extern int checksum_seed;
extern int protocol_version;
-@@ -221,3 +222,22 @@ int sum_end(char *sum)
+@@ -221,3 +222,24 @@ int sum_end(char *sum)
return MD4_DIGEST_LEN;
}
+ int i, x1, x2;
+ char *c = buf + checksum_len*2;
+
++ assert(c - buf < (int)sizeof buf);
++
+ *c = '\0';
+
+ for (i = checksum_len; --i >= 0; ) {
+
+ return buf;
+}
-diff --git a/flist.c b/flist.c
---- a/flist.c
-+++ b/flist.c
-@@ -72,6 +72,7 @@ extern int unsort_ndx;
- extern uid_t our_uid;
- extern struct stats stats;
- extern char *filesfrom_host;
+diff --git a/clientserver.c b/clientserver.c
+--- a/clientserver.c
++++ b/clientserver.c
+@@ -50,6 +50,7 @@ extern int logfile_format_has_i;
+ extern int logfile_format_has_o_or_i;
+ extern char *bind_address;
+ extern char *config_file;
+extern char *link_by_hash_dir;
- extern char *usermap, *groupmap;
-
- extern char curr_dir[MAXPATHLEN];
-@@ -903,7 +904,7 @@ static struct file_struct *recv_file_entry(int f, struct file_list *flist, int x
- extra_len += EXTRA_LEN;
- #endif
-
-- if (always_checksum && S_ISREG(mode))
-+ if ((always_checksum || link_by_hash_dir) && S_ISREG(mode))
- extra_len += SUM_EXTRA_CNT * EXTRA_LEN;
+ extern char *logfile_format;
+ extern char *files_from;
+ extern char *tmpdir;
+@@ -551,6 +552,9 @@ static int rsync_module(int f_in, int f_out, int i, const char *addr, const char
+ return -1;
+ }
- #if SIZEOF_INT64 >= 8
++ if (*lp_link_by_hash_dir(i))
++ link_by_hash_dir = lp_link_by_hash_dir(i);
++
+ if (am_daemon && am_server) {
+ rprintf(FLOG, "rsync allowed access on module %s from %s (%s)\n",
+ name, host, addr);
+diff --git a/compat.c b/compat.c
+--- a/compat.c
++++ b/compat.c
+@@ -55,6 +55,7 @@ extern char *partial_dir;
+ extern char *dest_option;
+ extern char *files_from;
+ extern char *filesfrom_host;
++extern char *link_by_hash_dir;
+ extern filter_rule_list filter_list;
+ extern int need_unsorted_flist;
+ #ifdef ICONV_OPTION
+@@ -328,4 +329,8 @@ void setup_protocol(int f_out,int f_in)
+ } else {
+ checksum_seed = read_int(f_in);
+ }
++ if (!am_sender && link_by_hash_dir && protocol_version < 30 && checksum_seed != 1) {
++ rprintf(FERROR, "You must specify --checksum-seed=1 when using --link-by-hash with an old version of rsync.\n");
++ exit_cleanup(RERR_PROTOCOL);
++ }
+ }
diff --git a/hashlink.c b/hashlink.c
new file mode 100644
--- /dev/null
+++ b/hashlink.c
-@@ -0,0 +1,325 @@
+@@ -0,0 +1,334 @@
+/*
+ Copyright (C) Cronosys, LLC 2004
+
+
+#include "rsync.h"
+
-+extern int checksum_len;
+extern char *link_by_hash_dir;
++extern char sender_file_sum[MAX_DIGEST_LEN];
+
+#ifdef HAVE_LINK
+
-+char *make_hash_name(struct file_struct *file)
++/* This function is always called after a file is received, so the
++ * sender_file_sum buffer has whatever the last checksum was for the
++ * transferred file. */
++static char *make_hash_name(void)
+{
-+ const char *hex = sum_as_hex(F_SUM(file));
++ const char *hex = sum_as_hex(sender_file_sum);
+ char *dst;
+
+ if (asprintf(&dst, "%s/%.8s/%s", link_by_hash_dir, hex, hex+8) < 0)
+}
+
+
-+void kill_hashfile(struct hashfile_struct *hashfile)
++static void kill_hashfile(struct hashfile_struct *hashfile)
+{
+ if (!hashfile)
+ return;
+}
+
+
-+void kill_hashfiles(struct hashfile_struct *hashfiles)
++static void kill_hashfiles(struct hashfile_struct *hashfiles)
+{
+ struct hashfile_struct *iter, *next;
+ if ((iter = hashfiles) != NULL) {
+}
+
+
-+struct hashfile_struct *find_hashfiles(char *hashname, int64 size, long *fnbr)
++static struct hashfile_struct *find_hashfiles(char *hashname, int64 size, long *fnbr)
+{
+ DIR *d;
+ struct dirent *di;
+}
+
+
-+struct hashfile_struct *compare_hashfiles(int fd,struct hashfile_struct *files)
++static struct hashfile_struct *compare_hashfiles(int fd,struct hashfile_struct *files)
+{
+ int amt, hamt;
+ char buffer[BUFSIZ], cmpbuffer[BUFSIZ];
+int link_by_hash(const char *fnametmp, const char *fname, struct file_struct *file)
+{
+ STRUCT_STAT st;
-+ char *hashname = make_hash_name(file);
++ char *hashname = make_hash_name();
+ int first = 0, rc;
+ char *linkname;
+ long last_fnbr;
+ first = 1;
+ if (asprintf(&linkname,"%s/0",hashname) < 0)
+ out_of_memory("link_by_hash");
-+ rprintf(FINFO, "(1) linkname = %s\n", linkname);
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "(1) linkname = %s\n", linkname);
+ } else {
+ struct hashfile_struct *hashfiles, *hashfile;
+
+ first = 1;
+ if (asprintf(&linkname,"%s/0",hashname) < 0)
+ out_of_memory("link_by_hash");
-+ rprintf(FINFO, "(2) linkname = %s\n", linkname);
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "(2) linkname = %s\n", linkname);
+ } else {
+ int fd;
+ /* Search for one identical to us. */
+ if (hashfile) {
+ first = 0;
+ linkname = strdup(hashfile->name);
-+ rprintf(FINFO, "(3) linkname = %s\n", linkname);
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "(3) linkname = %s\n", linkname);
+ kill_hashfile(hashfile);
+ } else {
+ first = 1;
+ if (asprintf(&linkname, "%s/%ld", hashname, last_fnbr + 1) < 0)
+ out_of_memory("link_by_hash");
-+ rprintf(FINFO, "(4) linkname = %s\n", linkname);
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "(4) linkname = %s\n", linkname);
+ }
+ }
+ }
+
+ if (!first) {
-+ rprintf(FINFO, "link-by-hash (existing): \"%s\" -> %s\n",
-+ linkname, full_fname(fname));
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "link-by-hash (existing): \"%s\" -> %s\n", linkname, full_fname(fname));
+ robust_unlink(fname);
+ rc = do_link(linkname, fname);
+ if (rc == -1) {
+ free(linkname);
+ if (asprintf(&linkname,"%s/%ld",hashname, last_fnbr + 1) < 0)
+ out_of_memory("link_by_hash");
-+ rprintf(FINFO, "(5) linkname = %s\n", linkname);
-+ rprintf(FINFO,"link-by-hash: max link count exceeded, starting new file \"%s\".\n", linkname);
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "(5) linkname = %s\n", linkname);
++ if (DEBUG_GTE(HASHLINK, 1))
++ rprintf(FINFO, "link-by-hash: max link count exceeded, starting new file \"%s\".\n", linkname);
+ } else {
+ rsyserr(FERROR, errno, "link \"%s\" -> \"%s\"",
+ linkname, full_fname(fname));
+ }
+
+ if (first) {
-+ rprintf(FINFO, "link-by-hash (new): %s -> \"%s\"\n",
-+ full_fname(fname),linkname);
++ if (DEBUG_GTE(HASHLINK, 2))
++ rprintf(FINFO, "link-by-hash (new): %s -> \"%s\"\n", full_fname(fname),linkname);
+
+ rc = robust_rename(fnametmp, fname, NULL, 0644);
+ if (rc != 0) {
+ return rc;
+}
+#endif
+diff --git a/loadparm.c b/loadparm.c
+--- a/loadparm.c
++++ b/loadparm.c
+@@ -119,6 +119,7 @@ typedef struct {
+ char *include;
+ char *include_from;
+ char *incoming_chmod;
++ char *link_by_hash_dir;
+ char *lock_file;
+ char *log_file;
+ char *log_format;
+@@ -195,6 +196,7 @@ static const all_vars Defaults = {
+ /* include; */ NULL,
+ /* include_from; */ NULL,
+ /* incoming_chmod; */ NULL,
++ /* link_by_hash_dir; */ NULL,
+ /* lock_file; */ DEFAULT_LOCK_FILE,
+ /* log_file; */ NULL,
+ /* log_format; */ "%o %h [%a] %m (%u) %f %l",
+@@ -336,6 +338,7 @@ static struct parm_struct parm_table[] =
+ {"include from", P_STRING, P_LOCAL, &Vars.l.include_from, NULL,0},
+ {"include", P_STRING, P_LOCAL, &Vars.l.include, NULL,0},
+ {"incoming chmod", P_STRING, P_LOCAL, &Vars.l.incoming_chmod, NULL,0},
++ {"link by hash dir", P_STRING, P_LOCAL, &Vars.l.link_by_hash_dir, NULL,0},
+ {"list", P_BOOL, P_LOCAL, &Vars.l.list, NULL,0},
+ {"lock file", P_STRING, P_LOCAL, &Vars.l.lock_file, NULL,0},
+ {"log file", P_STRING, P_LOCAL, &Vars.l.log_file, NULL,0},
+@@ -464,6 +467,7 @@ FN_LOCAL_STRING(lp_hosts_deny, hosts_deny)
+ FN_LOCAL_STRING(lp_include, include)
+ FN_LOCAL_STRING(lp_include_from, include_from)
+ FN_LOCAL_STRING(lp_incoming_chmod, incoming_chmod)
++FN_LOCAL_STRING(lp_link_by_hash_dir, link_by_hash_dir)
+ FN_LOCAL_STRING(lp_lock_file, lock_file)
+ FN_LOCAL_STRING(lp_log_file, log_file)
+ FN_LOCAL_STRING(lp_log_format, log_format)
diff --git a/log.c b/log.c
--- a/log.c
+++ b/log.c
char *config_file = NULL;
char *shell_cmd = NULL;
char *logfile_name = NULL;
-@@ -759,6 +760,7 @@ void usage(enum logcode F)
+@@ -208,7 +209,7 @@ static const char *debug_verbosity[] = {
+ /*2*/ "BIND,CMD,CONNECT,DEL,DELTASUM,DUP,FILTER,FLIST,ICONV",
+ /*3*/ "ACL,BACKUP,CONNECT2,DELTASUM2,DEL2,EXIT,FILTER2,FLIST2,FUZZY,GENR,OWN,RECV,SEND,TIME",
+ /*4*/ "CMD2,DELTASUM3,DEL3,EXIT2,FLIST3,ICONV2,OWN2,PROTO,TIME2",
+- /*5*/ "CHDIR,DELTASUM4,FLIST4,FUZZY2,HASH,HLINK",
++ /*5*/ "CHDIR,DELTASUM4,FLIST4,FUZZY2,HASH,HASHLINK,HLINK",
+ };
+
+ #define MAX_VERBOSITY ((int)(sizeof debug_verbosity / sizeof debug_verbosity[0]) - 1)
+@@ -278,6 +279,7 @@ static struct output_struct debug_words[COUNT_DEBUG+1] = {
+ DEBUG_WORD(FUZZY, W_REC, "Debug fuzzy scoring (levels 1-2)"),
+ DEBUG_WORD(GENR, W_REC, "Debug generator functions"),
+ DEBUG_WORD(HASH, W_SND|W_REC, "Debug hashtable code"),
++ DEBUG_WORD(HASHLINK, W_REC, "Debug hashlink code (levels 1-2)"),
+ DEBUG_WORD(HLINK, W_SND|W_REC, "Debug hard-link actions (levels 1-3)"),
+ DEBUG_WORD(ICONV, W_CLI|W_SRV, "Debug iconv character conversions (levels 1-2)"),
+ DEBUG_WORD(IO, W_CLI|W_SRV, "Debug I/O routines (levels 1-4)"),
+@@ -759,6 +761,7 @@ void usage(enum logcode F)
rprintf(F," --compare-dest=DIR also compare destination files relative to DIR\n");
rprintf(F," --copy-dest=DIR ... and include copies of unchanged files\n");
rprintf(F," --link-dest=DIR hardlink to files in DIR when unchanged\n");
rprintf(F," -z, --compress compress file data during the transfer\n");
rprintf(F," --compress-level=NUM explicitly set compression level\n");
rprintf(F," --skip-compress=LIST skip compressing files with a suffix in LIST\n");
-@@ -811,7 +813,7 @@ enum {OPT_VERSION = 1000, OPT_DAEMON, OPT_SENDER, OPT_EXCLUDE, OPT_EXCLUDE_FROM,
+@@ -811,7 +814,7 @@ enum {OPT_VERSION = 1000, OPT_DAEMON, OPT_SENDER, OPT_EXCLUDE, OPT_EXCLUDE_FROM,
OPT_FILTER, OPT_COMPARE_DEST, OPT_COPY_DEST, OPT_LINK_DEST, OPT_HELP,
OPT_INCLUDE, OPT_INCLUDE_FROM, OPT_MODIFY_WINDOW, OPT_MIN_SIZE, OPT_CHMOD,
OPT_READ_BATCH, OPT_WRITE_BATCH, OPT_ONLY_WRITE_BATCH, OPT_MAX_SIZE,
OPT_USERMAP, OPT_GROUPMAP, OPT_CHOWN, OPT_BWLIMIT,
OPT_SERVER, OPT_REFUSED_BASE = 9000};
-@@ -955,6 +957,7 @@ static struct poptOption long_options[] = {
+@@ -955,6 +958,7 @@ static struct poptOption long_options[] = {
{"compare-dest", 0, POPT_ARG_STRING, 0, OPT_COMPARE_DEST, 0, 0 },
{"copy-dest", 0, POPT_ARG_STRING, 0, OPT_COPY_DEST, 0, 0 },
{"link-dest", 0, POPT_ARG_STRING, 0, OPT_LINK_DEST, 0, 0 },
{"fuzzy", 'y', POPT_ARG_NONE, 0, 'y', 0, 0 },
{"no-fuzzy", 0, POPT_ARG_VAL, &fuzzy_basis, 0, 0, 0 },
{"no-y", 0, POPT_ARG_VAL, &fuzzy_basis, 0, 0, 0 },
-@@ -1785,6 +1788,21 @@ int parse_arguments(int *argc_p, const char ***argv_p)
+@@ -1299,6 +1303,9 @@ int parse_arguments(int *argc_p, const char ***argv_p)
+ iconv_opt = strdup(arg);
+ #endif
+
++ if (*lp_link_by_hash_dir(module_id))
++ set_refuse_options("link-by-hash");
++
+ /* TODO: Call poptReadDefaultConfig; handle errors. */
+
+ /* The context leaks in case of an error, but if there's a
+@@ -1785,6 +1792,21 @@ int parse_arguments(int *argc_p, const char ***argv_p)
return 0;
#endif
default:
/* A large opt value means that set_refuse_options()
* turned this option off. */
-@@ -2675,6 +2693,11 @@ void server_options(char **args, int *argc_p)
+@@ -2675,6 +2697,11 @@ void server_options(char **args, int *argc_p)
} else if (inplace)
args[ac++] = "--inplace";
if (files_from && (!am_sender || filesfrom_host)) {
if (filesfrom_host) {
args[ac++] = "--files-from";
-diff --git a/receiver.c b/receiver.c
---- a/receiver.c
-+++ b/receiver.c
-@@ -225,11 +225,13 @@ int open_tmpfile(char *fnametmp, const char *fname, struct file_struct *file)
- }
-
- static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
-- const char *fname, int fd, OFF_T total_size)
-+ const char *fname, int fd, OFF_T total_size,
-+ const char *md4)
- {
- static char file_sum1[MAX_DIGEST_LEN];
- struct map_struct *mapbuf;
- struct sum_struct sum;
-+ md_context mdfour_data;
- int32 len;
- OFF_T offset = 0;
- OFF_T offset2;
-@@ -265,6 +267,9 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
- } else
- mapbuf = NULL;
-
-+ if (md4)
-+ mdfour_begin(&mdfour_data);
-+
- sum_init(checksum_seed);
-
- if (append_mode > 0) {
-@@ -312,6 +317,8 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
- cleanup_got_literal = 1;
-
- sum_update(data, i);
-+ if (md4)
-+ mdfour_update(&mdfour_data, (uchar*)data, i);
-
- if (fd != -1 && write_file(fd,data,i) != i)
- goto report_write_error;
-@@ -339,6 +346,8 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
-
- see_token(map, len);
- sum_update(map, len);
-+ if (md4)
-+ mdfour_update(&mdfour_data, (uchar*)map, len);
- }
-
- if (updating_basis_or_equiv) {
-@@ -392,6 +401,9 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
- if (sum_end(file_sum1) != checksum_len)
- overflow_exit("checksum_len"); /* Impossible... */
-
-+ if (md4)
-+ mdfour_result(&mdfour_data, (uchar*)md4);
-+
- if (mapbuf)
- unmap_file(mapbuf);
-
-@@ -406,7 +418,7 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
-
- static void discard_receive_data(int f_in, OFF_T length)
- {
-- receive_data(f_in, NULL, -1, 0, NULL, -1, length);
-+ receive_data(f_in, NULL, -1, 0, NULL, -1, length, NULL);
- }
-
- static void handle_delayed_updates(char *local_name)
-@@ -841,7 +853,7 @@ int recv_files(int f_in, int f_out, char *local_name)
-
- /* recv file data */
- recv_ok = receive_data(f_in, fnamecmp, fd1, st.st_size,
-- fname, fd2, F_LENGTH(file));
-+ fname, fd2, F_LENGTH(file), F_SUM(file));
-
- log_item(log_code, file, iflags, NULL);
-
diff --git a/rsync.c b/rsync.c
--- a/rsync.c
+++ b/rsync.c
struct chmod_mode_struct;
struct flist_ndx_item {
+@@ -1263,7 +1271,8 @@ extern short info_levels[], debug_levels[];
+ #define DEBUG_FUZZY (DEBUG_FLIST+1)
+ #define DEBUG_GENR (DEBUG_FUZZY+1)
+ #define DEBUG_HASH (DEBUG_GENR+1)
+-#define DEBUG_HLINK (DEBUG_HASH+1)
++#define DEBUG_HASHLINK (DEBUG_HASH+1)
++#define DEBUG_HLINK (DEBUG_HASHLINK+1)
+ #define DEBUG_ICONV (DEBUG_HLINK+1)
+ #define DEBUG_IO (DEBUG_ICONV+1)
+ #define DEBUG_OWN (DEBUG_IO+1)
diff --git a/rsync.yo b/rsync.yo
--- a/rsync.yo
+++ b/rsync.yo
-z, --compress compress file data during the transfer
--compress-level=NUM explicitly set compression level
--skip-compress=LIST skip compressing files with suffix in LIST
+@@ -1830,6 +1831,19 @@ bf(--link-dest) from working properly for a non-super-user when bf(-o) was
+ specified (or implied by bf(-a)). You can work-around this bug by avoiding
+ the bf(-o) option when sending to an old rsync.
+
++dit(bf(--link-by-hash=DIR)) This option hard links the destination files into
++em(DIR), a link farm arranged by MD5 file hash (or sometimes MD4). The result
++is that the system will only store one copy of the unique contents of each
++file, regardless of the file's name.
++
++For a modern rsync (3.0.0 and newer), the link farm's directory hierarchy is
++determined by the file's MD5 hash. It is recommended that you don't use this
++option with any rsync older than that. However, if you really need to be able
++to interact with an older rsync on the sending side, you can use the options
++bf(--checksum-seed=1) and bf(--protocol=29) to force a consistent MD4 file
++checksum that will be used instead of MD5. Note that this MD4 checksum is not
++compatible with older versions of this patch (prior to 3.1.0).
++
+ dit(bf(-z, --compress)) With this option, rsync compresses the file data
+ as it is sent to the destination machine, which reduces the amount of data
+ being transmitted -- something that is useful over a slow connection.
+diff --git a/rsyncd.conf.yo b/rsyncd.conf.yo
+--- a/rsyncd.conf.yo
++++ b/rsyncd.conf.yo
+@@ -283,6 +283,13 @@ message telling them to try later. The default is 0, which means no limit.
+ A negative value disables the module.
+ See also the "lock file" parameter.
+
++dit(bf(link by hash dir)) When the "link by hash dir" parameter is set to a
++non-empty string, received files will be hard linked into em(DIR), a link farm
++arranged by MD5 file hash (or sometimes MD4). See the bf(--link-by-hash) option
++for a full explaination. If this parameter is set it will disable the
++bf(--link-by-hash) command-line option. The default is for this parameter to be
++unset.
++
+ dit(bf(log file)) When the "log file" parameter is set to a non-empty
+ string, the rsync daemon will log messages to the indicated file rather
+ than using syslog. This is particularly useful on systems (such as AIX)
git.samba.org / rsync-patches.git / commitdiff