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 | ||
cc3e685d WD |
35 | diff --git a/compat.c b/compat.c |
36 | --- a/compat.c | |
37 | +++ b/compat.c | |
cdcd2137 WD |
38 | @@ -41,6 +41,7 @@ extern int checksum_seed; |
39 | extern int basis_dir_cnt; | |
40 | extern int prune_empty_dirs; | |
41 | extern int protocol_version; | |
9bcaf4de | 42 | +extern int detect_renamed; |
cdcd2137 WD |
43 | extern int protect_args; |
44 | extern int preserve_uid; | |
45 | extern int preserve_gid; | |
ae306a29 | 46 | @@ -120,6 +121,7 @@ void set_allow_inc_recurse(void) |
9aab301c WD |
47 | allow_inc_recurse = 0; |
48 | else if (!am_sender | |
49 | && (delete_before || delete_after | |
50 | + || detect_renamed | |
51 | || delay_updates || prune_empty_dirs)) | |
51bc0e89 WD |
52 | allow_inc_recurse = 0; |
53 | else if (am_server && !local_server | |
cc3e685d WD |
54 | diff --git a/flist.c b/flist.c |
55 | --- a/flist.c | |
56 | +++ b/flist.c | |
f9df736a | 57 | @@ -63,6 +63,7 @@ extern int non_perishable_cnt; |
1fffd582 WD |
58 | extern int prune_empty_dirs; |
59 | extern int copy_links; | |
60 | extern int copy_unsafe_links; | |
61 | +extern int detect_renamed; | |
62 | extern int protocol_version; | |
63 | extern int sanitize_paths; | |
cc3e685d | 64 | extern int munge_symlinks; |
ae306a29 | 65 | @@ -121,6 +122,8 @@ static int64 tmp_dev, tmp_ino; |
7b80cd0e | 66 | #endif |
87a38eea | 67 | static char tmp_sum[MAX_DIGEST_LEN]; |
1fffd582 WD |
68 | |
69 | +struct file_list the_fattr_list; | |
70 | + | |
87a38eea | 71 | static char empty_sum[MAX_DIGEST_LEN]; |
a47d1f86 | 72 | static int flist_count_offset; /* for --delete --progress */ |
6cbbe66d | 73 | static int dir_count = 0; |
ae306a29 | 74 | @@ -288,6 +291,45 @@ static int is_excluded(const char *fname, int is_dir, int filter_level) |
73adde61 | 75 | return 0; |
1fffd582 WD |
76 | } |
77 | ||
78 | +static int fattr_compare(struct file_struct **file1, struct file_struct **file2) | |
79 | +{ | |
80 | + struct file_struct *f1 = *file1; | |
81 | + struct file_struct *f2 = *file2; | |
a47d1f86 | 82 | + int64 len1 = F_LENGTH(f1), len2 = F_LENGTH(f2); |
1fffd582 WD |
83 | + int diff; |
84 | + | |
a47d1f86 WD |
85 | + if (!f1->basename || !S_ISREG(f1->mode) || !len1) { |
86 | + if (!f2->basename || !S_ISREG(f2->mode) || !len2) | |
1fffd582 WD |
87 | + return 0; |
88 | + return 1; | |
89 | + } | |
a47d1f86 | 90 | + if (!f2->basename || !S_ISREG(f2->mode) || !len2) |
1fffd582 WD |
91 | + return -1; |
92 | + | |
93 | + /* Don't use diff for values that are longer than an int. */ | |
a47d1f86 WD |
94 | + if (len1 != len2) |
95 | + return len1 < len2 ? -1 : 1; | |
1fffd582 WD |
96 | + |
97 | + if (always_checksum) { | |
70891d26 | 98 | + diff = u_memcmp(F_SUM(f1), F_SUM(f2), checksum_len); |
1fffd582 WD |
99 | + if (diff) |
100 | + return diff; | |
101 | + } else if (f1->modtime != f2->modtime) | |
102 | + return f1->modtime < f2->modtime ? -1 : 1; | |
103 | + | |
104 | + diff = u_strcmp(f1->basename, f2->basename); | |
105 | + if (diff) | |
106 | + return diff; | |
107 | + | |
108 | + if (f1->dirname == f2->dirname) | |
109 | + return 0; | |
110 | + if (!f1->dirname) | |
111 | + return -1; | |
112 | + if (!f2->dirname) | |
113 | + return 1; | |
114 | + return u_strcmp(f1->dirname, f2->dirname); | |
115 | +} | |
116 | + | |
73adde61 | 117 | static void send_directory(int f, struct file_list *flist, |
fc068916 | 118 | char *fbuf, int len, int flags); |
1fffd582 | 119 | |
ae306a29 | 120 | @@ -2357,6 +2399,25 @@ struct file_list *recv_file_list(int f) |
6fa0767f | 121 | |
f2863bc0 | 122 | flist_sort_and_clean(flist, relative_paths); |
1fffd582 WD |
123 | |
124 | + if (detect_renamed) { | |
7bfcb297 WD |
125 | + int j = flist->used; |
126 | + the_fattr_list.used = j; | |
1fffd582 WD |
127 | + the_fattr_list.files = new_array(struct file_struct *, j); |
128 | + if (!the_fattr_list.files) | |
70891d26 | 129 | + out_of_memory("recv_file_list"); |
1fffd582 WD |
130 | + memcpy(the_fattr_list.files, flist->files, |
131 | + j * sizeof (struct file_struct *)); | |
132 | + qsort(the_fattr_list.files, j, | |
fc068916 | 133 | + sizeof the_fattr_list.files[0], (int (*)())fattr_compare); |
1fffd582 WD |
134 | + the_fattr_list.low = 0; |
135 | + while (j-- > 0) { | |
136 | + struct file_struct *fp = the_fattr_list.files[j]; | |
a47d1f86 | 137 | + if (fp->basename && S_ISREG(fp->mode) && F_LENGTH(fp)) |
1fffd582 WD |
138 | + break; |
139 | + } | |
140 | + the_fattr_list.high = j; | |
141 | + } | |
142 | + | |
6fa0767f WD |
143 | if (protocol_version < 30) { |
144 | /* Recv the io_error flag */ | |
145 | if (ignore_errors) | |
cc3e685d WD |
146 | diff --git a/generator.c b/generator.c |
147 | --- a/generator.c | |
148 | +++ b/generator.c | |
c0c7984e | 149 | @@ -80,6 +80,7 @@ extern char *basis_dir[]; |
1fffd582 WD |
150 | extern int compare_dest; |
151 | extern int copy_dest; | |
152 | extern int link_dest; | |
153 | +extern int detect_renamed; | |
154 | extern int whole_file; | |
155 | extern int list_only; | |
ccdb48f6 | 156 | extern int read_batch; |
4c107044 | 157 | @@ -98,6 +99,7 @@ extern char *backup_suffix; |
1fffd582 | 158 | extern int backup_suffix_len; |
fc068916 | 159 | extern struct file_list *cur_flist, *first_flist, *dir_flist; |
c0c7984e | 160 | extern struct filter_list_struct daemon_filter_list; |
fc068916 | 161 | +extern struct file_list the_fattr_list; |
d4dd2dd5 WD |
162 | |
163 | int ignore_perishable = 0; | |
164 | int non_perishable_cnt = 0; | |
4c107044 | 165 | @@ -105,6 +107,7 @@ int maybe_ATTRS_REPORT = 0; |
d16b5fd6 | 166 | |
fc068916 | 167 | static dev_t dev_zero; |
1fffd582 WD |
168 | static int deletion_count = 0; /* used to implement --max-delete */ |
169 | +static int unexplored_dirs = 1; | |
1071853f WD |
170 | static int deldelay_size = 0, deldelay_cnt = 0; |
171 | static char *deldelay_buf = NULL; | |
172 | static int deldelay_fd = -1; | |
4c107044 | 173 | @@ -115,7 +118,7 @@ static int need_retouch_dir_times; |
9c85142a | 174 | static int need_retouch_dir_perms; |
2dbc45e7 | 175 | static const char *solo_file = NULL; |
1fffd582 | 176 | |
d16b5fd6 WD |
177 | -/* For calling delete_item() and delete_dir_contents(). */ |
178 | +/* For calling delete_item(), delete_dir_contents(), and delete_in_dir(). */ | |
f9df736a | 179 | #define DEL_NO_UID_WRITE (1<<0) /* file/dir has our uid w/o write perm */ |
a5e6228a | 180 | #define DEL_RECURSE (1<<1) /* if dir, delete all contents */ |
d16b5fd6 | 181 | #define DEL_DIR_IS_EMPTY (1<<2) /* internal delete_FUNCTIONS use only */ |
4c107044 | 182 | @@ -124,6 +127,7 @@ static const char *solo_file = NULL; |
a5e6228a WD |
183 | #define DEL_FOR_SYMLINK (1<<5) /* making room for a replacement symlink */ |
184 | #define DEL_FOR_DEVICE (1<<6) /* making room for a replacement device */ | |
185 | #define DEL_FOR_SPECIAL (1<<7) /* making room for a replacement special */ | |
186 | +#define DEL_NO_DELETIONS (1<<9) /* just check for renames w/o deleting */ | |
1fffd582 | 187 | |
a5e6228a WD |
188 | #define DEL_MAKE_ROOM (DEL_FOR_FILE|DEL_FOR_DIR|DEL_FOR_SYMLINK|DEL_FOR_DEVICE|DEL_FOR_SPECIAL) |
189 | ||
4c107044 | 190 | @@ -148,11 +152,121 @@ static int is_backup_file(char *fn) |
1fffd582 WD |
191 | return k > 0 && strcmp(fn+k, backup_suffix) == 0; |
192 | } | |
193 | ||
194 | +/* Search for a regular file that matches either (1) the size & modified | |
195 | + * time (plus the basename, if possible) or (2) the size & checksum. If | |
196 | + * we find an exact match down to the dirname, return -1 because we found | |
197 | + * an up-to-date file in the transfer, not a renamed file. */ | |
a47d1f86 | 198 | +static int fattr_find(struct file_struct *f, char *fname) |
1fffd582 WD |
199 | +{ |
200 | + int low = the_fattr_list.low, high = the_fattr_list.high; | |
201 | + int mid, ok_match = -1, good_match = -1; | |
202 | + struct file_struct *fmid; | |
203 | + int diff; | |
204 | + | |
205 | + while (low <= high) { | |
206 | + mid = (low + high) / 2; | |
207 | + fmid = the_fattr_list.files[mid]; | |
a47d1f86 WD |
208 | + if (F_LENGTH(fmid) != F_LENGTH(f)) { |
209 | + if (F_LENGTH(fmid) < F_LENGTH(f)) | |
1fffd582 WD |
210 | + low = mid + 1; |
211 | + else | |
212 | + high = mid - 1; | |
213 | + continue; | |
214 | + } | |
215 | + if (always_checksum) { | |
9bcaf4de | 216 | + /* We use the FLAG_FILE_SENT flag to indicate when we |
a47d1f86 | 217 | + * have computed the checksum for an entry. */ |
9bcaf4de | 218 | + if (!(f->flags & FLAG_FILE_SENT)) { |
1fffd582 WD |
219 | + if (fmid->modtime == f->modtime |
220 | + && f_name_cmp(fmid, f) == 0) | |
221 | + return -1; /* assume we can't help */ | |
d7d6347c | 222 | + file_checksum(fname, F_SUM(f), F_LENGTH(f)); |
9bcaf4de | 223 | + f->flags |= FLAG_FILE_SENT; |
1fffd582 | 224 | + } |
70891d26 | 225 | + diff = u_memcmp(F_SUM(fmid), F_SUM(f), checksum_len); |
1fffd582 WD |
226 | + if (diff) { |
227 | + if (diff < 0) | |
228 | + low = mid + 1; | |
229 | + else | |
230 | + high = mid - 1; | |
231 | + continue; | |
232 | + } | |
233 | + } else { | |
234 | + if (fmid->modtime != f->modtime) { | |
235 | + if (fmid->modtime < f->modtime) | |
236 | + low = mid + 1; | |
237 | + else | |
238 | + high = mid - 1; | |
239 | + continue; | |
240 | + } | |
241 | + } | |
242 | + ok_match = mid; | |
243 | + diff = u_strcmp(fmid->basename, f->basename); | |
244 | + if (diff == 0) { | |
245 | + good_match = mid; | |
246 | + if (fmid->dirname == f->dirname) | |
247 | + return -1; /* file is up-to-date */ | |
248 | + if (!fmid->dirname) { | |
249 | + low = mid + 1; | |
250 | + continue; | |
251 | + } | |
252 | + if (!f->dirname) { | |
253 | + high = mid - 1; | |
254 | + continue; | |
255 | + } | |
256 | + diff = u_strcmp(fmid->dirname, f->dirname); | |
257 | + if (diff == 0) | |
258 | + return -1; /* file is up-to-date */ | |
259 | + } | |
260 | + if (diff < 0) | |
261 | + low = mid + 1; | |
262 | + else | |
263 | + high = mid - 1; | |
264 | + } | |
265 | + | |
266 | + return good_match >= 0 ? good_match : ok_match; | |
267 | +} | |
268 | + | |
a47d1f86 | 269 | +static void look_for_rename(struct file_struct *file, char *fname) |
1fffd582 WD |
270 | +{ |
271 | + struct file_struct *fp; | |
272 | + char *partialptr, *fn; | |
273 | + STRUCT_STAT st; | |
274 | + int ndx; | |
275 | + | |
6fa0767f | 276 | + if (!partial_dir || (ndx = fattr_find(file, fname)) < 0) |
1fffd582 WD |
277 | + return; |
278 | + | |
279 | + fp = the_fattr_list.files[ndx]; | |
280 | + fn = f_name(fp, NULL); | |
281 | + /* We don't provide an alternate-basis file if there is a basis file. */ | |
282 | + if (link_stat(fn, &st, 0) == 0) | |
283 | + return; | |
1fffd582 | 284 | + |
6fa0767f WD |
285 | + if (!dry_run) { |
286 | + if ((partialptr = partial_dir_fname(fn)) == NULL | |
287 | + || !handle_partial_dir(partialptr, PDIR_CREATE)) | |
288 | + return; | |
289 | + /* We only use the file if we can hard-link it into our tmp dir. */ | |
290 | + if (link(fname, partialptr) != 0) { | |
291 | + if (errno != EEXIST) | |
292 | + handle_partial_dir(partialptr, PDIR_DELETE); | |
293 | + return; | |
1fffd582 | 294 | + } |
1fffd582 WD |
295 | + } |
296 | + | |
6fa0767f WD |
297 | + /* I think this falls into the -vv category with "%s is uptodate", etc. */ |
298 | + if (verbose > 1) | |
299 | + rprintf(FINFO, "found renamed: %s => %s\n", fname, fn); | |
1fffd582 | 300 | +} |
87d0091c WD |
301 | + |
302 | /* Delete a file or directory. If DEL_RECURSE is set in the flags, this will | |
303 | * delete recursively. | |
1fffd582 | 304 | * |
f813befd | 305 | * Note that fbuf must point to a MAXPATHLEN buffer if the mode indicates it's |
15894839 WD |
306 | * a directory! (The buffer is used for recursion, but returned unchanged.) |
307 | + * | |
308 | + * Also note: --detect-rename may use this routine with DEL_NO_DELETIONS set! | |
309 | */ | |
f9df736a | 310 | static enum delret delete_item(char *fbuf, uint16 mode, uint16 flags) |
15894839 | 311 | { |
4c107044 | 312 | @@ -186,6 +300,8 @@ static enum delret delete_item(char *fbuf, uint16 mode, uint16 flags) |
15894839 WD |
313 | goto check_ret; |
314 | /* OK: try to delete the directory. */ | |
315 | } | |
316 | + if (flags & DEL_NO_DELETIONS) | |
317 | + return DR_SUCCESS; | |
318 | ||
a5e6228a | 319 | if (!(flags & DEL_MAKE_ROOM) && max_delete >= 0 && ++deletion_count > max_delete) |
15894839 | 320 | return DR_AT_LIMIT; |
4c107044 | 321 | @@ -241,6 +357,8 @@ static enum delret delete_item(char *fbuf, uint16 mode, uint16 flags) |
d16b5fd6 WD |
322 | * its contents, otherwise just checks for content. Returns DR_SUCCESS or |
323 | * DR_NOT_EMPTY. Note that fname must point to a MAXPATHLEN buffer! (The | |
324 | * buffer is used for recursion, but returned unchanged.) | |
1fffd582 | 325 | + * |
87d0091c | 326 | + * Note: --detect-rename may use this routine with DEL_NO_DELETIONS set! |
1fffd582 | 327 | */ |
f9df736a | 328 | static enum delret delete_dir_contents(char *fname, uint16 flags) |
1fffd582 | 329 | { |
4c107044 | 330 | @@ -260,7 +378,9 @@ static enum delret delete_dir_contents(char *fname, uint16 flags) |
a47d1f86 WD |
331 | save_filters = push_local_filters(fname, dlen); |
332 | ||
333 | non_perishable_cnt = 0; | |
7e27b6c0 | 334 | + file_extra_cnt += SUM_EXTRA_CNT; |
a47d1f86 | 335 | dirlist = get_dirlist(fname, dlen, 0); |
7e27b6c0 | 336 | + file_extra_cnt -= SUM_EXTRA_CNT; |
a47d1f86 WD |
337 | ret = non_perishable_cnt ? DR_NOT_EMPTY : DR_SUCCESS; |
338 | ||
9c85142a | 339 | if (!dirlist->used) |
4c107044 | 340 | @@ -300,7 +420,8 @@ static enum delret delete_dir_contents(char *fname, uint16 flags) |
f9df736a | 341 | if (S_ISDIR(fp->mode)) { |
a5e6228a WD |
342 | if (delete_dir_contents(fname, flags | DEL_RECURSE) != DR_SUCCESS) |
343 | ret = DR_NOT_EMPTY; | |
344 | - } | |
345 | + } else if (detect_renamed && S_ISREG(fp->mode)) | |
a47d1f86 | 346 | + look_for_rename(fp, fname); |
a5e6228a | 347 | if (delete_item(fname, fp->mode, flags) != DR_SUCCESS) |
d16b5fd6 WD |
348 | ret = DR_NOT_EMPTY; |
349 | } | |
4c107044 | 350 | @@ -465,13 +586,18 @@ static void do_delayed_deletions(char *delbuf) |
1fffd582 WD |
351 | * all the --delete-WHEN options. Note that the fbuf pointer must point to a |
352 | * MAXPATHLEN buffer with the name of the directory in it (the functions we | |
353 | * call will append names onto the end, but the old dir value will be restored | |
354 | - * on exit). */ | |
73adde61 | 355 | -static void delete_in_dir(char *fbuf, struct file_struct *file, dev_t *fs_dev) |
1fffd582 WD |
356 | + * on exit). |
357 | + * | |
358 | + * Note: --detect-rename may use this routine with DEL_NO_DELETIONS set! | |
359 | + */ | |
73adde61 | 360 | +static void delete_in_dir(char *fbuf, struct file_struct *file, dev_t *fs_dev, |
f9df736a | 361 | + int del_flags) |
1fffd582 | 362 | { |
1fffd582 WD |
363 | static int already_warned = 0; |
364 | struct file_list *dirlist; | |
365 | - char delbuf[MAXPATHLEN]; | |
6fa0767f | 366 | - int dlen, i; |
1fffd582 WD |
367 | + char *p, delbuf[MAXPATHLEN]; |
368 | + unsigned remainder; | |
6fa0767f | 369 | + int dlen, i, restore_dot = 0; |
f9df736a | 370 | int save_uid_ndx = uid_ndx; |
1fffd582 | 371 | |
73adde61 | 372 | if (!fbuf) { |
4c107044 | 373 | @@ -486,17 +612,22 @@ static void delete_in_dir(char *fbuf, struct file_struct *file, dev_t *fs_dev) |
1fffd582 WD |
374 | maybe_send_keepalive(); |
375 | ||
041d67b8 | 376 | if (io_error && !ignore_errors) { |
1fffd582 WD |
377 | - if (already_warned) |
378 | + if (!already_warned) { | |
379 | + rprintf(FINFO, | |
380 | + "IO error encountered -- skipping file deletion\n"); | |
381 | + already_warned = 1; | |
382 | + } | |
383 | + if (!detect_renamed) | |
384 | return; | |
385 | - rprintf(FINFO, | |
386 | - "IO error encountered -- skipping file deletion\n"); | |
387 | - already_warned = 1; | |
388 | - return; | |
f9df736a | 389 | + del_flags |= DEL_NO_DELETIONS; |
1fffd582 WD |
390 | } |
391 | ||
1fffd582 | 392 | dlen = strlen(fbuf); |
fc068916 | 393 | change_local_filter_dir(fbuf, dlen, F_DEPTH(file)); |
1fffd582 WD |
394 | |
395 | + if (detect_renamed) | |
396 | + unexplored_dirs--; | |
397 | + | |
398 | if (one_file_system) { | |
399 | if (file->flags & FLAG_TOP_DIR) | |
fc068916 | 400 | filesystem_dev = *fs_dev; |
4c107044 | 401 | @@ -509,6 +640,14 @@ static void delete_in_dir(char *fbuf, struct file_struct *file, dev_t *fs_dev) |
1fffd582 WD |
402 | |
403 | dirlist = get_dirlist(fbuf, dlen, 0); | |
404 | ||
405 | + p = fbuf + dlen; | |
6fa0767f WD |
406 | + if (dlen == 1 && *fbuf == '.') { |
407 | + restore_dot = 1; | |
408 | + p = fbuf; | |
409 | + } else if (dlen != 1 || *fbuf != '/') | |
1fffd582 WD |
410 | + *p++ = '/'; |
411 | + remainder = MAXPATHLEN - (p - fbuf); | |
412 | + | |
413 | /* If an item in dirlist is not found in flist, delete it | |
414 | * from the filesystem. */ | |
9c85142a | 415 | for (i = dirlist->used; i--; ) { |
4c107044 | 416 | @@ -521,19 +660,28 @@ static void delete_in_dir(char *fbuf, struct file_struct *file, dev_t *fs_dev) |
87d0091c | 417 | f_name(fp, NULL)); |
1fffd582 | 418 | continue; |
87d0091c | 419 | } |
1fffd582 WD |
420 | + if (detect_renamed && S_ISREG(fp->mode)) { |
421 | + strlcpy(p, fp->basename, remainder); | |
a47d1f86 | 422 | + look_for_rename(fp, fbuf); |
1fffd582 | 423 | + } |
73adde61 | 424 | if (flist_find(cur_flist, fp) < 0) { |
f9df736a WD |
425 | int flags = DEL_RECURSE; |
426 | if (!(fp->mode & S_IWUSR) && !am_root && (uid_t)F_OWNER(fp) == our_uid) | |
427 | flags |= DEL_NO_UID_WRITE; | |
1fffd582 | 428 | f_name(fp, delbuf); |
1071853f | 429 | - if (delete_during == 2) { |
a5e6228a | 430 | - if (!remember_delete(fp, delbuf, flags)) |
f9df736a WD |
431 | + if (delete_during == 2 && !(del_flags & DEL_NO_DELETIONS)) { |
432 | + if (!remember_delete(fp, delbuf, del_flags | flags)) | |
1071853f WD |
433 | break; |
434 | } else | |
a5e6228a | 435 | - delete_item(delbuf, fp->mode, flags); |
1fffd582 | 436 | - } |
f9df736a | 437 | + delete_item(delbuf, fp->mode, del_flags | flags); |
1fffd582 WD |
438 | + } else if (detect_renamed && S_ISDIR(fp->mode)) |
439 | + unexplored_dirs++; | |
440 | } | |
441 | ||
6fa0767f WD |
442 | + if (restore_dot) |
443 | + fbuf[0] = '.'; | |
1fffd582 WD |
444 | + fbuf[dlen] = '\0'; |
445 | + | |
446 | flist_free(dirlist); | |
1fffd582 | 447 | |
f9df736a | 448 | if (!save_uid_ndx) { |
4c107044 | 449 | @@ -571,9 +719,9 @@ static void do_delete_pass(void) |
1fffd582 WD |
450 | || !S_ISDIR(st.st_mode)) |
451 | continue; | |
452 | ||
73adde61 WD |
453 | - delete_in_dir(fbuf, file, &st.st_dev); |
454 | + delete_in_dir(fbuf, file, &st.st_dev, 0); | |
1fffd582 | 455 | } |
73adde61 WD |
456 | - delete_in_dir(NULL, NULL, &dev_zero); |
457 | + delete_in_dir(NULL, NULL, &dev_zero, 0); | |
1fffd582 WD |
458 | |
459 | if (do_progress && !am_server) | |
460 | rprintf(FINFO, " \r"); | |
4c107044 | 461 | @@ -1226,6 +1374,7 @@ static void list_file_entry(struct file_struct *f) |
cdcd2137 | 462 | } |
1fffd582 WD |
463 | } |
464 | ||
465 | +static struct bitbag *delayed_bits = NULL; | |
466 | static int phase = 0; | |
ffc18846 | 467 | static int dflt_perms; |
1fffd582 | 468 | |
4c107044 | 469 | @@ -1502,9 +1651,12 @@ static void recv_generator(char *fname, struct file_struct *file, int ndx, |
fc068916 | 470 | } |
c0c7984e | 471 | else if (delete_during && f_out != -1 && !phase |
65ecbe35 WD |
472 | && !(file->flags & FLAG_MISSING_DIR)) { |
473 | - if (file->flags & FLAG_CONTENT_DIR) | |
474 | - delete_in_dir(fname, file, &real_sx.st.st_dev); | |
475 | - else | |
476 | + if (file->flags & FLAG_CONTENT_DIR) { | |
477 | + if (detect_renamed && real_ret != 0) | |
478 | + unexplored_dirs++; | |
479 | + delete_in_dir(fname, file, &real_sx.st.st_dev, | |
480 | + delete_during < 0 ? DEL_NO_DELETIONS : 0); | |
481 | + } else | |
482 | change_local_filter_dir(fname, strlen(fname), F_DEPTH(file)); | |
65ecbe35 | 483 | } |
cbdf862c | 484 | goto cleanup; |
4c107044 | 485 | @@ -1782,8 +1934,14 @@ static void recv_generator(char *fname, struct file_struct *file, int ndx, |
ffc18846 | 486 | goto cleanup; |
ccdb48f6 | 487 | } |
81172142 | 488 | #endif |
1fffd582 WD |
489 | - if (stat_errno == ENOENT) |
490 | + if (stat_errno == ENOENT) { | |
491 | + if (detect_renamed && unexplored_dirs > 0 | |
a47d1f86 | 492 | + && F_LENGTH(file)) { |
1fffd582 WD |
493 | + bitbag_set_bit(delayed_bits, ndx); |
494 | + return; | |
495 | + } | |
496 | goto notify_others; | |
497 | + } | |
cc3e685d | 498 | rsyserr(FERROR_XFER, stat_errno, "recv_generator: failed to stat %s", |
1fffd582 | 499 | full_fname(fname)); |
ffc18846 | 500 | goto cleanup; |
4c107044 | 501 | @@ -2184,6 +2342,12 @@ void generate_files(int f_out, const char *local_name) |
fc068916 WD |
502 | if (verbose > 2) |
503 | rprintf(FINFO, "generator starting pid=%ld\n", (long)getpid()); | |
1fffd582 WD |
504 | |
505 | + if (detect_renamed) { | |
7bfcb297 | 506 | + delayed_bits = bitbag_create(cur_flist->used); |
1fffd582 WD |
507 | + if (!delete_before && !delete_during) |
508 | + delete_during = -1; | |
509 | + } | |
510 | + | |
9c85142a | 511 | if (delete_before && !solo_file && cur_flist->used > 0) |
73adde61 | 512 | do_delete_pass(); |
1071853f | 513 | if (delete_during == 2) { |
4c107044 | 514 | @@ -2194,7 +2358,7 @@ void generate_files(int f_out, const char *local_name) |
1071853f | 515 | } |
1fffd582 WD |
516 | do_progress = 0; |
517 | ||
fc068916 WD |
518 | - if (append_mode > 0 || whole_file < 0) |
519 | + if (append_mode > 0 || detect_renamed || whole_file < 0) | |
1fffd582 WD |
520 | whole_file = 0; |
521 | if (verbose >= 2) { | |
522 | rprintf(FINFO, "delta-transmission %s\n", | |
4c107044 WD |
523 | @@ -2236,7 +2400,7 @@ void generate_files(int f_out, const char *local_name) |
524 | dirdev = MAKEDEV(DEV_MAJOR(devp), DEV_MINOR(devp)); | |
525 | } else | |
526 | dirdev = MAKEDEV(0, 0); | |
65ecbe35 WD |
527 | - delete_in_dir(fbuf, fp, &dirdev); |
528 | + delete_in_dir(fbuf, fp, &dirdev, 0); | |
529 | } else | |
530 | change_local_filter_dir(fbuf, strlen(fbuf), F_DEPTH(fp)); | |
fc068916 | 531 | } |
4c107044 | 532 | @@ -2283,7 +2447,21 @@ void generate_files(int f_out, const char *local_name) |
2dbc45e7 | 533 | } while ((cur_flist = cur_flist->next) != NULL); |
fc068916 | 534 | |
2dbc45e7 | 535 | if (delete_during) |
73adde61 WD |
536 | - delete_in_dir(NULL, NULL, &dev_zero); |
537 | + delete_in_dir(NULL, NULL, &dev_zero, 0); | |
2dbc45e7 WD |
538 | + if (detect_renamed) { |
539 | + if (delete_during < 0) | |
540 | + delete_during = 0; | |
541 | + detect_renamed = 0; | |
1fffd582 | 542 | + |
2dbc45e7 | 543 | + for (i = -1; (i = bitbag_next_bit(delayed_bits, i)) >= 0; ) { |
9bcaf4de | 544 | + struct file_struct *file = cur_flist->files[i]; |
2dbc45e7 WD |
545 | + if (local_name) |
546 | + strlcpy(fbuf, local_name, sizeof fbuf); | |
547 | + else | |
548 | + f_name(file, fbuf); | |
549 | + recv_generator(fbuf, file, i, itemizing, code, f_out); | |
550 | + } | |
551 | + } | |
552 | phase++; | |
553 | if (verbose > 2) | |
554 | rprintf(FINFO, "generate_files phase=%d\n", phase); | |
cc3e685d WD |
555 | diff --git a/options.c b/options.c |
556 | --- a/options.c | |
557 | +++ b/options.c | |
c0c7984e | 558 | @@ -82,6 +82,7 @@ int am_generator = 0; |
a94141d9 | 559 | int am_starting_up = 1; |
1fffd582 WD |
560 | int relative_paths = -1; |
561 | int implied_dirs = 1; | |
562 | +int detect_renamed = 0; | |
563 | int numeric_ids = 0; | |
564 | int allow_8bit_chars = 0; | |
565 | int force_delete = 0; | |
c0c7984e | 566 | @@ -391,6 +392,7 @@ void usage(enum logcode F) |
1fffd582 WD |
567 | rprintf(F," --modify-window=NUM compare mod-times with reduced accuracy\n"); |
568 | rprintf(F," -T, --temp-dir=DIR create temporary files in directory DIR\n"); | |
569 | rprintf(F," -y, --fuzzy find similar file for basis if no dest file\n"); | |
570 | + rprintf(F," --detect-renamed try to find renamed files to speed up the transfer\n"); | |
571 | rprintf(F," --compare-dest=DIR also compare destination files relative to DIR\n"); | |
572 | rprintf(F," --copy-dest=DIR ... and include copies of unchanged files\n"); | |
573 | rprintf(F," --link-dest=DIR hardlink to files in DIR when unchanged\n"); | |
c0c7984e | 574 | @@ -578,6 +580,7 @@ static struct poptOption long_options[] = { |
1fffd582 WD |
575 | {"compare-dest", 0, POPT_ARG_STRING, 0, OPT_COMPARE_DEST, 0, 0 }, |
576 | {"copy-dest", 0, POPT_ARG_STRING, 0, OPT_COPY_DEST, 0, 0 }, | |
577 | {"link-dest", 0, POPT_ARG_STRING, 0, OPT_LINK_DEST, 0, 0 }, | |
578 | + {"detect-renamed", 0, POPT_ARG_NONE, &detect_renamed, 0, 0, 0 }, | |
c0c7984e WD |
579 | {"fuzzy", 'y', POPT_ARG_VAL, &fuzzy_basis, 1, 0, 0 }, |
580 | {"no-fuzzy", 0, POPT_ARG_VAL, &fuzzy_basis, 0, 0, 0 }, | |
581 | {"no-y", 0, POPT_ARG_VAL, &fuzzy_basis, 0, 0, 0 }, | |
582 | @@ -1590,7 +1593,7 @@ int parse_arguments(int *argc_p, const char ***argv_p) | |
1fffd582 WD |
583 | inplace = 1; |
584 | } | |
585 | ||
586 | - if (delay_updates && !partial_dir) | |
587 | + if ((delay_updates || detect_renamed) && !partial_dir) | |
588 | partial_dir = tmp_partialdir; | |
589 | ||
590 | if (inplace) { | |
c0c7984e | 591 | @@ -1599,6 +1602,7 @@ int parse_arguments(int *argc_p, const char ***argv_p) |
1fffd582 WD |
592 | snprintf(err_buf, sizeof err_buf, |
593 | "--%s cannot be used with --%s\n", | |
594 | append_mode ? "append" : "inplace", | |
595 | + detect_renamed ? "detect-renamed" : | |
596 | delay_updates ? "delay-updates" : "partial-dir"); | |
597 | return 0; | |
598 | } | |
ae306a29 | 599 | @@ -1959,6 +1963,8 @@ void server_options(char **args, int *argc_p) |
51bc0e89 WD |
600 | args[ac++] = "--super"; |
601 | if (size_only) | |
602 | args[ac++] = "--size-only"; | |
603 | + if (detect_renamed) | |
604 | + args[ac++] = "--detect-renamed"; | |
605 | } else { | |
606 | if (skip_compress) { | |
607 | if (asprintf(&arg, "--skip-compress=%s", skip_compress) < 0) | |
cc3e685d WD |
608 | diff --git a/rsync.yo b/rsync.yo |
609 | --- a/rsync.yo | |
610 | +++ b/rsync.yo | |
611 | @@ -385,6 +385,7 @@ to the detailed description below for a complete description. verb( | |
1fffd582 WD |
612 | --modify-window=NUM compare mod-times with reduced accuracy |
613 | -T, --temp-dir=DIR create temporary files in directory DIR | |
614 | -y, --fuzzy find similar file for basis if no dest file | |
615 | + --detect-renamed try to find renamed files to speed the xfer | |
616 | --compare-dest=DIR also compare received files relative to DIR | |
617 | --copy-dest=DIR ... and include copies of unchanged files | |
618 | --link-dest=DIR hardlink to files in DIR when unchanged | |
ae306a29 | 619 | @@ -1487,6 +1488,21 @@ Note that the use of the bf(--delete) option might get rid of any potential |
1fffd582 WD |
620 | fuzzy-match files, so either use bf(--delete-after) or specify some |
621 | filename exclusions if you need to prevent this. | |
622 | ||
6fa0767f WD |
623 | +dit(bf(--detect-renamed)) With this option, for each new source file |
624 | +(call it em(src/S)), rsync looks for a file em(dest/D) anywhere in the | |
625 | +destination that passes the quick check with em(src/S). If such a em(dest/D) | |
626 | +is found, rsync uses it as an alternate basis for transferring em(S). The | |
627 | +idea is that if em(src/S) was renamed from em(src/D) (as opposed to em(src/S) | |
628 | +passing the quick check with em(dest/D) by coincidence), the delta-transfer | |
629 | +algorithm will find that all the data matches between em(src/S) and em(dest/D), | |
630 | +and the transfer will be really fast. | |
631 | + | |
1fffd582 WD |
632 | +By default, alternate-basis files are hard-linked into a directory named |
633 | +".~tmp~" in each file's destination directory, but if you've specified | |
634 | +the bf(--partial-dir) option, that directory will be used instead. These | |
635 | +potential alternate-basis files will be removed as the transfer progresses. | |
636 | +This option conflicts with bf(--inplace) and bf(--append). | |
637 | + | |
638 | dit(bf(--compare-dest=DIR)) This option instructs rsync to use em(DIR) on | |
639 | the destination machine as an additional hierarchy to compare destination | |
640 | files against doing transfers (if the files are missing in the destination | |
cc3e685d WD |
641 | diff --git a/util.c b/util.c |
642 | --- a/util.c | |
643 | +++ b/util.c | |
ae306a29 | 644 | @@ -1129,6 +1129,32 @@ int handle_partial_dir(const char *fname, int create) |
1fffd582 WD |
645 | return 1; |
646 | } | |
647 | ||
648 | +/* We need to supply our own strcmp function for file list comparisons | |
649 | + * to ensure that signed/unsigned usage is consistent between machines. */ | |
650 | +int u_strcmp(const char *p1, const char *p2) | |
651 | +{ | |
652 | + for ( ; *p1; p1++, p2++) { | |
653 | + if (*p1 != *p2) | |
654 | + break; | |
655 | + } | |
656 | + | |
657 | + return (int)*(uchar*)p1 - (int)*(uchar*)p2; | |
658 | +} | |
659 | + | |
660 | +/* We need a memcmp function compares unsigned-byte values. */ | |
661 | +int u_memcmp(const void *p1, const void *p2, size_t len) | |
662 | +{ | |
663 | + const uchar *u1 = p1; | |
664 | + const uchar *u2 = p2; | |
665 | + | |
666 | + while (len--) { | |
667 | + if (*u1 != *u2) | |
668 | + return (int)*u1 - (int)*u2; | |
669 | + } | |
670 | + | |
671 | + return 0; | |
672 | +} | |
673 | + | |
674 | /** | |
675 | * Determine if a symlink points outside the current directory tree. | |
676 | * This is considered "unsafe" because e.g. when mirroring somebody |