A way to allow rsync to use a checksum cache on the sending side,

[rsync/rsync-patches.git] / checksum-updating.diff

This adds a sender optimization feature that allows a cache of checksums
to be created/updated and used when the client specifies the --checksum
option.

To use this patch, run these commands for a successful build:

    patch -p1 <patches/checksum-updating.diff
    ./configure                               (optional if already run)
    make

TODO:  when sending individual files (as opposed to an entire directory),
we should still update the .md[45]sums file(s) if we compute new checksum
info.  (Writing currently only occurs if we send an entire dir.)

--- old/clientserver.c
+++ new/clientserver.c
@@ -37,6 +37,7 @@ extern int sanitize_paths;
 extern int filesfrom_fd;
 extern int remote_protocol;
 extern int protocol_version;
+extern int checksum_updating;
 extern int io_timeout;
 extern int no_detach;
 extern int default_af_hint;
@@ -634,6 +635,8 @@ static int rsync_module(int f_in, int f_
 	else if (am_root < 0) /* Treat --fake-super from client as --super. */
 		am_root = 2;
 
+	checksum_updating = lp_checksum_updating(i);
+
 	if (filesfrom_fd == 0)
 		filesfrom_fd = f_in;
 
--- old/flist.c
+++ new/flist.c
@@ -25,6 +25,7 @@
 #include "io.h"
 
 extern int verbose;
+extern int dry_run;
 extern int list_only;
 extern int am_root;
 extern int am_server;
@@ -57,6 +58,7 @@ extern int implied_dirs;
 extern int file_extra_cnt;
 extern int ignore_perishable;
 extern int non_perishable_cnt;
+extern int checksum_updating;
 extern int prune_empty_dirs;
 extern int copy_links;
 extern int copy_unsafe_links;
@@ -101,6 +103,8 @@ static char tmp_sum[MAX_DIGEST_LEN];
 static char empty_sum[MAX_DIGEST_LEN];
 static int flist_count_offset; /* for --delete --progress */
 static int dir_count = 0;
+static struct file_list *checksum_flist = NULL;
+static int checksum_matches = 0;
 
 static void clean_flist(struct file_list *flist, int strip_root);
 static void output_flist(struct file_list *flist);
@@ -317,6 +321,216 @@ static void flist_done_allocating(struct
 		flist->pool_boundary = ptr;
 }
 
+/* The len count is the length of the basename + 1 for the null. */
+static void add_checksum(const char *dirname, const char *basename, int len,
+			 OFF_T file_length, time_t mtime, const char *sum,
+			 int flags)
+{
+	struct file_struct *file;
+	int alloc_len, extra_len;
+	char *bp;
+
+	if (len == 8+1 && *basename == '.'
+	 && (strcmp(basename, ".md5sums") == 0
+	  || strcmp(basename, ".md4sums") == 0))
+		return;
+
+	if (len < 0)
+		len = strlen(basename) + 1;
+
+	extra_len = (file_extra_cnt + (file_length > 0xFFFFFFFFu) + SUM_EXTRA_CNT)
+	          * EXTRA_LEN;
+#if EXTRA_ROUNDING > 0
+	if (extra_len & (EXTRA_ROUNDING * EXTRA_LEN))
+		extra_len = (extra_len | (EXTRA_ROUNDING * EXTRA_LEN)) + EXTRA_LEN;
+#endif
+	alloc_len = FILE_STRUCT_LEN + extra_len + len;
+	bp = pool_alloc(checksum_flist->file_pool, alloc_len, "add_checksum");
+
+	memset(bp, 0, extra_len + FILE_STRUCT_LEN);
+	bp += extra_len;
+	file = (struct file_struct *)bp;
+	bp += FILE_STRUCT_LEN;
+
+	memcpy(bp, basename, len);
+
+	file->flags = flags;
+	file->mode = S_IFREG;
+	file->modtime = mtime;
+	file->len32 = (uint32)file_length;
+	if (file_length > 0xFFFFFFFFu) {
+		file->flags |= FLAG_LENGTH64;
+		OPT_EXTRA(file, 0)->unum = (uint32)(file_length >> 32);
+	}
+	file->dirname = dirname;
+	bp = (char*)F_SUM(file);
+	memcpy(bp, sum, checksum_len);
+
+	flist_expand(checksum_flist, 1);
+	checksum_flist->files[checksum_flist->count++] = file;
+
+	checksum_flist->sorted = checksum_flist->files;
+}
+
+/* The direname value must remain unchanged during the lifespan of the
+ * created checksum_flist object because we use it directly. */
+static void read_checksums(const char *dirname)
+{
+	char line[MAXPATHLEN+1024], fbuf[MAXPATHLEN], sum[MAX_DIGEST_LEN];
+	const char *filename;
+	OFF_T file_length;
+	time_t mtime;
+	int len, dlen, i, flags;
+	char *cp;
+	FILE *fp;
+
+	if (checksum_flist) {
+		/* Reset the pool memory and empty the file-list array. */
+		pool_free_old(checksum_flist->file_pool,
+			      pool_boundary(checksum_flist->file_pool, 0));
+		checksum_flist->count = 0;
+	} else
+		checksum_flist = flist_new(FLIST_TEMP, "read_checksums");
+
+	checksum_flist->low = 0;
+	checksum_flist->high = -1;
+	checksum_matches = 0;
+
+	if (protocol_version >= 30)
+		filename = ".md5sums";
+	else
+		filename = ".md4sums";
+	if (dirname) {
+		dlen = strlcpy(fbuf, dirname, sizeof fbuf);
+		if (dlen >= (int)sizeof fbuf)
+			return;
+		fbuf[dlen++] = '/';
+	} else
+		dlen = 0;
+	strlcpy(fbuf+dlen, filename, sizeof fbuf - dlen);
+	if (!(fp = fopen(fbuf, "r")))
+		return;
+
+	while (fgets(line, sizeof line, fp)) {
+		for (i = 0, cp = line; i < checksum_len*2; i++, cp++) {
+			int x;
+			if (isDigit(cp))
+				x = *cp - '0';
+			else if (isAlpha(cp)) {
+				x = (*cp & 0xF) + 9;
+				if (x > 0xF) {
+					cp = "";
+					break;
+				}
+			} else {
+				cp = "";
+				break;
+			}
+			if (i & 1)
+				sum[i/2] |= x;
+			else
+				sum[i/2] = x << 4;
+		}
+
+		if (*cp != ' ')
+			continue;
+		while (*++cp == ' ') {}
+
+		file_length = 0;
+		while (isDigit(cp))
+			file_length = file_length * 10 + *cp++ - '0';
+
+		if (*cp != ' ')
+			continue;
+		while (*++cp == ' ') {}
+
+		mtime = 0;
+		while (isDigit(cp))
+			mtime = mtime * 10 + *cp++ - '0';
+
+		if (*cp != ' ')
+			continue;
+		while (*++cp == ' ') {}
+
+		len = strlen(cp);
+		while (len && (cp[len-1] == '\n' || cp[len-1] == '\r'))
+			len--;
+		if (!len)
+			continue;
+		cp[len++] = '\0'; /* len now counts the null */
+		if (strchr(cp, '/') || len > MAXPATHLEN)
+			continue;
+
+		strlcpy(fbuf+dlen, cp, sizeof fbuf - dlen);
+		if (is_excluded(fbuf, 0, ALL_FILTERS)) {
+			flags = FLAG_FILE_SENT;
+			checksum_matches++;
+		} else
+			flags = 0;
+
+		add_checksum(dirname, cp, len, file_length, mtime, sum, flags);
+	}
+	fclose(fp);
+
+	clean_flist(checksum_flist, 0);
+}
+
+static void write_checksums(const char *dirname)
+{
+	char buf[MAXPATHLEN+1024];
+	const char *filename;
+	int new_entries = checksum_flist->count > checksum_flist->high + 1;
+	int orphan_entires = checksum_flist->count != checksum_matches;
+	FILE *out_fp;
+	int i;
+
+	if (dry_run)
+		return;
+
+	for (i = checksum_flist->high + 1; i < checksum_flist->count; i++) {
+		struct file_struct *file = checksum_flist->sorted[i];
+		file->flags |= FLAG_FILE_SENT;
+	}
+
+	clean_flist(checksum_flist, 0);
+
+	if (protocol_version >= 30)
+		filename = ".md5sums";
+	else
+		filename = ".md4sums";
+	if (dirname) {
+		if (pathjoin(buf, sizeof buf, dirname, filename) >= sizeof buf)
+			return;
+	} else
+		strlcpy(buf, filename, sizeof buf);
+
+	if (checksum_flist->high - checksum_flist->low < 0) {
+		unlink(buf);
+		return;
+	}
+
+	if (!new_entries && !orphan_entires)
+		return;
+
+	if (!(out_fp = fopen(buf, "w")))
+		return;
+
+	for (i = checksum_flist->low; i <= checksum_flist->high; i++) {
+		struct file_struct *file = checksum_flist->sorted[i];
+		const char *cp = F_SUM(file);
+		const char *end = cp + checksum_len;
+		if (!(file->flags & FLAG_FILE_SENT))
+			continue;
+		while (cp != end)
+			fprintf(out_fp, "%02x", CVAL(cp++, 0));
+		fprintf(out_fp, " %10.0f %10ld %s\n",
+			(double)F_LENGTH(file), (long)file->modtime,
+			file->basename);
+	}
+
+	fclose(out_fp);
+}
+
 int push_pathname(const char *dir, int len)
 {
 	if (dir == pathname)
@@ -973,34 +1187,24 @@ static struct file_struct *recv_file_ent
 	return file;
 }
 
-/**
- * Create a file_struct for a named file by reading its stat()
- * information and performing extensive checks against global
- * options.
- *
- * @return the new file, or NULL if there was an error or this file
- * should be excluded.
+/* Create a file_struct for a named file by reading its stat() information
+ * and performing extensive checks against global options.
  *
- * @todo There is a small optimization opportunity here to avoid
- * stat()ing the file in some circumstances, which has a certain cost.
- * We are called immediately after doing readdir(), and so we may
- * already know the d_type of the file.  We could for example avoid
- * statting directories if we're not recursing, but this is not a very
- * important case.  Some systems may not have d_type.
- **/
+ * Returns a pointer to the new file struct, or NULL if there was an error
+ * or this file should be excluded. */
 struct file_struct *make_file(const char *fname, struct file_list *flist,
 			      STRUCT_STAT *stp, int flags, int filter_level)
 {
 	static char *lastdir;
-	static int lastdir_len = -1;
+	static int lastdir_len = -2;
 	struct file_struct *file;
-	STRUCT_STAT st;
 	char thisname[MAXPATHLEN];
 	char linkname[MAXPATHLEN];
 	int alloc_len, basename_len, linkname_len;
 	int extra_len = file_extra_cnt * EXTRA_LEN;
 	const char *basename;
 	alloc_pool_t *pool;
+	STRUCT_STAT st;
 	char *bp;
 
 	if (strlcpy(thisname, fname, sizeof thisname)
@@ -1115,9 +1319,16 @@ struct file_struct *make_file(const char
 			memcpy(lastdir, thisname, len);
 			lastdir[len] = '\0';
 			lastdir_len = len;
+			if (always_checksum && am_sender && flist)
+				read_checksums(lastdir);
 		}
-	} else
+	} else {
 		basename = thisname;
+		if (always_checksum && am_sender && flist && lastdir_len == -2) {
+			lastdir_len = -1;
+			read_checksums(NULL);
+		}
+	}
 	basename_len = strlen(basename) + 1; /* count the '\0' */
 
 #ifdef SUPPORT_LINKS
@@ -1193,11 +1404,30 @@ struct file_struct *make_file(const char
 	}
 #endif
 
-	if (always_checksum && am_sender && S_ISREG(st.st_mode))
-		file_checksum(thisname, tmp_sum, st.st_size);
-
 	F_PATHNAME(file) = pathname;
 
+	if (always_checksum && am_sender && S_ISREG(st.st_mode)) {
+		int j;
+		if (flist && (j = flist_find(checksum_flist, file)) >= 0) {
+			struct file_struct *fp = checksum_flist->sorted[j];
+			if (fp->modtime == file->modtime && F_LENGTH(fp) == F_LENGTH(file)) {
+				memcpy(tmp_sum, F_SUM(fp), MAX_DIGEST_LEN);
+				fp->flags |= FLAG_FILE_SENT;
+				checksum_matches++;
+			} else {
+				clear_file(fp);
+				goto compute_checksum;
+			}
+		} else {
+		  compute_checksum:
+			file_checksum(thisname, tmp_sum, st.st_size);
+			if (checksum_updating && flist) {
+				add_checksum(file->dirname, basename, basename_len,
+					     st.st_size, st.st_mtime, tmp_sum, 0);
+			}
+		}
+	}
+
 	/* This code is only used by the receiver when it is building
 	 * a list of files for a delete pass. */
 	if (keep_dirlinks && linkname_len && flist) {
@@ -1241,14 +1471,14 @@ void unmake_file(struct file_struct *fil
 
 static struct file_struct *send_file_name(int f, struct file_list *flist,
 					  char *fname, STRUCT_STAT *stp,
-					  int flags, int filter_flags)
+					  int flags, int filter_level)
 {
 	struct file_struct *file;
 #if defined SUPPORT_ACLS || defined SUPPORT_XATTRS
 	statx sx;
 #endif
 
-	file = make_file(fname, flist, stp, flags, filter_flags);
+	file = make_file(fname, flist, stp, flags, filter_level);
 	if (!file)
 		return NULL;
 
@@ -1442,7 +1672,7 @@ static void send_directory(int f, struct
 	DIR *d;
 	int divert_dirs = (flags & FLAG_DIVERT_DIRS) != 0;
 	int start = flist->count;
-	int filter_flags = f == -2 ? SERVER_FILTERS : ALL_FILTERS;
+	int filter_level = f == -2 ? SERVER_FILTERS : ALL_FILTERS;
 
 	assert(flist != NULL);
 
@@ -1471,7 +1701,7 @@ static void send_directory(int f, struct
 			continue;
 		}
 
-		send_file_name(f, flist, fbuf, NULL, flags, filter_flags);
+		send_file_name(f, flist, fbuf, NULL, flags, filter_level);
 	}
 
 	fbuf[len] = '\0';
@@ -1483,6 +1713,9 @@ static void send_directory(int f, struct
 
 	closedir(d);
 
+	if (checksum_updating && always_checksum && am_sender && f >= 0)
+		write_checksums(fbuf);
+
 	if (f >= 0 && recurse && !divert_dirs) {
 		int i, end = flist->count - 1;
 		/* send_if_directory() bumps flist->count, so use "end". */
@@ -2206,7 +2439,7 @@ void flist_free(struct file_list *flist)
 
 	if (!flist->prev || !flist_cnt)
 		pool_destroy(flist->file_pool);
-	else
+	else if (flist->pool_boundary)
 		pool_free_old(flist->file_pool, flist->pool_boundary);
 
 	if (flist->sorted && flist->sorted != flist->files)
@@ -2225,6 +2458,7 @@ static void clean_flist(struct file_list
 	if (!flist)
 		return;
 	if (flist->count == 0) {
+		flist->low = 0;
 		flist->high = -1;
 		return;
 	}
--- old/loadparm.c
+++ new/loadparm.c
@@ -149,6 +149,7 @@ typedef struct
 	int syslog_facility;
 	int timeout;
 
+	BOOL checksum_updating;
 	BOOL fake_super;
 	BOOL ignore_errors;
 	BOOL ignore_nonreadable;
@@ -197,6 +198,7 @@ static service sDefault =
  /* syslog_facility; */		LOG_DAEMON,
  /* timeout; */			0,
 
+ /* checksum_updating; */	False,
  /* fake_super; */		False,
  /* ignore_errors; */		False,
  /* ignore_nonreadable; */	False,
@@ -313,6 +315,7 @@ static struct parm_struct parm_table[] =
  {"lock file",         P_STRING, P_LOCAL, &sDefault.lock_file,         NULL,0},
  {"log file",          P_STRING, P_LOCAL, &sDefault.log_file,          NULL,0},
  {"log format",        P_STRING, P_LOCAL, &sDefault.log_format,        NULL,0},
+ {"checksum updating", P_BOOL,   P_LOCAL, &sDefault.checksum_updating, NULL,0},
  {"max connections",   P_INTEGER,P_LOCAL, &sDefault.max_connections,   NULL,0},
  {"max verbosity",     P_INTEGER,P_LOCAL, &sDefault.max_verbosity,     NULL,0},
  {"name",              P_STRING, P_LOCAL, &sDefault.name,              NULL,0},
@@ -418,6 +421,7 @@ FN_LOCAL_BOOL(lp_fake_super, fake_super)
 FN_LOCAL_BOOL(lp_ignore_errors, ignore_errors)
 FN_LOCAL_BOOL(lp_ignore_nonreadable, ignore_nonreadable)
 FN_LOCAL_BOOL(lp_list, list)
+FN_LOCAL_BOOL(lp_checksum_updating, checksum_updating)
 FN_LOCAL_BOOL(lp_read_only, read_only)
 FN_LOCAL_BOOL(lp_strict_modes, strict_modes)
 FN_LOCAL_BOOL(lp_transfer_logging, transfer_logging)
--- old/options.c
+++ new/options.c
@@ -109,6 +109,7 @@ size_t bwlimit_writemax = 0;
 int ignore_existing = 0;
 int ignore_non_existing = 0;
 int need_messages_from_generator = 0;
+int checksum_updating = 0;
 int max_delete = -1;
 OFF_T max_size = 0;
 OFF_T min_size = 0;
@@ -302,6 +303,7 @@ void usage(enum logcode F)
   rprintf(F," -q, --quiet                 suppress non-error messages\n");
   rprintf(F,"     --no-motd               suppress daemon-mode MOTD (see manpage caveat)\n");
   rprintf(F," -c, --checksum              skip based on checksum, not mod-time & size\n");
+  rprintf(F,"     --checksum-updating     sender updates .md[45]sums files\n");
   rprintf(F," -a, --archive               archive mode; equals -rlptgoD (no -H,-A,-X)\n");
   rprintf(F,"     --no-OPTION             turn off an implied OPTION (e.g. --no-D)\n");
   rprintf(F," -r, --recursive             recurse into directories\n");
@@ -542,6 +544,7 @@ static struct poptOption long_options[] 
   {"checksum",        'c', POPT_ARG_VAL,    &always_checksum, 1, 0, 0 },
   {"no-checksum",      0,  POPT_ARG_VAL,    &always_checksum, 0, 0, 0 },
   {"no-c",             0,  POPT_ARG_VAL,    &always_checksum, 0, 0, 0 },
+  {"checksum-updating",0,  POPT_ARG_NONE,   &checksum_updating, 0, 0, 0 },
   {"block-size",      'B', POPT_ARG_LONG,   &block_size, 0, 0, 0 },
   {"compare-dest",     0,  POPT_ARG_STRING, 0, OPT_COMPARE_DEST, 0, 0 },
   {"copy-dest",        0,  POPT_ARG_STRING, 0, OPT_COPY_DEST, 0, 0 },
@@ -1896,7 +1899,9 @@ void server_options(char **args,int *arg
 				args[ac++] = basis_dir[i];
 			}
 		}
-	}
+	} else if (checksum_updating)
+		args[ac++] = "--checksum-updating";
+
 
 	if (append_mode)
 		args[ac++] = "--append";
--- old/rsync.h
+++ new/rsync.h
@@ -1070,6 +1070,12 @@ isDigit(const char *ptr)
 }
 
 static inline int
+isAlpha(const char *ptr)
+{
+	return isalpha(*(unsigned char *)ptr);
+}
+
+static inline int
 isPrint(const char *ptr)
 {
 	return isprint(*(unsigned char *)ptr);
--- old/rsync.yo
+++ new/rsync.yo
@@ -307,6 +307,7 @@ to the detailed description below for a 
  -q, --quiet                 suppress non-error messages
      --no-motd               suppress daemon-mode MOTD (see caveat)
  -c, --checksum              skip based on checksum, not mod-time & size
+     --checksum-updating     sender updates .md[45]sums files
  -a, --archive               archive mode; equals -rlptgoD (no -H,-A,-X)
      --no-OPTION             turn off an implied OPTION (e.g. --no-D)
  -r, --recursive             recurse into directories
@@ -502,9 +503,9 @@ uses a "quick check" that (by default) c
 of last modification match between the sender and receiver.  This option
 changes this to compare a 128-bit MD4 checksum for each file that has a
 matching size.  Generating the checksums means that both sides will expend
-a lot of disk I/O reading all the data in the files in the transfer (and
-this is prior to any reading that will be done to transfer changed files),
-so this can slow things down significantly.
+a lot of disk I/O reading the data in all the files in the transfer, so
+this can slow things down significantly (and this is prior to any reading
+that will be done to transfer the files that have changed).
 
 The sending side generates its checksums while it is doing the file-system
 scan that builds the list of the available files.  The receiver generates
@@ -512,12 +513,43 @@ its checksums when it is scanning for ch
 file that has the same size as the corresponding sender's file:  files with
 either a changed size or a changed checksum are selected for transfer.
 
+Starting with version 3.0.0, the sending side will look for a checksum
+summary file and use a pre-generated checksum that it reads out of the file
+(as long as it matches the file's size and modified time).  This allows a
+server to support the --checksum option to clients without having to
+recompute the checksums for each client.  See the bf(--checksum-updating)
+option for a way to have rsync create/update the checksum files.
+
 Note that rsync always verifies that each em(transferred) file was
 correctly reconstructed on the receiving side by checking a whole-file
 checksum that is generated when as the file is transferred, but that
 automatic after-the-transfer verification has nothing to do with this
 option's before-the-transfer "Does this file need to be updated?" check.
 
+dit(bf(--checksum-updating)) This option tells the sending side to create
+and/or update per-directory checksum files that are used by the
+bf(--checksum) option.  The file that is updated is either .md5sums (for
+protocols >= 30) or .md4sums (for older protocols).  If pre-transfer
+checksums are not being computed, this option has no effect.
+
+The checksum files stores the computed checksum, last-known size,
+modification time, and name for each file in the current directory.  If a
+later transfer finds that a file matches its prior size and modification
+time, the checksum is assumed to still be correct.  Otherwise it is
+recomputed and udpated in the file.
+
+To avoid transferring the system's checksum files, you can use an exclude
+(e.g. bf(--exclude=.md[45]sums)).  To make this easier to type, you can use
+a popt alias.  For instance, adding the following line in your ~/.popt file
+defines a bf(-cc) option that enables checksum updating and excludes the
+checksum files:
+
+verb(  rsync alias --cc --checksum-updating --exclude='.md[45]sums')
+
+An rsync daemon does not allow the client to control this setting, so see
+the "checksum updating" daemon config option for information on how to make
+a daemon maintain these checksum files.
+
 dit(bf(-a, --archive)) This is equivalent to bf(-rlptgoD). It is a quick
 way of saying you want recursion and want to preserve almost
 everything (with -H being a notable omission).
--- old/rsyncd.conf.yo
+++ new/rsyncd.conf.yo
@@ -198,6 +198,21 @@ locking on this file to ensure that the 
 exceeded for the modules sharing the lock file.
 The default is tt(/var/run/rsyncd.lock).
 
+dit(bf(checksum updating)) This option tells rsync to update/create the
+checksum information in the per-directory checksum files when users copy
+files using the bf(--checksum) option.  Any file that has changed since it
+was last checksummed (or is not mentioned) has its data updated in the
+.md4sums or .md5sums file (the file used depends on what protocol version
+is used for the transfer).  
+
+Note that this updating will occur even if the module is listed as being
+read-only.  If you want to hide these files (and you will almost always
+want to do), add ".md[45]sums" to the module's exclude setting.
+
+Note also that the client's command-line option, bf(--checksum-updating),
+has no effect on a daemon.  A daemon will only update/create checksum files
+if this config option is true.
+
 dit(bf(read only)) The "read only" option determines whether clients
 will be able to upload files or not. If "read only" is true then any
 attempted uploads will fail. If "read only" is false then uploads will