xref: /freebsd/usr.bin/mkuzip/mkuz_blockcache.c (revision 02e9120893770924227138ba49df1edb3896112a)
1 /*
2  * Copyright (c) 2016 Maxim Sobolev <sobomax@FreeBSD.org>
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  *
14  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
15  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
16  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
17  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
18  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
19  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
20  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
21  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
22  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
23  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
24  * SUCH DAMAGE.
25  */
26 
27 #include <sys/cdefs.h>
28 #include <err.h>
29 #include <stdio.h>
30 #include <stdlib.h>
31 #include <string.h>
32 #include <unistd.h>
33 
34 #if defined(MKUZ_DEBUG)
35 # include <assert.h>
36 # include <stdio.h>
37 #endif
38 
39 #include "mkuz_blockcache.h"
40 #include "mkuz_blk.h"
41 
42 struct mkuz_blkcache_itm {
43     struct mkuz_blk_info hit;
44     struct mkuz_blkcache_itm *next;
45 };
46 
47 static struct mkuz_blkcache {
48     struct mkuz_blkcache_itm first[256];
49 } blkcache;
50 
51 static int
52 verify_match(int fd, const struct mkuz_blk *cbp, struct mkuz_blkcache_itm *bcep)
53 {
54     void *vbuf;
55     ssize_t rlen;
56     int rval;
57 
58     rval = -1;
59     vbuf = malloc(cbp->info.len);
60     if (vbuf == NULL) {
61         goto e0;
62     }
63     if (lseek(fd, bcep->hit.offset, SEEK_SET) < 0) {
64         goto e1;
65     }
66     rlen = read(fd, vbuf, cbp->info.len);
67     if (rlen < 0 || (unsigned)rlen != cbp->info.len) {
68         goto e2;
69     }
70     rval = (memcmp(cbp->data, vbuf, cbp->info.len) == 0) ? 1 : 0;
71 e2:
72     lseek(fd, cbp->info.offset, SEEK_SET);
73 e1:
74     free(vbuf);
75 e0:
76     return (rval);
77 }
78 
79 #define I2J(x)	((intmax_t)(x))
80 #define U2J(x)	((uintmax_t)(x))
81 
82 static unsigned char
83 digest_fold(const unsigned char *mdigest)
84 {
85     int i;
86     unsigned char rval;
87 
88     rval = mdigest[0];
89     for (i = 1; i < 16; i++) {
90         rval = rval ^ mdigest[i];
91     }
92     return (rval);
93 }
94 
95 struct mkuz_blk_info *
96 mkuz_blkcache_regblock(int fd, const struct mkuz_blk *bp)
97 {
98     struct mkuz_blkcache_itm *bcep;
99     int rval;
100     unsigned char h;
101 
102 #if defined(MKUZ_DEBUG)
103     assert((unsigned)lseek(fd, 0, SEEK_CUR) == bp->info.offset);
104 #endif
105     h = digest_fold(bp->info.digest);
106     if (blkcache.first[h].hit.len == 0) {
107         bcep = &blkcache.first[h];
108     } else {
109         for (bcep = &blkcache.first[h]; bcep != NULL; bcep = bcep->next) {
110             if (bcep->hit.len != bp->info.len)
111                 continue;
112             if (memcmp(bp->info.digest, bcep->hit.digest,
113               sizeof(bp->info.digest)) == 0) {
114                 break;
115             }
116         }
117         if (bcep != NULL) {
118             rval = verify_match(fd, bp, bcep);
119             if (rval == 1) {
120 #if defined(MKUZ_DEBUG)
121                 fprintf(stderr, "cache hit %jd, %jd, %jd, %jd\n",
122                   I2J(bcep->hit.blkno), I2J(bcep->hit.offset),
123                   I2J(bp->info.offset), I2J(bp->info.len));
124 #endif
125                 return (&bcep->hit);
126             }
127             if (rval == 0) {
128 #if defined(MKUZ_DEBUG)
129                 fprintf(stderr, "block MD5 collision, you should try lottery, "
130                   "man!\n");
131 #endif
132                 return (NULL);
133             }
134             warn("verify_match");
135             return (NULL);
136         }
137         bcep = malloc(sizeof(struct mkuz_blkcache_itm));
138         if (bcep == NULL)
139             return (NULL);
140         memset(bcep, '\0', sizeof(struct mkuz_blkcache_itm));
141         bcep->next = blkcache.first[h].next;
142         blkcache.first[h].next = bcep;
143     }
144     bcep->hit = bp->info;
145     return (NULL);
146 }
147