trace(1): resolve all level-5 LLVM warnings
[minix3.git] / lib / libm / src / math_private.h
blobf9e0dc2f77371aa6217a191a677d9695754dc855
1 /*
2 * ====================================================
3 * Copyright (C) 1993 by Sun Microsystems, Inc. All rights reserved.
5 * Developed at SunPro, a Sun Microsystems, Inc. business.
6 * Permission to use, copy, modify, and distribute this
7 * software is freely granted, provided that this notice
8 * is preserved.
9 * ====================================================
13 * from: @(#)fdlibm.h 5.1 93/09/24
14 * $NetBSD: math_private.h,v 1.22 2015/03/26 11:59:38 justin Exp $
17 #ifndef _MATH_PRIVATE_H_
18 #define _MATH_PRIVATE_H_
20 #include <sys/types.h>
22 /* The original fdlibm code used statements like:
23 n0 = ((*(int*)&one)>>29)^1; * index of high word *
24 ix0 = *(n0+(int*)&x); * high word of x *
25 ix1 = *((1-n0)+(int*)&x); * low word of x *
26 to dig two 32 bit words out of the 64 bit IEEE floating point
27 value. That is non-ANSI, and, moreover, the gcc instruction
28 scheduler gets it wrong. We instead use the following macros.
29 Unlike the original code, we determine the endianness at compile
30 time, not at run time; I don't see much benefit to selecting
31 endianness at run time. */
33 /* A union which permits us to convert between a double and two 32 bit
34 ints. */
37 * The ARM ports are little endian except for the FPA word order which is
38 * big endian.
41 #if (BYTE_ORDER == BIG_ENDIAN) || (defined(__arm__) && !defined(__VFP_FP__))
43 typedef union
45 double value;
46 struct
48 u_int32_t msw;
49 u_int32_t lsw;
50 } parts;
51 struct {
52 u_int64_t w;
53 } xparts;
54 } ieee_double_shape_type;
56 #endif
58 #if (BYTE_ORDER == LITTLE_ENDIAN) && \
59 !(defined(__arm__) && !defined(__VFP_FP__))
61 typedef union
63 double value;
64 struct
66 u_int32_t lsw;
67 u_int32_t msw;
68 } parts;
69 struct {
70 u_int64_t w;
71 } xparts;
72 } ieee_double_shape_type;
74 #endif
76 /* Get two 32 bit ints from a double. */
78 #define EXTRACT_WORDS(ix0,ix1,d) \
79 do { \
80 ieee_double_shape_type ew_u; \
81 ew_u.value = (d); \
82 (ix0) = ew_u.parts.msw; \
83 (ix1) = ew_u.parts.lsw; \
84 } while (/*CONSTCOND*/0)
86 /* Get a 64-bit int from a double. */
87 #define EXTRACT_WORD64(ix,d) \
88 do { \
89 ieee_double_shape_type ew_u; \
90 ew_u.value = (d); \
91 (ix) = ew_u.xparts.w; \
92 } while (/*CONSTCOND*/0)
95 /* Get the more significant 32 bit int from a double. */
97 #define GET_HIGH_WORD(i,d) \
98 do { \
99 ieee_double_shape_type gh_u; \
100 gh_u.value = (d); \
101 (i) = gh_u.parts.msw; \
102 } while (/*CONSTCOND*/0)
104 /* Get the less significant 32 bit int from a double. */
106 #define GET_LOW_WORD(i,d) \
107 do { \
108 ieee_double_shape_type gl_u; \
109 gl_u.value = (d); \
110 (i) = gl_u.parts.lsw; \
111 } while (/*CONSTCOND*/0)
113 /* Set a double from two 32 bit ints. */
115 #define INSERT_WORDS(d,ix0,ix1) \
116 do { \
117 ieee_double_shape_type iw_u; \
118 iw_u.parts.msw = (ix0); \
119 iw_u.parts.lsw = (ix1); \
120 (d) = iw_u.value; \
121 } while (/*CONSTCOND*/0)
123 /* Set a double from a 64-bit int. */
124 #define INSERT_WORD64(d,ix) \
125 do { \
126 ieee_double_shape_type iw_u; \
127 iw_u.xparts.w = (ix); \
128 (d) = iw_u.value; \
129 } while (/*CONSTCOND*/0)
132 /* Set the more significant 32 bits of a double from an int. */
134 #define SET_HIGH_WORD(d,v) \
135 do { \
136 ieee_double_shape_type sh_u; \
137 sh_u.value = (d); \
138 sh_u.parts.msw = (v); \
139 (d) = sh_u.value; \
140 } while (/*CONSTCOND*/0)
142 /* Set the less significant 32 bits of a double from an int. */
144 #define SET_LOW_WORD(d,v) \
145 do { \
146 ieee_double_shape_type sl_u; \
147 sl_u.value = (d); \
148 sl_u.parts.lsw = (v); \
149 (d) = sl_u.value; \
150 } while (/*CONSTCOND*/0)
152 /* A union which permits us to convert between a float and a 32 bit
153 int. */
155 typedef union
157 float value;
158 u_int32_t word;
159 } ieee_float_shape_type;
161 /* Get a 32 bit int from a float. */
163 #define GET_FLOAT_WORD(i,d) \
164 do { \
165 ieee_float_shape_type gf_u; \
166 gf_u.value = (d); \
167 (i) = gf_u.word; \
168 } while (/*CONSTCOND*/0)
170 /* Set a float from a 32 bit int. */
172 #define SET_FLOAT_WORD(d,i) \
173 do { \
174 ieee_float_shape_type sf_u; \
175 sf_u.word = (i); \
176 (d) = sf_u.value; \
177 } while (/*CONSTCOND*/0)
180 * Attempt to get strict C99 semantics for assignment with non-C99 compilers.
182 #if FLT_EVAL_METHOD == 0 || __GNUC__ == 0
183 #define STRICT_ASSIGN(type, lval, rval) ((lval) = (rval))
184 #else
185 #define STRICT_ASSIGN(type, lval, rval) do { \
186 volatile type __lval; \
188 if (sizeof(type) >= sizeof(long double)) \
189 (lval) = (rval); \
190 else { \
191 __lval = (rval); \
192 (lval) = __lval; \
194 } while (/*CONSTCOND*/0)
195 #endif
197 #ifdef _COMPLEX_H
200 * Quoting from ISO/IEC 9899:TC2:
202 * 6.2.5.13 Types
203 * Each complex type has the same representation and alignment requirements as
204 * an array type containing exactly two elements of the corresponding real type;
205 * the first element is equal to the real part, and the second element to the
206 * imaginary part, of the complex number.
208 typedef union {
209 float complex z;
210 float parts[2];
211 } float_complex;
213 typedef union {
214 double complex z;
215 double parts[2];
216 } double_complex;
218 typedef union {
219 long double complex z;
220 long double parts[2];
221 } long_double_complex;
223 #define REAL_PART(z) ((z).parts[0])
224 #define IMAG_PART(z) ((z).parts[1])
226 #endif /* _COMPLEX_H */
228 /* ieee style elementary functions */
229 extern double __ieee754_sqrt __P((double));
230 extern double __ieee754_acos __P((double));
231 extern double __ieee754_acosh __P((double));
232 extern double __ieee754_log __P((double));
233 extern double __ieee754_atanh __P((double));
234 extern double __ieee754_asin __P((double));
235 extern double __ieee754_atan2 __P((double,double));
236 extern double __ieee754_exp __P((double));
237 extern double __ieee754_cosh __P((double));
238 extern double __ieee754_fmod __P((double,double));
239 extern double __ieee754_pow __P((double,double));
240 extern double __ieee754_lgamma_r __P((double,int *));
241 extern double __ieee754_gamma_r __P((double,int *));
242 extern double __ieee754_lgamma __P((double));
243 extern double __ieee754_gamma __P((double));
244 extern double __ieee754_log10 __P((double));
245 extern double __ieee754_log2 __P((double));
246 extern double __ieee754_sinh __P((double));
247 extern double __ieee754_hypot __P((double,double));
248 extern double __ieee754_j0 __P((double));
249 extern double __ieee754_j1 __P((double));
250 extern double __ieee754_y0 __P((double));
251 extern double __ieee754_y1 __P((double));
252 extern double __ieee754_jn __P((int,double));
253 extern double __ieee754_yn __P((int,double));
254 extern double __ieee754_remainder __P((double,double));
255 extern int32_t __ieee754_rem_pio2 __P((double,double*));
256 extern double __ieee754_scalb __P((double,double));
258 /* fdlibm kernel function */
259 extern double __kernel_standard __P((double,double,int));
260 extern double __kernel_sin __P((double,double,int));
261 extern double __kernel_cos __P((double,double));
262 extern double __kernel_tan __P((double,double,int));
263 extern int __kernel_rem_pio2 __P((double*,double*,int,int,int,const int32_t*));
266 /* ieee style elementary float functions */
267 extern float __ieee754_sqrtf __P((float));
268 extern float __ieee754_acosf __P((float));
269 extern float __ieee754_acoshf __P((float));
270 extern float __ieee754_logf __P((float));
271 extern float __ieee754_atanhf __P((float));
272 extern float __ieee754_asinf __P((float));
273 extern float __ieee754_atan2f __P((float,float));
274 extern float __ieee754_expf __P((float));
275 extern float __ieee754_coshf __P((float));
276 extern float __ieee754_fmodf __P((float,float));
277 extern float __ieee754_powf __P((float,float));
278 extern float __ieee754_lgammaf_r __P((float,int *));
279 extern float __ieee754_gammaf_r __P((float,int *));
280 extern float __ieee754_lgammaf __P((float));
281 extern float __ieee754_gammaf __P((float));
282 extern float __ieee754_log10f __P((float));
283 extern float __ieee754_log2f __P((float));
284 extern float __ieee754_sinhf __P((float));
285 extern float __ieee754_hypotf __P((float,float));
286 extern float __ieee754_j0f __P((float));
287 extern float __ieee754_j1f __P((float));
288 extern float __ieee754_y0f __P((float));
289 extern float __ieee754_y1f __P((float));
290 extern float __ieee754_jnf __P((int,float));
291 extern float __ieee754_ynf __P((int,float));
292 extern float __ieee754_remainderf __P((float,float));
293 extern int32_t __ieee754_rem_pio2f __P((float,float*));
294 extern float __ieee754_scalbf __P((float,float));
296 /* float versions of fdlibm kernel functions */
297 extern float __kernel_sinf __P((float,float,int));
298 extern float __kernel_cosf __P((float,float));
299 extern float __kernel_tanf __P((float,float,int));
300 extern int __kernel_rem_pio2f __P((float*,float*,int,int,int,const int32_t*));
302 /* ieee style elementary long double functions */
303 extern long double __ieee754_fmodl(long double, long double);
304 extern long double __ieee754_sqrtl(long double);
307 * TRUNC() is a macro that sets the trailing 27 bits in the mantissa of an
308 * IEEE double variable to zero. It must be expression-like for syntactic
309 * reasons, and we implement this expression using an inline function
310 * instead of a pure macro to avoid depending on the gcc feature of
311 * statement-expressions.
313 #define TRUNC(d) (_b_trunc(&(d)))
315 static __inline void
316 _b_trunc(volatile double *_dp)
318 uint32_t _lw;
320 GET_LOW_WORD(_lw, *_dp);
321 SET_LOW_WORD(*_dp, _lw & 0xf8000000);
324 struct Double {
325 double a;
326 double b;
330 * Functions internal to the math package, yet not static.
332 double __exp__D(double, double);
333 struct Double __log__D(double);
335 #endif /* _MATH_PRIVATE_H_ */