patch -p1 <patches/link-by-hash.diff
./prepare-source
- ./configure
+ ./configure (optional if already run)
make
+based-on: 40afd365cc8ca968fd16e161d24df5b8a8a520cc
diff --git a/Makefile.in b/Makefile.in
--- a/Makefile.in
+++ b/Makefile.in
diff --git a/flist.c b/flist.c
--- a/flist.c
+++ b/flist.c
-@@ -70,6 +70,7 @@ extern int need_unsorted_flist;
- extern int unsort_ndx;
+@@ -70,6 +70,7 @@ extern int unsort_ndx;
+ extern uid_t our_uid;
extern struct stats stats;
extern char *filesfrom_host;
+extern char *link_by_hash_dir;
extern char curr_dir[MAXPATHLEN];
-@@ -824,7 +825,7 @@ static struct file_struct *recv_file_entry(struct file_list *flist,
- extra_len += (S_ISDIR(mode) ? 2 : 1) * EXTRA_LEN;
+@@ -854,7 +855,7 @@ 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;
- if (file_length > 0xFFFFFFFFu && S_ISREG(mode))
+ #if SIZEOF_INT64 >= 8
diff --git a/hashlink.c b/hashlink.c
new file mode 100644
--- /dev/null
+++ b/hashlink.c
-@@ -0,0 +1,336 @@
+@@ -0,0 +1,340 @@
+/*
+ Copyright (C) Cronosys, LLC 2004
+
+
+char *make_hash_name(struct file_struct *file)
+{
-+ char hash[33], *dst;
++ char hash[4*2 + 1 + 12*2 + 1], *dst;
+ uchar c, *src = (uchar*)F_SUM(file);
+ int i;
+
+ c = *src & 0x0f;
+ *(dst++) = (c >= 10) ? (c - 10 + 'a') : (c + '0');
+ }
-+ *dst = 0;
++ *dst = '\0';
++
++ if (asprintf(&dst, "%s/%s", link_by_hash_dir, hash) < 0)
++ out_of_memory("make_hash_name");
+
-+ asprintf(&dst,"%s/%s",link_by_hash_dir,hash);
+ return dst;
+}
+
+ *fnbr = this_fnbr;
+
+ hashfile = new_array(struct hashfile_struct, 1);
-+ asprintf(&hashfile->name,"%s/%s",hashname,
-+ di->d_name);
++ if (asprintf(&hashfile->name,"%s/%s",hashname, di->d_name) < 0)
++ out_of_memory("find_hashfiles");
+ if (do_stat(hashfile->name,&st) == -1) {
+ rsyserr(FERROR, errno, "stat failed: %s", hashfile->name);
+ kill_hashfile(hashfile);
+ }
+
+ first = 1;
-+ asprintf(&linkname,"%s/0",hashname);
++ if (asprintf(&linkname,"%s/0",hashname) < 0)
++ out_of_memory("link_by_hash");
+ rprintf(FINFO, "(1) linkname = %s\n", linkname);
+ } else {
+ struct hashfile_struct *hashfiles, *hashfile;
+
+ if (hashfiles == NULL) {
+ first = 1;
-+ asprintf(&linkname,"%s/0",hashname);
++ if (asprintf(&linkname,"%s/0",hashname) < 0)
++ out_of_memory("link_by_hash");
+ rprintf(FINFO, "(2) linkname = %s\n", linkname);
+ } else {
+ int fd;
+ kill_hashfile(hashfile);
+ } else {
+ first = 1;
-+ asprintf(&linkname, "%s/%ld", hashname,
-+ last_fnbr + 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 (errno == EMLINK) {
+ first = 1;
+ free(linkname);
-+ asprintf(&linkname,"%s/%ld",hashname,
-+ last_fnbr + 1);
++ 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);
+ } else {
diff --git a/options.c b/options.c
--- a/options.c
+++ b/options.c
-@@ -156,6 +156,7 @@ char *backup_suffix = NULL;
+@@ -155,6 +155,7 @@ char *backup_suffix = NULL;
char *tmpdir = NULL;
char *partial_dir = NULL;
char *basis_dir[MAX_BASIS_DIRS+1];
OPT_SERVER, OPT_REFUSED_BASE = 9000};
static struct poptOption long_options[] = {
-@@ -570,6 +572,7 @@ static struct poptOption long_options[] = {
+@@ -577,6 +579,7 @@ 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 },
+ {"link-by-hash", 0, POPT_ARG_STRING, 0, OPT_LINK_BY_HASH, 0, 0},
- {"fuzzy", 'y', POPT_ARG_NONE, &fuzzy_basis, 0, 0, 0 },
- {"compress", 'z', POPT_ARG_NONE, 0, 'z', 0, 0 },
- {"no-compress", 0, POPT_ARG_VAL, &do_compression, 0, 0, 0 },
-@@ -1244,6 +1247,21 @@ int parse_arguments(int *argc_p, const char ***argv_p, int frommain)
+ {"fuzzy", 'y', POPT_ARG_VAL, &fuzzy_basis, 1, 0, 0 },
+ {"no-fuzzy", 0, POPT_ARG_VAL, &fuzzy_basis, 0, 0, 0 },
+ {"no-y", 0, POPT_ARG_VAL, &fuzzy_basis, 0, 0, 0 },
+@@ -1259,6 +1262,21 @@ int parse_arguments(int *argc_p, const char ***argv_p)
return 0;
#endif
+#ifdef HAVE_LINK
+ arg = poptGetOptArg(pc);
+ if (sanitize_paths)
-+ arg = sanitize_path(NULL, arg, NULL, 0);
++ arg = sanitize_path(NULL, arg, NULL, 0, SP_DEFAULT);
+ link_by_hash_dir = (char *)arg;
+ break;
+#else
default:
/* A large opt value means that set_refuse_options()
* turned this option off. */
-@@ -1997,6 +2015,11 @@ void server_options(char **args, int *argc_p)
+@@ -2049,6 +2067,11 @@ void server_options(char **args, int *argc_p)
} else if (inplace)
args[ac++] = "--inplace";
diff --git a/receiver.c b/receiver.c
--- a/receiver.c
+++ b/receiver.c
-@@ -162,12 +162,14 @@ int open_tmpfile(char *fnametmp, const char *fname, struct file_struct *file)
+@@ -183,12 +183,14 @@ 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,
int32 len, sum_len;
OFF_T offset = 0;
OFF_T offset2;
-@@ -187,6 +189,9 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
+@@ -208,6 +210,9 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
} else
mapbuf = NULL;
sum_init(checksum_seed);
if (append_mode > 0) {
-@@ -231,6 +236,8 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
+@@ -252,6 +257,8 @@ 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 (fd != -1 && write_file(fd,data,i) != i)
goto report_write_error;
-@@ -257,6 +264,8 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
+@@ -279,6 +286,8 @@ 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 (updating_basis_or_equiv) {
-@@ -299,6 +308,8 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
+@@ -323,6 +332,8 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
}
sum_len = sum_end(file_sum1);
if (mapbuf)
unmap_file(mapbuf);
-@@ -314,7 +325,7 @@ static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r,
+@@ -338,7 +349,7 @@ 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)
{
}
static void handle_delayed_updates(char *local_name)
-@@ -676,7 +687,7 @@ int recv_files(int f_in, char *local_name)
+@@ -740,7 +751,7 @@ int recv_files(int f_in, char *local_name)
/* recv file data */
recv_ok = receive_data(f_in, fnamecmp, fd1, st.st_size,
diff --git a/rsync.c b/rsync.c
--- a/rsync.c
+++ b/rsync.c
-@@ -49,6 +49,7 @@ extern int inplace;
+@@ -47,6 +47,7 @@ extern int inplace;
extern int flist_eof;
extern int keep_dirlinks;
extern int make_backups;
extern struct file_list *cur_flist, *first_flist, *dir_flist;
extern struct chmod_mode_struct *daemon_chmod_modes;
#ifdef ICONV_OPTION
-@@ -536,8 +537,15 @@ int finish_transfer(const char *fname, const char *fnametmp,
+@@ -583,8 +584,15 @@ int finish_transfer(const char *fname, const char *fnametmp,
/* move tmp file over real file */
if (verbose > 2)
rprintf(FINFO, "renaming %s to %s\n", fnametmp, fname);
diff --git a/rsync.h b/rsync.h
--- a/rsync.h
+++ b/rsync.h
-@@ -818,6 +818,14 @@ struct stats {
+@@ -850,6 +850,14 @@ struct stats {
int num_transferred_files;
};
+
struct chmod_mode_struct;
- #define EMPTY_ITEM_LIST {NULL, 0, 0}
+ struct flist_ndx_item {
diff --git a/rsync.yo b/rsync.yo
--- a/rsync.yo
+++ b/rsync.yo
-@@ -388,6 +388,7 @@ to the detailed description below for a complete description. verb(
+@@ -405,6 +405,7 @@ to the detailed description below for a complete description. verb(
--compare-dest=DIR also compare received files relative to DIR
--copy-dest=DIR ... and include copies of unchanged files
--link-dest=DIR hardlink to files in DIR when unchanged