xref: /titanic_44/usr/src/uts/common/fs/zfs/lzjb.c (revision 967a0a51c3e8ee257644dcad360e72dd4cf555f8)
1 /*
2  * CDDL HEADER START
3  *
4  * The contents of this file are subject to the terms of the
5  * Common Development and Distribution License (the "License").
6  * You may not use this file except in compliance with the License.
7  *
8  * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9  * or http://www.opensolaris.org/os/licensing.
10  * See the License for the specific language governing permissions
11  * and limitations under the License.
12  *
13  * When distributing Covered Code, include this CDDL HEADER in each
14  * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15  * If applicable, add the following below this CDDL HEADER, with the
16  * fields enclosed by brackets "[]" replaced with your own identifying
17  * information: Portions Copyright [yyyy] [name of copyright owner]
18  *
19  * CDDL HEADER END
20  */
21 
22 /*
23  * Copyright 2009 Sun Microsystems, Inc.  All rights reserved.
24  * Use is subject to license terms.
25  */
26 
27 /*
28  * We keep our own copy of this algorithm for 2 main reasons:
29  *	1. If we didn't, anyone modifying common/os/compress.c would
30  *         directly break our on disk format
31  *	2. Our version of lzjb does not have a number of checks that the
32  *         common/os version needs and uses
33  *	3. We initialize the lempel to ensure deterministic results,
34  *	   so that identical blocks can always be deduplicated.
35  * In particular, we are adding the "feature" that compress() can
36  * take a destination buffer size and return -1 if the data will not
37  * compress to d_len or less.
38  */
39 
40 #include <sys/types.h>
41 
42 #define	MATCH_BITS	6
43 #define	MATCH_MIN	3
44 #define	MATCH_MAX	((1 << MATCH_BITS) + (MATCH_MIN - 1))
45 #define	OFFSET_MASK	((1 << (16 - MATCH_BITS)) - 1)
46 #define	LEMPEL_SIZE	1024
47 
48 /*ARGSUSED*/
49 size_t
50 lzjb_compress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n)
51 {
52 	uchar_t *src = s_start;
53 	uchar_t *dst = d_start;
54 	uchar_t *cpy, *copymap;
55 	int copymask = 1 << (NBBY - 1);
56 	int mlen, offset, hash;
57 	uint16_t *hp;
58 	uint16_t lempel[LEMPEL_SIZE] = { 0 };
59 
60 	while (src < (uchar_t *)s_start + s_len) {
61 		if ((copymask <<= 1) == (1 << NBBY)) {
62 			if (dst >= (uchar_t *)d_start + d_len - 1 - 2 * NBBY)
63 				return (s_len);
64 			copymask = 1;
65 			copymap = dst;
66 			*dst++ = 0;
67 		}
68 		if (src > (uchar_t *)s_start + s_len - MATCH_MAX) {
69 			*dst++ = *src++;
70 			continue;
71 		}
72 		hash = (src[0] << 16) + (src[1] << 8) + src[2];
73 		hash += hash >> 9;
74 		hash += hash >> 5;
75 		hp = &lempel[hash & (LEMPEL_SIZE - 1)];
76 		offset = (intptr_t)(src - *hp) & OFFSET_MASK;
77 		*hp = (uint16_t)(uintptr_t)src;
78 		cpy = src - offset;
79 		if (cpy >= (uchar_t *)s_start && cpy != src &&
80 		    src[0] == cpy[0] && src[1] == cpy[1] && src[2] == cpy[2]) {
81 			*copymap |= copymask;
82 			for (mlen = MATCH_MIN; mlen < MATCH_MAX; mlen++)
83 				if (src[mlen] != cpy[mlen])
84 					break;
85 			*dst++ = ((mlen - MATCH_MIN) << (NBBY - MATCH_BITS)) |
86 			    (offset >> NBBY);
87 			*dst++ = (uchar_t)offset;
88 			src += mlen;
89 		} else {
90 			*dst++ = *src++;
91 		}
92 	}
93 	return (dst - (uchar_t *)d_start);
94 }
95 
96 /*ARGSUSED*/
97 int
98 lzjb_decompress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n)
99 {
100 	uchar_t *src = s_start;
101 	uchar_t *dst = d_start;
102 	uchar_t *d_end = (uchar_t *)d_start + d_len;
103 	uchar_t *cpy, copymap;
104 	int copymask = 1 << (NBBY - 1);
105 
106 	while (dst < d_end) {
107 		if ((copymask <<= 1) == (1 << NBBY)) {
108 			copymask = 1;
109 			copymap = *src++;
110 		}
111 		if (copymap & copymask) {
112 			int mlen = (src[0] >> (NBBY - MATCH_BITS)) + MATCH_MIN;
113 			int offset = ((src[0] << NBBY) | src[1]) & OFFSET_MASK;
114 			src += 2;
115 			if ((cpy = dst - offset) < (uchar_t *)d_start)
116 				return (-1);
117 			while (--mlen >= 0 && dst < d_end)
118 				*dst++ = *cpy++;
119 		} else {
120 			*dst++ = *src++;
121 		}
122 	}
123 	return (0);
124 }
125