Commit | Line | Data |
---|---|---|
1fffd582 WD |
1 | This patch adds the --detect-renamed option which makes rsync notice files |
2 | that either (1) match in size & modify-time (plus the basename, if possible) | |
3 | or (2) match in size & checksum (when --checksum was also specified) and use | |
4 | each match as an alternate basis file to speed up the transfer. | |
5 | ||
6 | The algorithm attempts to scan the receiving-side's files in an efficient | |
7 | manner. If --delete[-before] is enabled, we'll take advantage of the | |
8 | pre-transfer delete pass to prepare any alternate-basis-file matches we | |
9 | might find. If --delete-before is not enabled, rsync does the rename scan | |
10 | during the regular file-sending scan (scanning each directory right before | |
11 | the generator starts updating files from that dir). In this latter mode, | |
12 | rsync might delay the updating of a file (if no alternate-basis match was | |
13 | yet found) until the full scan of the receiving side is complete, at which | |
14 | point any delayed files are processed. | |
15 | ||
16 | I chose to hard-link the alternate-basis files into a ".~tmp~" subdir that | |
17 | takes advantage of rsync's pre-existing partial-dir logic. This uses less | |
18 | memory than trying to keep track of the matches internally, and also allows | |
19 | any deletions or file-updates to occur normally without interfering with | |
20 | these alternate-basis discoveries. | |
21 | ||
03019e41 | 22 | To use this patch, run these commands for a successful build: |
1fffd582 | 23 | |
03019e41 WD |
24 | patch -p1 <patches/detect-renamed.diff |
25 | ./configure (optional if already run) | |
1fffd582 WD |
26 | make |
27 | ||
28 | TODO: | |
29 | ||
30 | We need to never return a match from fattr_find() that has a basis | |
31 | file. This will ensure that we don't try to give a renamed file to | |
32 | a file that can't use it, while missing out on giving it to a file | |
33 | that could use it. | |
34 | ||
35 | --- old/flist.c | |
36 | +++ new/flist.c | |
a47d1f86 | 37 | @@ -55,6 +55,7 @@ extern int non_perishable_cnt; |
1fffd582 WD |
38 | extern int prune_empty_dirs; |
39 | extern int copy_links; | |
40 | extern int copy_unsafe_links; | |
41 | +extern int detect_renamed; | |
42 | extern int protocol_version; | |
43 | extern int sanitize_paths; | |
03019e41 | 44 | extern struct stats stats; |
a47d1f86 | 45 | @@ -80,6 +81,8 @@ static int64 tmp_dev, tmp_ino; |
7b80cd0e | 46 | #endif |
70891d26 | 47 | static char tmp_sum[MD4_SUM_LENGTH]; |
1fffd582 WD |
48 | |
49 | +struct file_list the_fattr_list; | |
50 | + | |
51 | static char empty_sum[MD4_SUM_LENGTH]; | |
a47d1f86 | 52 | static int flist_count_offset; /* for --delete --progress */ |
1fffd582 | 53 | |
a47d1f86 | 54 | @@ -262,6 +265,45 @@ static mode_t from_wire_mode(int mode) |
03019e41 | 55 | return mode; |
1fffd582 WD |
56 | } |
57 | ||
58 | +static int fattr_compare(struct file_struct **file1, struct file_struct **file2) | |
59 | +{ | |
60 | + struct file_struct *f1 = *file1; | |
61 | + struct file_struct *f2 = *file2; | |
a47d1f86 | 62 | + int64 len1 = F_LENGTH(f1), len2 = F_LENGTH(f2); |
1fffd582 WD |
63 | + int diff; |
64 | + | |
a47d1f86 WD |
65 | + if (!f1->basename || !S_ISREG(f1->mode) || !len1) { |
66 | + if (!f2->basename || !S_ISREG(f2->mode) || !len2) | |
1fffd582 WD |
67 | + return 0; |
68 | + return 1; | |
69 | + } | |
a47d1f86 | 70 | + if (!f2->basename || !S_ISREG(f2->mode) || !len2) |
1fffd582 WD |
71 | + return -1; |
72 | + | |
73 | + /* Don't use diff for values that are longer than an int. */ | |
a47d1f86 WD |
74 | + if (len1 != len2) |
75 | + return len1 < len2 ? -1 : 1; | |
1fffd582 WD |
76 | + |
77 | + if (always_checksum) { | |
70891d26 | 78 | + diff = u_memcmp(F_SUM(f1), F_SUM(f2), checksum_len); |
1fffd582 WD |
79 | + if (diff) |
80 | + return diff; | |
81 | + } else if (f1->modtime != f2->modtime) | |
82 | + return f1->modtime < f2->modtime ? -1 : 1; | |
83 | + | |
84 | + diff = u_strcmp(f1->basename, f2->basename); | |
85 | + if (diff) | |
86 | + return diff; | |
87 | + | |
88 | + if (f1->dirname == f2->dirname) | |
89 | + return 0; | |
90 | + if (!f1->dirname) | |
91 | + return -1; | |
92 | + if (!f2->dirname) | |
93 | + return 1; | |
94 | + return u_strcmp(f1->dirname, f2->dirname); | |
95 | +} | |
96 | + | |
97 | static void send_directory(int f, struct file_list *flist, | |
98 | char *fbuf, int len); | |
99 | ||
a47d1f86 | 100 | @@ -1501,6 +1543,25 @@ struct file_list *recv_file_list(int f) |
1fffd582 WD |
101 | |
102 | clean_flist(flist, relative_paths, 1); | |
103 | ||
104 | + if (detect_renamed) { | |
105 | + int j = flist->count; | |
106 | + the_fattr_list.count = j; | |
107 | + the_fattr_list.files = new_array(struct file_struct *, j); | |
108 | + if (!the_fattr_list.files) | |
70891d26 | 109 | + out_of_memory("recv_file_list"); |
1fffd582 WD |
110 | + memcpy(the_fattr_list.files, flist->files, |
111 | + j * sizeof (struct file_struct *)); | |
112 | + qsort(the_fattr_list.files, j, | |
113 | + sizeof the_fattr_list.files[0], (int (*)())fattr_compare); | |
114 | + the_fattr_list.low = 0; | |
115 | + while (j-- > 0) { | |
116 | + struct file_struct *fp = the_fattr_list.files[j]; | |
a47d1f86 | 117 | + if (fp->basename && S_ISREG(fp->mode) && F_LENGTH(fp)) |
1fffd582 WD |
118 | + break; |
119 | + } | |
120 | + the_fattr_list.high = j; | |
121 | + } | |
122 | + | |
123 | if (f >= 0) { | |
124 | recv_uid_list(f, flist); | |
125 | ||
126 | --- old/generator.c | |
127 | +++ new/generator.c | |
81172142 | 128 | @@ -76,6 +76,7 @@ extern char *basis_dir[]; |
1fffd582 WD |
129 | extern int compare_dest; |
130 | extern int copy_dest; | |
131 | extern int link_dest; | |
132 | +extern int detect_renamed; | |
133 | extern int whole_file; | |
134 | extern int list_only; | |
03019e41 | 135 | extern int new_root_dir; |
7b80cd0e | 136 | @@ -91,6 +92,7 @@ extern char *backup_dir; |
1fffd582 WD |
137 | extern char *backup_suffix; |
138 | extern int backup_suffix_len; | |
139 | extern struct file_list *the_file_list; | |
140 | +extern struct file_list the_fattr_list; | |
141 | extern struct filter_list_struct server_filter_list; | |
142 | ||
d16b5fd6 | 143 | int ignore_perishable = 0; |
7b80cd0e WD |
144 | @@ -98,12 +100,14 @@ int non_perishable_cnt = 0; |
145 | int maybe_ATTRS_REPORT = 0; | |
d16b5fd6 | 146 | |
1fffd582 WD |
147 | static int deletion_count = 0; /* used to implement --max-delete */ |
148 | +static int unexplored_dirs = 1; | |
1071853f WD |
149 | static int deldelay_size = 0, deldelay_cnt = 0; |
150 | static char *deldelay_buf = NULL; | |
151 | static int deldelay_fd = -1; | |
81172142 | 152 | static BOOL solo_file = 0; |
1fffd582 | 153 | |
d16b5fd6 WD |
154 | -/* For calling delete_item() and delete_dir_contents(). */ |
155 | +/* For calling delete_item(), delete_dir_contents(), and delete_in_dir(). */ | |
156 | +#define DEL_NO_DELETIONS (1<<0) | |
87d0091c | 157 | #define DEL_RECURSE (1<<1) /* recurse */ |
d16b5fd6 | 158 | #define DEL_DIR_IS_EMPTY (1<<2) /* internal delete_FUNCTIONS use only */ |
1fffd582 | 159 | |
a47d1f86 | 160 | @@ -125,11 +129,120 @@ static int is_backup_file(char *fn) |
1fffd582 WD |
161 | return k > 0 && strcmp(fn+k, backup_suffix) == 0; |
162 | } | |
163 | ||
164 | +/* Search for a regular file that matches either (1) the size & modified | |
165 | + * time (plus the basename, if possible) or (2) the size & checksum. If | |
166 | + * we find an exact match down to the dirname, return -1 because we found | |
167 | + * an up-to-date file in the transfer, not a renamed file. */ | |
a47d1f86 | 168 | +static int fattr_find(struct file_struct *f, char *fname) |
1fffd582 WD |
169 | +{ |
170 | + int low = the_fattr_list.low, high = the_fattr_list.high; | |
171 | + int mid, ok_match = -1, good_match = -1; | |
172 | + struct file_struct *fmid; | |
173 | + int diff; | |
174 | + | |
175 | + while (low <= high) { | |
176 | + mid = (low + high) / 2; | |
177 | + fmid = the_fattr_list.files[mid]; | |
a47d1f86 WD |
178 | + if (F_LENGTH(fmid) != F_LENGTH(f)) { |
179 | + if (F_LENGTH(fmid) < F_LENGTH(f)) | |
1fffd582 WD |
180 | + low = mid + 1; |
181 | + else | |
182 | + high = mid - 1; | |
183 | + continue; | |
184 | + } | |
185 | + if (always_checksum) { | |
a47d1f86 WD |
186 | + /* We use the FLAG_SENT flag to indicate when we |
187 | + * have computed the checksum for an entry. */ | |
188 | + if (!(f->flags & FLAG_SENT)) { | |
1fffd582 WD |
189 | + if (fmid->modtime == f->modtime |
190 | + && f_name_cmp(fmid, f) == 0) | |
191 | + return -1; /* assume we can't help */ | |
a47d1f86 WD |
192 | + file_checksum(fname, (char*)F_SUM(f), F_LENGTH(f)); |
193 | + f->flags |= FLAG_SENT; | |
1fffd582 | 194 | + } |
70891d26 | 195 | + diff = u_memcmp(F_SUM(fmid), F_SUM(f), checksum_len); |
1fffd582 WD |
196 | + if (diff) { |
197 | + if (diff < 0) | |
198 | + low = mid + 1; | |
199 | + else | |
200 | + high = mid - 1; | |
201 | + continue; | |
202 | + } | |
203 | + } else { | |
204 | + if (fmid->modtime != f->modtime) { | |
205 | + if (fmid->modtime < f->modtime) | |
206 | + low = mid + 1; | |
207 | + else | |
208 | + high = mid - 1; | |
209 | + continue; | |
210 | + } | |
211 | + } | |
212 | + ok_match = mid; | |
213 | + diff = u_strcmp(fmid->basename, f->basename); | |
214 | + if (diff == 0) { | |
215 | + good_match = mid; | |
216 | + if (fmid->dirname == f->dirname) | |
217 | + return -1; /* file is up-to-date */ | |
218 | + if (!fmid->dirname) { | |
219 | + low = mid + 1; | |
220 | + continue; | |
221 | + } | |
222 | + if (!f->dirname) { | |
223 | + high = mid - 1; | |
224 | + continue; | |
225 | + } | |
226 | + diff = u_strcmp(fmid->dirname, f->dirname); | |
227 | + if (diff == 0) | |
228 | + return -1; /* file is up-to-date */ | |
229 | + } | |
230 | + if (diff < 0) | |
231 | + low = mid + 1; | |
232 | + else | |
233 | + high = mid - 1; | |
234 | + } | |
235 | + | |
236 | + return good_match >= 0 ? good_match : ok_match; | |
237 | +} | |
238 | + | |
a47d1f86 | 239 | +static void look_for_rename(struct file_struct *file, char *fname) |
1fffd582 WD |
240 | +{ |
241 | + struct file_struct *fp; | |
242 | + char *partialptr, *fn; | |
243 | + STRUCT_STAT st; | |
244 | + int ndx; | |
245 | + | |
a47d1f86 | 246 | + if ((ndx = fattr_find(file, fname)) < 0) |
1fffd582 WD |
247 | + return; |
248 | + | |
249 | + fp = the_fattr_list.files[ndx]; | |
250 | + fn = f_name(fp, NULL); | |
251 | + /* We don't provide an alternate-basis file if there is a basis file. */ | |
252 | + if (link_stat(fn, &st, 0) == 0) | |
253 | + return; | |
254 | + if ((partialptr = partial_dir_fname(fn)) == NULL | |
255 | + || !handle_partial_dir(partialptr, PDIR_CREATE)) | |
256 | + return; | |
257 | + | |
258 | + /* We only use the file if we can hard-link it into our tmp dir. */ | |
259 | + if (link(fname, partialptr) == 0) { | |
260 | + if (verbose > 2) { | |
261 | + rprintf(FINFO, "found renamed: %s => %s\n", | |
262 | + fname, partialptr); | |
263 | + } | |
264 | + return; | |
265 | + } | |
266 | + | |
267 | + if (errno != EEXIST) | |
268 | + handle_partial_dir(partialptr, PDIR_DELETE); | |
269 | +} | |
87d0091c WD |
270 | + |
271 | /* Delete a file or directory. If DEL_RECURSE is set in the flags, this will | |
272 | * delete recursively. | |
1fffd582 | 273 | * |
f813befd | 274 | * Note that fbuf must point to a MAXPATHLEN buffer if the mode indicates it's |
15894839 WD |
275 | * a directory! (The buffer is used for recursion, but returned unchanged.) |
276 | + * | |
277 | + * Also note: --detect-rename may use this routine with DEL_NO_DELETIONS set! | |
278 | */ | |
f813befd | 279 | static enum delret delete_item(char *fbuf, int mode, char *replace, int flags) |
15894839 | 280 | { |
a47d1f86 | 281 | @@ -151,6 +264,8 @@ static enum delret delete_item(char *fbu |
15894839 WD |
282 | goto check_ret; |
283 | /* OK: try to delete the directory. */ | |
284 | } | |
285 | + if (flags & DEL_NO_DELETIONS) | |
286 | + return DR_SUCCESS; | |
287 | ||
288 | if (!replace && max_delete >= 0 && ++deletion_count > max_delete) | |
289 | return DR_AT_LIMIT; | |
a47d1f86 | 290 | @@ -197,6 +312,8 @@ static enum delret delete_item(char *fbu |
d16b5fd6 WD |
291 | * its contents, otherwise just checks for content. Returns DR_SUCCESS or |
292 | * DR_NOT_EMPTY. Note that fname must point to a MAXPATHLEN buffer! (The | |
293 | * buffer is used for recursion, but returned unchanged.) | |
1fffd582 | 294 | + * |
87d0091c | 295 | + * Note: --detect-rename may use this routine with DEL_NO_DELETIONS set! |
1fffd582 | 296 | */ |
87d0091c | 297 | static enum delret delete_dir_contents(char *fname, int flags) |
1fffd582 | 298 | { |
a47d1f86 WD |
299 | @@ -216,7 +333,9 @@ static enum delret delete_dir_contents(c |
300 | save_filters = push_local_filters(fname, dlen); | |
301 | ||
302 | non_perishable_cnt = 0; | |
7e27b6c0 | 303 | + file_extra_cnt += SUM_EXTRA_CNT; |
a47d1f86 | 304 | dirlist = get_dirlist(fname, dlen, 0); |
7e27b6c0 | 305 | + file_extra_cnt -= SUM_EXTRA_CNT; |
a47d1f86 WD |
306 | ret = non_perishable_cnt ? DR_NOT_EMPTY : DR_SUCCESS; |
307 | ||
308 | if (!dirlist->count) | |
309 | @@ -253,6 +372,8 @@ static enum delret delete_dir_contents(c | |
d16b5fd6 WD |
310 | if (S_ISDIR(fp->mode) |
311 | && delete_dir_contents(fname, flags | DEL_RECURSE) != DR_SUCCESS) | |
312 | ret = DR_NOT_EMPTY; | |
313 | + if (detect_renamed && S_ISREG(fp->mode)) | |
a47d1f86 | 314 | + look_for_rename(fp, fname); |
d16b5fd6 WD |
315 | if (delete_item(fname, fp->mode, NULL, flags) != DR_SUCCESS) |
316 | ret = DR_NOT_EMPTY; | |
317 | } | |
a47d1f86 | 318 | @@ -405,15 +526,19 @@ static void do_delayed_deletions(char *d |
1fffd582 WD |
319 | * all the --delete-WHEN options. Note that the fbuf pointer must point to a |
320 | * MAXPATHLEN buffer with the name of the directory in it (the functions we | |
321 | * call will append names onto the end, but the old dir value will be restored | |
322 | - * on exit). */ | |
323 | + * on exit). | |
324 | + * | |
325 | + * Note: --detect-rename may use this routine with DEL_NO_DELETIONS set! | |
326 | + */ | |
327 | static void delete_in_dir(struct file_list *flist, char *fbuf, | |
328 | - struct file_struct *file, STRUCT_STAT *stp) | |
329 | + struct file_struct *file, STRUCT_STAT *stp, int flags) | |
330 | { | |
331 | static int min_depth = MAXPATHLEN, cur_depth = -1; | |
332 | static void *filt_array[MAXPATHLEN/2+1]; | |
333 | static int already_warned = 0; | |
334 | struct file_list *dirlist; | |
335 | - char delbuf[MAXPATHLEN]; | |
336 | + char *p, delbuf[MAXPATHLEN]; | |
337 | + unsigned remainder; | |
338 | int dlen, i; | |
339 | ||
340 | if (!flist) { | |
a47d1f86 | 341 | @@ -427,6 +552,8 @@ static void delete_in_dir(struct file_li |
1fffd582 WD |
342 | if (verbose > 2) |
343 | rprintf(FINFO, "delete_in_dir(%s)\n", fbuf); | |
344 | ||
87d0091c | 345 | + flags |= DEL_RECURSE; |
1fffd582 WD |
346 | + |
347 | if (allowed_lull) | |
348 | maybe_send_keepalive(); | |
349 | ||
a47d1f86 | 350 | @@ -434,12 +561,14 @@ static void delete_in_dir(struct file_li |
1fffd582 WD |
351 | return; /* Impossible... */ |
352 | ||
041d67b8 | 353 | if (io_error && !ignore_errors) { |
1fffd582 WD |
354 | - if (already_warned) |
355 | + if (!already_warned) { | |
356 | + rprintf(FINFO, | |
357 | + "IO error encountered -- skipping file deletion\n"); | |
358 | + already_warned = 1; | |
359 | + } | |
360 | + if (!detect_renamed) | |
361 | return; | |
362 | - rprintf(FINFO, | |
363 | - "IO error encountered -- skipping file deletion\n"); | |
364 | - already_warned = 1; | |
365 | - return; | |
366 | + flags |= DEL_NO_DELETIONS; | |
367 | } | |
368 | ||
a47d1f86 WD |
369 | while (cur_depth >= F_DEPTH(file) && cur_depth >= min_depth) |
370 | @@ -450,6 +579,9 @@ static void delete_in_dir(struct file_li | |
1fffd582 WD |
371 | dlen = strlen(fbuf); |
372 | filt_array[cur_depth] = push_local_filters(fbuf, dlen); | |
373 | ||
374 | + if (detect_renamed) | |
375 | + unexplored_dirs--; | |
376 | + | |
377 | if (one_file_system) { | |
378 | if (file->flags & FLAG_TOP_DIR) | |
379 | filesystem_dev = stp->st_dev; | |
a47d1f86 | 380 | @@ -459,6 +591,11 @@ static void delete_in_dir(struct file_li |
1fffd582 WD |
381 | |
382 | dirlist = get_dirlist(fbuf, dlen, 0); | |
383 | ||
384 | + p = fbuf + dlen; | |
385 | + if (dlen != 1 || *fbuf != '/') | |
386 | + *p++ = '/'; | |
387 | + remainder = MAXPATHLEN - (p - fbuf); | |
388 | + | |
389 | /* If an item in dirlist is not found in flist, delete it | |
390 | * from the filesystem. */ | |
391 | for (i = dirlist->count; i--; ) { | |
a47d1f86 | 392 | @@ -471,16 +608,23 @@ static void delete_in_dir(struct file_li |
87d0091c | 393 | f_name(fp, NULL)); |
1fffd582 | 394 | continue; |
87d0091c | 395 | } |
1fffd582 WD |
396 | + if (detect_renamed && S_ISREG(fp->mode)) { |
397 | + strlcpy(p, fp->basename, remainder); | |
a47d1f86 | 398 | + look_for_rename(fp, fbuf); |
1fffd582 WD |
399 | + } |
400 | if (flist_find(flist, fp) < 0) { | |
401 | f_name(fp, delbuf); | |
1071853f | 402 | - if (delete_during == 2) { |
a47d1f86 | 403 | + if (delete_during == 2 && !(flags & DEL_NO_DELETIONS)) { |
1071853f WD |
404 | if (!remember_delete(fp, delbuf)) |
405 | break; | |
406 | } else | |
f813befd | 407 | - delete_item(delbuf, fp->mode, NULL, DEL_RECURSE); |
1fffd582 | 408 | - } |
f813befd | 409 | + delete_item(delbuf, fp->mode, NULL, flags); |
1fffd582 WD |
410 | + } else if (detect_renamed && S_ISDIR(fp->mode)) |
411 | + unexplored_dirs++; | |
412 | } | |
413 | ||
414 | + fbuf[dlen] = '\0'; | |
415 | + | |
416 | flist_free(dirlist); | |
417 | } | |
418 | ||
a47d1f86 | 419 | @@ -510,9 +654,9 @@ static void do_delete_pass(struct file_l |
1fffd582 WD |
420 | || !S_ISDIR(st.st_mode)) |
421 | continue; | |
422 | ||
423 | - delete_in_dir(flist, fbuf, file, &st); | |
424 | + delete_in_dir(flist, fbuf, file, &st, 0); | |
425 | } | |
426 | - delete_in_dir(NULL, NULL, NULL, NULL); | |
427 | + delete_in_dir(NULL, NULL, NULL, NULL, 0); | |
428 | ||
429 | if (do_progress && !am_server) | |
430 | rprintf(FINFO, " \r"); | |
a47d1f86 | 431 | @@ -1047,6 +1191,7 @@ static int try_dests_non(struct file_str |
9a70b743 | 432 | return j; |
1fffd582 WD |
433 | } |
434 | ||
435 | +static struct bitbag *delayed_bits = NULL; | |
436 | static int phase = 0; | |
437 | ||
438 | /* Acts on the_file_list->file's ndx'th item, whose name is fname. If a dir, | |
a47d1f86 | 439 | @@ -1232,8 +1377,12 @@ static void recv_generator(char *fname, |
9a70b743 WD |
440 | if (real_ret != 0 && one_file_system) |
441 | real_st.st_dev = filesystem_dev; | |
1fffd582 | 442 | if (delete_during && f_out != -1 && !phase && dry_run < 2 |
70891d26 | 443 | - && (file->flags & FLAG_XFER_DIR)) |
9a70b743 | 444 | - delete_in_dir(the_file_list, fname, file, &real_st); |
70891d26 | 445 | + && (file->flags & FLAG_XFER_DIR)) { |
9a70b743 | 446 | + if (detect_renamed && real_ret != 0) |
1fffd582 | 447 | + unexplored_dirs++; |
9a70b743 | 448 | + delete_in_dir(the_file_list, fname, file, &real_st, |
1fffd582 WD |
449 | + delete_during < 0 ? DEL_NO_DELETIONS : 0); |
450 | + } | |
451 | return; | |
452 | } | |
453 | ||
a47d1f86 WD |
454 | @@ -1497,8 +1646,14 @@ static void recv_generator(char *fname, |
455 | if (preserve_hard_links && F_HLINK_NOT_LAST(file)) | |
1fffd582 | 456 | return; |
81172142 | 457 | #endif |
1fffd582 WD |
458 | - if (stat_errno == ENOENT) |
459 | + if (stat_errno == ENOENT) { | |
460 | + if (detect_renamed && unexplored_dirs > 0 | |
a47d1f86 | 461 | + && F_LENGTH(file)) { |
1fffd582 WD |
462 | + bitbag_set_bit(delayed_bits, ndx); |
463 | + return; | |
464 | + } | |
465 | goto notify_others; | |
466 | + } | |
467 | rsyserr(FERROR, stat_errno, "recv_generator: failed to stat %s", | |
468 | full_fname(fname)); | |
469 | return; | |
a47d1f86 | 470 | @@ -1688,6 +1843,12 @@ void generate_files(int f_out, struct fi |
1fffd582 WD |
471 | (long)getpid(), flist->count); |
472 | } | |
473 | ||
474 | + if (detect_renamed) { | |
475 | + delayed_bits = bitbag_create(flist->count); | |
476 | + if (!delete_before && !delete_during) | |
477 | + delete_during = -1; | |
478 | + } | |
479 | + | |
480 | if (delete_before && !local_name && flist->count > 0) | |
481 | do_delete_pass(flist); | |
1071853f | 482 | if (delete_during == 2) { |
a47d1f86 | 483 | @@ -1698,7 +1859,7 @@ void generate_files(int f_out, struct fi |
1071853f | 484 | } |
1fffd582 WD |
485 | do_progress = 0; |
486 | ||
487 | - if (append_mode || whole_file < 0) | |
488 | + if (append_mode || detect_renamed || whole_file < 0) | |
489 | whole_file = 0; | |
490 | if (verbose >= 2) { | |
491 | rprintf(FINFO, "delta-transmission %s\n", | |
a47d1f86 | 492 | @@ -1754,7 +1915,22 @@ void generate_files(int f_out, struct fi |
1fffd582 | 493 | } |
7b80cd0e | 494 | recv_generator(NULL, NULL, 0, 0, code, -1); |
1fffd582 WD |
495 | if (delete_during) |
496 | - delete_in_dir(NULL, NULL, NULL, NULL); | |
497 | + delete_in_dir(NULL, NULL, NULL, NULL, 0); | |
498 | + | |
499 | + if (detect_renamed) { | |
500 | + if (delete_during < 0) | |
501 | + delete_during = 0; | |
502 | + detect_renamed = 0; | |
503 | + | |
504 | + for (i = -1; (i = bitbag_next_bit(delayed_bits, i)) >= 0; ) { | |
505 | + struct file_struct *file = flist->files[i]; | |
506 | + if (local_name) | |
507 | + strlcpy(fbuf, local_name, sizeof fbuf); | |
508 | + else | |
509 | + f_name(file, fbuf); | |
a47d1f86 | 510 | + recv_generator(fbuf, file, i, itemizing, code, f_out); |
1fffd582 WD |
511 | + } |
512 | + } | |
513 | ||
514 | phase++; | |
515 | csum_length = SUM_LENGTH; | |
516 | --- old/options.c | |
517 | +++ new/options.c | |
03019e41 | 518 | @@ -78,6 +78,7 @@ int am_generator = 0; |
a94141d9 | 519 | int am_starting_up = 1; |
1fffd582 WD |
520 | int relative_paths = -1; |
521 | int implied_dirs = 1; | |
522 | +int detect_renamed = 0; | |
523 | int numeric_ids = 0; | |
524 | int allow_8bit_chars = 0; | |
525 | int force_delete = 0; | |
a47d1f86 | 526 | @@ -342,6 +343,7 @@ void usage(enum logcode F) |
1fffd582 WD |
527 | rprintf(F," --modify-window=NUM compare mod-times with reduced accuracy\n"); |
528 | rprintf(F," -T, --temp-dir=DIR create temporary files in directory DIR\n"); | |
529 | rprintf(F," -y, --fuzzy find similar file for basis if no dest file\n"); | |
530 | + rprintf(F," --detect-renamed try to find renamed files to speed up the transfer\n"); | |
531 | rprintf(F," --compare-dest=DIR also compare destination files relative to DIR\n"); | |
532 | rprintf(F," --copy-dest=DIR ... and include copies of unchanged files\n"); | |
533 | rprintf(F," --link-dest=DIR hardlink to files in DIR when unchanged\n"); | |
a47d1f86 | 534 | @@ -496,6 +498,7 @@ static struct poptOption long_options[] |
1fffd582 WD |
535 | {"compare-dest", 0, POPT_ARG_STRING, 0, OPT_COMPARE_DEST, 0, 0 }, |
536 | {"copy-dest", 0, POPT_ARG_STRING, 0, OPT_COPY_DEST, 0, 0 }, | |
537 | {"link-dest", 0, POPT_ARG_STRING, 0, OPT_LINK_DEST, 0, 0 }, | |
538 | + {"detect-renamed", 0, POPT_ARG_NONE, &detect_renamed, 0, 0, 0 }, | |
539 | {"fuzzy", 'y', POPT_ARG_NONE, &fuzzy_basis, 0, 0, 0 }, | |
540 | {"compress", 'z', POPT_ARG_NONE, 0, 'z', 0, 0 }, | |
541 | {"compress-level", 0, POPT_ARG_INT, &def_compress_level, 'z', 0, 0 }, | |
a47d1f86 | 542 | @@ -1355,7 +1358,7 @@ int parse_arguments(int *argc, const cha |
1fffd582 WD |
543 | inplace = 1; |
544 | } | |
545 | ||
546 | - if (delay_updates && !partial_dir) | |
547 | + if ((delay_updates || detect_renamed) && !partial_dir) | |
548 | partial_dir = tmp_partialdir; | |
549 | ||
550 | if (inplace) { | |
a47d1f86 | 551 | @@ -1364,6 +1367,7 @@ int parse_arguments(int *argc, const cha |
1fffd582 WD |
552 | snprintf(err_buf, sizeof err_buf, |
553 | "--%s cannot be used with --%s\n", | |
554 | append_mode ? "append" : "inplace", | |
555 | + detect_renamed ? "detect-renamed" : | |
556 | delay_updates ? "delay-updates" : "partial-dir"); | |
557 | return 0; | |
558 | } | |
a47d1f86 | 559 | @@ -1674,6 +1678,8 @@ void server_options(char **args,int *arg |
a94141d9 WD |
560 | args[ac++] = "--super"; |
561 | if (size_only) | |
562 | args[ac++] = "--size-only"; | |
563 | + if (detect_renamed) | |
564 | + args[ac++] = "--detect-renamed"; | |
565 | } | |
566 | ||
567 | if (modify_window_set) { | |
1fffd582 WD |
568 | --- old/rsync.yo |
569 | +++ new/rsync.yo | |
f813befd | 570 | @@ -364,6 +364,7 @@ to the detailed description below for a |
1fffd582 WD |
571 | --modify-window=NUM compare mod-times with reduced accuracy |
572 | -T, --temp-dir=DIR create temporary files in directory DIR | |
573 | -y, --fuzzy find similar file for basis if no dest file | |
574 | + --detect-renamed try to find renamed files to speed the xfer | |
575 | --compare-dest=DIR also compare received files relative to DIR | |
576 | --copy-dest=DIR ... and include copies of unchanged files | |
577 | --link-dest=DIR hardlink to files in DIR when unchanged | |
f813befd | 578 | @@ -1272,6 +1273,15 @@ Note that the use of the bf(--delete) op |
1fffd582 WD |
579 | fuzzy-match files, so either use bf(--delete-after) or specify some |
580 | filename exclusions if you need to prevent this. | |
581 | ||
582 | +dit(bf(--detect-renamed)) This option tells rsync to scan the receiving | |
583 | +side for files that have been renamed, and to use any that are found as | |
584 | +alternate basis files to help speed up the transfer. | |
585 | +By default, alternate-basis files are hard-linked into a directory named | |
586 | +".~tmp~" in each file's destination directory, but if you've specified | |
587 | +the bf(--partial-dir) option, that directory will be used instead. These | |
588 | +potential alternate-basis files will be removed as the transfer progresses. | |
589 | +This option conflicts with bf(--inplace) and bf(--append). | |
590 | + | |
591 | dit(bf(--compare-dest=DIR)) This option instructs rsync to use em(DIR) on | |
592 | the destination machine as an additional hierarchy to compare destination | |
593 | files against doing transfers (if the files are missing in the destination | |
594 | --- old/util.c | |
595 | +++ new/util.c | |
f813befd | 596 | @@ -1027,6 +1027,32 @@ int handle_partial_dir(const char *fname |
1fffd582 WD |
597 | return 1; |
598 | } | |
599 | ||
600 | +/* We need to supply our own strcmp function for file list comparisons | |
601 | + * to ensure that signed/unsigned usage is consistent between machines. */ | |
602 | +int u_strcmp(const char *p1, const char *p2) | |
603 | +{ | |
604 | + for ( ; *p1; p1++, p2++) { | |
605 | + if (*p1 != *p2) | |
606 | + break; | |
607 | + } | |
608 | + | |
609 | + return (int)*(uchar*)p1 - (int)*(uchar*)p2; | |
610 | +} | |
611 | + | |
612 | +/* We need a memcmp function compares unsigned-byte values. */ | |
613 | +int u_memcmp(const void *p1, const void *p2, size_t len) | |
614 | +{ | |
615 | + const uchar *u1 = p1; | |
616 | + const uchar *u2 = p2; | |
617 | + | |
618 | + while (len--) { | |
619 | + if (*u1 != *u2) | |
620 | + return (int)*u1 - (int)*u2; | |
621 | + } | |
622 | + | |
623 | + return 0; | |
624 | +} | |
625 | + | |
626 | /** | |
627 | * Determine if a symlink points outside the current directory tree. | |
628 | * This is considered "unsafe" because e.g. when mirroring somebody |