1/*-
2 * Copyright (c) 1992, 1993
3 * The Regents of the University of California. All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Peter McIlroy.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * 3. All advertising materials mentioning features or use of this software
17 * must display the following acknowledgement:
18 * This product includes software developed by the University of
19 * California, Berkeley and its contributors.
20 * 4. Neither the name of the University nor the names of its contributors
21 * may be used to endorse or promote products derived from this software
22 * without specific prior written permission.
23 *
24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * SUCH DAMAGE.
35 */
36
37/* $Id$ */
38
39#if defined(LIBC_SCCS) && !defined(lint)
40static char sccsid[] = "@(#)merge.c 8.2 (Berkeley) 2/14/94";
41#endif /* LIBC_SCCS and not lint */
42
43/*
44 * Hybrid exponential search/linear search merge sort with hybrid
45 * natural/pairwise first pass. Requires about .3% more comparisons
46 * for random data than LSMS with pairwise first pass alone.
47 * It works for objects as small as two bytes.
48 */
49
50#define NATURAL
51#define THRESHOLD 16 /* Best choice for natural merge cut-off. */
52
53/* #define NATURAL to get hybrid natural merge.
54 * (The default is pairwise merging.)
55 */
56
57#include <sys/types.h>
58
59#include <errno.h>
60#include <stdlib.h>
61#include <string.h>
62
63#include "php.h"
64
65#ifdef PHP_WIN32
66#include <winsock2.h> /* Includes definition for u_char */
67#endif
68
69static void setup(u_char *list1, u_char *list2, size_t n, size_t size, int (*cmp)(const void *, const void * TSRMLS_DC) TSRMLS_DC);
70static void insertionsort(u_char *a, size_t n, size_t size, int (*cmp)(const void *, const void * TSRMLS_DC) TSRMLS_DC);
71
72#define ISIZE sizeof(int)
73#define PSIZE sizeof(u_char *)
74#define ICOPY_LIST(src, dst, last) \
75 do \
76 *(int*)dst = *(int*)src, src += ISIZE, dst += ISIZE; \
77 while(src < last)
78#define ICOPY_ELT(src, dst, i) \
79 do \
80 *(int*) dst = *(int*) src, src += ISIZE, dst += ISIZE; \
81 while (i -= ISIZE)
82
83#define CCOPY_LIST(src, dst, last) \
84 do \
85 *dst++ = *src++; \
86 while (src < last)
87#define CCOPY_ELT(src, dst, i) \
88 do \
89 *dst++ = *src++; \
90 while (i -= 1)
91
92/*
93 * Find the next possible pointer head. (Trickery for forcing an array
94 * to do double duty as a linked list when objects do not align with word
95 * boundaries.
96 */
97/* Assumption: PSIZE is a power of 2. */
98#define EVAL(p) (u_char **) \
99 ((u_char *)0 + \
100 (((u_char *)p + PSIZE - 1 - (u_char *) 0) & ~(PSIZE - 1)))
101
102/* {{{ php_mergesort
103 * Arguments are as for qsort.
104 */
105PHPAPI int php_mergesort(void *base, size_t nmemb, size_t size, int (*cmp)(const void *, const void * TSRMLS_DC) TSRMLS_DC)
106{
107 register unsigned int i;
108 register int sense;
109 int big, iflag;
110 register u_char *f1, *f2, *t, *b, *tp2, *q, *l1, *l2;
111 u_char *list2, *list1, *p2, *p, *last, **p1;
112
113 if (size < PSIZE / 2) { /* Pointers must fit into 2 * size. */
114 errno = EINVAL;
115 return (-1);
116 }
117
118 if (nmemb == 0)
119 return (0);
120
121 /*
122 * XXX
123 * Stupid subtraction for the Cray.
124 */
125 iflag = 0;
126 if (!(size % ISIZE) && !(((char *)base - (char *)0) % ISIZE))
127 iflag = 1;
128
129 if ((list2 = malloc(nmemb * size + PSIZE)) == NULL)
130 return (-1);
131
132 list1 = base;
133 setup(list1, list2, nmemb, size, cmp TSRMLS_CC);
134 last = list2 + nmemb * size;
135 i = big = 0;
136 while (*EVAL(list2) != last) {
137 l2 = list1;
138 p1 = EVAL(list1);
139 for (tp2 = p2 = list2; p2 != last; p1 = EVAL(l2)) {
140 p2 = *EVAL(p2);
141 f1 = l2;
142 f2 = l1 = list1 + (p2 - list2);
143 if (p2 != last)
144 p2 = *EVAL(p2);
145 l2 = list1 + (p2 - list2);
146 while (f1 < l1 && f2 < l2) {
147 if ((*cmp)(f1, f2 TSRMLS_CC) <= 0) {
148 q = f2;
149 b = f1, t = l1;
150 sense = -1;
151 } else {
152 q = f1;
153 b = f2, t = l2;
154 sense = 0;
155 }
156 if (!big) { /* here i = 0 */
157 while ((b += size) < t && cmp(q, b TSRMLS_CC) >sense)
158 if (++i == 6) {
159 big = 1;
160 goto EXPONENTIAL;
161 }
162 } else {
163EXPONENTIAL: for (i = size; ; i <<= 1)
164 if ((p = (b + i)) >= t) {
165 if ((p = t - size) > b &&
166 (*cmp)(q, p TSRMLS_CC) <= sense)
167 t = p;
168 else
169 b = p;
170 break;
171 } else if ((*cmp)(q, p TSRMLS_CC) <= sense) {
172 t = p;
173 if (i == size)
174 big = 0;
175 goto FASTCASE;
176 } else
177 b = p;
178 while (t > b+size) {
179 i = (((t - b) / size) >> 1) * size;
180 if ((*cmp)(q, p = b + i TSRMLS_CC) <= sense)
181 t = p;
182 else
183 b = p;
184 }
185 goto COPY;
186FASTCASE: while (i > size)
187 if ((*cmp)(q,
188 p = b + (i >>= 1) TSRMLS_CC) <= sense)
189 t = p;
190 else
191 b = p;
192COPY: b = t;
193 }
194 i = size;
195 if (q == f1) {
196 if (iflag) {
197 ICOPY_LIST(f2, tp2, b);
198 ICOPY_ELT(f1, tp2, i);
199 } else {
200 CCOPY_LIST(f2, tp2, b);
201 CCOPY_ELT(f1, tp2, i);
202 }
203 } else {
204 if (iflag) {
205 ICOPY_LIST(f1, tp2, b);
206 ICOPY_ELT(f2, tp2, i);
207 } else {
208 CCOPY_LIST(f1, tp2, b);
209 CCOPY_ELT(f2, tp2, i);
210 }
211 }
212 }
213 if (f2 < l2) {
214 if (iflag)
215 ICOPY_LIST(f2, tp2, l2);
216 else
217 CCOPY_LIST(f2, tp2, l2);
218 } else if (f1 < l1) {
219 if (iflag)
220 ICOPY_LIST(f1, tp2, l1);
221 else
222 CCOPY_LIST(f1, tp2, l1);
223 }
224 *p1 = l2;
225 }
226 tp2 = list1; /* swap list1, list2 */
227 list1 = list2;
228 list2 = tp2;
229 last = list2 + nmemb*size;
230 }
231 if (base == list2) {
232 memmove(list2, list1, nmemb*size);
233 list2 = list1;
234 }
235 free(list2);
236 return (0);
237}
238/* }}} */
239
240#define swap(a, b) { \
241 s = b; \
242 i = size; \
243 do { \
244 tmp = *a; *a++ = *s; *s++ = tmp; \
245 } while (--i); \
246 a -= size; \
247 }
248#define reverse(bot, top) { \
249 s = top; \
250 do { \
251 i = size; \
252 do { \
253 tmp = *bot; *bot++ = *s; *s++ = tmp; \
254 } while (--i); \
255 s -= size2; \
256 } while(bot < s); \
257}
258
259/* {{{ setup
260 * Optional hybrid natural/pairwise first pass. Eats up list1 in runs of
261 * increasing order, list2 in a corresponding linked list. Checks for runs
262 * when THRESHOLD/2 pairs compare with same sense. (Only used when NATURAL
263 * is defined. Otherwise simple pairwise merging is used.)
264 */
265static void setup(u_char *list1, u_char *list2, size_t n, size_t size, int (*cmp)(const void *, const void * TSRMLS_DC) TSRMLS_DC)
266{
267 int i, length, size2, tmp, sense;
268 u_char *f1, *f2, *s, *l2, *last, *p2;
269
270 size2 = size*2;
271 if (n <= 5) {
272 insertionsort(list1, n, size, cmp TSRMLS_CC);
273 *EVAL(list2) = (u_char*) list2 + n*size;
274 return;
275 }
276 /*
277 * Avoid running pointers out of bounds; limit n to evens
278 * for simplicity.
279 */
280 i = 4 + (n & 1);
281 insertionsort(list1 + (n - i) * size, i, size, cmp TSRMLS_CC);
282 last = list1 + size * (n - i);
283 *EVAL(list2 + (last - list1)) = list2 + n * size;
284
285#ifdef NATURAL
286 p2 = list2;
287 f1 = list1;
288 sense = (cmp(f1, f1 + size TSRMLS_CC) > 0);
289 for (; f1 < last; sense = !sense) {
290 length = 2;
291 /* Find pairs with same sense. */
292 for (f2 = f1 + size2; f2 < last; f2 += size2) {
293 if ((cmp(f2, f2+ size TSRMLS_CC) > 0) != sense)
294 break;
295 length += 2;
296 }
297 if (length < THRESHOLD) { /* Pairwise merge */
298 do {
299 p2 = *EVAL(p2) = f1 + size2 - list1 + list2;
300 if (sense > 0)
301 swap (f1, f1 + size);
302 } while ((f1 += size2) < f2);
303 } else { /* Natural merge */
304 l2 = f2;
305 for (f2 = f1 + size2; f2 < l2; f2 += size2) {
306 if ((cmp(f2-size, f2 TSRMLS_CC) > 0) != sense) {
307 p2 = *EVAL(p2) = f2 - list1 + list2;
308 if (sense > 0)
309 reverse(f1, f2-size);
310 f1 = f2;
311 }
312 }
313 if (sense > 0)
314 reverse (f1, f2-size);
315 f1 = f2;
316 if (f2 < last || cmp(f2 - size, f2 TSRMLS_CC) > 0)
317 p2 = *EVAL(p2) = f2 - list1 + list2;
318 else
319 p2 = *EVAL(p2) = list2 + n*size;
320 }
321 }
322#else /* pairwise merge only. */
323 for (f1 = list1, p2 = list2; f1 < last; f1 += size2) {
324 p2 = *EVAL(p2) = p2 + size2;
325 if (cmp (f1, f1 + size TSRMLS_CC) > 0)
326 swap(f1, f1 + size);
327 }
328#endif /* NATURAL */
329}
330/* }}} */
331
332/* {{{ insertionsort
333 * This is to avoid out-of-bounds addresses in sorting the
334 * last 4 elements.
335 */
336static void insertionsort(u_char *a, size_t n, size_t size, int (*cmp)(const void *, const void * TSRMLS_DC) TSRMLS_DC)
337{
338 u_char *ai, *s, *t, *u, tmp;
339 int i;
340
341 for (ai = a+size; --n >= 1; ai += size)
342 for (t = ai; t > a; t -= size) {
343 u = t - size;
344 if (cmp(u, t TSRMLS_CC) <= 0)
345 break;
346 swap(u, t);
347 }
348}
349/* }}} */
350
351/*
352 * Local variables:
353 * tab-width: 4
354 * c-basic-offset: 4
355 * End:
356 * vim600: fdm=marker
357 * vim: noet sw=4 ts=4
358 */
359