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 (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved. 24 */ 25 26 /* 27 * We keep our own copy of this algorithm for 3 main reasons: 28 * 1. If we didn't, anyone modifying common/os/compress.c would 29 * directly break our on disk format 30 * 2. Our version of lzjb does not have a number of checks that the 31 * common/os version needs and uses 32 * 3. We initialize the lempel to ensure deterministic results, 33 * so that identical blocks can always be deduplicated. 34 * In particular, we are adding the "feature" that compress() can 35 * take a destination buffer size and returns the compressed length, or the 36 * source length if compression would overflow the destination buffer. 37 */ 38 39 #include <sys/types.h> 40 41 #define MATCH_BITS 6 42 #define MATCH_MIN 3 43 #define MATCH_MAX ((1 << MATCH_BITS) + (MATCH_MIN - 1)) 44 #define OFFSET_MASK ((1 << (16 - MATCH_BITS)) - 1) 45 #define LEMPEL_SIZE 1024 46 47 /*ARGSUSED*/ 48 size_t 49 lzjb_compress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n) 50 { 51 uchar_t *src = s_start; 52 uchar_t *dst = d_start; 53 uchar_t *cpy, *copymap; 54 int copymask = 1 << (NBBY - 1); 55 int mlen, offset, hash; 56 uint16_t *hp; 57 uint16_t lempel[LEMPEL_SIZE] = { 0 }; 58 59 while (src < (uchar_t *)s_start + s_len) { 60 if ((copymask <<= 1) == (1 << NBBY)) { 61 if (dst >= (uchar_t *)d_start + d_len - 1 - 2 * NBBY) 62 return (s_len); 63 copymask = 1; 64 copymap = dst; 65 *dst++ = 0; 66 } 67 if (src > (uchar_t *)s_start + s_len - MATCH_MAX) { 68 *dst++ = *src++; 69 continue; 70 } 71 hash = (src[0] << 16) + (src[1] << 8) + src[2]; 72 hash += hash >> 9; 73 hash += hash >> 5; 74 hp = &lempel[hash & (LEMPEL_SIZE - 1)]; 75 offset = (intptr_t)(src - *hp) & OFFSET_MASK; 76 *hp = (uint16_t)(uintptr_t)src; 77 cpy = src - offset; 78 if (cpy >= (uchar_t *)s_start && cpy != src && 79 src[0] == cpy[0] && src[1] == cpy[1] && src[2] == cpy[2]) { 80 *copymap |= copymask; 81 for (mlen = MATCH_MIN; mlen < MATCH_MAX; mlen++) 82 if (src[mlen] != cpy[mlen]) 83 break; 84 *dst++ = ((mlen - MATCH_MIN) << (NBBY - MATCH_BITS)) | 85 (offset >> NBBY); 86 *dst++ = (uchar_t)offset; 87 src += mlen; 88 } else { 89 *dst++ = *src++; 90 } 91 } 92 return (dst - (uchar_t *)d_start); 93 } 94 95 /*ARGSUSED*/ 96 int 97 lzjb_decompress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n) 98 { 99 uchar_t *src = s_start; 100 uchar_t *dst = d_start; 101 uchar_t *d_end = (uchar_t *)d_start + d_len; 102 uchar_t *cpy, copymap; 103 int copymask = 1 << (NBBY - 1); 104 105 while (dst < d_end) { 106 if ((copymask <<= 1) == (1 << NBBY)) { 107 copymask = 1; 108 copymap = *src++; 109 } 110 if (copymap & copymask) { 111 int mlen = (src[0] >> (NBBY - MATCH_BITS)) + MATCH_MIN; 112 int offset = ((src[0] << NBBY) | src[1]) & OFFSET_MASK; 113 src += 2; 114 if ((cpy = dst - offset) < (uchar_t *)d_start) 115 return (-1); 116 while (--mlen >= 0 && dst < d_end) 117 *dst++ = *cpy++; 118 } else { 119 *dst++ = *src++; 120 } 121 } 122 return (0); 123 } 124