| 1 | /* |
| 2 | Copyright (C) Andrew Tridgell 1996 |
| 3 | Copyright (C) Paul Mackerras 1996 |
| 4 | |
| 5 | This program is free software; you can redistribute it and/or modify |
| 6 | it under the terms of the GNU General Public License as published by |
| 7 | the Free Software Foundation; either version 2 of the License, or |
| 8 | (at your option) any later version. |
| 9 | |
| 10 | This program is distributed in the hope that it will be useful, |
| 11 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 12 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 13 | GNU General Public License for more details. |
| 14 | |
| 15 | You should have received a copy of the GNU General Public License |
| 16 | along with this program; if not, write to the Free Software |
| 17 | Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. |
| 18 | */ |
| 19 | |
| 20 | #include "rsync.h" |
| 21 | |
| 22 | extern int csum_length; |
| 23 | |
| 24 | extern int verbose; |
| 25 | extern int am_server; |
| 26 | |
| 27 | extern int remote_version; |
| 28 | |
| 29 | typedef unsigned short tag; |
| 30 | |
| 31 | #define TABLESIZE (1<<16) |
| 32 | #define NULL_TAG ((tag)-1) |
| 33 | |
| 34 | static int false_alarms; |
| 35 | static int tag_hits; |
| 36 | static int matches; |
| 37 | static int data_transfer; |
| 38 | |
| 39 | static int total_false_alarms; |
| 40 | static int total_tag_hits; |
| 41 | static int total_matches; |
| 42 | static int64 total_data_transfer; |
| 43 | |
| 44 | |
| 45 | struct target { |
| 46 | tag t; |
| 47 | int i; |
| 48 | }; |
| 49 | |
| 50 | static struct target *targets; |
| 51 | |
| 52 | static tag *tag_table; |
| 53 | |
| 54 | #define gettag2(s1,s2) (((s1) + (s2)) & 0xFFFF) |
| 55 | #define gettag(sum) gettag2((sum)&0xFFFF,(sum)>>16) |
| 56 | |
| 57 | static int compare_targets(struct target *t1,struct target *t2) |
| 58 | { |
| 59 | return((int)t1->t - (int)t2->t); |
| 60 | } |
| 61 | |
| 62 | |
| 63 | static void build_hash_table(struct sum_struct *s) |
| 64 | { |
| 65 | int i; |
| 66 | |
| 67 | if (!tag_table) |
| 68 | tag_table = (tag *)malloc(sizeof(tag)*TABLESIZE); |
| 69 | |
| 70 | targets = (struct target *)malloc(sizeof(targets[0])*s->count); |
| 71 | if (!tag_table || !targets) |
| 72 | out_of_memory("build_hash_table"); |
| 73 | |
| 74 | for (i=0;i<s->count;i++) { |
| 75 | targets[i].i = i; |
| 76 | targets[i].t = gettag(s->sums[i].sum1); |
| 77 | } |
| 78 | |
| 79 | qsort(targets,s->count,sizeof(targets[0]),(int (*)())compare_targets); |
| 80 | |
| 81 | for (i=0;i<TABLESIZE;i++) |
| 82 | tag_table[i] = NULL_TAG; |
| 83 | |
| 84 | for (i=s->count-1;i>=0;i--) { |
| 85 | tag_table[targets[i].t] = i; |
| 86 | } |
| 87 | } |
| 88 | |
| 89 | |
| 90 | static off_t last_match; |
| 91 | |
| 92 | |
| 93 | static void matched(int f,struct sum_struct *s,struct map_struct *buf, |
| 94 | off_t offset,int i) |
| 95 | { |
| 96 | off_t n = offset - last_match; |
| 97 | int j; |
| 98 | |
| 99 | if (verbose > 2 && i >= 0) |
| 100 | fprintf(FINFO,"match at %d last_match=%d j=%d len=%d n=%d\n", |
| 101 | (int)offset,(int)last_match,i,(int)s->sums[i].len,(int)n); |
| 102 | |
| 103 | send_token(f,i,buf,last_match,n,i==-1?0:s->sums[i].len); |
| 104 | data_transfer += n; |
| 105 | |
| 106 | if (n > 0) |
| 107 | write_flush(f); |
| 108 | |
| 109 | if (i >= 0) |
| 110 | n += s->sums[i].len; |
| 111 | |
| 112 | for (j=0;j<n;j+=CHUNK_SIZE) { |
| 113 | int n1 = MIN(CHUNK_SIZE,n-j); |
| 114 | sum_update(map_ptr(buf,last_match+j,n1),n1); |
| 115 | } |
| 116 | |
| 117 | |
| 118 | if (i >= 0) |
| 119 | last_match = offset + s->sums[i].len; |
| 120 | else |
| 121 | last_match = offset; |
| 122 | } |
| 123 | |
| 124 | |
| 125 | static void hash_search(int f,struct sum_struct *s, |
| 126 | struct map_struct *buf,off_t len) |
| 127 | { |
| 128 | off_t offset; |
| 129 | int j,k; |
| 130 | int end; |
| 131 | char sum2[SUM_LENGTH]; |
| 132 | uint32 s1, s2, sum; |
| 133 | signed char *map; |
| 134 | |
| 135 | if (verbose > 2) |
| 136 | fprintf(FINFO,"hash search b=%d len=%d\n",s->n,(int)len); |
| 137 | |
| 138 | k = MIN(len, s->n); |
| 139 | |
| 140 | map = (signed char *)map_ptr(buf,0,k); |
| 141 | |
| 142 | sum = get_checksum1((char *)map, k); |
| 143 | s1 = sum & 0xFFFF; |
| 144 | s2 = sum >> 16; |
| 145 | if (verbose > 3) |
| 146 | fprintf(FINFO, "sum=%.8x k=%d\n", sum, k); |
| 147 | |
| 148 | offset = 0; |
| 149 | |
| 150 | end = len + 1 - s->sums[s->count-1].len; |
| 151 | |
| 152 | if (verbose > 3) |
| 153 | fprintf(FINFO,"hash search s->n=%d len=%d count=%d\n", |
| 154 | s->n,(int)len,s->count); |
| 155 | |
| 156 | do { |
| 157 | tag t = gettag2(s1,s2); |
| 158 | int done_csum2 = 0; |
| 159 | |
| 160 | j = tag_table[t]; |
| 161 | if (verbose > 4) |
| 162 | fprintf(FINFO,"offset=%d sum=%08x\n",(int)offset,sum); |
| 163 | |
| 164 | if (j == NULL_TAG) { |
| 165 | goto null_tag; |
| 166 | } |
| 167 | |
| 168 | sum = (s1 & 0xffff) | (s2 << 16); |
| 169 | tag_hits++; |
| 170 | for (; j<s->count && targets[j].t == t; j++) { |
| 171 | int i = targets[j].i; |
| 172 | |
| 173 | if (sum != s->sums[i].sum1) continue; |
| 174 | |
| 175 | if (verbose > 3) |
| 176 | fprintf(FINFO,"potential match at %d target=%d %d sum=%08x\n", |
| 177 | (int)offset,j,i,sum); |
| 178 | |
| 179 | if (!done_csum2) { |
| 180 | int l = MIN(s->n,len-offset); |
| 181 | map = (signed char *)map_ptr(buf,offset,l); |
| 182 | get_checksum2((char *)map,l,sum2); |
| 183 | done_csum2 = 1; |
| 184 | } |
| 185 | |
| 186 | if (memcmp(sum2,s->sums[i].sum2,csum_length) != 0) { |
| 187 | false_alarms++; |
| 188 | continue; |
| 189 | } |
| 190 | |
| 191 | matched(f,s,buf,offset,i); |
| 192 | offset += s->sums[i].len - 1; |
| 193 | k = MIN((len-offset), s->n); |
| 194 | map = (signed char *)map_ptr(buf,offset,k); |
| 195 | sum = get_checksum1((char *)map, k); |
| 196 | s1 = sum & 0xFFFF; |
| 197 | s2 = sum >> 16; |
| 198 | matches++; |
| 199 | break; |
| 200 | } |
| 201 | |
| 202 | null_tag: |
| 203 | /* Trim off the first byte from the checksum */ |
| 204 | map = (signed char *)map_ptr(buf,offset,k+1); |
| 205 | s1 -= map[0] + CHAR_OFFSET; |
| 206 | s2 -= k * (map[0]+CHAR_OFFSET); |
| 207 | |
| 208 | /* Add on the next byte (if there is one) to the checksum */ |
| 209 | if (k < (len-offset)) { |
| 210 | s1 += (map[k]+CHAR_OFFSET); |
| 211 | s2 += s1; |
| 212 | } else { |
| 213 | --k; |
| 214 | } |
| 215 | |
| 216 | } while (++offset < end); |
| 217 | |
| 218 | matched(f,s,buf,len,-1); |
| 219 | map_ptr(buf,len-1,1); |
| 220 | } |
| 221 | |
| 222 | |
| 223 | void match_sums(int f,struct sum_struct *s,struct map_struct *buf,off_t len) |
| 224 | { |
| 225 | char file_sum[MD4_SUM_LENGTH]; |
| 226 | |
| 227 | last_match = 0; |
| 228 | false_alarms = 0; |
| 229 | tag_hits = 0; |
| 230 | matches=0; |
| 231 | data_transfer=0; |
| 232 | |
| 233 | sum_init(); |
| 234 | |
| 235 | if (len > 0 && s->count>0) { |
| 236 | build_hash_table(s); |
| 237 | |
| 238 | if (verbose > 2) |
| 239 | fprintf(FINFO,"built hash table\n"); |
| 240 | |
| 241 | hash_search(f,s,buf,len); |
| 242 | |
| 243 | if (verbose > 2) |
| 244 | fprintf(FINFO,"done hash search\n"); |
| 245 | } else { |
| 246 | matched(f,s,buf,len,-1); |
| 247 | } |
| 248 | |
| 249 | sum_end(file_sum); |
| 250 | |
| 251 | if (remote_version >= 14) { |
| 252 | if (verbose > 2) |
| 253 | fprintf(FINFO,"sending file_sum\n"); |
| 254 | write_buf(f,file_sum,MD4_SUM_LENGTH); |
| 255 | } |
| 256 | |
| 257 | if (targets) { |
| 258 | free(targets); |
| 259 | targets=NULL; |
| 260 | } |
| 261 | |
| 262 | if (verbose > 2) |
| 263 | fprintf(FINFO, "false_alarms=%d tag_hits=%d matches=%d\n", |
| 264 | false_alarms, tag_hits, matches); |
| 265 | |
| 266 | total_tag_hits += tag_hits; |
| 267 | total_false_alarms += false_alarms; |
| 268 | total_matches += matches; |
| 269 | total_data_transfer += data_transfer; |
| 270 | } |
| 271 | |
| 272 | void match_report(void) |
| 273 | { |
| 274 | if (verbose <= 1) |
| 275 | return; |
| 276 | |
| 277 | fprintf(FINFO, |
| 278 | "total: matches=%d tag_hits=%d false_alarms=%d data=%ld\n", |
| 279 | total_matches,total_tag_hits, |
| 280 | total_false_alarms,(long)total_data_transfer); |
| 281 | } |