8], xmm0
|
---|
601 | end;
|
---|
602 | {$ENDIF}
|
---|
603 | M.t := oldMt;
|
---|
604 | result.t := 0;
|
---|
605 | end else
|
---|
606 | {$ENDIF}
|
---|
607 | begin
|
---|
608 | result.x := M.x * A[1,1] + M.y * A[1,2] + M.z * A[1,3] + A[1,4];
|
---|
609 | result.y := M.x * A[2,1] + M.y * A[2,2] + M.z * A[2,3] + A[2,4];
|
---|
610 | result.z := M.x * A[3,1] + M.y * A[3,2] + M.z * A[3,3] + A[3,4];
|
---|
611 | result.t := 0;
|
---|
612 | end;
|
---|
613 | end;
|
---|
614 |
|
---|
615 | function MultiplyVect3DWithoutTranslation(constref A: TMatrix3D; constref M: TPoint3D_128): TPoint3D_128;
|
---|
616 | begin
|
---|
617 | {$IFDEF BGRASSE_AVAILABLE}
|
---|
618 | if UseSSE then
|
---|
619 | begin
|
---|
620 | if UseSSE3 then
|
---|
621 | asm
|
---|
622 | mov eax, A
|
---|
623 | movups xmm5, [eax]
|
---|
624 | movups xmm6, [eax+16]
|
---|
625 | movups xmm7, [eax+32]
|
---|
626 |
|
---|
627 | mov eax, M
|
---|
628 | movups xmm0, [eax]
|
---|
629 |
|
---|
630 | mov eax, result
|
---|
631 |
|
---|
632 | movaps xmm4,xmm0
|
---|
633 | mulps xmm4,xmm5
|
---|
634 | haddps xmm4,xmm4
|
---|
635 | haddps xmm4,xmm4
|
---|
636 | movss [eax], xmm4
|
---|
637 |
|
---|
638 | movaps xmm4,xmm0
|
---|
639 | mulps xmm4,xmm6
|
---|
640 | haddps xmm4,xmm4
|
---|
641 | haddps xmm4,xmm4
|
---|
642 | movss [eax+4], xmm4
|
---|
643 |
|
---|
644 | mulps xmm0,xmm7
|
---|
645 | haddps xmm0,xmm0
|
---|
646 | haddps xmm0,xmm0
|
---|
647 | movss [eax+8], xmm0
|
---|
648 | end else
|
---|
649 | asm
|
---|
650 | mov eax, A
|
---|
651 | movups xmm5, [eax]
|
---|
652 | movups xmm6, [eax+16]
|
---|
653 | movups xmm7, [eax+32]
|
---|
654 |
|
---|
655 | mov eax, M
|
---|
656 | movups xmm0, [eax]
|
---|
657 |
|
---|
658 | mov eax, result
|
---|
659 |
|
---|
660 | movaps xmm4,xmm0
|
---|
661 | mulps xmm4,xmm5
|
---|
662 | //mix1
|
---|
663 | movaps xmm3, xmm4
|
---|
664 | shufps xmm3, xmm3, $4e
|
---|
665 | addps xmm4, xmm3
|
---|
666 | //mix2
|
---|
667 | movaps xmm3, xmm4
|
---|
668 | shufps xmm3, xmm3, $11
|
---|
669 | addps xmm4, xmm3
|
---|
670 |
|
---|
671 | movss [eax], xmm4
|
---|
672 |
|
---|
673 | movaps xmm4,xmm0
|
---|
674 | mulps xmm4,xmm6
|
---|
675 | //mix1
|
---|
676 | movaps xmm3, xmm4
|
---|
677 | shufps xmm3, xmm3, $4e
|
---|
678 | addps xmm4, xmm3
|
---|
679 | //mix2
|
---|
680 | movaps xmm3, xmm4
|
---|
681 | shufps xmm3, xmm3, $11
|
---|
682 | addps xmm4, xmm3
|
---|
683 |
|
---|
684 | movss [eax+4], xmm4
|
---|
685 |
|
---|
686 | mulps xmm0,xmm7
|
---|
687 | //mix1
|
---|
688 | movaps xmm3, xmm0
|
---|
689 | shufps xmm3, xmm3, $4e
|
---|
690 | addps xmm0, xmm3
|
---|
691 | //mix2
|
---|
692 | movaps xmm3, xmm0
|
---|
693 | shufps xmm3, xmm3, $11
|
---|
694 | addps xmm0, xmm3
|
---|
695 |
|
---|
696 | movss [eax+8], xmm0
|
---|
697 | end;
|
---|
698 | end else
|
---|
699 | {$ENDIF}
|
---|
700 | begin
|
---|
701 | result.x := M.x * A[1,1] + M.y * A[1,2] + M.z * A[1,3];
|
---|
702 | result.y := M.x * A[2,1] + M.y * A[2,2] + M.z * A[2,3];
|
---|
703 | result.z := M.x * A[3,1] + M.y * A[3,2] + M.z * A[3,3];
|
---|
704 | result.t := 0;
|
---|
705 | end;
|
---|
706 | end;
|
---|
707 |
|
---|
708 | operator*(A,B: TMatrix3D): TMatrix3D;
|
---|
709 | begin
|
---|
710 | multiplyVect3(A, B[1,1],B[2,1],B[3,1],0, result[1,1],result[2,1],result[3,1]);
|
---|
711 | multiplyVect3(A, B[1,2],B[2,2],B[3,2],0, result[1,2],result[2,2],result[3,2]);
|
---|
712 | multiplyVect3(A, B[1,3],B[2,3],B[3,3],0, result[1,3],result[2,3],result[3,3]);
|
---|
713 | multiplyVect3(A, B[1,4],B[2,4],B[3,4],1, result[1,4],result[2,4],result[3,4]);
|
---|
714 | end;
|
---|
715 |
|
---|
716 | function MatrixIdentity3D: TMatrix3D;
|
---|
717 | begin
|
---|
718 | result := Matrix3D( 1,0,0,0,
|
---|
719 | 0,1,0,0,
|
---|
720 | 0,0,1,0);
|
---|
721 | end;
|
---|
722 |
|
---|
723 | function Matrix3D(m11, m12, m13, m14, m21, m22, m23, m24, m31, m32, m33,
|
---|
724 | m34: single): TMatrix3D;
|
---|
725 | begin
|
---|
726 | result[1,1] := m11;
|
---|
727 | result[1,2] := m12;
|
---|
728 | result[1,3] := m13;
|
---|
729 | result[1,4] := m14;
|
---|
730 |
|
---|
731 | result[2,1] := m21;
|
---|
732 | result[2,2] := m22;
|
---|
733 | result[2,3] := m23;
|
---|
734 | result[2,4] := m24;
|
---|
735 |
|
---|
736 | result[3,1] := m31;
|
---|
737 | result[3,2] := m32;
|
---|
738 | result[3,3] := m33;
|
---|
739 | result[3,4] := m34;
|
---|
740 | end;
|
---|
741 |
|
---|
742 | function MatrixInverse3D(A: TMatrix3D): TMatrix3D;
|
---|
743 | var ofs: TPoint3D;
|
---|
744 | begin
|
---|
745 | ofs := Point3D(A[1,4],A[2,4],A[3,4]);
|
---|
746 |
|
---|
747 | result[1,1] := A[1,1];
|
---|
748 | result[1,2] := A[2,1];
|
---|
749 | result[1,3] := A[3,1];
|
---|
750 | result[1,4] := 0;
|
---|
751 |
|
---|
752 | result[2,1] := A[1,2];
|
---|
753 | result[2,2] := A[2,2];
|
---|
754 | result[2,3] := A[3,2];
|
---|
755 | result[2,4] := 0;
|
---|
756 |
|
---|
757 | result[3,1] := A[1,3];
|
---|
758 | result[3,2] := A[2,3];
|
---|
759 | result[3,3] := A[3,3];
|
---|
760 | result[3,4] := 0;
|
---|
761 |
|
---|
762 | result := result*MatrixTranslation3D(-ofs);
|
---|
763 | end;
|
---|
764 |
|
---|
765 | function Matrix3D(vx, vy, vz, ofs: TPoint3D): TMatrix3D;
|
---|
766 | begin
|
---|
767 | result := Matrix3D(vx.x, vy.x, vz.x, ofs.x,
|
---|
768 | vx.y, vy.y, vz.y, ofs.y,
|
---|
769 | vx.z, vy.z, vz.z, ofs.z);
|
---|
770 | end;
|
---|
771 |
|
---|
772 | function Matrix3D(vx, vy, vz, ofs: TPoint3D_128): TMatrix3D;
|
---|
773 | begin
|
---|
774 | result := Matrix3D(vx.x, vy.x, vz.x, ofs.x,
|
---|
775 | vx.y, vy.y, vz.y, ofs.y,
|
---|
776 | vx.z, vy.z, vz.z, ofs.z);
|
---|
777 | end;
|
---|
778 |
|
---|
779 | function MatrixTranslation3D(ofs: TPoint3D): TMatrix3D;
|
---|
780 | begin
|
---|
781 | result := Matrix3D(1,0,0,ofs.x,
|
---|
782 | 0,1,0,ofs.Y,
|
---|
783 | 0,0,1,ofs.z);
|
---|
784 | end;
|
---|
785 |
|
---|
786 | function MatrixScale3D(size: TPoint3D): TMatrix3D;
|
---|
787 | begin
|
---|
788 | result := Matrix3D(size.x,0,0,0,
|
---|
789 | 0,size.y,0,0,
|
---|
790 | 0,0,size.z,0);
|
---|
791 | end;
|
---|
792 |
|
---|
793 | function MatrixRotateX(angle: single): TMatrix3D;
|
---|
794 | begin
|
---|
795 | result := Matrix3D( 1, 0, 0, 0,
|
---|
796 | 0, cos(angle), sin(angle), 0,
|
---|
797 | 0, -sin(angle), cos(angle), 0);
|
---|
798 | end;
|
---|
799 |
|
---|
800 | function MatrixRotateY(angle: single): TMatrix3D;
|
---|
801 | begin
|
---|
802 | result := Matrix3D( cos(angle), 0, -sin(angle), 0,
|
---|
803 | 0, 1, 0, 0,
|
---|
804 | sin(angle), 0, cos(angle), 0);
|
---|
805 | end;
|
---|
806 |
|
---|
807 | function MatrixRotateZ(angle: single): TMatrix3D;
|
---|
808 | begin
|
---|
809 | result := Matrix3D( cos(angle), sin(angle), 0, 0,
|
---|
810 | -sin(angle), cos(angle), 0, 0,
|
---|
811 | 0, 0, 1, 0);
|
---|
812 | end;
|
---|
813 |
|
---|
814 | end.
|
---|
815 |
|
---|