Home | History | Annotate | Line # | Download | only in stdlib
qsort.c revision 1.3
      1 /*-
      2  * Copyright (c) 1980, 1983, 1990 The Regents of the University of California.
      3  * All rights reserved.
      4  *
      5  * Redistribution and use in source and binary forms, with or without
      6  * modification, are permitted provided that the following conditions
      7  * are met:
      8  * 1. Redistributions of source code must retain the above copyright
      9  *    notice, this list of conditions and the following disclaimer.
     10  * 2. Redistributions in binary form must reproduce the above copyright
     11  *    notice, this list of conditions and the following disclaimer in the
     12  *    documentation and/or other materials provided with the distribution.
     13  * 3. All advertising materials mentioning features or use of this software
     14  *    must display the following acknowledgement:
     15  *	This product includes software developed by the University of
     16  *	California, Berkeley and its contributors.
     17  * 4. Neither the name of the University nor the names of its contributors
     18  *    may be used to endorse or promote products derived from this software
     19  *    without specific prior written permission.
     20  *
     21  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     22  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     23  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     24  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     25  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     26  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     27  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     28  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     29  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     30  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     31  * SUCH DAMAGE.
     32  */
     33 
     34 #if defined(LIBC_SCCS) && !defined(lint)
     35 /*static char *sccsid = "from: @(#)qsort.c	5.9 (Berkeley) 2/23/91";*/
     36 static char *rcsid = "$Id: qsort.c,v 1.3 1993/08/26 00:48:06 jtc Exp $";
     37 #endif /* LIBC_SCCS and not lint */
     38 
     39 #include <sys/types.h>
     40 #include <stdlib.h>
     41 
     42 /*
     43  * MTHRESH is the smallest partition for which we compare for a median
     44  * value instead of using the middle value.
     45  */
     46 #define	MTHRESH	6
     47 
     48 /*
     49  * THRESH is the minimum number of entries in a partition for continued
     50  * partitioning.
     51  */
     52 #define	THRESH	4
     53 
     54 void
     55 qsort(bot, nmemb, size, compar)
     56 	void *bot;
     57 	size_t nmemb, size;
     58 	int (*compar) __P((const void *, const void *));
     59 {
     60 	static void insertion_sort(), quick_sort();
     61 
     62 	if (nmemb <= 1)
     63 		return;
     64 
     65 	if (nmemb >= THRESH)
     66 		quick_sort(bot, nmemb, size, compar);
     67 	else
     68 		insertion_sort(bot, nmemb, size, compar);
     69 }
     70 
     71 /*
     72  * Swap two areas of size number of bytes.  Although qsort(3) permits random
     73  * blocks of memory to be sorted, sorting pointers is almost certainly the
     74  * common case (and, were it not, could easily be made so).  Regardless, it
     75  * isn't worth optimizing; the SWAP's get sped up by the cache, and pointer
     76  * arithmetic gets lost in the time required for comparison function calls.
     77  */
     78 #define	SWAP(a, b) { \
     79 	cnt = size; \
     80 	do { \
     81 		ch = *a; \
     82 		*a++ = *b; \
     83 		*b++ = ch; \
     84 	} while (--cnt); \
     85 }
     86 
     87 /*
     88  * Knuth, Vol. 3, page 116, Algorithm Q, step b, argues that a single pass
     89  * of straight insertion sort after partitioning is complete is better than
     90  * sorting each small partition as it is created.  This isn't correct in this
     91  * implementation because comparisons require at least one (and often two)
     92  * function calls and are likely to be the dominating expense of the sort.
     93  * Doing a final insertion sort does more comparisons than are necessary
     94  * because it compares the "edges" and medians of the partitions which are
     95  * known to be already sorted.
     96  *
     97  * This is also the reasoning behind selecting a small THRESH value (see
     98  * Knuth, page 122, equation 26), since the quicksort algorithm does less
     99  * comparisons than the insertion sort.
    100  */
    101 #define	SORT(bot, n) { \
    102 	if (n > 1) \
    103 		if (n == 2) { \
    104 			t1 = bot + size; \
    105 			if (compar(t1, bot) < 0) \
    106 				SWAP(t1, bot); \
    107 		} else \
    108 			insertion_sort(bot, n, size, compar); \
    109 }
    110 
    111 static void
    112 quick_sort(bot, nmemb, size, compar)
    113 	register char *bot;
    114 	register int size;
    115 	int nmemb, (*compar)();
    116 {
    117 	register int cnt;
    118 	register u_char ch;
    119 	register char *top, *mid, *t1, *t2;
    120 	register int n1, n2;
    121 	char *bsv;
    122 	static void insertion_sort();
    123 
    124 	/* bot and nmemb must already be set. */
    125 partition:
    126 
    127 	/* find mid and top elements */
    128 	mid = bot + size * (nmemb >> 1);
    129 	top = bot + (nmemb - 1) * size;
    130 
    131 	/*
    132 	 * Find the median of the first, last and middle element (see Knuth,
    133 	 * Vol. 3, page 123, Eq. 28).  This test order gets the equalities
    134 	 * right.
    135 	 */
    136 	if (nmemb >= MTHRESH) {
    137 		n1 = compar(bot, mid);
    138 		n2 = compar(mid, top);
    139 		if (n1 < 0 && n2 > 0)
    140 			t1 = compar(bot, top) < 0 ? top : bot;
    141 		else if (n1 > 0 && n2 < 0)
    142 			t1 = compar(bot, top) > 0 ? top : bot;
    143 		else
    144 			t1 = mid;
    145 
    146 		/* if mid element not selected, swap selection there */
    147 		if (t1 != mid) {
    148 			SWAP(t1, mid);
    149 			mid -= size;
    150 		}
    151 	}
    152 
    153 	/* Standard quicksort, Knuth, Vol. 3, page 116, Algorithm Q. */
    154 #define	didswap	n1
    155 #define	newbot	t1
    156 #define	replace	t2
    157 	didswap = 0;
    158 	for (bsv = bot;;) {
    159 		for (; bot < mid && compar(bot, mid) <= 0; bot += size);
    160 		while (top > mid) {
    161 			if (compar(mid, top) <= 0) {
    162 				top -= size;
    163 				continue;
    164 			}
    165 			newbot = bot + size;	/* value of bot after swap */
    166 			if (bot == mid)		/* top <-> mid, mid == top */
    167 				replace = mid = top;
    168 			else {			/* bot <-> top */
    169 				replace = top;
    170 				top -= size;
    171 			}
    172 			goto swap;
    173 		}
    174 		if (bot == mid)
    175 			break;
    176 
    177 		/* bot <-> mid, mid == bot */
    178 		replace = mid;
    179 		newbot = mid = bot;		/* value of bot after swap */
    180 		top -= size;
    181 
    182 swap:		SWAP(bot, replace);
    183 		bot = newbot;
    184 		didswap = 1;
    185 	}
    186 
    187 	/*
    188 	 * Quicksort behaves badly in the presence of data which is already
    189 	 * sorted (see Knuth, Vol. 3, page 119) going from O N lg N to O N^2.
    190 	 * To avoid this worst case behavior, if a re-partitioning occurs
    191 	 * without swapping any elements, it is not further partitioned and
    192 	 * is insert sorted.  This wins big with almost sorted data sets and
    193 	 * only loses if the data set is very strangely partitioned.  A fix
    194 	 * for those data sets would be to return prematurely if the insertion
    195 	 * sort routine is forced to make an excessive number of swaps, and
    196 	 * continue the partitioning.
    197 	 */
    198 	if (!didswap) {
    199 		insertion_sort(bsv, nmemb, size, compar);
    200 		return;
    201 	}
    202 
    203 	/*
    204 	 * Re-partition or sort as necessary.  Note that the mid element
    205 	 * itself is correctly positioned and can be ignored.
    206 	 */
    207 #define	nlower	n1
    208 #define	nupper	n2
    209 	bot = bsv;
    210 	nlower = (mid - bot) / size;	/* size of lower partition */
    211 	mid += size;
    212 	nupper = nmemb - nlower - 1;	/* size of upper partition */
    213 
    214 	/*
    215 	 * If must call recursively, do it on the smaller partition; this
    216 	 * bounds the stack to lg N entries.
    217 	 */
    218 	if (nlower > nupper) {
    219 		if (nupper >= THRESH)
    220 			quick_sort(mid, nupper, size, compar);
    221 		else {
    222 			SORT(mid, nupper);
    223 			if (nlower < THRESH) {
    224 				SORT(bot, nlower);
    225 				return;
    226 			}
    227 		}
    228 		nmemb = nlower;
    229 	} else {
    230 		if (nlower >= THRESH)
    231 			quick_sort(bot, nlower, size, compar);
    232 		else {
    233 			SORT(bot, nlower);
    234 			if (nupper < THRESH) {
    235 				SORT(mid, nupper);
    236 				return;
    237 			}
    238 		}
    239 		bot = mid;
    240 		nmemb = nupper;
    241 	}
    242 	goto partition;
    243 	/* NOTREACHED */
    244 }
    245 
    246 static void
    247 insertion_sort(bot, nmemb, size, compar)
    248 	char *bot;
    249 	register int size;
    250 	int nmemb, (*compar)();
    251 {
    252 	register int cnt;
    253 	register u_char ch;
    254 	register char *s1, *s2, *t1, *t2, *top;
    255 
    256 	/*
    257 	 * A simple insertion sort (see Knuth, Vol. 3, page 81, Algorithm
    258 	 * S).  Insertion sort has the same worst case as most simple sorts
    259 	 * (O N^2).  It gets used here because it is (O N) in the case of
    260 	 * sorted data.
    261 	 */
    262 	top = bot + nmemb * size;
    263 	for (t1 = bot + size; t1 < top;) {
    264 		for (t2 = t1; (t2 -= size) >= bot && compar(t1, t2) < 0;);
    265 		if (t1 != (t2 += size)) {
    266 			/* Bubble bytes up through each element. */
    267 			for (cnt = size; cnt--; ++t1) {
    268 				ch = *t1;
    269 				for (s1 = s2 = t1; (s2 -= size) >= t2; s1 = s2)
    270 					*s1 = *s2;
    271 				*s1 = ch;
    272 			}
    273 		} else
    274 			t1 += size;
    275 	}
    276 }
    277