Annotation of OpenXM_contrib/gmp/gmp-impl.h, Revision 1.1.1.1
1.1 maekawa 1: /* Include file for internal GNU MP types and definitions.
2:
3: Copyright (C) 1991, 1993, 1994, 1995, 1996 Free Software Foundation, Inc.
4:
5: This file is part of the GNU MP Library.
6:
7: The GNU MP Library is free software; you can redistribute it and/or modify
8: it under the terms of the GNU Library General Public License as published by
9: the Free Software Foundation; either version 2 of the License, or (at your
10: option) any later version.
11:
12: The GNU MP Library is distributed in the hope that it will be useful, but
13: WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
14: or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Library General Public
15: License for more details.
16:
17: You should have received a copy of the GNU Library General Public License
18: along with the GNU MP Library; see the file COPYING.LIB. If not, write to
19: the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
20: MA 02111-1307, USA. */
21:
22: /* When using gcc, make sure to use its builtin alloca. */
23: #if ! defined (alloca) && defined (__GNUC__)
24: #define alloca __builtin_alloca
25: #define HAVE_ALLOCA
26: #endif
27:
28: /* When using cc, do whatever necessary to allow use of alloca. For many
29: machines, this means including alloca.h. IBM's compilers need a #pragma
30: in "each module that needs to use alloca". */
31: #if ! defined (alloca)
32: /* We need lots of variants for MIPS, to cover all versions and perversions
33: of OSes for MIPS. */
34: #if defined (__mips) || defined (MIPSEL) || defined (MIPSEB) \
35: || defined (_MIPSEL) || defined (_MIPSEB) || defined (__sgi) \
36: || defined (__alpha) || defined (__sparc) || defined (sparc) \
37: || defined (__ksr__)
38: #include <alloca.h>
39: #define HAVE_ALLOCA
40: #endif
41: #if defined (_IBMR2)
42: #pragma alloca
43: #define HAVE_ALLOCA
44: #endif
45: #if defined (__DECC)
46: #define alloca(x) __ALLOCA(x)
47: #define HAVE_ALLOCA
48: #endif
49: #endif
50:
51: #if ! defined (HAVE_ALLOCA) || USE_STACK_ALLOC
52: #include "stack-alloc.h"
53: #else
54: #define TMP_DECL(m)
55: #define TMP_ALLOC(x) alloca(x)
56: #define TMP_MARK(m)
57: #define TMP_FREE(m)
58: #endif
59:
60: #ifndef NULL
61: #define NULL ((void *) 0)
62: #endif
63:
64: #if ! defined (__GNUC__)
65: #define inline /* Empty */
66: #endif
67:
68: #define ABS(x) (x >= 0 ? x : -x)
69: #define MIN(l,o) ((l) < (o) ? (l) : (o))
70: #define MAX(h,i) ((h) > (i) ? (h) : (i))
71:
72: /* Field access macros. */
73: #define SIZ(x) ((x)->_mp_size)
74: #define ABSIZ(x) ABS (SIZ (x))
75: #define PTR(x) ((x)->_mp_d)
76: #define EXP(x) ((x)->_mp_exp)
77: #define PREC(x) ((x)->_mp_prec)
78: #define ALLOC(x) ((x)->_mp_alloc)
79:
80: #include "gmp-mparam.h"
81: /* #include "longlong.h" */
82:
83: #if defined (__STDC__) || defined (__cplusplus)
84: void *malloc (size_t);
85: void *realloc (void *, size_t);
86: void free (void *);
87:
88: extern void * (*_mp_allocate_func) (size_t);
89: extern void * (*_mp_reallocate_func) (void *, size_t, size_t);
90: extern void (*_mp_free_func) (void *, size_t);
91:
92: void *_mp_default_allocate (size_t);
93: void *_mp_default_reallocate (void *, size_t, size_t);
94: void _mp_default_free (void *, size_t);
95:
96: #else
97:
98: #define const /* Empty */
99: #define signed /* Empty */
100:
101: void *malloc ();
102: void *realloc ();
103: void free ();
104:
105: extern void * (*_mp_allocate_func) ();
106: extern void * (*_mp_reallocate_func) ();
107: extern void (*_mp_free_func) ();
108:
109: void *_mp_default_allocate ();
110: void *_mp_default_reallocate ();
111: void _mp_default_free ();
112: #endif
113:
114: /* Copy NLIMBS *limbs* from SRC to DST. */
115: #define MPN_COPY_INCR(DST, SRC, NLIMBS) \
116: do { \
117: mp_size_t __i; \
118: for (__i = 0; __i < (NLIMBS); __i++) \
119: (DST)[__i] = (SRC)[__i]; \
120: } while (0)
121: #define MPN_COPY_DECR(DST, SRC, NLIMBS) \
122: do { \
123: mp_size_t __i; \
124: for (__i = (NLIMBS) - 1; __i >= 0; __i--) \
125: (DST)[__i] = (SRC)[__i]; \
126: } while (0)
127: #define MPN_COPY MPN_COPY_INCR
128:
129: /* Zero NLIMBS *limbs* AT DST. */
130: #define MPN_ZERO(DST, NLIMBS) \
131: do { \
132: mp_size_t __i; \
133: for (__i = 0; __i < (NLIMBS); __i++) \
134: (DST)[__i] = 0; \
135: } while (0)
136:
137: #define MPN_NORMALIZE(DST, NLIMBS) \
138: do { \
139: while (NLIMBS > 0) \
140: { \
141: if ((DST)[(NLIMBS) - 1] != 0) \
142: break; \
143: NLIMBS--; \
144: } \
145: } while (0)
146: #define MPN_NORMALIZE_NOT_ZERO(DST, NLIMBS) \
147: do { \
148: while (1) \
149: { \
150: if ((DST)[(NLIMBS) - 1] != 0) \
151: break; \
152: NLIMBS--; \
153: } \
154: } while (0)
155:
156: /* Initialize the MP_INT X with space for NLIMBS limbs.
157: X should be a temporary variable, and it will be automatically
158: cleared out when the running function returns.
159: We use __x here to make it possible to accept both mpz_ptr and mpz_t
160: arguments. */
161: #define MPZ_TMP_INIT(X, NLIMBS) \
162: do { \
163: mpz_ptr __x = (X); \
164: __x->_mp_alloc = (NLIMBS); \
165: __x->_mp_d = (mp_ptr) TMP_ALLOC ((NLIMBS) * BYTES_PER_MP_LIMB); \
166: } while (0)
167:
168: #define MPN_MUL_N_RECURSE(prodp, up, vp, size, tspace) \
169: do { \
170: if ((size) < KARATSUBA_THRESHOLD) \
171: impn_mul_n_basecase (prodp, up, vp, size); \
172: else \
173: impn_mul_n (prodp, up, vp, size, tspace); \
174: } while (0);
175: #define MPN_SQR_N_RECURSE(prodp, up, size, tspace) \
176: do { \
177: if ((size) < KARATSUBA_THRESHOLD) \
178: impn_sqr_n_basecase (prodp, up, size); \
179: else \
180: impn_sqr_n (prodp, up, size, tspace); \
181: } while (0);
182:
183: /* Structure for conversion between internal binary format and
184: strings in base 2..36. */
185: struct bases
186: {
187: /* Number of digits in the conversion base that always fits in an mp_limb_t.
188: For example, for base 10 on a machine where a mp_limb_t has 32 bits this
189: is 9, since 10**9 is the largest number that fits into a mp_limb_t. */
190: int chars_per_limb;
191:
192: /* log(2)/log(conversion_base) */
193: float chars_per_bit_exactly;
194:
195: /* base**chars_per_limb, i.e. the biggest number that fits a word, built by
196: factors of base. Exception: For 2, 4, 8, etc, big_base is log2(base),
197: i.e. the number of bits used to represent each digit in the base. */
198: mp_limb_t big_base;
199:
200: /* A BITS_PER_MP_LIMB bit approximation to 1/big_base, represented as a
201: fixed-point number. Instead of dividing by big_base an application can
202: choose to multiply by big_base_inverted. */
203: mp_limb_t big_base_inverted;
204: };
205:
206: extern const struct bases __mp_bases[];
207: extern mp_size_t __gmp_default_fp_limb_precision;
208:
209: /* Divide the two-limb number in (NH,,NL) by D, with DI being the largest
210: limb not larger than (2**(2*BITS_PER_MP_LIMB))/D - (2**BITS_PER_MP_LIMB).
211: If this would yield overflow, DI should be the largest possible number
212: (i.e., only ones). For correct operation, the most significant bit of D
213: has to be set. Put the quotient in Q and the remainder in R. */
214: #define udiv_qrnnd_preinv(q, r, nh, nl, d, di) \
215: do { \
216: mp_limb_t _q, _ql, _r; \
217: mp_limb_t _xh, _xl; \
218: umul_ppmm (_q, _ql, (nh), (di)); \
219: _q += (nh); /* DI is 2**BITS_PER_MP_LIMB too small */\
220: umul_ppmm (_xh, _xl, _q, (d)); \
221: sub_ddmmss (_xh, _r, (nh), (nl), _xh, _xl); \
222: if (_xh != 0) \
223: { \
224: sub_ddmmss (_xh, _r, _xh, _r, 0, (d)); \
225: _q += 1; \
226: if (_xh != 0) \
227: { \
228: sub_ddmmss (_xh, _r, _xh, _r, 0, (d)); \
229: _q += 1; \
230: } \
231: } \
232: if (_r >= (d)) \
233: { \
234: _r -= (d); \
235: _q += 1; \
236: } \
237: (r) = _r; \
238: (q) = _q; \
239: } while (0)
240: /* Like udiv_qrnnd_preinv, but for for any value D. DNORM is D shifted left
241: so that its most significant bit is set. LGUP is ceil(log2(D)). */
242: #define udiv_qrnnd_preinv2gen(q, r, nh, nl, d, di, dnorm, lgup) \
243: do { \
244: mp_limb_t n2, n10, n1, nadj, q1; \
245: mp_limb_t _xh, _xl; \
246: n2 = ((nh) << (BITS_PER_MP_LIMB - (lgup))) + ((nl) >> 1 >> (l - 1));\
247: n10 = (nl) << (BITS_PER_MP_LIMB - (lgup)); \
248: n1 = ((mp_limb_signed_t) n10 >> (BITS_PER_MP_LIMB - 1)); \
249: nadj = n10 + (n1 & (dnorm)); \
250: umul_ppmm (_xh, _xl, di, n2 - n1); \
251: add_ssaaaa (_xh, _xl, _xh, _xl, 0, nadj); \
252: q1 = ~(n2 + _xh); \
253: umul_ppmm (_xh, _xl, q1, d); \
254: add_ssaaaa (_xh, _xl, _xh, _xl, nh, nl); \
255: _xh -= (d); \
256: (r) = _xl + ((d) & _xh); \
257: (q) = _xh - q1; \
258: } while (0)
259: /* Exactly like udiv_qrnnd_preinv, but branch-free. It is not clear which
260: version to use. */
261: #define udiv_qrnnd_preinv2norm(q, r, nh, nl, d, di) \
262: do { \
263: mp_limb_t n2, n10, n1, nadj, q1; \
264: mp_limb_t _xh, _xl; \
265: n2 = (nh); \
266: n10 = (nl); \
267: n1 = ((mp_limb_signed_t) n10 >> (BITS_PER_MP_LIMB - 1)); \
268: nadj = n10 + (n1 & (d)); \
269: umul_ppmm (_xh, _xl, di, n2 - n1); \
270: add_ssaaaa (_xh, _xl, _xh, _xl, 0, nadj); \
271: q1 = ~(n2 + _xh); \
272: umul_ppmm (_xh, _xl, q1, d); \
273: add_ssaaaa (_xh, _xl, _xh, _xl, nh, nl); \
274: _xh -= (d); \
275: (r) = _xl + ((d) & _xh); \
276: (q) = _xh - q1; \
277: } while (0)
278:
279: #if defined (__GNUC__)
280: /* Define stuff for longlong.h. */
281: typedef unsigned int UQItype __attribute__ ((mode (QI)));
282: typedef int SItype __attribute__ ((mode (SI)));
283: typedef unsigned int USItype __attribute__ ((mode (SI)));
284: typedef int DItype __attribute__ ((mode (DI)));
285: typedef unsigned int UDItype __attribute__ ((mode (DI)));
286: #else
287: typedef unsigned char UQItype;
288: typedef long SItype;
289: typedef unsigned long USItype;
290: #endif
291:
292: typedef mp_limb_t UWtype;
293: typedef unsigned int UHWtype;
294: #define W_TYPE_SIZE BITS_PER_MP_LIMB
295:
296: /* Internal mpn calls */
297: #define impn_mul_n_basecase __MPN(impn_mul_n_basecase)
298: #define impn_mul_n __MPN(impn_mul_n)
299: #define impn_sqr_n_basecase __MPN(impn_sqr_n_basecase)
300: #define impn_sqr_n __MPN(impn_sqr_n)
301:
302: /* Define ieee_double_extract and _GMP_IEEE_FLOATS. */
303:
304: #if defined (_LITTLE_ENDIAN) || defined (__LITTLE_ENDIAN__) \
305: || defined (__alpha) \
306: || (defined (__arm__) && defined (__ARMWEL__)) \
307: || defined (__clipper__) \
308: || defined (__cris) \
309: || defined (__i386__) \
310: || defined (__i860__) \
311: || defined (__i960__) \
312: || defined (MIPSEL) || defined (_MIPSEL) \
313: || defined (__ns32000__) \
314: || defined (__WINNT) || defined (_WIN32)
315: #define _GMP_IEEE_FLOATS 1
316: union ieee_double_extract
317: {
318: struct
319: {
320: unsigned int manl:32;
321: unsigned int manh:20;
322: unsigned int exp:11;
323: unsigned int sig:1;
324: } s;
325: double d;
326: };
327: #else /* Need this as an #else since the tests aren't made exclusive. */
328: #if defined (_BIG_ENDIAN) \
329: || defined (__a29k__) || defined (_AM29K) \
330: || defined (__arm__) \
331: || (defined (__convex__) && defined (_IEEE_FLOAT_)) \
332: || defined (__i370__) || defined (__mvs__) \
333: || defined (__mc68000__) || defined (__mc68020__) || defined (__NeXT__)\
334: || defined(mc68020) \
335: || defined (__m88000__) \
336: || defined (MIPSEB) || defined (_MIPSEB) \
337: || defined (__hppa) \
338: || defined (__pyr__) \
339: || defined (__ibm032__) \
340: || defined (_IBMR2) || defined (_ARCH_PPC) \
341: || defined (__sh__) \
342: || defined (__sparc) || defined (sparc) \
343: || defined (__we32k__)
344: #define _GMP_IEEE_FLOATS 1
345: union ieee_double_extract
346: {
347: struct
348: {
349: unsigned int sig:1;
350: unsigned int exp:11;
351: unsigned int manh:20;
352: unsigned int manl:32;
353: } s;
354: double d;
355: };
356: #endif
357: #endif
358:
359: #define MP_BASE_AS_DOUBLE (2.0 * ((mp_limb_t) 1 << (BITS_PER_MP_LIMB - 1)))
360: #if BITS_PER_MP_LIMB == 64
361: #define LIMBS_PER_DOUBLE 2
362: #else
363: #define LIMBS_PER_DOUBLE 3
364: #endif
365:
366: double __gmp_scale2 _PROTO ((double, int));
367: int __gmp_extract_double _PROTO((mp_ptr, double));
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>