Coverage Report

Created: 2026-04-12 08:01

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/src/opus/celt/mathops.c
Line
Count
Source
1
/* Copyright (c) 2002-2008 Jean-Marc Valin
2
   Copyright (c) 2007-2008 CSIRO
3
   Copyright (c) 2007-2009 Xiph.Org Foundation
4
   Copyright (c) 2024 Arm Limited
5
   Written by Jean-Marc Valin */
6
/**
7
   @file mathops.h
8
   @brief Various math functions
9
*/
10
/*
11
   Redistribution and use in source and binary forms, with or without
12
   modification, are permitted provided that the following conditions
13
   are met:
14
15
   - Redistributions of source code must retain the above copyright
16
   notice, this list of conditions and the following disclaimer.
17
18
   - Redistributions in binary form must reproduce the above copyright
19
   notice, this list of conditions and the following disclaimer in the
20
   documentation and/or other materials provided with the distribution.
21
22
   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
23
   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
24
   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
25
   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
26
   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
27
   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
28
   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
29
   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
30
   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
31
   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
32
   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33
*/
34
35
#ifdef HAVE_CONFIG_H
36
#include "config.h"
37
#endif
38
39
#include "float_cast.h"
40
#include "mathops.h"
41
42
/*Compute floor(sqrt(_val)) with exact arithmetic.
43
  _val must be greater than 0.
44
  This has been tested on all possible 32-bit inputs greater than 0.*/
45
0
unsigned isqrt32(opus_uint32 _val){
46
0
  unsigned b;
47
0
  unsigned g;
48
0
  int      bshift;
49
  /*Uses the second method from
50
     http://www.azillionmonkeys.com/qed/sqroot.html
51
    The main idea is to search for the largest binary digit b such that
52
     (g+b)*(g+b) <= _val, and add it to the solution g.*/
53
0
  g=0;
54
0
  bshift=(EC_ILOG(_val)-1)>>1;
55
0
  b=1U<<bshift;
56
0
  do{
57
0
    opus_uint32 t;
58
0
    t=(((opus_uint32)g<<1)+b)<<bshift;
59
0
    if(t<=_val){
60
0
      g+=b;
61
0
      _val-=t;
62
0
    }
63
0
    b>>=1;
64
0
    bshift--;
65
0
  }
66
0
  while(bshift>=0);
67
0
  return g;
68
0
}
69
70
#ifdef FIXED_POINT
71
72
opus_val32 frac_div32_q29(opus_val32 a, opus_val32 b)
73
68.9M
{
74
68.9M
   opus_val16 rcp;
75
68.9M
   opus_val32 result, rem;
76
68.9M
   int shift = celt_ilog2(b)-29;
77
68.9M
   a = VSHR32(a,shift);
78
68.9M
   b = VSHR32(b,shift);
79
   /* 16-bit reciprocal */
80
68.9M
   rcp = ROUND16(celt_rcp(ROUND16(b,16)),3);
81
68.9M
   result = MULT16_32_Q15(rcp, a);
82
68.9M
   rem = PSHR32(a,2)-MULT32_32_Q31(result, b);
83
68.9M
   result = ADD32(result, SHL32(MULT16_32_Q15(rcp, rem),2));
84
68.9M
   return result;
85
68.9M
}
86
87
66.1M
opus_val32 frac_div32(opus_val32 a, opus_val32 b) {
88
66.1M
   opus_val32 result = frac_div32_q29(a,b);
89
66.1M
   if (result >= 536870912)       /*  2^29 */
90
1.45M
      return 2147483647;          /*  2^31 - 1 */
91
64.7M
   else if (result <= -536870912) /* -2^29 */
92
0
      return -2147483647;         /* -2^31 */
93
64.7M
   else
94
64.7M
      return SHL32(result, 2);
95
66.1M
}
96
97
/** Reciprocal sqrt approximation in the range [0.25,1) (Q16 in, Q14 out) */
98
opus_val16 celt_rsqrt_norm(opus_val32 x)
99
607M
{
100
607M
   opus_val16 n;
101
607M
   opus_val16 r;
102
607M
   opus_val16 r2;
103
607M
   opus_val16 y;
104
   /* Range of n is [-16384,32767] ([-0.5,1) in Q15). */
105
607M
   n = x-32768;
106
   /* Get a rough initial guess for the root.
107
      The optimal minimax quadratic approximation (using relative error) is
108
       r = 1.437799046117536+n*(-0.823394375837328+n*0.4096419668459485).
109
      Coefficients here, and the final result r, are Q14.*/
110
607M
   r = ADD16(23557, MULT16_16_Q15(n, ADD16(-13490, MULT16_16_Q15(n, 6713))));
111
   /* We want y = x*r*r-1 in Q15, but x is 32-bit Q16 and r is Q14.
112
      We can compute the result from n and r using Q15 multiplies with some
113
       adjustment, carefully done to avoid overflow.
114
      Range of y is [-1564,1594]. */
115
607M
   r2 = MULT16_16_Q15(r, r);
116
607M
   y = SHL16(SUB16(ADD16(MULT16_16_Q15(r2, n), r2), 16384), 1);
117
   /* Apply a 2nd-order Householder iteration: r += r*y*(y*0.375-0.5).
118
      This yields the Q14 reciprocal square root of the Q16 x, with a maximum
119
       relative error of 1.04956E-4, a (relative) RMSE of 2.80979E-5, and a
120
       peak absolute error of 2.26591/16384. */
121
607M
   return ADD16(r, MULT16_16_Q15(r, MULT16_16_Q15(y,
122
607M
              SUB16(MULT16_16_Q15(y, 12288), 16384))));
123
607M
}
124
125
/** Reciprocal sqrt approximation in the range [0.25,1) (Q31 in, Q29 out) */
126
opus_val32 celt_rsqrt_norm32(opus_val32 x)
127
533M
{
128
533M
   opus_int32 tmp;
129
   /* Use the first-order Newton-Raphson method to refine the root estimate.
130
    * r = r * (1.5 - 0.5*x*r*r) */
131
533M
   opus_int32 r_q29 = SHL32(celt_rsqrt_norm(SHR32(x, 31-16)), 15);
132
   /* Split evaluation in steps to avoid exploding macro expansion. */
133
533M
   tmp = MULT32_32_Q31(r_q29, r_q29);
134
533M
   tmp = MULT32_32_Q31(1073741824 /* Q31 */, tmp);
135
533M
   tmp = MULT32_32_Q31(x, tmp);
136
533M
   return SHL32(MULT32_32_Q31(r_q29, SUB32(201326592 /* Q27 */, tmp)), 4);
137
533M
}
138
139
/** Sqrt approximation (QX input, QX/2 output) */
140
opus_val32 celt_sqrt(opus_val32 x)
141
367M
{
142
367M
   int k;
143
367M
   opus_val16 n;
144
367M
   opus_val32 rt;
145
   /* These coeffs are optimized in fixed-point to minimize both RMS and max error
146
      of sqrt(x) over .25<x<1 without exceeding 32767.
147
      The RMS error is 3.4e-5 and the max is 8.2e-5. */
148
367M
   static const opus_val16 C[6] = {23171, 11574, -2901, 1592, -1002, 336};
149
367M
   if (x==0)
150
86.0M
      return 0;
151
281M
   else if (x>=1073741824)
152
121k
      return 32767;
153
281M
   k = (celt_ilog2(x)>>1)-7;
154
281M
   x = VSHR32(x, 2*k);
155
281M
   n = x-32768;
156
281M
   rt = ADD32(C[0], MULT16_16_Q15(n, ADD16(C[1], MULT16_16_Q15(n, ADD16(C[2],
157
281M
              MULT16_16_Q15(n, ADD16(C[3], MULT16_16_Q15(n, ADD16(C[4], MULT16_16_Q15(n, (C[5])))))))))));
158
281M
   rt = VSHR32(rt,7-k);
159
281M
   return rt;
160
367M
}
161
162
/* Perform fixed-point arithmetic to approximate the square root. When the input
163
 * is in Qx format, the output will be in Q(x/2 + 16) format. */
164
opus_val32 celt_sqrt32(opus_val32 x)
165
857M
{
166
857M
   int k;
167
857M
   opus_int32 x_frac;
168
857M
   if (x==0)
169
467M
      return 0;
170
389M
   else if (x>=1073741824)
171
0
      return 2147483647; /* 2^31 -1 */
172
389M
   k = (celt_ilog2(x)>>1);
173
389M
   x_frac = VSHR32(x, 2*(k-14)-1);
174
389M
   x_frac = MULT32_32_Q31(celt_rsqrt_norm32(x_frac), x_frac);
175
389M
   if (k < 12) return PSHR32(x_frac, 12-k);
176
370M
   else return SHL32(x_frac, k-12);
177
389M
}
178
179
#define L1 32767
180
#define L2 -7651
181
#define L3 8277
182
#define L4 -626
183
184
static OPUS_INLINE opus_val16 _celt_cos_pi_2(opus_val16 x)
185
24.1M
{
186
24.1M
   opus_val16 x2;
187
188
24.1M
   x2 = MULT16_16_P15(x,x);
189
24.1M
   return ADD16(1,MIN16(32766,ADD32(SUB16(L1,x2), MULT16_16_P15(x2, ADD32(L2, MULT16_16_P15(x2, ADD32(L3, MULT16_16_P15(L4, x2
190
24.1M
                                                                                ))))))));
191
24.1M
}
192
193
#undef L1
194
#undef L2
195
#undef L3
196
#undef L4
197
198
opus_val16 celt_cos_norm(opus_val32 x)
199
24.1M
{
200
24.1M
   x = x&0x0001ffff;
201
24.1M
   if (x>SHL32(EXTEND32(1), 16))
202
0
      x = SUB32(SHL32(EXTEND32(1), 17),x);
203
24.1M
   if (x&0x00007fff)
204
24.1M
   {
205
24.1M
      if (x<SHL32(EXTEND32(1), 15))
206
24.1M
      {
207
24.1M
         return _celt_cos_pi_2(EXTRACT16(x));
208
24.1M
      } else {
209
0
         return NEG16(_celt_cos_pi_2(EXTRACT16(65536-x)));
210
0
      }
211
24.1M
   } else {
212
0
      if (x&0x0000ffff)
213
0
         return 0;
214
0
      else if (x&0x0001ffff)
215
0
         return -32767;
216
0
      else
217
0
         return 32767;
218
0
   }
219
24.1M
}
220
221
/* Calculates the cosine of (PI*0.5*x) where the input x ranges from -1 to 1 and
222
 * is in Q30 format. The output will also be in Q31 format. */
223
opus_val32 celt_cos_norm32(opus_val32 x)
224
0
{
225
0
   static const opus_val32 COS_NORM_COEFF_A0 = 134217720;   /* Q27 */
226
0
   static const opus_val32 COS_NORM_COEFF_A1 = -662336704;  /* Q29 */
227
0
   static const opus_val32 COS_NORM_COEFF_A2 = 544710848;   /* Q31 */
228
0
   static const opus_val32 COS_NORM_COEFF_A3 = -178761936;  /* Q33 */
229
0
   static const opus_val32 COS_NORM_COEFF_A4 = 29487206;    /* Q35 */
230
0
   opus_int32 x_sq_q29, tmp;
231
   /* The expected x is in the range of [-1.0f, 1.0f] */
232
0
   celt_sig_assert((x >= -1073741824) && (x <= 1073741824));
233
   /* Make cos(+/- pi/2) exactly zero. */
234
0
   if (ABS32(x) == 1<<30) return 0;
235
0
   x_sq_q29 = MULT32_32_Q31(x, x);
236
   /* Split evaluation in steps to avoid exploding macro expansion. */
237
0
   tmp = ADD32(COS_NORM_COEFF_A3, MULT32_32_Q31(x_sq_q29, COS_NORM_COEFF_A4));
238
0
   tmp = ADD32(COS_NORM_COEFF_A2, MULT32_32_Q31(x_sq_q29, tmp));
239
0
   tmp = ADD32(COS_NORM_COEFF_A1, MULT32_32_Q31(x_sq_q29, tmp));
240
0
   return SHL32(ADD32(COS_NORM_COEFF_A0, MULT32_32_Q31(x_sq_q29, tmp)), 4);
241
0
}
242
243
/* Computes a 16 bit approximate reciprocal (1/x) for a normalized Q15 input,
244
 * resulting in a Q15 output. */
245
opus_val16 celt_rcp_norm16(opus_val16 x)
246
1.23G
{
247
1.23G
   opus_val16 r;
248
   /* Start with a linear approximation:
249
      r = 1.8823529411764706-0.9411764705882353*n.
250
      The coefficients and the result are Q14 in the range [15420,30840].*/
251
1.23G
   r = ADD16(30840, MULT16_16_Q15(-15420, x));
252
   /* Perform two Newton iterations:
253
      r -= r*((r*n)+(r-1.Q15))
254
         = r*((r*n)+(r-1.Q15)). */
255
1.23G
   r = SUB16(r, MULT16_16_Q15(r,
256
1.23G
             ADD16(MULT16_16_Q15(r, x), ADD16(r, -32768))));
257
   /* We subtract an extra 1 in the second iteration to avoid overflow; it also
258
       neatly compensates for truncation error in the rest of the process. */
259
1.23G
   return SUB16(r, ADD16(1, MULT16_16_Q15(r,
260
1.23G
                ADD16(MULT16_16_Q15(r, x), ADD16(r, -32768)))));
261
1.23G
}
262
263
/* Computes a 32 bit approximated reciprocal (1/x) for a normalized Q31 input,
264
 * resulting in a Q30 output. The expected input range is [0.5f, 1.0f) in Q31
265
 * and the expected output range is [1.0f, 2.0f) in Q30. */
266
opus_val32 celt_rcp_norm32(opus_val32 x)
267
933M
{
268
933M
   opus_val32 r_q30;
269
933M
   celt_sig_assert(x >= 1073741824);
270
933M
   r_q30 = SHL32(EXTEND32(celt_rcp_norm16(SHR32(x, 15)-32768)), 16);
271
   /* Solving f(y) = a - 1/y using the Newton Method
272
    * Note: f(y)' = 1/y^2
273
    * r = r - f(r)/f(r)' = r - (x * r*r - r)
274
    *   = r - r*(r*x - 1)
275
    * where
276
    *   - r means 1/y's approximation.
277
    *   - x means a, the input of function.
278
    * Please note that:
279
    *   - It adds 1 to avoid overflow
280
    *   - -1.0f in Q30 is -1073741824. */
281
933M
   return SUB32(r_q30, ADD32(SHL32(
282
933M
                MULT32_32_Q31(ADD32(MULT32_32_Q31(r_q30, x), -1073741824),
283
933M
                              r_q30), 1), 1));
284
933M
}
285
286
/** Reciprocal approximation (Q15 input, Q16 output) */
287
opus_val32 celt_rcp(opus_val32 x)
288
299M
{
289
299M
   int i;
290
299M
   opus_val16 r;
291
299M
   celt_sig_assert(x>0);
292
299M
   i = celt_ilog2(x);
293
294
   /* Compute the reciprocal of a Q15 number in the range [0, 1). */
295
299M
   r = celt_rcp_norm16(VSHR32(x,i-15)-32768);
296
297
   /* r is now the Q15 solution to 2/(n+1), with a maximum relative error
298
       of 7.05346E-5, a (relative) RMSE of 2.14418E-5, and a peak absolute
299
       error of 1.24665/32768. */
300
299M
   return VSHR32(EXTEND32(r),i-16);
301
299M
}
302
303
#endif
304
305
#ifndef DISABLE_FLOAT_API
306
307
void celt_float2int16_c(const float * OPUS_RESTRICT in, short * OPUS_RESTRICT out, int cnt)
308
0
{
309
0
   int i;
310
0
   for (i = 0; i < cnt; i++)
311
0
   {
312
0
      out[i] = FLOAT2INT16(in[i]);
313
0
   }
314
0
}
315
316
int opus_limit2_checkwithin1_c(float * samples, int cnt)
317
0
{
318
0
   int i;
319
0
   if (cnt <= 0)
320
0
   {
321
0
      return 1;
322
0
   }
323
324
0
   for (i = 0; i < cnt; i++)
325
0
   {
326
0
      float clippedVal = samples[i];
327
0
      clippedVal = FMAX(-2.0f, clippedVal);
328
0
      clippedVal = FMIN(2.0f, clippedVal);
329
0
      samples[i] = clippedVal;
330
0
   }
331
332
   /* C implementation can't provide quick hint. Assume it might exceed -1/+1. */
333
0
   return 0;
334
0
}
335
336
#endif /* DISABLE_FLOAT_API */