X-Git-Url: https://mattmccutchen.net/rsync/rsync-patches.git/blobdiff_plain/f6c3b30060864f00f4df7078fe5ae3bc10eb1caa..cad12f627670106269929565f075db42ca2e06c0:/link-by-hash.diff diff --git a/link-by-hash.diff b/link-by-hash.diff index 66d442c..49bcf6b 100644 --- a/link-by-hash.diff +++ b/link-by-hash.diff @@ -11,7 +11,7 @@ will only store one copy of the unique contents of each file, regardless of the file's name. ---- orig/Makefile.in 2004-07-04 08:59:17 +--- orig/Makefile.in 2004-08-13 07:18:58 +++ Makefile.in 2004-07-03 20:20:15 @@ -35,7 +35,7 @@ OBJS1=rsync.o generator.o receiver.o cle main.o checksum.o match.o syscall.o log.o backup.o @@ -22,9 +22,9 @@ the file's name. DAEMON_OBJ = params.o loadparm.o clientserver.o access.o connection.o authenticate.o popt_OBJS=popt/findme.o popt/popt.o popt/poptconfig.o \ popt/popthelp.o popt/poptparse.o ---- orig/hashlink.c 2004-07-02 21:41:54 -+++ hashlink.c 2004-07-02 21:41:54 -@@ -0,0 +1,342 @@ +--- orig/hashlink.c 2004-09-24 16:44:25 ++++ hashlink.c 2004-09-24 16:44:25 +@@ -0,0 +1,340 @@ +/* + Copyright (C) Cronosys, LLC 2004 + @@ -111,11 +111,11 @@ the file's name. + long this_fnbr; + + *fnbr = 0; -+ ++ + /* Build a list of potential candidates and open + * them. */ + if ((d = opendir(hashname)) == NULL) { -+ rsyserr(FERROR, errno, "opendir \"%s\"", hashname); ++ rsyserr(FERROR, errno, "opendir failed: \"%s\"", hashname); + free(hashname); + return NULL; + } @@ -134,7 +134,7 @@ the file's name. + asprintf(&hashfile->name,"%s/%s",hashname, + di->d_name); + if (do_stat(hashfile->name,&st) == -1) { -+ rsyserr(FERROR, errno, "%s: %s", hashfile->name); ++ rsyserr(FERROR, errno, "stat failed: %s", hashfile->name); + kill_hashfile(hashfile); + continue; + } @@ -145,7 +145,7 @@ the file's name. + hashfile->nlink = st.st_nlink; + hashfile->fd = open(hashfile->name,O_RDONLY|O_BINARY); + if (hashfile->fd == -1) { -+ rsyserr(FERROR, errno, "%s", hashfile->name); ++ rsyserr(FERROR, errno, "open failed: %s", hashfile->name); + kill_hashfile(hashfile); + continue; + } @@ -210,11 +210,10 @@ the file's name. + /* There are no matches. */ + return NULL; + } -+ + } + + if (amt == -1) { -+ rsyserr(FERROR, errno, "%s"); ++ rsyserr(FERROR, errno, "read failed in compare_hashfiles()"); + kill_hashfiles(files); + return NULL; + } @@ -250,7 +249,7 @@ the file's name. +int link_by_hash(char *fnametmp,char *fname,struct file_struct *file) +{ + STRUCT_STAT st; -+ char *hashname = make_hash_name(file); ++ char *hashname = make_hash_name(file); + int first = 0, rc; + char *linkname; + long last_fnbr; @@ -266,7 +265,7 @@ the file's name. + dirname = strdup(hashname); + *strrchr(dirname,'/') = 0; + if (do_mkdir(dirname, 0755) == -1 && errno != EEXIST) { -+ rsyserr(FERROR, errno, "mkdir %s", dirname); ++ rsyserr(FERROR, errno, "mkdir failed: %s", dirname); + free(hashname); + free(dirname); + return robust_rename(fnametmp,fname,0644); @@ -274,7 +273,7 @@ the file's name. + free(dirname); + + if (do_mkdir(hashname, 0755) == -1 && errno != EEXIST) { -+ rsyserr(FERROR, errno, "mkdir %s", hashname); ++ rsyserr(FERROR, errno, "mkdir failed: %s", hashname); + free(hashname); + return robust_rename(fnametmp,fname,0644); + } @@ -282,13 +281,11 @@ the file's name. + first = 1; + asprintf(&linkname,"%s/0",hashname); + rprintf(FINFO, "(1) linkname = %s\n", linkname); -+ + } else { + struct hashfile_struct *hashfiles, *hashfile; -+ int fd; + + if (do_stat(fnametmp,&st) == -1) { -+ rsyserr(FERROR, errno, "%s", fname); ++ rsyserr(FERROR, errno, "stat failed: %s", fname); + return -1; + } + hashfiles = find_hashfiles(hashname, st.st_size, &last_fnbr); @@ -298,15 +295,16 @@ the file's name. + asprintf(&linkname,"%s/0",hashname); + rprintf(FINFO, "(2) linkname = %s\n", linkname); + } else { -+ ++ int fd; + /* Search for one identical to us. */ + if ((fd = open(fnametmp,O_RDONLY|O_BINARY)) == -1) { -+ rsyserr(FERROR, errno, "%s", fnametmp); ++ rsyserr(FERROR, errno, "open failed: %s", fnametmp); + kill_hashfiles(hashfiles); + return -1; + } + hashfile = compare_hashfiles(fd, hashfiles); + hashfiles = NULL; ++ close(fd); + + if (hashfile) { + first = 0; @@ -325,6 +323,7 @@ the file's name. + if (!first) { + rprintf(FINFO, "link-by-hash (existing): \"%s\" -> %s\n", + linkname, full_fname(fname)); ++ robust_unlink(fname); + rc = do_link(linkname, fname); + if (rc == -1) { + if (errno == EMLINK) { @@ -337,7 +336,6 @@ the file's name. + } else { + rsyserr(FERROR, errno, "link \"%s\" -> \"%s\"", + linkname, full_fname(fname)); -+ robust_unlink(fname); + rc = robust_rename(fnametmp,fname,0644); + } + } else { @@ -367,9 +365,9 @@ the file's name. +} + +#endif ---- orig/options.c 2004-07-16 20:07:22 -+++ options.c 2004-07-03 20:20:15 -@@ -125,6 +125,7 @@ char *log_format = NULL; +--- orig/options.c 2004-09-23 17:42:07 ++++ options.c 2004-08-13 18:13:18 +@@ -126,6 +126,7 @@ char *log_format = NULL; char *password_file = NULL; char *rsync_path = RSYNC_PATH; char *backup_dir = NULL; @@ -377,15 +375,15 @@ the file's name. char backup_dir_buf[MAXPATHLEN]; int rsync_port = RSYNC_PORT; int link_dest = 0; -@@ -276,6 +277,7 @@ void usage(enum logcode F) - rprintf(F," -T --temp-dir=DIR create temporary files in directory DIR\n"); +@@ -279,6 +280,7 @@ void usage(enum logcode F) + rprintf(F," -T, --temp-dir=DIR create temporary files in directory DIR\n"); rprintf(F," --compare-dest=DIR also compare destination files relative to DIR\n"); rprintf(F," --link-dest=DIR create hardlinks to DIR for unchanged files\n"); + rprintf(F," --link-by-hash=DIR create hardlinks by hash to DIR for regular files\n"); rprintf(F," -P equivalent to --partial --progress\n"); rprintf(F," -z, --compress compress file data\n"); rprintf(F," -C, --cvs-exclude auto ignore files in the same way CVS does\n"); -@@ -316,7 +318,7 @@ void usage(enum logcode F) +@@ -319,7 +321,7 @@ void usage(enum logcode F) enum {OPT_VERSION = 1000, OPT_SENDER, OPT_EXCLUDE, OPT_EXCLUDE_FROM, OPT_DELETE_AFTER, OPT_DELETE_EXCLUDED, OPT_LINK_DEST, OPT_INCLUDE, OPT_INCLUDE_FROM, OPT_MODIFY_WINDOW, @@ -394,7 +392,7 @@ the file's name. OPT_REFUSED_BASE = 9000}; static struct poptOption long_options[] = { -@@ -375,6 +377,7 @@ static struct poptOption long_options[] +@@ -378,6 +380,7 @@ static struct poptOption long_options[] {"temp-dir", 'T', POPT_ARG_STRING, &tmpdir, 0, 0, 0 }, {"compare-dest", 0, POPT_ARG_STRING, &compare_dest, 0, 0, 0 }, {"link-dest", 0, POPT_ARG_STRING, &compare_dest, OPT_LINK_DEST, 0, 0 }, @@ -402,14 +400,16 @@ the file's name. /* TODO: Should this take an optional int giving the compression level? */ {"compress", 'z', POPT_ARG_NONE, &do_compression, 0, 0, 0 }, {"daemon", 0, POPT_ARG_NONE, &daemon_opt, 0, 0, 0 }, -@@ -601,6 +604,19 @@ int parse_arguments(int *argc, const cha +@@ -616,6 +619,21 @@ int parse_arguments(int *argc, const cha return 0; #endif + case OPT_LINK_BY_HASH: +#if HAVE_LINK -+ link_by_hash_dir = (char *)poptGetOptArg(pc); -+ checksum_seed = FIXED_CHECKSUM_SEED; ++ arg = poptGetOptArg(pc); ++ if (sanitize_paths) ++ arg = sanitize_path(NULL, arg, NULL, 0); ++ link_by_hash_dir = (char *)arg; + break; +#else + snprintf(err_buf, sizeof err_buf, @@ -422,7 +422,7 @@ the file's name. default: /* A large opt value means that set_refuse_options() * turned this option off (opt-BASE is its index). */ -@@ -1007,6 +1023,11 @@ void server_options(char **args,int *arg +@@ -1087,6 +1105,11 @@ void server_options(char **args,int *arg args[ac++] = compare_dest; } @@ -434,32 +434,34 @@ the file's name. if (files_from && (!am_sender || remote_filesfrom_file)) { if (remote_filesfrom_file) { args[ac++] = "--files-from"; ---- orig/receiver.c 2004-07-16 20:07:22 -+++ receiver.c 2004-07-16 20:16:37 -@@ -37,6 +37,7 @@ extern int cvs_exclude; - extern int io_error; +--- orig/receiver.c 2004-09-21 09:40:27 ++++ receiver.c 2004-07-20 21:44:05 +@@ -39,6 +39,7 @@ extern int io_error; extern char *tmpdir; + extern char *partial_dir; extern char *compare_dest; +extern char *link_by_hash_dir; extern int make_backups; extern int do_progress; extern char *backup_dir; -@@ -197,10 +198,11 @@ static int get_tmpname(char *fnametmp, c +@@ -202,12 +203,13 @@ static int get_tmpname(char *fnametmp, c - static int receive_data(int f_in,struct map_struct *mapbuf,int fd,char *fname, -- OFF_T total_size) -+ OFF_T total_size,char *md4) + static int receive_data(int f_in, char *fname_r, int fd_r, OFF_T size_r, +- char *fname, int fd, OFF_T total_size) ++ char *fname, int fd, OFF_T total_size, char *md4) { - int i; + static char file_sum1[MD4_SUM_LENGTH]; + static char file_sum2[MD4_SUM_LENGTH]; + struct map_struct *mapbuf; struct sum_struct sum; + struct mdfour mdfour_data; unsigned int len; OFF_T offset = 0; OFF_T offset2; -@@ -211,6 +213,9 @@ static int receive_data(int f_in,struct - - read_sum_head(f_in, &sum); +@@ -227,6 +229,9 @@ static int receive_data(int f_in, char * + } else + mapbuf = NULL; + if (md4) + mdfour_begin(&mdfour_data); @@ -467,16 +469,16 @@ the file's name. sum_init(checksum_seed); while ((i = recv_token(f_in, &data)) != 0) { -@@ -227,6 +232,8 @@ static int receive_data(int f_in,struct +@@ -243,6 +248,8 @@ static int receive_data(int f_in, char * cleanup_got_literal = 1; sum_update(data,i); + if (md4) + mdfour_update(&mdfour_data,data,i); - if (fd != -1 && write_file(fd,data,i) != i) { - rsyserr(FERROR, errno, "write failed on %s", -@@ -254,6 +261,8 @@ static int receive_data(int f_in,struct + if (fd != -1 && write_file(fd,data,i) != i) + goto report_write_error; +@@ -267,6 +274,8 @@ static int receive_data(int f_in, char * see_token(map, len); sum_update(map,len); @@ -484,91 +486,66 @@ the file's name. + mdfour_update(&mdfour_data,map,len); } - if (!inplace || offset != offset2) { -@@ -290,6 +299,8 @@ static int receive_data(int f_in,struct + if (inplace) { +@@ -306,6 +315,8 @@ static int receive_data(int f_in, char * } sum_end(file_sum1); + if (md4) + mdfour_result(&mdfour_data, (unsigned char*)md4); - read_buf(f_in,file_sum2,MD4_SUM_LENGTH); - if (verbose > 2) -@@ -389,7 +400,7 @@ int recv_files(int f_in, struct file_lis - if (fd1 != -1 && do_fstat(fd1,&st) != 0) { - rsyserr(FERROR, errno, "fstat %s failed", - full_fname(fnamecmp)); -- receive_data(f_in,NULL,-1,NULL,file->length); -+ receive_data(f_in,NULL,-1,NULL,file->length,NULL); - close(fd1); - continue; - } -@@ -402,7 +413,7 @@ int recv_files(int f_in, struct file_lis - */ - rprintf(FERROR,"recv_files: %s is a directory\n", - full_fname(fnamecmp)); -- receive_data(f_in, NULL, -1, NULL, file->length); -+ receive_data(f_in,NULL,-1,NULL,file->length,NULL); - close(fd1); - continue; - } -@@ -435,7 +446,7 @@ int recv_files(int f_in, struct file_lis - if (fd2 == -1) { - rsyserr(FERROR, errno, "open %s failed", - full_fname(fnamecmp)); -- receive_data(f_in,mapbuf,-1,NULL,file->length); -+ receive_data(f_in,mapbuf,-1,NULL,file->length,NULL); - if (mapbuf) - unmap_file(mapbuf); - if (fd1 != -1) -@@ -472,7 +483,7 @@ int recv_files(int f_in, struct file_lis - if (fd2 == -1) { - rsyserr(FERROR, errno, "mkstemp %s failed", - full_fname(fnametmp)); -- receive_data(f_in,mapbuf,-1,NULL,file->length); -+ receive_data(f_in,mapbuf,-1,NULL,file->length,NULL); - if (mapbuf) - unmap_file(mapbuf); - if (fd1 != -1) -@@ -487,7 +498,11 @@ int recv_files(int f_in, struct file_lis - rprintf(FINFO, "%s\n", fname); + if (mapbuf) + unmap_file(mapbuf); +@@ -321,7 +332,7 @@ static int receive_data(int f_in, char * + + 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); + } + + +@@ -542,8 +553,12 @@ int recv_files(int f_in, struct file_lis + rprintf(FINFO, "%s\n", safe_fname(fname)); /* recv file data */ -- recv_ok = receive_data(f_in,mapbuf,fd2,fname,file->length); +#ifdef HAVE_LINK + if (link_by_hash_dir) + file->u.sum = (char*)malloc(MD4_SUM_LENGTH); +#endif -+ recv_ok = receive_data(f_in,mapbuf,fd2,fname,file->length,file->u.sum); + recv_ok = receive_data(f_in, fnamecmp, fd1, st.st_size, +- fname, fd2, file->length); ++ fname, fd2, file->length, file->u.sum); log_recv(file, &initial_stats); ---- orig/rsync.c 2004-07-16 20:07:23 -+++ rsync.c 2004-07-16 20:16:53 -@@ -33,6 +33,7 @@ extern int preserve_gid; - extern int force_delete; +--- orig/rsync.c 2004-09-07 21:45:30 ++++ rsync.c 2004-08-13 18:14:34 +@@ -34,6 +34,7 @@ extern int force_delete; extern int recurse; + extern int keep_dirlinks; extern int make_backups; +extern char *link_by_hash_dir; extern char *backup_dir; extern int inplace; -@@ -250,6 +251,12 @@ void finish_transfer(char *fname, char * +@@ -254,7 +255,12 @@ void finish_transfer(char *fname, char * /* move tmp file over real file */ if (verbose > 2) rprintf(FINFO, "renaming %s to %s\n", fnametmp, fname); +- ret = robust_rename(fnametmp, fname, file->mode & INITACCESSPERMS); +#ifdef HAVE_LINK + if (link_by_hash_dir) + ret = link_by_hash(fnametmp, fname, file); + else +#endif + ret = robust_rename(fnametmp, fname, file->mode & INITACCESSPERMS); - ret = robust_rename(fnametmp, fname, file->mode & INITACCESSPERMS); if (ret < 0) { rsyserr(FERROR, errno, "%s %s -> \"%s\"", ---- orig/rsync.h 2004-07-16 20:07:23 + ret == -2 ? "copy" : "rename", +--- orig/rsync.h 2004-09-22 08:47:31 +++ rsync.h 2004-07-03 20:20:15 -@@ -524,6 +524,14 @@ struct stats { +@@ -526,6 +526,14 @@ struct stats { int current_file_index; };