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.
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.
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.
22 To use this patch, run these commands for a successful build:
24 patch -p1 <patches/detect-renamed.diff
25 ./configure (optional if already run)
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
37 @@ -51,6 +51,7 @@ extern int implied_dirs;
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;
44 extern struct stats stats;
45 @@ -68,6 +69,8 @@ int checksum_len;
46 dev_t filesystem_dev; /* used to implement -x */
47 unsigned int file_struct_len;
49 +struct file_list the_fattr_list;
51 static char empty_sum[MD4_SUM_LENGTH];
52 static int flist_count_offset;
54 @@ -250,6 +253,44 @@ static mode_t from_wire_mode(int mode)
58 +static int fattr_compare(struct file_struct **file1, struct file_struct **file2)
60 + struct file_struct *f1 = *file1;
61 + struct file_struct *f2 = *file2;
64 + if (!f1->basename || !S_ISREG(f1->mode) || !f1->length) {
65 + if (!f2->basename || !S_ISREG(f2->mode) || !f2->length)
69 + if (!f2->basename || !S_ISREG(f2->mode) || !f2->length)
72 + /* Don't use diff for values that are longer than an int. */
73 + if (f1->length != f2->length)
74 + return f1->length < f2->length ? -1 : 1;
76 + if (always_checksum) {
77 + diff = u_memcmp(f1->u.sum, f2->u.sum, checksum_len);
80 + } else if (f1->modtime != f2->modtime)
81 + return f1->modtime < f2->modtime ? -1 : 1;
83 + diff = u_strcmp(f1->basename, f2->basename);
87 + if (f1->dirname == f2->dirname)
93 + return u_strcmp(f1->dirname, f2->dirname);
96 static void send_directory(int f, struct file_list *flist,
99 @@ -1379,6 +1420,25 @@ struct file_list *recv_file_list(int f)
101 clean_flist(flist, relative_paths, 1);
103 + if (detect_renamed) {
104 + int j = flist->count;
105 + the_fattr_list.count = j;
106 + the_fattr_list.files = new_array(struct file_struct *, j);
107 + if (!the_fattr_list.files)
109 + memcpy(the_fattr_list.files, flist->files,
110 + j * sizeof (struct file_struct *));
111 + qsort(the_fattr_list.files, j,
112 + sizeof the_fattr_list.files[0], (int (*)())fattr_compare);
113 + the_fattr_list.low = 0;
115 + struct file_struct *fp = the_fattr_list.files[j];
116 + if (fp->basename && S_ISREG(fp->mode) && fp->length)
119 + the_fattr_list.high = j;
123 recv_uid_list(f, flist);
127 @@ -76,6 +76,7 @@ extern char *basis_dir[];
128 extern int compare_dest;
129 extern int copy_dest;
130 extern int link_dest;
131 +extern int detect_renamed;
132 extern int whole_file;
133 extern int list_only;
134 extern int new_root_dir;
135 @@ -91,12 +92,15 @@ extern char *backup_dir;
136 extern char *backup_suffix;
137 extern int backup_suffix_len;
138 extern struct file_list *the_file_list;
139 +extern struct file_list the_fattr_list;
140 extern struct filter_list_struct server_filter_list;
142 static int deletion_count = 0; /* used to implement --max-delete */
143 +static int unexplored_dirs = 1;
145 -/* For calling delete_item() */
146 +/* For calling delete_item() and delete_in_dir() */
147 #define DEL_RECURSE (1<<1) /* recurse */
148 +#define DEL_NO_DELETIONS (1<<2)
151 TYPE_DIR, TYPE_SPECIAL, TYPE_DEVICE, TYPE_SYMLINK
152 @@ -116,6 +120,113 @@ static int is_backup_file(char *fn)
153 return k > 0 && strcmp(fn+k, backup_suffix) == 0;
156 +/* Search for a regular file that matches either (1) the size & modified
157 + * time (plus the basename, if possible) or (2) the size & checksum. If
158 + * we find an exact match down to the dirname, return -1 because we found
159 + * an up-to-date file in the transfer, not a renamed file. */
160 +static int fattr_find(struct file_struct *f, char *fname, alloc_pool_t pool)
162 + int low = the_fattr_list.low, high = the_fattr_list.high;
163 + int mid, ok_match = -1, good_match = -1;
164 + struct file_struct *fmid;
167 + while (low <= high) {
168 + mid = (low + high) / 2;
169 + fmid = the_fattr_list.files[mid];
170 + if (fmid->length != f->length) {
171 + if (fmid->length < f->length)
177 + if (always_checksum) {
179 + if (fmid->modtime == f->modtime
180 + && f_name_cmp(fmid, f) == 0)
181 + return -1; /* assume we can't help */
182 + f->u.sum = pool_alloc(pool, MD4_SUM_LENGTH,
184 + file_checksum(fname, f->u.sum, f->length);
186 + diff = u_memcmp(fmid->u.sum, f->u.sum, checksum_len);
195 + if (fmid->modtime != f->modtime) {
196 + if (fmid->modtime < f->modtime)
204 + diff = u_strcmp(fmid->basename, f->basename);
207 + if (fmid->dirname == f->dirname)
208 + return -1; /* file is up-to-date */
209 + if (!fmid->dirname) {
217 + diff = u_strcmp(fmid->dirname, f->dirname);
219 + return -1; /* file is up-to-date */
227 + return good_match >= 0 ? good_match : ok_match;
230 +static void look_for_rename(struct file_struct *file, char *fname,
233 + struct file_struct *fp;
234 + char *partialptr, *fn;
238 + if ((ndx = fattr_find(file, fname, pool)) < 0)
241 + fp = the_fattr_list.files[ndx];
242 + fn = f_name(fp, NULL);
243 + /* We don't provide an alternate-basis file if there is a basis file. */
244 + if (link_stat(fn, &st, 0) == 0)
246 + if ((partialptr = partial_dir_fname(fn)) == NULL
247 + || !handle_partial_dir(partialptr, PDIR_CREATE))
250 + /* We only use the file if we can hard-link it into our tmp dir. */
251 + if (link(fname, partialptr) == 0) {
253 + rprintf(FINFO, "found renamed: %s => %s\n",
254 + fname, partialptr);
259 + if (errno != EEXIST)
260 + handle_partial_dir(partialptr, PDIR_DELETE);
263 /* Delete a file or directory. If DEL_RECURSE is set in the flags, this will
264 * delete recursively.
266 @@ -185,6 +296,8 @@ static enum delret delete_item(char *fna
267 /* Prep directory is to be deleted, so delete all its contents. Note
268 * that fname must point to a MAXPATHLEN buffer! (The buffer is used
269 * for recursion, but returned unchanged.)
271 + * Note: --detect-rename may use this routine with DEL_NO_DELETIONS set!
273 static enum delret delete_dir_contents(char *fname, int flags)
275 @@ -221,6 +334,16 @@ static enum delret delete_dir_contents(c
278 strlcpy(p, fp->basename, remainder);
279 + if (detect_renamed) {
280 + if (S_ISDIR(fp->mode) && flags & DEL_NO_DELETIONS) {
281 + delete_dir_contents(fname, flags);
284 + if (S_ISREG(fp->mode))
285 + look_for_rename(fp, fname, dirlist->file_pool);
286 + if (flags & DEL_NO_DELETIONS)
289 result = delete_item(fname, fp->mode, NULL, flags);
290 if (result != DR_SUCCESS && ret == DR_SUCCESS)
291 ret = result == DR_PINNED ? result : DR_NOT_EMPTY;
292 @@ -239,15 +362,19 @@ static enum delret delete_dir_contents(c
293 * all the --delete-WHEN options. Note that the fbuf pointer must point to a
294 * MAXPATHLEN buffer with the name of the directory in it (the functions we
295 * call will append names onto the end, but the old dir value will be restored
299 + * Note: --detect-rename may use this routine with DEL_NO_DELETIONS set!
301 static void delete_in_dir(struct file_list *flist, char *fbuf,
302 - struct file_struct *file, STRUCT_STAT *stp)
303 + struct file_struct *file, STRUCT_STAT *stp, int flags)
305 static int min_depth = MAXPATHLEN, cur_depth = -1;
306 static void *filt_array[MAXPATHLEN/2+1];
307 static int already_warned = 0;
308 struct file_list *dirlist;
309 - char delbuf[MAXPATHLEN];
310 + char *p, delbuf[MAXPATHLEN];
311 + unsigned remainder;
315 @@ -261,6 +388,8 @@ static void delete_in_dir(struct file_li
317 rprintf(FINFO, "delete_in_dir(%s)\n", fbuf);
319 + flags |= DEL_RECURSE;
322 maybe_send_keepalive();
324 @@ -268,12 +397,14 @@ static void delete_in_dir(struct file_li
325 return; /* Impossible... */
327 if (io_error && !(lp_ignore_errors(module_id) || ignore_errors)) {
328 - if (already_warned)
329 + if (!already_warned) {
331 + "IO error encountered -- skipping file deletion\n");
332 + already_warned = 1;
334 + if (!detect_renamed)
337 - "IO error encountered -- skipping file deletion\n");
338 - already_warned = 1;
340 + flags |= DEL_NO_DELETIONS;
343 while (cur_depth >= file->dir.depth && cur_depth >= min_depth)
344 @@ -284,6 +415,9 @@ static void delete_in_dir(struct file_li
346 filt_array[cur_depth] = push_local_filters(fbuf, dlen);
348 + if (detect_renamed)
351 if (one_file_system) {
352 if (file->flags & FLAG_TOP_DIR)
353 filesystem_dev = stp->st_dev;
354 @@ -293,6 +427,11 @@ static void delete_in_dir(struct file_li
356 dirlist = get_dirlist(fbuf, dlen, 0);
359 + if (dlen != 1 || *fbuf != '/')
361 + remainder = MAXPATHLEN - (p - fbuf);
363 /* If an item in dirlist is not found in flist, delete it
364 * from the filesystem. */
365 for (i = dirlist->count; i--; ) {
366 @@ -305,12 +444,19 @@ static void delete_in_dir(struct file_li
370 + if (detect_renamed && S_ISREG(fp->mode)) {
371 + strlcpy(p, fp->basename, remainder);
372 + look_for_rename(fp, fbuf, dirlist->file_pool);
374 if (flist_find(flist, fp) < 0) {
376 - delete_item(delbuf, fp->mode, NULL, DEL_RECURSE);
378 + delete_item(delbuf, fp->mode, NULL, flags);
379 + } else if (detect_renamed && S_ISDIR(fp->mode))
388 @@ -340,9 +486,9 @@ static void do_delete_pass(struct file_l
389 || !S_ISDIR(st.st_mode))
392 - delete_in_dir(flist, fbuf, file, &st);
393 + delete_in_dir(flist, fbuf, file, &st, 0);
395 - delete_in_dir(NULL, NULL, NULL, NULL);
396 + delete_in_dir(NULL, NULL, NULL, NULL, 0);
398 if (do_progress && !am_server)
399 rprintf(FINFO, " \r");
400 @@ -871,6 +1017,7 @@ static int try_dests_non(struct file_str
404 +static struct bitbag *delayed_bits = NULL;
405 static int phase = 0;
407 /* Acts on the_file_list->file's ndx'th item, whose name is fname. If a dir,
408 @@ -1056,8 +1203,12 @@ static void recv_generator(char *fname,
409 if (real_ret != 0 && one_file_system)
410 real_st.st_dev = filesystem_dev;
411 if (delete_during && f_out != -1 && !phase && dry_run < 2
412 - && (file->flags & FLAG_DEL_HERE))
413 - delete_in_dir(the_file_list, fname, file, &real_st);
414 + && (file->flags & FLAG_DEL_HERE)) {
415 + if (detect_renamed && real_ret != 0)
417 + delete_in_dir(the_file_list, fname, file, &real_st,
418 + delete_during < 0 ? DEL_NO_DELETIONS : 0);
423 @@ -1309,8 +1460,14 @@ static void recv_generator(char *fname,
424 && hard_link_check(file, ndx, fname, statret, &st,
425 itemizing, code, HL_SKIP))
427 - if (stat_errno == ENOENT)
428 + if (stat_errno == ENOENT) {
429 + if (detect_renamed && unexplored_dirs > 0
431 + bitbag_set_bit(delayed_bits, ndx);
436 rsyserr(FERROR, stat_errno, "recv_generator: failed to stat %s",
439 @@ -1496,11 +1653,17 @@ void generate_files(int f_out, struct fi
440 (long)getpid(), flist->count);
443 + if (detect_renamed) {
444 + delayed_bits = bitbag_create(flist->count);
445 + if (!delete_before && !delete_during)
446 + delete_during = -1;
449 if (delete_before && !local_name && flist->count > 0)
450 do_delete_pass(flist);
453 - if (append_mode || whole_file < 0)
454 + if (append_mode || detect_renamed || whole_file < 0)
457 rprintf(FINFO, "delta-transmission %s\n",
458 @@ -1555,7 +1718,23 @@ void generate_files(int f_out, struct fi
460 recv_generator(NULL, NULL, 0, 0, 0, code, -1);
462 - delete_in_dir(NULL, NULL, NULL, NULL);
463 + delete_in_dir(NULL, NULL, NULL, NULL, 0);
465 + if (detect_renamed) {
466 + if (delete_during < 0)
468 + detect_renamed = 0;
470 + for (i = -1; (i = bitbag_next_bit(delayed_bits, i)) >= 0; ) {
471 + struct file_struct *file = flist->files[i];
473 + strlcpy(fbuf, local_name, sizeof fbuf);
475 + f_name(file, fbuf);
476 + recv_generator(fbuf, file, i, itemizing,
477 + maybe_ATTRS_REPORT, code, f_out);
482 csum_length = SUM_LENGTH;
485 @@ -78,6 +78,7 @@ int am_generator = 0;
486 int am_starting_up = 1;
487 int relative_paths = -1;
488 int implied_dirs = 1;
489 +int detect_renamed = 0;
491 int allow_8bit_chars = 0;
492 int force_delete = 0;
493 @@ -346,6 +347,7 @@ void usage(enum logcode F)
494 rprintf(F," --modify-window=NUM compare mod-times with reduced accuracy\n");
495 rprintf(F," -T, --temp-dir=DIR create temporary files in directory DIR\n");
496 rprintf(F," -y, --fuzzy find similar file for basis if no dest file\n");
497 + rprintf(F," --detect-renamed try to find renamed files to speed up the transfer\n");
498 rprintf(F," --compare-dest=DIR also compare destination files relative to DIR\n");
499 rprintf(F," --copy-dest=DIR ... and include copies of unchanged files\n");
500 rprintf(F," --link-dest=DIR hardlink to files in DIR when unchanged\n");
501 @@ -499,6 +501,7 @@ static struct poptOption long_options[]
502 {"compare-dest", 0, POPT_ARG_STRING, 0, OPT_COMPARE_DEST, 0, 0 },
503 {"copy-dest", 0, POPT_ARG_STRING, 0, OPT_COPY_DEST, 0, 0 },
504 {"link-dest", 0, POPT_ARG_STRING, 0, OPT_LINK_DEST, 0, 0 },
505 + {"detect-renamed", 0, POPT_ARG_NONE, &detect_renamed, 0, 0, 0 },
506 {"fuzzy", 'y', POPT_ARG_NONE, &fuzzy_basis, 0, 0, 0 },
507 {"compress", 'z', POPT_ARG_NONE, 0, 'z', 0, 0 },
508 {"compress-level", 0, POPT_ARG_INT, &def_compress_level, 'z', 0, 0 },
509 @@ -1362,7 +1365,7 @@ int parse_arguments(int *argc, const cha
513 - if (delay_updates && !partial_dir)
514 + if ((delay_updates || detect_renamed) && !partial_dir)
515 partial_dir = tmp_partialdir;
518 @@ -1371,6 +1374,7 @@ int parse_arguments(int *argc, const cha
519 snprintf(err_buf, sizeof err_buf,
520 "--%s cannot be used with --%s\n",
521 append_mode ? "append" : "inplace",
522 + detect_renamed ? "detect-renamed" :
523 delay_updates ? "delay-updates" : "partial-dir");
526 @@ -1674,6 +1678,8 @@ void server_options(char **args,int *arg
527 args[ac++] = "--super";
529 args[ac++] = "--size-only";
530 + if (detect_renamed)
531 + args[ac++] = "--detect-renamed";
534 if (modify_window_set) {
537 @@ -363,6 +363,7 @@ to the detailed description below for a
538 --modify-window=NUM compare mod-times with reduced accuracy
539 -T, --temp-dir=DIR create temporary files in directory DIR
540 -y, --fuzzy find similar file for basis if no dest file
541 + --detect-renamed try to find renamed files to speed the xfer
542 --compare-dest=DIR also compare received files relative to DIR
543 --copy-dest=DIR ... and include copies of unchanged files
544 --link-dest=DIR hardlink to files in DIR when unchanged
545 @@ -1264,6 +1265,15 @@ Note that the use of the bf(--delete) op
546 fuzzy-match files, so either use bf(--delete-after) or specify some
547 filename exclusions if you need to prevent this.
549 +dit(bf(--detect-renamed)) This option tells rsync to scan the receiving
550 +side for files that have been renamed, and to use any that are found as
551 +alternate basis files to help speed up the transfer.
552 +By default, alternate-basis files are hard-linked into a directory named
553 +".~tmp~" in each file's destination directory, but if you've specified
554 +the bf(--partial-dir) option, that directory will be used instead. These
555 +potential alternate-basis files will be removed as the transfer progresses.
556 +This option conflicts with bf(--inplace) and bf(--append).
558 dit(bf(--compare-dest=DIR)) This option instructs rsync to use em(DIR) on
559 the destination machine as an additional hierarchy to compare destination
560 files against doing transfers (if the files are missing in the destination
563 @@ -1025,6 +1025,32 @@ int handle_partial_dir(const char *fname
567 +/* We need to supply our own strcmp function for file list comparisons
568 + * to ensure that signed/unsigned usage is consistent between machines. */
569 +int u_strcmp(const char *p1, const char *p2)
571 + for ( ; *p1; p1++, p2++) {
576 + return (int)*(uchar*)p1 - (int)*(uchar*)p2;
579 +/* We need a memcmp function compares unsigned-byte values. */
580 +int u_memcmp(const void *p1, const void *p2, size_t len)
582 + const uchar *u1 = p1;
583 + const uchar *u2 = p2;
587 + return (int)*u1 - (int)*u2;
594 * Determine if a symlink points outside the current directory tree.
595 * This is considered "unsafe" because e.g. when mirroring somebody