Subversion Repository Public Repository

Divide-Dependencies

This repository has no backups
This repository's network speed is throttled to 100KB/sec

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions
// are met:
//  * Redistributions of source code must retain the above copyright
//    notice, this list of conditions and the following disclaimer.
//  * Redistributions in binary form must reproduce the above copyright
//    notice, this list of conditions and the following disclaimer in the
//    documentation and/or other materials provided with the distribution.
//  * Neither the name of NVIDIA CORPORATION nor the names of its
//    contributors may be used to endorse or promote products derived
//    from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
// EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
// PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR
// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
// OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Copyright (c) 2018 NVIDIA Corporation. All rights reserved.


// Copyright (c) 2004-2008 AGEIA Technologies, Inc. All rights reserved.
// Copyright (c) 2001-2004 NovodeX AG. All rights reserved.

#pragma once

// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
// factory implementation
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

template <>
inline Simd4fFactory<const float&>::operator Simd4f() const
{
	return _mm_set1_ps(v);
}

inline Simd4fFactory<detail::FourTuple>::operator Simd4f() const
{
	return reinterpret_cast<const Simd4f&>(v);
}

template <>
inline Simd4fFactory<detail::IntType<0> >::operator Simd4f() const
{
	return _mm_setzero_ps();
}

template <>
inline Simd4fFactory<detail::IntType<1> >::operator Simd4f() const
{
	return _mm_set1_ps(1.0f);
}

template <>
inline Simd4fFactory<detail::IntType<int(0x80000000)> >::operator Simd4f() const
{
	return _mm_castsi128_ps(_mm_set1_epi32(0x80000000));
}

template <>
inline Simd4fFactory<detail::IntType<int(0xffffffff)> >::operator Simd4f() const
{
	return _mm_castsi128_ps(_mm_set1_epi32(-1));
}

template <>
inline Simd4fFactory<const float*>::operator Simd4f() const
{
	return _mm_loadu_ps(v);
}

template <>
inline Simd4fFactory<detail::AlignedPointer<float> >::operator Simd4f() const
{
	return _mm_load_ps(v.ptr);
}

template <>
inline Simd4fFactory<detail::OffsetPointer<float> >::operator Simd4f() const
{
	return _mm_load_ps(reinterpret_cast<const float*>(reinterpret_cast<const char*>(v.ptr) + v.offset));
}

// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
// expression template
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

template <>
inline ComplementExpr<Simd4f>::operator Simd4f() const
{
	return _mm_andnot_ps(v, _mm_castsi128_ps(_mm_set1_epi32(-1)));
}

Simd4f operator&(const ComplementExpr<Simd4f>& complement, const Simd4f& v)
{
	return _mm_andnot_ps(complement.v, v);
}

Simd4f operator&(const Simd4f& v, const ComplementExpr<Simd4f>& complement)
{
	return _mm_andnot_ps(complement.v, v);
}

// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
// operator implementations
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Simd4f operator==(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_cmpeq_ps(v0, v1);
}

Simd4f operator<(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_cmplt_ps(v0, v1);
}

Simd4f operator<=(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_cmple_ps(v0, v1);
}

Simd4f operator>(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_cmpgt_ps(v0, v1);
}

Simd4f operator>=(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_cmpge_ps(v0, v1);
}

ComplementExpr<Simd4f> operator~(const Simd4f& v)
{
	return ComplementExpr<Simd4f>(v);
}

Simd4f operator&(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_and_ps(v0, v1);
}

Simd4f operator|(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_or_ps(v0, v1);
}

Simd4f operator^(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_xor_ps(v0, v1);
}

Simd4f operator<<(const Simd4f& v, int shift)
{
	return _mm_castsi128_ps(_mm_slli_epi32(_mm_castps_si128(v), shift));
}

Simd4f operator>>(const Simd4f& v, int shift)
{
	return _mm_castsi128_ps(_mm_srli_epi32(_mm_castps_si128(v), shift));
}

Simd4f operator+(const Simd4f& v)
{
	return v;
}

Simd4f operator+(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_add_ps(v0, v1);
}

Simd4f operator-(const Simd4f& v)
{
	return _mm_sub_ps(_mm_setzero_ps(), v);
}

Simd4f operator-(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_sub_ps(v0, v1);
}

Simd4f operator*(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_mul_ps(v0, v1);
}

Simd4f operator/(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_div_ps(v0, v1);
}

// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
// function implementations
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Simd4f simd4f(const Simd4i& v)
{
	return _mm_castsi128_ps(v);
}

float (&array(Simd4f& v))[4]
{
	return reinterpret_cast<float(&)[4]>(v);
}

const float (&array(const Simd4f& v))[4]
{
	return reinterpret_cast<const float(&)[4]>(v);
}

void store(float* ptr, Simd4f const& v)
{
	_mm_storeu_ps(ptr, v);
}

void storeAligned(float* ptr, Simd4f const& v)
{
	_mm_store_ps(ptr, v);
}

void storeAligned(float* ptr, unsigned int offset, Simd4f const& v)
{
	_mm_store_ps(reinterpret_cast<float*>(reinterpret_cast<char*>(ptr) + offset), v);
}

template <size_t i>
Simd4f splat(Simd4f const& v)
{
	return _mm_shuffle_ps(v, v, _MM_SHUFFLE(i, i, i, i));
}

Simd4f select(Simd4f const& mask, Simd4f const& v0, Simd4f const& v1)
{
	return _mm_xor_ps(v1, _mm_and_ps(mask, _mm_xor_ps(v1, v0)));
}

Simd4f abs(const Simd4f& v)
{
	return _mm_andnot_ps(_mm_castsi128_ps(_mm_set1_epi32(0x80000000)), v);
}

Simd4f floor(const Simd4f& v)
{
	// SSE 4.1: return _mm_floor_ps(v);
	Simd4i i = _mm_cvttps_epi32(v);
	return _mm_cvtepi32_ps(_mm_sub_epi32(i, _mm_srli_epi32(i, 31)));
}

Simd4f max(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_max_ps(v0, v1);
}

Simd4f min(const Simd4f& v0, const Simd4f& v1)
{
	return _mm_min_ps(v0, v1);
}

Simd4f recip(const Simd4f& v)
{
	return _mm_rcp_ps(v);
}

template <int n>
Simd4f recipT(const Simd4f& v)
{
	Simd4f two = simd4f(2.0f);
	Simd4f recipV = recip(v);
	for(int i = 0; i < n; ++i)
		recipV = recipV * (two - v * recipV);
	return recipV;
}

Simd4f sqrt(const Simd4f& v)
{
	return _mm_sqrt_ps(v);
}

Simd4f rsqrt(const Simd4f& v)
{
	return _mm_rsqrt_ps(v);
}

template <int n>
Simd4f rsqrtT(const Simd4f& v)
{
	Simd4f halfV = v * simd4f(0.5f);
	Simd4f threeHalf = simd4f(1.5f);
	Simd4f rsqrtV = rsqrt(v);
	for(int i = 0; i < n; ++i)
		rsqrtV = rsqrtV * (threeHalf - halfV * rsqrtV * rsqrtV);
	return rsqrtV;
}

Simd4f exp2(const Simd4f& v)
{
	// http://www.netlib.org/cephes/

	Simd4f limit = simd4f(127.4999f);
	Simd4f x = min(max(-limit, v), limit);

	// separate into integer and fractional part

	Simd4f fx = x + simd4f(0.5f);
	Simd4i ix = _mm_sub_epi32(_mm_cvttps_epi32(fx), _mm_srli_epi32(_mm_castps_si128(fx), 31));
	fx = x - Simd4f(_mm_cvtepi32_ps(ix));

	// exp2(fx) ~ 1 + 2*P(fx) / (Q(fx) - P(fx))

	Simd4f fx2 = fx * fx;

	Simd4f px = fx * (simd4f(1.51390680115615096133e+3f) +
	                  fx2 * (simd4f(2.02020656693165307700e+1f) + fx2 * simd4f(2.30933477057345225087e-2f)));
	Simd4f qx = simd4f(4.36821166879210612817e+3f) + fx2 * (simd4f(2.33184211722314911771e+2f) + fx2);

	Simd4f exp2fx = px * recip(qx - px);
	exp2fx = simd4f(_1) + exp2fx + exp2fx;

	// exp2(ix)

	Simd4f exp2ix = _mm_castsi128_ps(_mm_slli_epi32(_mm_add_epi32(ix, _mm_set1_epi32(0x7f)), 23));

	return exp2fx * exp2ix;
}

Simd4f log2(const Simd4f& v)
{
	// todo: fast approximate implementation like exp2
	Simd4f scale = simd4f(1.44269504088896341f); // 1/ln(2)
	const float* ptr = array(v);
	return simd4f(::logf(ptr[0]), ::logf(ptr[1]), ::logf(ptr[2]), ::logf(ptr[3])) * scale;
}

Simd4f dot3(const Simd4f& v0, const Simd4f& v1)
{
	Simd4f tmp = v0 * v1;
	return splat<0>(tmp) + splat<1>(tmp) + splat<2>(tmp);
}

Simd4f cross3(const Simd4f& v0, const Simd4f& v1)
{
	Simd4f t0 = _mm_shuffle_ps(v0, v0, 0xc9); // w z y x -> w x z y
	Simd4f t1 = _mm_shuffle_ps(v1, v1, 0xc9);
	Simd4f tmp = v0 * t1 - t0 * v1;
	return _mm_shuffle_ps(tmp, tmp, 0xc9);
}

void transpose(Simd4f& x, Simd4f& y, Simd4f& z, Simd4f& w)
{
	_MM_TRANSPOSE4_PS(x, y, z, w);
}

int allEqual(const Simd4f& v0, const Simd4f& v1)
{
	return allTrue(v0 == v1);
}

int allEqual(const Simd4f& v0, const Simd4f& v1, Simd4f& outMask)
{
	return allTrue(outMask = v0 == v1);
}

int anyEqual(const Simd4f& v0, const Simd4f& v1)
{
	return anyTrue(v0 == v1);
}

int anyEqual(const Simd4f& v0, const Simd4f& v1, Simd4f& outMask)
{
	return anyTrue(outMask = v0 == v1);
}

int allGreater(const Simd4f& v0, const Simd4f& v1)
{
	return allTrue(v0 > v1);
}

int allGreater(const Simd4f& v0, const Simd4f& v1, Simd4f& outMask)
{
	return allTrue(outMask = v0 > v1);
}

int anyGreater(const Simd4f& v0, const Simd4f& v1)
{
	return anyTrue(v0 > v1);
}

int anyGreater(const Simd4f& v0, const Simd4f& v1, Simd4f& outMask)
{
	return anyTrue(outMask = v0 > v1);
}

int allGreaterEqual(const Simd4f& v0, const Simd4f& v1)
{
	return allTrue(v0 >= v1);
}

int allGreaterEqual(const Simd4f& v0, const Simd4f& v1, Simd4f& outMask)
{
	return allTrue(outMask = v0 >= v1);
}

int anyGreaterEqual(const Simd4f& v0, const Simd4f& v1)
{
	return anyTrue(v0 >= v1);
}

int anyGreaterEqual(const Simd4f& v0, const Simd4f& v1, Simd4f& outMask)
{
	return anyTrue(outMask = v0 >= v1);
}

int allTrue(const Simd4f& v)
{
	return _mm_movemask_ps(v) == 0xf;
}

int anyTrue(const Simd4f& v)
{
	return _mm_movemask_ps(v);
}

Commits for Divide-Dependencies/physx/APEX_1.4/module/clothing/embedded/LowLevelCloth/src/sse2/Simd4f.h

Diff revisions: vs.
Revision Author Commited Message
105 IonutCava picture IonutCava Tue 16 Apr, 2019 19:55:41 +0000

Forgot to actually add physx