/* * Generate and receive file lists. * * Copyright (C) 1996 Andrew Tridgell * Copyright (C) 1996 Paul Mackerras * Copyright (C) 2001, 2002 Martin Pool * Copyright (C) 2002-2014 Wayne Davison * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 3 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License along * with this program; if not, visit the http://fsf.org website. */ #include "rsync.h" #include "ifuncs.h" #include "rounding.h" #include "inums.h" #include "io.h" extern int am_root; extern int am_server; extern int am_daemon; extern int am_sender; extern int am_generator; extern int inc_recurse; extern int always_checksum; extern int module_id; extern int ignore_errors; extern int numeric_ids; extern int recurse; extern int use_qsort; extern int xfer_dirs; extern int filesfrom_fd; extern int one_file_system; extern int copy_dirlinks; extern int preserve_uid; extern int preserve_gid; extern int preserve_acls; extern int preserve_xattrs; extern int preserve_links; extern int preserve_hard_links; extern int preserve_devices; extern int preserve_specials; extern int delete_during; extern int missing_args; extern int eol_nulls; extern int relative_paths; extern int implied_dirs; extern int ignore_perishable; extern int non_perishable_cnt; extern int prune_empty_dirs; extern int copy_links; extern int copy_unsafe_links; extern int protocol_version; extern int sanitize_paths; extern int munge_symlinks; extern int use_safe_inc_flist; extern int need_unsorted_flist; extern int sender_symlink_iconv; extern int output_needs_newline; extern int sender_keeps_checksum; extern int unsort_ndx; extern uid_t our_uid; extern struct stats stats; extern char *filesfrom_host; extern char *usermap, *groupmap; extern char curr_dir[MAXPATHLEN]; extern struct chmod_mode_struct *chmod_modes; extern filter_rule_list filter_list; extern filter_rule_list daemon_filter_list; #ifdef ICONV_OPTION extern int filesfrom_convert; extern iconv_t ic_send, ic_recv; #endif #define PTR_SIZE (sizeof (struct file_struct *)) int io_error; int checksum_len; dev_t filesystem_dev; /* used to implement -x */ struct file_list *cur_flist, *first_flist, *dir_flist; int send_dir_ndx = -1, send_dir_depth = -1; int flist_cnt = 0; /* how many (non-tmp) file list objects exist */ int file_total = 0; /* total of all active items over all file-lists */ int file_old_total = 0; /* total of active items that will soon be gone */ int flist_eof = 0; /* all the file-lists are now known */ #define NORMAL_NAME 0 #define SLASH_ENDING_NAME 1 #define DOTDIR_NAME 2 #define MISSING_NAME 3 /* Starting from protocol version 26, we always use 64-bit ino_t and dev_t * internally, even if this platform does not allow files to have 64-bit inums. * The only exception is if we're on a platform with no 64-bit type at all. * * Because we use read_longint() to get these off the wire, if you transfer * devices or (for protocols < 30) hardlinks with dev or inum > 2**32 to a * machine with no 64-bit types then you will get an overflow error. * * Note that if you transfer devices from a 64-bit-devt machine (say, Solaris) * to a 32-bit-devt machine (say, Linux-2.2/x86) then the device numbers will * be truncated. But it's a kind of silly thing to do anyhow. */ /* The tmp_* vars are used as a cache area by make_file() to store data * that the sender doesn't need to remember in its file list. The data * will survive just long enough to be used by send_file_entry(). */ static dev_t tmp_rdev; #ifdef SUPPORT_HARD_LINKS static int64 tmp_dev = -1, tmp_ino; #endif static char tmp_sum[MAX_DIGEST_LEN]; static char empty_sum[MAX_DIGEST_LEN]; static int flist_count_offset; /* for --delete --progress */ static void flist_sort_and_clean(struct file_list *flist, int strip_root); static void output_flist(struct file_list *flist); void init_flist(void) { if (DEBUG_GTE(FLIST, 4)) { rprintf(FINFO, "FILE_STRUCT_LEN=%d, EXTRA_LEN=%d\n", (int)FILE_STRUCT_LEN, (int)EXTRA_LEN); } checksum_len = protocol_version < 21 ? 2 : protocol_version < 30 ? MD4_DIGEST_LEN : MD5_DIGEST_LEN; } static int show_filelist_p(void) { return INFO_GTE(FLIST, 1) && xfer_dirs && !am_server && !inc_recurse; } static void start_filelist_progress(char *kind) { rprintf(FCLIENT, "%s ... ", kind); output_needs_newline = 1; rflush(FINFO); } static void emit_filelist_progress(int count) { rprintf(FCLIENT, " %d files...\r", count); } static void maybe_emit_filelist_progress(int count) { if (INFO_GTE(FLIST, 2) && show_filelist_p() && (count % 100) == 0) emit_filelist_progress(count); } static void finish_filelist_progress(const struct file_list *flist) { if (INFO_GTE(FLIST, 2)) { /* This overwrites the progress line */ rprintf(FINFO, "%d file%sto consider\n", flist->used, flist->used == 1 ? " " : "s "); } else { output_needs_newline = 0; rprintf(FINFO, "done\n"); } } void show_flist_stats(void) { /* Nothing yet */ } /* Stat either a symlink or its referent, depending on the settings of * copy_links, copy_unsafe_links, etc. Returns -1 on error, 0 on success. * * If path is the name of a symlink, then the linkbuf buffer (which must hold * MAXPATHLEN chars) will be set to the symlink's target string. * * The stat structure pointed to by stp will contain information about the * link or the referent as appropriate, if they exist. */ static int readlink_stat(const char *path, STRUCT_STAT *stp, char *linkbuf) { #ifdef SUPPORT_LINKS if (link_stat(path, stp, copy_dirlinks) < 0) return -1; if (S_ISLNK(stp->st_mode)) { int llen = do_readlink(path, linkbuf, MAXPATHLEN - 1); if (llen < 0) return -1; linkbuf[llen] = '\0'; if (copy_unsafe_links && unsafe_symlink(linkbuf, path)) { if (INFO_GTE(SYMSAFE, 1)) { rprintf(FINFO,"copying unsafe symlink \"%s\" -> \"%s\"\n", path, linkbuf); } return x_stat(path, stp, NULL); } if (munge_symlinks && am_sender && llen > SYMLINK_PREFIX_LEN && strncmp(linkbuf, SYMLINK_PREFIX, SYMLINK_PREFIX_LEN) == 0) { memmove(linkbuf, linkbuf + SYMLINK_PREFIX_LEN, llen - SYMLINK_PREFIX_LEN + 1); } } return 0; #else return x_stat(path, stp, NULL); #endif } int link_stat(const char *path, STRUCT_STAT *stp, int follow_dirlinks) { #ifdef SUPPORT_LINKS if (copy_links) return x_stat(path, stp, NULL); if (x_lstat(path, stp, NULL) < 0) return -1; if (follow_dirlinks && S_ISLNK(stp->st_mode)) { STRUCT_STAT st; if (x_stat(path, &st, NULL) == 0 && S_ISDIR(st.st_mode)) *stp = st; } return 0; #else return x_stat(path, stp, NULL); #endif } static inline int is_daemon_excluded(const char *fname, int is_dir) { if (daemon_filter_list.head && check_filter(&daemon_filter_list, FLOG, fname, is_dir) < 0) { errno = ENOENT; return 1; } return 0; } static inline int path_is_daemon_excluded(char *path, int ignore_filename) { if (daemon_filter_list.head) { char *slash = path; while ((slash = strchr(slash+1, '/')) != NULL) { int ret; *slash = '\0'; ret = check_filter(&daemon_filter_list, FLOG, path, 1); *slash = '/'; if (ret < 0) { errno = ENOENT; return 1; } } if (!ignore_filename && check_filter(&daemon_filter_list, FLOG, path, 1) < 0) { errno = ENOENT; return 1; } } return 0; } /* This function is used to check if a file should be included/excluded * from the list of files based on its name and type etc. The value of * filter_level is set to either SERVER_FILTERS or ALL_FILTERS. */ static int is_excluded(const char *fname, int is_dir, int filter_level) { #if 0 /* This currently never happens, so avoid a useless compare. */ if (filter_level == NO_FILTERS) return 0; #endif if (is_daemon_excluded(fname, is_dir)) return 1; if (filter_level != ALL_FILTERS) return 0; if (filter_list.head && check_filter(&filter_list, FINFO, fname, is_dir) < 0) return 1; return 0; } static void send_directory(int f, struct file_list *flist, char *fbuf, int len, int flags); static const char *pathname, *orig_dir; static int pathname_len; /* Make sure flist can hold at least flist->used + extra entries. */ static void flist_expand(struct file_list *flist, int extra) { struct file_struct **new_ptr; if (flist->used + extra <= flist->malloced) return; if (flist->malloced < FLIST_START) flist->malloced = FLIST_START; else if (flist->malloced >= FLIST_LINEAR) flist->malloced += FLIST_LINEAR; else flist->malloced *= 2; /* In case count jumped or we are starting the list * with a known size just set it. */ if (flist->malloced < flist->used + extra) flist->malloced = flist->used + extra; new_ptr = realloc_array(flist->files, struct file_struct *, flist->malloced); if (DEBUG_GTE(FLIST, 1) && flist->malloced != FLIST_START) { rprintf(FCLIENT, "[%s] expand file_list pointer array to %s bytes, did%s move\n", who_am_i(), big_num(sizeof flist->files[0] * flist->malloced), (new_ptr == flist->files) ? " not" : ""); } flist->files = new_ptr; if (!flist->files) out_of_memory("flist_expand"); } static void flist_done_allocating(struct file_list *flist) { void *ptr = pool_boundary(flist->file_pool, 8*1024); if (flist->pool_boundary == ptr) flist->pool_boundary = NULL; /* list didn't use any pool memory */ else flist->pool_boundary = ptr; } /* Call this with EITHER (1) "file, NULL, 0" to chdir() to the file's * F_PATHNAME(), or (2) "NULL, dir, dirlen" to chdir() to the supplied dir, * with dir == NULL taken to be the starting directory, and dirlen < 0 * indicating that strdup(dir) should be called and then the -dirlen length * value checked to ensure that it is not daemon-excluded. */ int change_pathname(struct file_struct *file, const char *dir, int dirlen) { if (dirlen < 0) { char *cpy = strdup(dir); if (*cpy != '/') change_dir(orig_dir, CD_SKIP_CHDIR); if (path_is_daemon_excluded(cpy, 0)) goto chdir_error; dir = cpy; dirlen = -dirlen; } else { if (file) { if (pathname == F_PATHNAME(file)) return 1; dir = F_PATHNAME(file); if (dir) dirlen = strlen(dir); } else if (pathname == dir) return 1; if (dir && *dir != '/') change_dir(orig_dir, CD_SKIP_CHDIR); } pathname = dir; pathname_len = dirlen; if (!dir) dir = orig_dir; if (!change_dir(dir, CD_NORMAL)) { chdir_error: io_error |= IOERR_GENERAL; rsyserr(FERROR_XFER, errno, "change_dir %s failed", full_fname(dir)); if (dir != orig_dir) change_dir(orig_dir, CD_NORMAL); pathname = NULL; pathname_len = 0; return 0; } return 1; } static void send_file_entry(int f, const char *fname, struct file_struct *file, #ifdef SUPPORT_LINKS const char *symlink_name, int symlink_len, #endif int ndx, int first_ndx) { static time_t modtime; static mode_t mode; #ifdef SUPPORT_HARD_LINKS static int64 dev; #endif static dev_t rdev; static uint32 rdev_major; static uid_t uid; static gid_t gid; static const char *user_name, *group_name; static char lastname[MAXPATHLEN]; #ifdef SUPPORT_HARD_LINKS int first_hlink_ndx = -1; #endif int l1, l2; int xflags; /* Initialize starting value of xflags and adjust counts. */ if (S_ISREG(file->mode)) xflags = 0; else if (S_ISDIR(file->mode)) { stats.num_dirs++; if (protocol_version >= 30) { if (file->flags & FLAG_CONTENT_DIR) xflags = file->flags & FLAG_TOP_DIR; else if (file->flags & FLAG_IMPLIED_DIR) xflags = XMIT_TOP_DIR | XMIT_NO_CONTENT_DIR; else xflags = XMIT_NO_CONTENT_DIR; } else xflags = file->flags & FLAG_TOP_DIR; /* FLAG_TOP_DIR == XMIT_TOP_DIR */ } else { if (S_ISLNK(file->mode)) stats.num_symlinks++; else if (IS_DEVICE(file->mode)) stats.num_devices++; else if (IS_SPECIAL(file->mode)) stats.num_specials++; xflags = 0; } if (file->mode == mode) xflags |= XMIT_SAME_MODE; else mode = file->mode; if (preserve_devices && IS_DEVICE(mode)) { if (protocol_version < 28) { if (tmp_rdev == rdev) xflags |= XMIT_SAME_RDEV_pre28; else rdev = tmp_rdev; } else { rdev = tmp_rdev; if ((uint32)major(rdev) == rdev_major) xflags |= XMIT_SAME_RDEV_MAJOR; else rdev_major = major(rdev); if (protocol_version < 30 && (uint32)minor(rdev) <= 0xFFu) xflags |= XMIT_RDEV_MINOR_8_pre30; } } else if (preserve_specials && IS_SPECIAL(mode) && protocol_version < 31) { /* Special files don't need an rdev number, so just make * the historical transmission of the value efficient. */ if (protocol_version < 28) xflags |= XMIT_SAME_RDEV_pre28; else { rdev = MAKEDEV(major(rdev), 0); xflags |= XMIT_SAME_RDEV_MAJOR; if (protocol_version < 30) xflags |= XMIT_RDEV_MINOR_8_pre30; } } else if (protocol_version < 28) rdev = MAKEDEV(0, 0); if (!preserve_uid || ((uid_t)F_OWNER(file) == uid && *lastname)) xflags |= XMIT_SAME_UID; else { uid = F_OWNER(file); if (!numeric_ids) { user_name = add_uid(uid); if (inc_recurse && user_name) xflags |= XMIT_USER_NAME_FOLLOWS; } } if (!preserve_gid || ((gid_t)F_GROUP(file) == gid && *lastname)) xflags |= XMIT_SAME_GID; else { gid = F_GROUP(file); if (!numeric_ids) { group_name = add_gid(gid); if (inc_recurse && group_name) xflags |= XMIT_GROUP_NAME_FOLLOWS; } } if (file->modtime == modtime) xflags |= XMIT_SAME_TIME; else modtime = file->modtime; if (NSEC_BUMP(file) && protocol_version >= 31) xflags |= XMIT_MOD_NSEC; #ifdef SUPPORT_HARD_LINKS if (tmp_dev != -1) { if (protocol_version >= 30) { struct ht_int64_node *np = idev_find(tmp_dev, tmp_ino); first_hlink_ndx = (int32)(long)np->data - 1; if (first_hlink_ndx < 0) { np->data = (void*)(long)(first_ndx + ndx + 1); xflags |= XMIT_HLINK_FIRST; } if (DEBUG_GTE(HLINK, 1)) { if (first_hlink_ndx >= 0) { rprintf(FINFO, "[%s] #%d hard-links #%d (%sabbrev)\n", who_am_i(), first_ndx + ndx, first_hlink_ndx, first_hlink_ndx >= first_ndx ? "" : "un"); } else if (DEBUG_GTE(HLINK, 3)) { rprintf(FINFO, "[%s] dev:inode for #%d is %s:%s\n", who_am_i(), first_ndx + ndx, big_num(tmp_dev), big_num(tmp_ino)); } } } else { if (tmp_dev == dev) { if (protocol_version >= 28) xflags |= XMIT_SAME_DEV_pre30; } else dev = tmp_dev; } xflags |= XMIT_HLINKED; } #endif for (l1 = 0; lastname[l1] && (fname[l1] == lastname[l1]) && (l1 < 255); l1++) {} l2 = strlen(fname+l1); if (l1 > 0) xflags |= XMIT_SAME_NAME; if (l2 > 255) xflags |= XMIT_LONG_NAME; /* We must make sure we don't send a zero flag byte or the * other end will terminate the flist transfer. Note that * the use of XMIT_TOP_DIR on a non-dir has no meaning, so * it's harmless way to add a bit to the first flag byte. */ if (protocol_version >= 28) { if (!xflags && !S_ISDIR(mode)) xflags |= XMIT_TOP_DIR; if ((xflags & 0xFF00) || !xflags) { xflags |= XMIT_EXTENDED_FLAGS; write_shortint(f, xflags); } else write_byte(f, xflags); } else { if (!(xflags & 0xFF)) xflags |= S_ISDIR(mode) ? XMIT_LONG_NAME : XMIT_TOP_DIR; write_byte(f, xflags); } if (xflags & XMIT_SAME_NAME) write_byte(f, l1); if (xflags & XMIT_LONG_NAME) write_varint30(f, l2); else write_byte(f, l2); write_buf(f, fname + l1, l2); #ifdef SUPPORT_HARD_LINKS if (first_hlink_ndx >= 0) { write_varint(f, first_hlink_ndx); if (first_hlink_ndx >= first_ndx) goto the_end; } #endif write_varlong30(f, F_LENGTH(file), 3); if (!(xflags & XMIT_SAME_TIME)) { if (protocol_version >= 30) write_varlong(f, modtime, 4); else write_int(f, modtime); } if (xflags & XMIT_MOD_NSEC) write_varint(f, F_MOD_NSEC(file)); if (!(xflags & XMIT_SAME_MODE)) write_int(f, to_wire_mode(mode)); if (preserve_uid && !(xflags & XMIT_SAME_UID)) { if (protocol_version < 30) write_int(f, uid); else { write_varint(f, uid); if (xflags & XMIT_USER_NAME_FOLLOWS) { int len = strlen(user_name); write_byte(f, len); write_buf(f, user_name, len); } } } if (preserve_gid && !(xflags & XMIT_SAME_GID)) { if (protocol_version < 30) write_int(f, gid); else { write_varint(f, gid); if (xflags & XMIT_GROUP_NAME_FOLLOWS) { int len = strlen(group_name); write_byte(f, len); write_buf(f, group_name, len); } } } if ((preserve_devices && IS_DEVICE(mode)) || (preserve_specials && IS_SPECIAL(mode) && protocol_version < 31)) { if (protocol_version < 28) { if (!(xflags & XMIT_SAME_RDEV_pre28)) write_int(f, (int)rdev); } else { if (!(xflags & XMIT_SAME_RDEV_MAJOR)) write_varint30(f, major(rdev)); if (protocol_version >= 30) write_varint(f, minor(rdev)); else if (xflags & XMIT_RDEV_MINOR_8_pre30) write_byte(f, minor(rdev)); else write_int(f, minor(rdev)); } } #ifdef SUPPORT_LINKS if (symlink_len) { write_varint30(f, symlink_len); write_buf(f, symlink_name, symlink_len); } #endif #ifdef SUPPORT_HARD_LINKS if (tmp_dev != -1 && protocol_version < 30) { /* Older protocols expect the dev number to be transmitted * 1-incremented so that it is never zero. */ if (protocol_version < 26) { /* 32-bit dev_t and ino_t */ write_int(f, (int32)(dev+1)); write_int(f, (int32)tmp_ino); } else { /* 64-bit dev_t and ino_t */ if (!(xflags & XMIT_SAME_DEV_pre30)) write_longint(f, dev+1); write_longint(f, tmp_ino); } } #endif if (always_checksum && (S_ISREG(mode) || protocol_version < 28)) { const char *sum; if (S_ISREG(mode)) sum = tmp_sum; else { /* Prior to 28, we sent a useless set of nulls. */ sum = empty_sum; } write_buf(f, sum, checksum_len); } #ifdef SUPPORT_HARD_LINKS the_end: #endif strlcpy(lastname, fname, MAXPATHLEN); if (S_ISREG(mode) || S_ISLNK(mode)) stats.total_size += F_LENGTH(file); } static struct file_struct *recv_file_entry(int f, struct file_list *flist, int xflags) { static int64 modtime; static mode_t mode; #ifdef SUPPORT_HARD_LINKS static int64 dev; #endif static dev_t rdev; static uint32 rdev_major; static uid_t uid; static gid_t gid; static uint16 gid_flags; static char lastname[MAXPATHLEN], *lastdir; static int lastdir_depth, lastdir_len = -1; static unsigned int del_hier_name_len = 0; static int in_del_hier = 0; char thisname[MAXPATHLEN]; unsigned int l1 = 0, l2 = 0; int alloc_len, basename_len, linkname_len; int extra_len = file_extra_cnt * EXTRA_LEN; int first_hlink_ndx = -1; int64 file_length; uint32 modtime_nsec; const char *basename; struct file_struct *file; alloc_pool_t *pool; char *bp; if (xflags & XMIT_SAME_NAME) l1 = read_byte(f); if (xflags & XMIT_LONG_NAME) l2 = read_varint30(f); else l2 = read_byte(f); if (l2 >= MAXPATHLEN - l1) { rprintf(FERROR, "overflow: xflags=0x%x l1=%d l2=%d lastname=%s [%s]\n", xflags, l1, l2, lastname, who_am_i()); overflow_exit("recv_file_entry"); } strlcpy(thisname, lastname, l1 + 1); read_sbuf(f, &thisname[l1], l2); thisname[l1 + l2] = 0; /* Abuse basename_len for a moment... */ basename_len = strlcpy(lastname, thisname, MAXPATHLEN); #ifdef ICONV_OPTION if (ic_recv != (iconv_t)-1) { xbuf outbuf, inbuf; INIT_CONST_XBUF(outbuf, thisname); INIT_XBUF(inbuf, lastname, basename_len, (size_t)-1); if (iconvbufs(ic_recv, &inbuf, &outbuf, ICB_INIT) < 0) { io_error |= IOERR_GENERAL; rprintf(FERROR_UTF8, "[%s] cannot convert filename: %s (%s)\n", who_am_i(), lastname, strerror(errno)); outbuf.len = 0; } thisname[outbuf.len] = '\0'; } #endif if (*thisname) clean_fname(thisname, 0); if (sanitize_paths) sanitize_path(thisname, thisname, "", 0, SP_DEFAULT); if ((basename = strrchr(thisname, '/')) != NULL) { int len = basename++ - thisname; if (len != lastdir_len || memcmp(thisname, lastdir, len) != 0) { lastdir = new_array(char, len + 1); memcpy(lastdir, thisname, len); lastdir[len] = '\0'; lastdir_len = len; lastdir_depth = count_dir_elements(lastdir); } } else basename = thisname; basename_len = strlen(basename) + 1; /* count the '\0' */ #ifdef SUPPORT_HARD_LINKS if (protocol_version >= 30 && BITS_SETnUNSET(xflags, XMIT_HLINKED, XMIT_HLINK_FIRST)) { first_hlink_ndx = read_varint(f); if (first_hlink_ndx < 0 || first_hlink_ndx >= flist->ndx_start + flist->used) { rprintf(FERROR, "hard-link reference out of range: %d (%d)\n", first_hlink_ndx, flist->ndx_start + flist->used); exit_cleanup(RERR_PROTOCOL); } if (DEBUG_GTE(HLINK, 1)) { rprintf(FINFO, "[%s] #%d hard-links #%d (%sabbrev)\n", who_am_i(), flist->used+flist->ndx_start, first_hlink_ndx, first_hlink_ndx >= flist->ndx_start ? "" : "un"); } if (first_hlink_ndx >= flist->ndx_start) { struct file_struct *first = flist->files[first_hlink_ndx - flist->ndx_start]; file_length = F_LENGTH(first); modtime = first->modtime; modtime_nsec = F_MOD_NSEC(first); mode = first->mode; if (preserve_uid) uid = F_OWNER(first); if (preserve_gid) gid = F_GROUP(first); if (preserve_devices && IS_DEVICE(mode)) { uint32 *devp = F_RDEV_P(first); rdev = MAKEDEV(DEV_MAJOR(devp), DEV_MINOR(devp)); extra_len += DEV_EXTRA_CNT * EXTRA_LEN; } if (preserve_links && S_ISLNK(mode)) linkname_len = strlen(F_SYMLINK(first)) + 1; else linkname_len = 0; goto create_object; } } #endif file_length = read_varlong30(f, 3); if (!(xflags & XMIT_SAME_TIME)) { if (protocol_version >= 30) { modtime = read_varlong(f, 4); #if SIZEOF_TIME_T < SIZEOF_INT64 if (!am_generator && (int64)(time_t)modtime != modtime) { rprintf(FERROR_XFER, "Time value of %s truncated on receiver.\n", lastname); } #endif } else modtime = read_int(f); } if (xflags & XMIT_MOD_NSEC) modtime_nsec = read_varint(f); else modtime_nsec = 0; if (!(xflags & XMIT_SAME_MODE)) mode = from_wire_mode(read_int(f)); if (chmod_modes && !S_ISLNK(mode) && mode) mode = tweak_mode(mode, chmod_modes); if (preserve_uid && !(xflags & XMIT_SAME_UID)) { if (protocol_version < 30) uid = (uid_t)read_int(f); else { uid = (uid_t)read_varint(f); if (xflags & XMIT_USER_NAME_FOLLOWS) uid = recv_user_name(f, uid); else if (inc_recurse && am_root && (!numeric_ids || usermap)) uid = match_uid(uid); } } if (preserve_gid && !(xflags & XMIT_SAME_GID)) { if (protocol_version < 30) gid = (gid_t)read_int(f); else { gid = (gid_t)read_varint(f); gid_flags = 0; if (xflags & XMIT_GROUP_NAME_FOLLOWS) gid = recv_group_name(f, gid, &gid_flags); else if (inc_recurse && (!am_root || !numeric_ids || groupmap)) gid = match_gid(gid, &gid_flags); } } if ((preserve_devices && IS_DEVICE(mode)) || (preserve_specials && IS_SPECIAL(mode) && protocol_version < 31)) { if (protocol_version < 28) { if (!(xflags & XMIT_SAME_RDEV_pre28)) rdev = (dev_t)read_int(f); } else { uint32 rdev_minor; if (!(xflags & XMIT_SAME_RDEV_MAJOR)) rdev_major = read_varint30(f); if (protocol_version >= 30) rdev_minor = read_varint(f); else if (xflags & XMIT_RDEV_MINOR_8_pre30) rdev_minor = read_byte(f); else rdev_minor = read_int(f); rdev = MAKEDEV(rdev_major, rdev_minor); } if (IS_DEVICE(mode)) extra_len += DEV_EXTRA_CNT * EXTRA_LEN; file_length = 0; } else if (protocol_version < 28) rdev = MAKEDEV(0, 0); #ifdef SUPPORT_LINKS if (preserve_links && S_ISLNK(mode)) { linkname_len = read_varint30(f) + 1; /* count the '\0' */ if (linkname_len <= 0 || linkname_len > MAXPATHLEN) { rprintf(FERROR, "overflow: linkname_len=%d\n", linkname_len - 1); overflow_exit("recv_file_entry"); } #ifdef ICONV_OPTION /* We don't know how much extra room we need to convert * the as-yet-unread symlink data, so let's hope that a * double-size buffer is plenty. */ if (sender_symlink_iconv) linkname_len *= 2; #endif if (munge_symlinks) linkname_len += SYMLINK_PREFIX_LEN; } else #endif linkname_len = 0; #ifdef SUPPORT_HARD_LINKS create_object: if (preserve_hard_links) { if (protocol_version < 28 && S_ISREG(mode)) xflags |= XMIT_HLINKED; if (xflags & XMIT_HLINKED) extra_len += (inc_recurse+1) * EXTRA_LEN; } #endif #ifdef SUPPORT_ACLS /* Directories need an extra int32 for the default ACL. */ if (preserve_acls && S_ISDIR(mode)) extra_len += EXTRA_LEN; #endif if (always_checksum && S_ISREG(mode)) extra_len += SUM_EXTRA_CNT * EXTRA_LEN; #if SIZEOF_INT64 >= 8 if (file_length > 0xFFFFFFFFu && S_ISREG(mode)) extra_len += EXTRA_LEN; #endif #ifdef HAVE_UTIMENSAT if (modtime_nsec) extra_len += EXTRA_LEN; #endif if (file_length < 0) { rprintf(FERROR, "Offset underflow: file-length is negative\n"); exit_cleanup(RERR_UNSUPPORTED); } if (inc_recurse && S_ISDIR(mode)) { if (one_file_system) { /* Room to save the dir's device for -x */ extra_len += DEV_EXTRA_CNT * EXTRA_LEN; } pool = dir_flist->file_pool; } else pool = flist->file_pool; #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 + basename_len + linkname_len; bp = pool_alloc(pool, alloc_len, "recv_file_entry"); memset(bp, 0, extra_len + FILE_STRUCT_LEN); bp += extra_len; file = (struct file_struct *)bp; bp += FILE_STRUCT_LEN; memcpy(bp, basename, basename_len); #ifdef SUPPORT_HARD_LINKS if (xflags & XMIT_HLINKED) file->flags |= FLAG_HLINKED; #endif file->modtime = (time_t)modtime; #ifdef HAVE_UTIMENSAT if (modtime_nsec) { file->flags |= FLAG_MOD_NSEC; OPT_EXTRA(file, 0)->unum = modtime_nsec; } #endif file->len32 = (uint32)file_length; #if SIZEOF_INT64 >= 8 if (file_length > 0xFFFFFFFFu && S_ISREG(mode)) { #if SIZEOF_CAPITAL_OFF_T < 8 rprintf(FERROR, "Offset overflow: attempted 64-bit file-length\n"); exit_cleanup(RERR_UNSUPPORTED); #else file->flags |= FLAG_LENGTH64; OPT_EXTRA(file, NSEC_BUMP(file))->unum = (uint32)(file_length >> 32); #endif } #endif file->mode = mode; if (preserve_uid) F_OWNER(file) = uid; if (preserve_gid) { F_GROUP(file) = gid; file->flags |= gid_flags; } if (unsort_ndx) F_NDX(file) = flist->used + flist->ndx_start; if (basename != thisname) { file->dirname = lastdir; F_DEPTH(file) = lastdir_depth + 1; } else F_DEPTH(file) = 1; if (S_ISDIR(mode)) { if (basename_len == 1+1 && *basename == '.') /* +1 for '\0' */ F_DEPTH(file)--; if (protocol_version >= 30) { if (!(xflags & XMIT_NO_CONTENT_DIR)) { if (xflags & XMIT_TOP_DIR) file->flags |= FLAG_TOP_DIR; file->flags |= FLAG_CONTENT_DIR; } else if (xflags & XMIT_TOP_DIR) file->flags |= FLAG_IMPLIED_DIR; } else if (xflags & XMIT_TOP_DIR) { in_del_hier = recurse; del_hier_name_len = F_DEPTH(file) == 0 ? 0 : l1 + l2; if (relative_paths && del_hier_name_len > 2 && lastname[del_hier_name_len-1] == '.' && lastname[del_hier_name_len-2] == '/') del_hier_name_len -= 2; file->flags |= FLAG_TOP_DIR | FLAG_CONTENT_DIR; } else if (in_del_hier) { if (!relative_paths || !del_hier_name_len || (l1 >= del_hier_name_len && lastname[del_hier_name_len] == '/')) file->flags |= FLAG_CONTENT_DIR; else in_del_hier = 0; } } if (preserve_devices && IS_DEVICE(mode)) { uint32 *devp = F_RDEV_P(file); DEV_MAJOR(devp) = major(rdev); DEV_MINOR(devp) = minor(rdev); } #ifdef SUPPORT_LINKS if (linkname_len) { bp += basename_len; if (first_hlink_ndx >= flist->ndx_start) { struct file_struct *first = flist->files[first_hlink_ndx - flist->ndx_start]; memcpy(bp, F_SYMLINK(first), linkname_len); } else { if (munge_symlinks) { strlcpy(bp, SYMLINK_PREFIX, linkname_len); bp += SYMLINK_PREFIX_LEN; linkname_len -= SYMLINK_PREFIX_LEN; } #ifdef ICONV_OPTION if (sender_symlink_iconv) { xbuf outbuf, inbuf; alloc_len = linkname_len; linkname_len /= 2; /* Read the symlink data into the end of our double-sized * buffer and then convert it into the right spot. */ INIT_XBUF(inbuf, bp + alloc_len - linkname_len, linkname_len - 1, (size_t)-1); read_sbuf(f, inbuf.buf, inbuf.len); INIT_XBUF(outbuf, bp, 0, alloc_len); if (iconvbufs(ic_recv, &inbuf, &outbuf, ICB_INIT) < 0) { io_error |= IOERR_GENERAL; rprintf(FERROR_XFER, "[%s] cannot convert symlink data for: %s (%s)\n", who_am_i(), full_fname(thisname), strerror(errno)); bp = (char*)file->basename; *bp++ = '\0'; outbuf.len = 0; } bp[outbuf.len] = '\0'; } else #endif read_sbuf(f, bp, linkname_len - 1); if (sanitize_paths && !munge_symlinks && *bp) sanitize_path(bp, bp, "", lastdir_depth, SP_DEFAULT); } } #endif #ifdef SUPPORT_HARD_LINKS if (preserve_hard_links && xflags & XMIT_HLINKED) { if (protocol_version >= 30) { if (xflags & XMIT_HLINK_FIRST) { F_HL_GNUM(file) = flist->ndx_start + flist->used; } else F_HL_GNUM(file) = first_hlink_ndx; } else { static int32 cnt = 0; struct ht_int64_node *np; int64 ino; int32 ndx; if (protocol_version < 26) { dev = read_int(f); ino = read_int(f); } else { if (!(xflags & XMIT_SAME_DEV_pre30)) dev = read_longint(f); ino = read_longint(f); } np = idev_find(dev, ino); ndx = (int32)(long)np->data - 1; if (ndx < 0) { ndx = cnt++; np->data = (void*)(long)cnt; } F_HL_GNUM(file) = ndx; } } #endif if (always_checksum && (S_ISREG(mode) || protocol_version < 28)) { if (S_ISREG(mode)) bp = F_SUM(file); else { /* Prior to 28, we get a useless set of nulls. */ bp = tmp_sum; } if (first_hlink_ndx >= flist->ndx_start) { struct file_struct *first = flist->files[first_hlink_ndx - flist->ndx_start]; memcpy(bp, F_SUM(first), checksum_len); } else read_buf(f, bp, checksum_len); } #ifdef SUPPORT_ACLS if (preserve_acls && !S_ISLNK(mode)) receive_acl(f, file); #endif #ifdef SUPPORT_XATTRS if (preserve_xattrs) receive_xattr(f, file); #endif if (S_ISREG(mode) || S_ISLNK(mode)) stats.total_size += file_length; return file; } /* Create a file_struct for a named file by reading its stat() information * and performing extensive checks against global options. * * Returns a pointer to the new file struct, or NULL if there was an error * or this file should be excluded. * * Note: Any error (here or in send_file_name) that results in the omission of * an existent source file from the file list should set * "io_error |= IOERR_GENERAL" to avoid deletion of the file from the * destination if --delete is on. */ 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; struct file_struct *file; 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) >= sizeof thisname) { io_error |= IOERR_GENERAL; rprintf(FERROR_XFER, "skipping overly long name: %s\n", fname); return NULL; } clean_fname(thisname, 0); if (sanitize_paths) sanitize_path(thisname, thisname, "", 0, SP_DEFAULT); if (stp && (S_ISDIR(stp->st_mode) || IS_MISSING_FILE(*stp))) { /* This is needed to handle a "symlink/." with a --relative * dir, or a request to delete a specific file. */ st = *stp; *linkname = '\0'; /* make IBM code checker happy */ } else if (readlink_stat(thisname, &st, linkname) != 0) { int save_errno = errno; /* See if file is excluded before reporting an error. */ if (filter_level != NO_FILTERS && (is_excluded(thisname, 0, filter_level) || is_excluded(thisname, 1, filter_level))) { if (ignore_perishable && save_errno != ENOENT) non_perishable_cnt++; return NULL; } if (save_errno == ENOENT) { #ifdef SUPPORT_LINKS /* When our options tell us to follow a symlink that * points nowhere, tell the user about the symlink * instead of giving a "vanished" message. We only * dereference a symlink if one of the --copy*links * options was specified, so there's no need for the * extra lstat() if one of these options isn't on. */ if ((copy_links || copy_unsafe_links || copy_dirlinks) && x_lstat(thisname, &st, NULL) == 0 && S_ISLNK(st.st_mode)) { io_error |= IOERR_GENERAL; rprintf(FERROR_XFER, "symlink has no referent: %s\n", full_fname(thisname)); } else #endif { enum logcode c = am_daemon && protocol_version < 28 ? FERROR : FWARNING; io_error |= IOERR_VANISHED; rprintf(c, "file has vanished: %s\n", full_fname(thisname)); } } else { io_error |= IOERR_GENERAL; rsyserr(FERROR_XFER, save_errno, "readlink_stat(%s) failed", full_fname(thisname)); } return NULL; } else if (IS_MISSING_FILE(st)) { io_error |= IOERR_GENERAL; rprintf(FINFO, "skipping file with bogus (zero) st_mode: %s\n", full_fname(thisname)); return NULL; } if (filter_level == NO_FILTERS) goto skip_filters; if (S_ISDIR(st.st_mode)) { if (!xfer_dirs) { rprintf(FINFO, "skipping directory %s\n", thisname); return NULL; } /* -x only affects dirs because we need to avoid recursing * into a mount-point directory, not to avoid copying a * symlinked file if -L (or similar) was specified. */ if (one_file_system && st.st_dev != filesystem_dev && BITS_SETnUNSET(flags, FLAG_CONTENT_DIR, FLAG_TOP_DIR)) { if (one_file_system > 1) { if (INFO_GTE(MOUNT, 1)) { rprintf(FINFO, "[%s] skipping mount-point dir %s\n", who_am_i(), thisname); } return NULL; } flags |= FLAG_MOUNT_DIR; flags &= ~FLAG_CONTENT_DIR; } } else flags &= ~FLAG_CONTENT_DIR; if (is_excluded(thisname, S_ISDIR(st.st_mode) != 0, filter_level)) { if (ignore_perishable) non_perishable_cnt++; return NULL; } if (lp_ignore_nonreadable(module_id)) { #ifdef SUPPORT_LINKS if (!S_ISLNK(st.st_mode)) #endif if (access(thisname, R_OK) != 0) return NULL; } skip_filters: /* Only divert a directory in the main transfer. */ if (flist) { if (flist->prev && S_ISDIR(st.st_mode) && flags & FLAG_DIVERT_DIRS) { /* Room for parent/sibling/next-child info. */ extra_len += DIRNODE_EXTRA_CNT * EXTRA_LEN; if (relative_paths) extra_len += PTR_EXTRA_CNT * EXTRA_LEN; pool = dir_flist->file_pool; } else pool = flist->file_pool; } else { #ifdef SUPPORT_ACLS /* Directories need an extra int32 for the default ACL. */ if (preserve_acls && S_ISDIR(st.st_mode)) extra_len += EXTRA_LEN; #endif pool = NULL; } if (DEBUG_GTE(FLIST, 2)) { rprintf(FINFO, "[%s] make_file(%s,*,%d)\n", who_am_i(), thisname, filter_level); } if ((basename = strrchr(thisname, '/')) != NULL) { int len = basename++ - thisname; if (len != lastdir_len || memcmp(thisname, lastdir, len) != 0) { lastdir = new_array(char, len + 1); memcpy(lastdir, thisname, len); lastdir[len] = '\0'; lastdir_len = len; } } else basename = thisname; basename_len = strlen(basename) + 1; /* count the '\0' */ #ifdef SUPPORT_LINKS linkname_len = S_ISLNK(st.st_mode) ? strlen(linkname) + 1 : 0; #else linkname_len = 0; #endif #ifdef ST_MTIME_NSEC if (st.ST_MTIME_NSEC && protocol_version >= 31) extra_len += EXTRA_LEN; #endif #if SIZEOF_CAPITAL_OFF_T >= 8 if (st.st_size > 0xFFFFFFFFu && S_ISREG(st.st_mode)) extra_len += EXTRA_LEN; #endif if (always_checksum && am_sender && S_ISREG(st.st_mode)) { file_checksum(thisname, tmp_sum, st.st_size); if (sender_keeps_checksum) extra_len += 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 + basename_len + linkname_len; if (pool) bp = pool_alloc(pool, alloc_len, "make_file"); else { if (!(bp = new_array(char, alloc_len))) out_of_memory("make_file"); } memset(bp, 0, extra_len + FILE_STRUCT_LEN); bp += extra_len; file = (struct file_struct *)bp; bp += FILE_STRUCT_LEN; memcpy(bp, basename, basename_len); #ifdef SUPPORT_HARD_LINKS if (preserve_hard_links && flist && flist->prev) { if (protocol_version >= 28 ? (!S_ISDIR(st.st_mode) && st.st_nlink > 1) : S_ISREG(st.st_mode)) { tmp_dev = (int64)st.st_dev; tmp_ino = (int64)st.st_ino; } else tmp_dev = -1; } #endif #ifdef HAVE_STRUCT_STAT_ST_RDEV if (IS_DEVICE(st.st_mode)) { tmp_rdev = st.st_rdev; st.st_size = 0; } else if (IS_SPECIAL(st.st_mode)) st.st_size = 0; #endif file->flags = flags; file->modtime = st.st_mtime; #ifdef ST_MTIME_NSEC if (st.ST_MTIME_NSEC && protocol_version >= 31) { file->flags |= FLAG_MOD_NSEC; OPT_EXTRA(file, 0)->unum = st.ST_MTIME_NSEC; } #endif file->len32 = (uint32)st.st_size; #if SIZEOF_CAPITAL_OFF_T >= 8 if (st.st_size > 0xFFFFFFFFu && S_ISREG(st.st_mode)) { file->flags |= FLAG_LENGTH64; OPT_EXTRA(file, NSEC_BUMP(file))->unum = (uint32)(st.st_size >> 32); } #endif file->mode = st.st_mode; if (preserve_uid) F_OWNER(file) = st.st_uid; if (preserve_gid) F_GROUP(file) = st.st_gid; if (am_generator && st.st_uid == our_uid) file->flags |= FLAG_OWNED_BY_US; if (basename != thisname) file->dirname = lastdir; #ifdef SUPPORT_LINKS if (linkname_len) memcpy(bp + basename_len, linkname, linkname_len); #endif if (am_sender) F_PATHNAME(file) = pathname; else if (!pool) F_DEPTH(file) = extra_len / EXTRA_LEN; if (basename_len == 0+1) { if (!pool) unmake_file(file); return NULL; } if (sender_keeps_checksum && S_ISREG(st.st_mode)) memcpy(F_SUM(file), tmp_sum, checksum_len); if (unsort_ndx) F_NDX(file) = stats.num_dirs; return file; } /* Only called for temporary file_struct entries created by make_file(). */ void unmake_file(struct file_struct *file) { free(REQ_EXTRA(file, F_DEPTH(file))); } static struct file_struct *send_file_name(int f, struct file_list *flist, const char *fname, STRUCT_STAT *stp, int flags, int filter_level) { struct file_struct *file; file = make_file(fname, flist, stp, flags, filter_level); if (!file) return NULL; if (chmod_modes && !S_ISLNK(file->mode) && file->mode) file->mode = tweak_mode(file->mode, chmod_modes); if (f >= 0) { char fbuf[MAXPATHLEN]; #ifdef SUPPORT_LINKS const char *symlink_name; int symlink_len; #ifdef ICONV_OPTION char symlink_buf[MAXPATHLEN]; #endif #endif #if defined SUPPORT_ACLS || defined SUPPORT_XATTRS stat_x sx; init_stat_x(&sx); #endif #ifdef SUPPORT_LINKS if (preserve_links && S_ISLNK(file->mode)) { symlink_name = F_SYMLINK(file); symlink_len = strlen(symlink_name); if (symlink_len == 0) { io_error |= IOERR_GENERAL; f_name(file, fbuf); rprintf(FERROR_XFER, "skipping symlink with 0-length value: %s\n", full_fname(fbuf)); return NULL; } } else { symlink_name = NULL; symlink_len = 0; } #endif #ifdef ICONV_OPTION if (ic_send != (iconv_t)-1) { xbuf outbuf, inbuf; INIT_CONST_XBUF(outbuf, fbuf); if (file->dirname) { INIT_XBUF_STRLEN(inbuf, (char*)file->dirname); outbuf.size -= 2; /* Reserve room for '/' & 1 more char. */ if (iconvbufs(ic_send, &inbuf, &outbuf, ICB_INIT) < 0) goto convert_error; outbuf.size += 2; fbuf[outbuf.len++] = '/'; } INIT_XBUF_STRLEN(inbuf, (char*)file->basename); if (iconvbufs(ic_send, &inbuf, &outbuf, ICB_INIT) < 0) { convert_error: io_error |= IOERR_GENERAL; rprintf(FERROR_XFER, "[%s] cannot convert filename: %s (%s)\n", who_am_i(), f_name(file, fbuf), strerror(errno)); return NULL; } fbuf[outbuf.len] = '\0'; #ifdef SUPPORT_LINKS if (symlink_len && sender_symlink_iconv) { INIT_XBUF(inbuf, (char*)symlink_name, symlink_len, (size_t)-1); INIT_CONST_XBUF(outbuf, symlink_buf); if (iconvbufs(ic_send, &inbuf, &outbuf, ICB_INIT) < 0) { io_error |= IOERR_GENERAL; f_name(file, fbuf); rprintf(FERROR_XFER, "[%s] cannot convert symlink data for: %s (%s)\n", who_am_i(), full_fname(fbuf), strerror(errno)); return NULL; } symlink_buf[outbuf.len] = '\0'; symlink_name = symlink_buf; symlink_len = outbuf.len; } #endif } else #endif f_name(file, fbuf); #ifdef SUPPORT_ACLS if (preserve_acls && !S_ISLNK(file->mode)) { sx.st.st_mode = file->mode; if (get_acl(fname, &sx) < 0) { io_error |= IOERR_GENERAL; return NULL; } } #endif #ifdef SUPPORT_XATTRS if (preserve_xattrs) { sx.st.st_mode = file->mode; if (get_xattr(fname, &sx) < 0) { io_error |= IOERR_GENERAL; return NULL; } } #endif send_file_entry(f, fbuf, file, #ifdef SUPPORT_LINKS symlink_name, symlink_len, #endif flist->used, flist->ndx_start); #ifdef SUPPORT_ACLS if (preserve_acls && !S_ISLNK(file->mode)) { send_acl(f, &sx); free_acl(&sx); } #endif #ifdef SUPPORT_XATTRS if (preserve_xattrs) { F_XATTR(file) = send_xattr(f, &sx); free_xattr(&sx); } #endif } maybe_emit_filelist_progress(flist->used + flist_count_offset); flist_expand(flist, 1); flist->files[flist->used++] = file; return file; } static void send_if_directory(int f, struct file_list *flist, struct file_struct *file, char *fbuf, unsigned int ol, int flags) { char is_dot_dir = fbuf[ol-1] == '.' && (ol == 1 || fbuf[ol-2] == '/'); if (S_ISDIR(file->mode) && !(file->flags & FLAG_MOUNT_DIR) && f_name(file, fbuf)) { void *save_filters; unsigned int len = strlen(fbuf); if (len > 1 && fbuf[len-1] == '/') fbuf[--len] = '\0'; save_filters = push_local_filters(fbuf, len); send_directory(f, flist, fbuf, len, flags); pop_local_filters(save_filters); fbuf[ol] = '\0'; if (is_dot_dir) fbuf[ol-1] = '.'; } } static int file_compare(const void *file1, const void *file2) { return f_name_cmp(*(struct file_struct **)file1, *(struct file_struct **)file2); } /* The guts of a merge-sort algorithm. This was derived from the glibc * version, but I (Wayne) changed the merge code to do less copying and * to require only half the amount of temporary memory. */ static void fsort_tmp(struct file_struct **fp, size_t num, struct file_struct **tmp) { struct file_struct **f1, **f2, **t; size_t n1, n2; n1 = num / 2; n2 = num - n1; f1 = fp; f2 = fp + n1; if (n1 > 1) fsort_tmp(f1, n1, tmp); if (n2 > 1) fsort_tmp(f2, n2, tmp); while (f_name_cmp(*f1, *f2) <= 0) { if (!--n1) return; f1++; } t = tmp; memcpy(t, f1, n1 * PTR_SIZE); *f1++ = *f2++, n2--; while (n1 > 0 && n2 > 0) { if (f_name_cmp(*t, *f2) <= 0) *f1++ = *t++, n1--; else *f1++ = *f2++, n2--; } if (n1 > 0) memcpy(f1, t, n1 * PTR_SIZE); } /* This file-struct sorting routine makes sure that any identical names in * the file list stay in the same order as they were in the original list. * This is particularly vital in inc_recurse mode where we expect a sort * on the flist to match the exact order of a sort on the dir_flist. */ static void fsort(struct file_struct **fp, size_t num) { if (num <= 1) return; if (use_qsort) qsort(fp, num, PTR_SIZE, file_compare); else { struct file_struct **tmp = new_array(struct file_struct *, (num+1) / 2); fsort_tmp(fp, num, tmp); free(tmp); } } /* We take an entire set of sibling dirs from the sorted flist and link them * into the tree, setting the appropriate parent/child/sibling pointers. */ static void add_dirs_to_tree(int parent_ndx, struct file_list *from_flist, int dir_cnt) { int i; int32 *dp = NULL; int32 *parent_dp = parent_ndx < 0 ? NULL : F_DIR_NODE_P(dir_flist->sorted[parent_ndx]); flist_expand(dir_flist, dir_cnt); dir_flist->sorted = dir_flist->files; for (i = 0; dir_cnt; i++) { struct file_struct *file = from_flist->sorted[i]; if (!S_ISDIR(file->mode)) continue; dir_flist->files[dir_flist->used++] = file; dir_cnt--; if (file->basename[0] == '.' && file->basename[1] == '\0') continue; if (dp) DIR_NEXT_SIBLING(dp) = dir_flist->used - 1; else if (parent_dp) DIR_FIRST_CHILD(parent_dp) = dir_flist->used - 1; else send_dir_ndx = dir_flist->used - 1; dp = F_DIR_NODE_P(file); DIR_PARENT(dp) = parent_ndx; DIR_FIRST_CHILD(dp) = -1; } if (dp) DIR_NEXT_SIBLING(dp) = -1; } static void interpret_stat_error(const char *fname, int is_dir) { if (errno == ENOENT) { io_error |= IOERR_VANISHED; rprintf(FWARNING, "%s has vanished: %s\n", is_dir ? "directory" : "file", full_fname(fname)); } else { io_error |= IOERR_GENERAL; rsyserr(FERROR_XFER, errno, "link_stat %s failed", full_fname(fname)); } } /* This function is normally called by the sender, but the receiving side also * calls it from get_dirlist() with f set to -1 so that we just construct the * file list in memory without sending it over the wire. Also, get_dirlist() * might call this with f set to -2, which also indicates that local filter * rules should be ignored. */ static void send_directory(int f, struct file_list *flist, char *fbuf, int len, int flags) { struct dirent *di; unsigned remainder; char *p; DIR *d; int divert_dirs = (flags & FLAG_DIVERT_DIRS) != 0; int start = flist->used; int filter_level = f == -2 ? SERVER_FILTERS : ALL_FILTERS; assert(flist != NULL); if (!(d = opendir(fbuf))) { if (errno == ENOENT) { if (am_sender) /* Can abuse this for vanished error w/ENOENT: */ interpret_stat_error(fbuf, True); return; } io_error |= IOERR_GENERAL; rsyserr(FERROR_XFER, errno, "opendir %s failed", full_fname(fbuf)); return; } p = fbuf + len; if (len == 1 && *fbuf == '/') remainder = MAXPATHLEN - 1; else if (len < MAXPATHLEN-1) { *p++ = '/'; *p = '\0'; remainder = MAXPATHLEN - (len + 1); } else remainder = 0; for (errno = 0, di = readdir(d); di; errno = 0, di = readdir(d)) { unsigned name_len; char *dname = d_name(di); if (dname[0] == '.' && (dname[1] == '\0' || (dname[1] == '.' && dname[2] == '\0'))) continue; name_len = strlcpy(p, dname, remainder); if (name_len >= remainder) { char save = fbuf[len]; fbuf[len] = '\0'; io_error |= IOERR_GENERAL; rprintf(FERROR_XFER, "filename overflows max-path len by %u: %s/%s\n", name_len - remainder + 1, fbuf, dname); fbuf[len] = save; continue; } if (dname[0] == '\0') { io_error |= IOERR_GENERAL; rprintf(FERROR_XFER, "cannot send file with empty name in %s\n", full_fname(fbuf)); continue; } send_file_name(f, flist, fbuf, NULL, flags, filter_level); } fbuf[len] = '\0'; if (errno) { io_error |= IOERR_GENERAL; rsyserr(FERROR_XFER, errno, "readdir(%s)", full_fname(fbuf)); } closedir(d); if (f >= 0 && recurse && !divert_dirs) { int i, end = flist->used - 1; /* send_if_directory() bumps flist->used, so use "end". */ for (i = start; i <= end; i++) send_if_directory(f, flist, flist->files[i], fbuf, len, flags); } } static void send_implied_dirs(int f, struct file_list *flist, char *fname, char *start, char *limit, int flags, char name_type) { static char lastpath[MAXPATHLEN] = ""; static int lastpath_len = 0; static struct file_struct *lastpath_struct = NULL; struct file_struct *file; item_list *relname_list; relnamecache **rnpp; int len, need_new_dir, depth = 0; filter_rule_list save_filter_list = filter_list; flags = (flags | FLAG_IMPLIED_DIR) & ~(FLAG_TOP_DIR | FLAG_CONTENT_DIR); filter_list.head = filter_list.tail = NULL; /* Don't filter implied dirs. */ if (inc_recurse) { if (lastpath_struct && F_PATHNAME(lastpath_struct) == pathname && lastpath_len == limit - fname && strncmp(lastpath, fname, lastpath_len) == 0) need_new_dir = 0; else need_new_dir = 1; } else { char *tp = fname, *lp = lastpath; /* Skip any initial directories in our path that we * have in common with lastpath. */ assert(start == fname); for ( ; ; tp++, lp++) { if (tp == limit) { if (*lp == '/' || *lp == '\0') goto done; break; } if (*lp != *tp) break; if (*tp == '/') { start = tp; depth++; } } need_new_dir = 1; } if (need_new_dir) { int save_copy_links = copy_links; int save_xfer_dirs = xfer_dirs; char *slash; copy_links = xfer_dirs = 1; *limit = '\0'; for (slash = start; (slash = strchr(slash+1, '/')) != NULL; ) { *slash = '\0'; file = send_file_name(f, flist, fname, NULL, flags, ALL_FILTERS); depth++; if (!inc_recurse && file && S_ISDIR(file->mode)) change_local_filter_dir(fname, strlen(fname), depth); *slash = '/'; } file = send_file_name(f, flist, fname, NULL, flags, ALL_FILTERS); if (inc_recurse) { if (file && !S_ISDIR(file->mode)) file = NULL; lastpath_struct = file; } else if (file && S_ISDIR(file->mode)) change_local_filter_dir(fname, strlen(fname), ++depth); strlcpy(lastpath, fname, sizeof lastpath); lastpath_len = limit - fname; *limit = '/'; copy_links = save_copy_links; xfer_dirs = save_xfer_dirs; if (!inc_recurse) goto done; } if (!lastpath_struct) goto done; /* dir must have vanished */ len = strlen(limit+1); memcpy(&relname_list, F_DIR_RELNAMES_P(lastpath_struct), sizeof relname_list); if (!relname_list) { if (!(relname_list = new0(item_list))) out_of_memory("send_implied_dirs"); memcpy(F_DIR_RELNAMES_P(lastpath_struct), &relname_list, sizeof relname_list); } rnpp = EXPAND_ITEM_LIST(relname_list, relnamecache *, 32); if (!(*rnpp = (relnamecache*)new_array(char, sizeof (relnamecache) + len))) out_of_memory("send_implied_dirs"); (*rnpp)->name_type = name_type; strlcpy((*rnpp)->fname, limit+1, len + 1); done: filter_list = save_filter_list; } static NORETURN void fatal_unsafe_io_error(void) { /* This (sadly) can only happen when pushing data because * the sender does not know about what kind of delete * is in effect on the receiving side when pulling. */ rprintf(FERROR_XFER, "FATAL I/O ERROR: dying to avoid a --delete-%s issue with a pre-3.0.7 receiver.\n", delete_during == 2 ? "delay" : "during"); exit_cleanup(RERR_UNSUPPORTED); } static void send1extra(int f, struct file_struct *file, struct file_list *flist) { char fbuf[MAXPATHLEN]; item_list *relname_list; int len, dlen, flags = FLAG_DIVERT_DIRS | FLAG_CONTENT_DIR; size_t j; f_name(file, fbuf); dlen = strlen(fbuf); if (!change_pathname(file, NULL, 0)) exit_cleanup(RERR_FILESELECT); change_local_filter_dir(fbuf, dlen, send_dir_depth); if (file->flags & FLAG_CONTENT_DIR) { if (one_file_system) { STRUCT_STAT st; if (link_stat(fbuf, &st, copy_dirlinks) != 0) { interpret_stat_error(fbuf, True); return; } filesystem_dev = st.st_dev; } send_directory(f, flist, fbuf, dlen, flags); } if (!relative_paths) return; memcpy(&relname_list, F_DIR_RELNAMES_P(file), sizeof relname_list); if (!relname_list) return; for (j = 0; j < relname_list->count; j++) { char *slash; relnamecache *rnp = ((relnamecache**)relname_list->items)[j]; char name_type = rnp->name_type; fbuf[dlen] = '/'; len = strlcpy(fbuf + dlen + 1, rnp->fname, sizeof fbuf - dlen - 1); free(rnp); if (len >= (int)sizeof fbuf) continue; /* Impossible... */ slash = strchr(fbuf+dlen+1, '/'); if (slash) { send_implied_dirs(f, flist, fbuf, fbuf+dlen+1, slash, flags, name_type); continue; } if (name_type != NORMAL_NAME) { STRUCT_STAT st; if (name_type == MISSING_NAME) memset(&st, 0, sizeof st); else if (link_stat(fbuf, &st, 1) != 0) { interpret_stat_error(fbuf, True); continue; } send_file_name(f, flist, fbuf, &st, FLAG_TOP_DIR | flags, ALL_FILTERS); } else send_file_name(f, flist, fbuf, NULL, FLAG_TOP_DIR | flags, ALL_FILTERS); } free(relname_list); } void send_extra_file_list(int f, int at_least) { struct file_list *flist; int64 start_write; uint16 prev_flags; int save_io_error = io_error; if (flist_eof) return; if (at_least < 0) at_least = file_total - file_old_total + 1; /* Keep sending data until we have the requested number of * files in the upcoming file-lists. */ while (file_total - file_old_total < at_least) { struct file_struct *file = dir_flist->sorted[send_dir_ndx]; int dir_ndx, dstart = stats.num_dirs; const char *pathname = F_PATHNAME(file); int32 *dp; flist = flist_new(0, "send_extra_file_list"); start_write = stats.total_written; if (unsort_ndx) dir_ndx = F_NDX(file); else dir_ndx = send_dir_ndx; write_ndx(f, NDX_FLIST_OFFSET - dir_ndx); flist->parent_ndx = dir_ndx; send1extra(f, file, flist); prev_flags = file->flags; dp = F_DIR_NODE_P(file); /* If there are any duplicate directory names that follow, we * send all the dirs together in one file-list. The dir_flist * tree links all the child subdirs onto the last dup dir. */ while ((dir_ndx = DIR_NEXT_SIBLING(dp)) >= 0 && dir_flist->sorted[dir_ndx]->flags & FLAG_DUPLICATE) { send_dir_ndx = dir_ndx; file = dir_flist->sorted[dir_ndx]; /* Try to avoid some duplicate scanning of identical dirs. */ if (F_PATHNAME(file) == pathname && prev_flags & FLAG_CONTENT_DIR) file->flags &= ~FLAG_CONTENT_DIR; send1extra(f, file, flist); prev_flags = file->flags; dp = F_DIR_NODE_P(file); } if (io_error == save_io_error || ignore_errors) write_byte(f, 0); else if (use_safe_inc_flist) { write_shortint(f, XMIT_EXTENDED_FLAGS|XMIT_IO_ERROR_ENDLIST); write_varint(f, io_error); } else { if (delete_during) fatal_unsafe_io_error(); write_byte(f, 0); } if (need_unsorted_flist) { if (!(flist->sorted = new_array(struct file_struct *, flist->used))) out_of_memory("send_extra_file_list"); memcpy(flist->sorted, flist->files, flist->used * sizeof (struct file_struct*)); } else flist->sorted = flist->files; flist_sort_and_clean(flist, 0); add_dirs_to_tree(send_dir_ndx, flist, stats.num_dirs - dstart); flist_done_allocating(flist); file_total += flist->used; stats.flist_size += stats.total_written - start_write; stats.num_files += flist->used; if (DEBUG_GTE(FLIST, 3)) output_flist(flist); if (DIR_FIRST_CHILD(dp) >= 0) { send_dir_ndx = DIR_FIRST_CHILD(dp); send_dir_depth++; } else { while (DIR_NEXT_SIBLING(dp) < 0) { if ((send_dir_ndx = DIR_PARENT(dp)) < 0) { write_ndx(f, NDX_FLIST_EOF); flist_eof = 1; if (DEBUG_GTE(FLIST, 3)) rprintf(FINFO, "[%s] flist_eof=1\n", who_am_i()); change_local_filter_dir(NULL, 0, 0); goto finish; } send_dir_depth--; file = dir_flist->sorted[send_dir_ndx]; dp = F_DIR_NODE_P(file); } send_dir_ndx = DIR_NEXT_SIBLING(dp); } } finish: if (io_error != save_io_error && protocol_version == 30 && !ignore_errors) send_msg_int(MSG_IO_ERROR, io_error); } struct file_list *send_file_list(int f, int argc, char *argv[]) { static const char *lastdir; static int lastdir_len = -1; int len, dirlen; STRUCT_STAT st; char *p, *dir; struct file_list *flist; struct timeval start_tv, end_tv; int64 start_write; int use_ff_fd = 0; int disable_buffering, reenable_multiplex = -1; int flags = recurse ? FLAG_CONTENT_DIR : 0; int reading_remotely = filesfrom_host != NULL; int rl_flags = (reading_remotely ? 0 : RL_DUMP_COMMENTS) #ifdef ICONV_OPTION | (filesfrom_convert ? RL_CONVERT : 0) #endif | (eol_nulls || reading_remotely ? RL_EOL_NULLS : 0); int implied_dot_dir = 0; rprintf(FLOG, "building file list\n"); if (show_filelist_p()) start_filelist_progress("building file list"); else if (inc_recurse && INFO_GTE(FLIST, 1) && !am_server) rprintf(FCLIENT, "sending incremental file list\n"); start_write = stats.total_written; gettimeofday(&start_tv, NULL); if (relative_paths && protocol_version >= 30) implied_dirs = 1; /* We send flagged implied dirs */ #ifdef SUPPORT_HARD_LINKS if (preserve_hard_links && protocol_version >= 30 && !cur_flist) init_hard_links(); #endif flist = cur_flist = flist_new(0, "send_file_list"); if (inc_recurse) { dir_flist = flist_new(FLIST_TEMP, "send_file_list"); flags |= FLAG_DIVERT_DIRS; } else dir_flist = cur_flist; disable_buffering = io_start_buffering_out(f); if (filesfrom_fd >= 0) { if (argv[0] && !change_dir(argv[0], CD_NORMAL)) { rsyserr(FERROR_XFER, errno, "change_dir %s failed", full_fname(argv[0])); exit_cleanup(RERR_FILESELECT); } if (protocol_version < 31) { /* Older protocols send the files-from data w/o packaging * it in multiplexed I/O packets, so temporarily switch * to buffered I/O to match this behavior. */ reenable_multiplex = io_end_multiplex_in(MPLX_TO_BUFFERED); } use_ff_fd = 1; } if (!orig_dir) orig_dir = strdup(curr_dir); while (1) { char fbuf[MAXPATHLEN], *fn, name_type; if (use_ff_fd) { if (read_line(filesfrom_fd, fbuf, sizeof fbuf, rl_flags) == 0) break; sanitize_path(fbuf, fbuf, "", 0, SP_KEEP_DOT_DIRS); } else { if (argc-- == 0) break; strlcpy(fbuf, *argv++, MAXPATHLEN); if (sanitize_paths) sanitize_path(fbuf, fbuf, "", 0, SP_KEEP_DOT_DIRS); } len = strlen(fbuf); if (relative_paths) { /* We clean up fbuf below. */ name_type = NORMAL_NAME; } else if (!len || fbuf[len - 1] == '/') { if (len == 2 && fbuf[0] == '.') { /* Turn "./" into just "." rather than "./." */ fbuf[--len] = '\0'; } else { if (len + 1 >= MAXPATHLEN) overflow_exit("send_file_list"); fbuf[len++] = '.'; fbuf[len] = '\0'; } name_type = DOTDIR_NAME; } else if (len > 1 && fbuf[len-1] == '.' && fbuf[len-2] == '.' && (len == 2 || fbuf[len-3] == '/')) { if (len + 2 >= MAXPATHLEN) overflow_exit("send_file_list"); fbuf[len++] = '/'; fbuf[len++] = '.'; fbuf[len] = '\0'; name_type = DOTDIR_NAME; } else if (fbuf[len-1] == '.' && (len == 1 || fbuf[len-2] == '/')) name_type = DOTDIR_NAME; else name_type = NORMAL_NAME; dir = NULL; if (!relative_paths) { p = strrchr(fbuf, '/'); if (p) { *p = '\0'; if (p == fbuf) dir = "/"; else dir = fbuf; len -= p - fbuf + 1; fn = p + 1; } else fn = fbuf; } else { if ((p = strstr(fbuf, "/./")) != NULL) { *p = '\0'; if (p == fbuf) dir = "/"; else { dir = fbuf; clean_fname(dir, 0); } fn = p + 3; while (*fn == '/') fn++; if (!*fn) *--fn = '\0'; /* ensure room for '.' */ } else fn = fbuf; /* A leading ./ can be used in relative mode to affect * the dest dir without its name being in the path. */ if (*fn == '.' && fn[1] == '/' && fn[2] && !implied_dot_dir) implied_dot_dir = -1; len = clean_fname(fn, CFN_KEEP_TRAILING_SLASH | CFN_DROP_TRAILING_DOT_DIR); if (len == 1) { if (fn[0] == '/') { fn = "/."; len = 2; name_type = DOTDIR_NAME; } else if (fn[0] == '.') name_type = DOTDIR_NAME; } else if (fn[len-1] == '/') { fn[--len] = '\0'; if (len == 1 && *fn == '.') name_type = DOTDIR_NAME; else name_type = SLASH_ENDING_NAME; } /* Reject a ".." dir in the active part of the path. */ for (p = fn; (p = strstr(p, "..")) != NULL; p += 2) { if ((p[2] == '/' || p[2] == '\0') && (p == fn || p[-1] == '/')) { rprintf(FERROR, "found \"..\" dir in relative path: %s\n", fn); exit_cleanup(RERR_SYNTAX); } } } if (!*fn) { len = 1; fn = "."; name_type = DOTDIR_NAME; } dirlen = dir ? strlen(dir) : 0; if (dirlen != lastdir_len || memcmp(lastdir, dir, dirlen) != 0) { if (!change_pathname(NULL, dir, -dirlen)) goto bad_path; lastdir = pathname; lastdir_len = pathname_len; } else if (!change_pathname(NULL, lastdir, lastdir_len)) { bad_path: if (implied_dot_dir < 0) implied_dot_dir = 0; continue; } if (implied_dot_dir < 0) { implied_dot_dir = 1; send_file_name(f, flist, ".", NULL, (flags | FLAG_IMPLIED_DIR) & ~FLAG_CONTENT_DIR, ALL_FILTERS); } if (fn != fbuf) memmove(fbuf, fn, len + 1); if (link_stat(fbuf, &st, copy_dirlinks || name_type != NORMAL_NAME) != 0 || (name_type != DOTDIR_NAME && is_daemon_excluded(fbuf, S_ISDIR(st.st_mode))) || (relative_paths && path_is_daemon_excluded(fbuf, 1))) { if (errno != ENOENT || missing_args == 0) { /* This is a transfer error, but inhibit deletion * only if we might be omitting an existing file. */ if (errno != ENOENT) io_error |= IOERR_GENERAL; rsyserr(FERROR_XFER, errno, "link_stat %s failed", full_fname(fbuf)); continue; } else if (missing_args == 1) { /* Just ignore the arg. */ continue; } else /* (missing_args == 2) */ { /* Send the arg as a "missing" entry with * mode 0, which tells the generator to delete it. */ memset(&st, 0, sizeof st); } } /* A dot-dir should not be excluded! */ if (name_type != DOTDIR_NAME && st.st_mode != 0 && is_excluded(fbuf, S_ISDIR(st.st_mode) != 0, ALL_FILTERS)) continue; if (S_ISDIR(st.st_mode) && !xfer_dirs) { rprintf(FINFO, "skipping directory %s\n", fbuf); continue; } if (inc_recurse && relative_paths && *fbuf) { if ((p = strchr(fbuf+1, '/')) != NULL) { if (p - fbuf == 1 && *fbuf == '.') { if ((fn = strchr(p+1, '/')) != NULL) p = fn; } else fn = p; send_implied_dirs(f, flist, fbuf, fbuf, p, flags, IS_MISSING_FILE(st) ? MISSING_NAME : name_type); if (fn == p) continue; } } else if (implied_dirs && (p=strrchr(fbuf,'/')) && p != fbuf) { /* Send the implied directories at the start of the * source spec, so we get their permissions right. */ send_implied_dirs(f, flist, fbuf, fbuf, p, flags, 0); } if (one_file_system) filesystem_dev = st.st_dev; if (recurse || (xfer_dirs && name_type != NORMAL_NAME)) { struct file_struct *file; file = send_file_name(f, flist, fbuf, &st, FLAG_TOP_DIR | FLAG_CONTENT_DIR | flags, NO_FILTERS); if (!file) continue; if (inc_recurse) { if (name_type == DOTDIR_NAME) { if (send_dir_depth < 0) { send_dir_depth = 0; change_local_filter_dir(fbuf, len, send_dir_depth); } send_directory(f, flist, fbuf, len, flags); } } else send_if_directory(f, flist, file, fbuf, len, flags); } else send_file_name(f, flist, fbuf, &st, flags, NO_FILTERS); } if (reenable_multiplex >= 0) io_start_multiplex_in(reenable_multiplex); gettimeofday(&end_tv, NULL); stats.flist_buildtime = (int64)(end_tv.tv_sec - start_tv.tv_sec) * 1000 + (end_tv.tv_usec - start_tv.tv_usec) / 1000; if (stats.flist_buildtime == 0) stats.flist_buildtime = 1; start_tv = end_tv; /* Indicate end of file list */ if (io_error == 0 || ignore_errors) write_byte(f, 0); else if (use_safe_inc_flist) { write_shortint(f, XMIT_EXTENDED_FLAGS|XMIT_IO_ERROR_ENDLIST); write_varint(f, io_error); } else { if (delete_during && inc_recurse) fatal_unsafe_io_error(); write_byte(f, 0); } #ifdef SUPPORT_HARD_LINKS if (preserve_hard_links && protocol_version >= 30 && !inc_recurse) idev_destroy(); #endif if (show_filelist_p()) finish_filelist_progress(flist); gettimeofday(&end_tv, NULL); stats.flist_xfertime = (int64)(end_tv.tv_sec - start_tv.tv_sec) * 1000 + (end_tv.tv_usec - start_tv.tv_usec) / 1000; /* When converting names, both sides keep an unsorted file-list array * because the names will differ on the sending and receiving sides * (both sides will use the unsorted index number for each item). */ /* Sort the list without removing any duplicates. This allows the * receiving side to ask for whatever name it kept. For incremental * recursion mode, the sender marks duplicate dirs so that it can * send them together in a single file-list. */ if (need_unsorted_flist) { if (!(flist->sorted = new_array(struct file_struct *, flist->used))) out_of_memory("send_file_list"); memcpy(flist->sorted, flist->files, flist->used * sizeof (struct file_struct*)); } else flist->sorted = flist->files; flist_sort_and_clean(flist, 0); file_total += flist->used; file_old_total += flist->used; if (numeric_ids <= 0 && !inc_recurse) send_id_list(f); /* send the io_error flag */ if (protocol_version < 30) write_int(f, ignore_errors ? 0 : io_error); else if (!use_safe_inc_flist && io_error && !ignore_errors) send_msg_int(MSG_IO_ERROR, io_error); if (disable_buffering) io_end_buffering_out(IOBUF_FREE_BUFS); stats.flist_size = stats.total_written - start_write; stats.num_files = flist->used; if (DEBUG_GTE(FLIST, 3)) output_flist(flist); if (DEBUG_GTE(FLIST, 2)) rprintf(FINFO, "send_file_list done\n"); if (inc_recurse) { send_dir_depth = 1; add_dirs_to_tree(-1, flist, stats.num_dirs); if (!file_total || strcmp(flist->sorted[flist->low]->basename, ".") != 0) flist->parent_ndx = -1; flist_done_allocating(flist); if (send_dir_ndx < 0) { write_ndx(f, NDX_FLIST_EOF); flist_eof = 1; if (DEBUG_GTE(FLIST, 3)) rprintf(FINFO, "[%s] flist_eof=1\n", who_am_i()); } else if (file_total == 1) { /* If we're creating incremental file-lists and there * was just 1 item in the first file-list, send 1 more * file-list to check if this is a 1-file xfer. */ send_extra_file_list(f, 1); } } else { flist_eof = 1; if (DEBUG_GTE(FLIST, 3)) rprintf(FINFO, "[%s] flist_eof=1\n", who_am_i()); } return flist; } struct file_list *recv_file_list(int f) { struct file_list *flist; int dstart, flags; int64 start_read; if (!first_flist) { if (show_filelist_p()) start_filelist_progress("receiving file list"); else if (inc_recurse && INFO_GTE(FLIST, 1) && !am_server) rprintf(FCLIENT, "receiving incremental file list\n"); rprintf(FLOG, "receiving file list\n"); if (usermap) parse_name_map(usermap, True); if (groupmap) parse_name_map(groupmap, False); } start_read = stats.total_read; #ifdef SUPPORT_HARD_LINKS if (preserve_hard_links && !first_flist) init_hard_links(); #endif flist = flist_new(0, "recv_file_list"); if (inc_recurse) { if (flist->ndx_start == 1) dir_flist = flist_new(FLIST_TEMP, "recv_file_list"); dstart = dir_flist->used; } else { dir_flist = flist; dstart = 0; } while ((flags = read_byte(f)) != 0) { struct file_struct *file; if (protocol_version >= 28 && (flags & XMIT_EXTENDED_FLAGS)) flags |= read_byte(f) << 8; if (flags == (XMIT_EXTENDED_FLAGS|XMIT_IO_ERROR_ENDLIST)) { int err; if (!use_safe_inc_flist) { rprintf(FERROR, "Invalid flist flag: %x\n", flags); exit_cleanup(RERR_PROTOCOL); } err = read_varint(f); if (!ignore_errors) io_error |= err; break; } flist_expand(flist, 1); file = recv_file_entry(f, flist, flags); if (S_ISREG(file->mode)) { /* Already counted */ } else if (S_ISDIR(file->mode)) { if (inc_recurse) { flist_expand(dir_flist, 1); dir_flist->files[dir_flist->used++] = file; } stats.num_dirs++; } else if (S_ISLNK(file->mode)) stats.num_symlinks++; else if (IS_DEVICE(file->mode)) stats.num_symlinks++; else stats.num_specials++; flist->files[flist->used++] = file; maybe_emit_filelist_progress(flist->used); if (DEBUG_GTE(FLIST, 2)) { char *name = f_name(file, NULL); rprintf(FINFO, "recv_file_name(%s)\n", NS(name)); } } file_total += flist->used; if (DEBUG_GTE(FLIST, 2)) rprintf(FINFO, "received %d names\n", flist->used); if (show_filelist_p()) finish_filelist_progress(flist); if (need_unsorted_flist) { /* Create an extra array of index pointers that we can sort for * the generator's use (for wading through the files in sorted * order and for calling flist_find()). We keep the "files" * list unsorted for our exchange of index numbers with the * other side (since their names may not sort the same). */ if (!(flist->sorted = new_array(struct file_struct *, flist->used))) out_of_memory("recv_file_list"); memcpy(flist->sorted, flist->files, flist->used * sizeof (struct file_struct*)); if (inc_recurse && dir_flist->used > dstart) { static int dir_flist_malloced = 0; if (dir_flist_malloced < dir_flist->malloced) { dir_flist->sorted = realloc_array(dir_flist->sorted, struct file_struct *, dir_flist->malloced); dir_flist_malloced = dir_flist->malloced; } memcpy(dir_flist->sorted + dstart, dir_flist->files + dstart, (dir_flist->used - dstart) * sizeof (struct file_struct*)); fsort(dir_flist->sorted + dstart, dir_flist->used - dstart); } } else { flist->sorted = flist->files; if (inc_recurse && dir_flist->used > dstart) { dir_flist->sorted = dir_flist->files; fsort(dir_flist->sorted + dstart, dir_flist->used - dstart); } } if (inc_recurse) flist_done_allocating(flist); else if (f >= 0) { recv_id_list(f, flist); flist_eof = 1; if (DEBUG_GTE(FLIST, 3)) rprintf(FINFO, "[%s] flist_eof=1\n", who_am_i()); } flist_sort_and_clean(flist, relative_paths); if (protocol_version < 30) { /* Recv the io_error flag */ int err = read_int(f); if (!ignore_errors) io_error |= err; } else if (inc_recurse && flist->ndx_start == 1) { if (!file_total || strcmp(flist->sorted[flist->low]->basename, ".") != 0) flist->parent_ndx = -1; } if (DEBUG_GTE(FLIST, 3)) output_flist(flist); if (DEBUG_GTE(FLIST, 2)) rprintf(FINFO, "recv_file_list done\n"); stats.flist_size += stats.total_read - start_read; stats.num_files += flist->used; return flist; } /* This is only used once by the receiver if the very first file-list * has exactly one item in it. */ void recv_additional_file_list(int f) { struct file_list *flist; int ndx = read_ndx(f); if (ndx == NDX_FLIST_EOF) { flist_eof = 1; if (DEBUG_GTE(FLIST, 3)) rprintf(FINFO, "[%s] flist_eof=1\n", who_am_i()); change_local_filter_dir(NULL, 0, 0); } else { ndx = NDX_FLIST_OFFSET - ndx; if (ndx < 0 || ndx >= dir_flist->used) { ndx = NDX_FLIST_OFFSET - ndx; rprintf(FERROR, "[%s] Invalid dir index: %d (%d - %d)\n", who_am_i(), ndx, NDX_FLIST_OFFSET, NDX_FLIST_OFFSET - dir_flist->used + 1); exit_cleanup(RERR_PROTOCOL); } if (DEBUG_GTE(FLIST, 3)) { rprintf(FINFO, "[%s] receiving flist for dir %d\n", who_am_i(), ndx); } flist = recv_file_list(f); flist->parent_ndx = ndx; } } /* Search for an identically-named item in the file list. Note that the * items must agree in their directory-ness, or no match is returned. */ int flist_find(struct file_list *flist, struct file_struct *f) { int low = flist->low, high = flist->high; int diff, mid, mid_up; while (low <= high) { mid = (low + high) / 2; if (F_IS_ACTIVE(flist->sorted[mid])) mid_up = mid; else { /* Scan for the next non-empty entry using the cached * distance values. If the value isn't fully up-to- * date, update it. */ mid_up = mid + F_DEPTH(flist->sorted[mid]); if (!F_IS_ACTIVE(flist->sorted[mid_up])) { do { mid_up += F_DEPTH(flist->sorted[mid_up]); } while (!F_IS_ACTIVE(flist->sorted[mid_up])); F_DEPTH(flist->sorted[mid]) = mid_up - mid; } if (mid_up > high) { /* If there's nothing left above us, set high to * a non-empty entry below us and continue. */ high = mid - (int)flist->sorted[mid]->len32; if (!F_IS_ACTIVE(flist->sorted[high])) { do { high -= (int)flist->sorted[high]->len32; } while (!F_IS_ACTIVE(flist->sorted[high])); flist->sorted[mid]->len32 = mid - high; } continue; } } diff = f_name_cmp(flist->sorted[mid_up], f); if (diff == 0) { if (protocol_version < 29 && S_ISDIR(flist->sorted[mid_up]->mode) != S_ISDIR(f->mode)) return -1; return mid_up; } if (diff < 0) low = mid_up + 1; else high = mid - 1; } return -1; } /* Search for an identically-named item in the file list. Differs from * flist_find in that an item that agrees with "f" in directory-ness is * preferred but one that does not is still found. */ int flist_find_ignore_dirness(struct file_list *flist, struct file_struct *f) { mode_t save_mode; int ndx; /* First look for an item that agrees in directory-ness. */ ndx = flist_find(flist, f); if (ndx >= 0) return ndx; /* Temporarily flip f->mode to look for an item of opposite * directory-ness. */ save_mode = f->mode; f->mode = S_ISDIR(f->mode) ? S_IFREG : S_IFDIR; ndx = flist_find(flist, f); f->mode = save_mode; return ndx; } /* * Free up any resources a file_struct has allocated * and clear the file. */ void clear_file(struct file_struct *file) { /* The +1 zeros out the first char of the basename. */ memset(file, 0, FILE_STRUCT_LEN + 1); /* In an empty entry, F_DEPTH() is an offset to the next non-empty * entry. Likewise for len32 in the opposite direction. We assume * that we're alone for now since flist_find() will adjust the counts * it runs into that aren't up-to-date. */ file->len32 = F_DEPTH(file) = 1; } /* Allocate a new file list. */ struct file_list *flist_new(int flags, char *msg) { struct file_list *flist; if (!(flist = new0(struct file_list))) out_of_memory(msg); if (flags & FLIST_TEMP) { if (!(flist->file_pool = pool_create(SMALL_EXTENT, 0, out_of_memory, POOL_INTERN))) out_of_memory(msg); } else { /* This is a doubly linked list with prev looping back to * the end of the list, but the last next pointer is NULL. */ if (!first_flist) { flist->file_pool = pool_create(NORMAL_EXTENT, 0, out_of_memory, POOL_INTERN); if (!flist->file_pool) out_of_memory(msg); flist->ndx_start = flist->flist_num = inc_recurse ? 1 : 0; first_flist = cur_flist = flist->prev = flist; } else { struct file_list *prev = first_flist->prev; flist->file_pool = first_flist->file_pool; flist->ndx_start = prev->ndx_start + prev->used + 1; flist->flist_num = prev->flist_num + 1; flist->prev = prev; prev->next = first_flist->prev = flist; } flist->pool_boundary = pool_boundary(flist->file_pool, 0); flist_cnt++; } return flist; } /* Free up all elements in a flist. */ void flist_free(struct file_list *flist) { if (!flist->prev) { /* Was FLIST_TEMP dir-list. */ } else if (flist == flist->prev) { first_flist = cur_flist = NULL; file_total = 0; flist_cnt = 0; } else { if (flist == cur_flist) cur_flist = flist->next; if (flist == first_flist) first_flist = first_flist->next; else { flist->prev->next = flist->next; if (!flist->next) flist->next = first_flist; } flist->next->prev = flist->prev; file_total -= flist->used; flist_cnt--; } if (!flist->prev || !flist_cnt) pool_destroy(flist->file_pool); else pool_free_old(flist->file_pool, flist->pool_boundary); if (flist->sorted && flist->sorted != flist->files) free(flist->sorted); free(flist->files); free(flist); } /* This routine ensures we don't have any duplicate names in our file list. * duplicate names can cause corruption because of the pipelining. */ static void flist_sort_and_clean(struct file_list *flist, int strip_root) { char fbuf[MAXPATHLEN]; int i, prev_i; if (!flist) return; if (flist->used == 0) { flist->high = -1; flist->low = 0; return; } fsort(flist->sorted, flist->used); if (!am_sender || inc_recurse) { for (i = prev_i = 0; i < flist->used; i++) { if (F_IS_ACTIVE(flist->sorted[i])) { prev_i = i; break; } } flist->low = prev_i; } else { i = prev_i = flist->used - 1; flist->low = 0; } while (++i < flist->used) { int j; struct file_struct *file = flist->sorted[i]; if (!F_IS_ACTIVE(file)) continue; if (f_name_cmp(file, flist->sorted[prev_i]) == 0) j = prev_i; else if (protocol_version >= 29 && S_ISDIR(file->mode)) { int save_mode = file->mode; /* Make sure that this directory doesn't duplicate a * non-directory earlier in the list. */ flist->high = prev_i; file->mode = S_IFREG; j = flist_find(flist, file); file->mode = save_mode; } else j = -1; if (j >= 0) { int keep, drop; /* If one is a dir and the other is not, we want to * keep the dir because it might have contents in the * list. Otherwise keep the first one. */ if (S_ISDIR(file->mode)) { struct file_struct *fp = flist->sorted[j]; if (!S_ISDIR(fp->mode)) keep = i, drop = j; else { if (am_sender) file->flags |= FLAG_DUPLICATE; else { /* Make sure we merge our vital flags. */ fp->flags |= file->flags & (FLAG_TOP_DIR|FLAG_CONTENT_DIR); fp->flags &= file->flags | ~FLAG_IMPLIED_DIR; } keep = j, drop = i; } } else keep = j, drop = i; if (!am_sender) { if (DEBUG_GTE(DUP, 1)) { rprintf(FINFO, "removing duplicate name %s from file list (%d)\n", f_name(file, fbuf), drop + flist->ndx_start); } clear_file(flist->sorted[drop]); } if (keep == i) { if (flist->low == drop) { for (j = drop + 1; j < i && !F_IS_ACTIVE(flist->sorted[j]); j++) {} flist->low = j; } prev_i = i; } } else prev_i = i; } flist->high = prev_i; if (strip_root) { /* We need to strip off the leading slashes for relative * paths, but this must be done _after_ the sorting phase. */ for (i = flist->low; i <= flist->high; i++) { struct file_struct *file = flist->sorted[i]; if (!file->dirname) continue; while (*file->dirname == '/') file->dirname++; if (!*file->dirname) file->dirname = NULL; } } if (prune_empty_dirs && !am_sender) { int j, prev_depth = 0; prev_i = 0; /* It's OK that this isn't really true. */ for (i = flist->low; i <= flist->high; i++) { struct file_struct *fp, *file = flist->sorted[i]; /* This temporarily abuses the F_DEPTH() value for a * directory that is in a chain that might get pruned. * We restore the old value if it gets a reprieve. */ if (S_ISDIR(file->mode) && F_DEPTH(file)) { /* Dump empty dirs when coming back down. */ for (j = prev_depth; j >= F_DEPTH(file); j--) { fp = flist->sorted[prev_i]; if (F_DEPTH(fp) >= 0) break; prev_i = -F_DEPTH(fp)-1; clear_file(fp); } prev_depth = F_DEPTH(file); if (is_excluded(f_name(file, fbuf), 1, ALL_FILTERS)) { /* Keep dirs through this dir. */ for (j = prev_depth-1; ; j--) { fp = flist->sorted[prev_i]; if (F_DEPTH(fp) >= 0) break; prev_i = -F_DEPTH(fp)-1; F_DEPTH(fp) = j; } } else F_DEPTH(file) = -prev_i-1; prev_i = i; } else { /* Keep dirs through this non-dir. */ for (j = prev_depth; ; j--) { fp = flist->sorted[prev_i]; if (F_DEPTH(fp) >= 0) break; prev_i = -F_DEPTH(fp)-1; F_DEPTH(fp) = j; } } } /* Dump all remaining empty dirs. */ while (1) { struct file_struct *fp = flist->sorted[prev_i]; if (F_DEPTH(fp) >= 0) break; prev_i = -F_DEPTH(fp)-1; clear_file(fp); } for (i = flist->low; i <= flist->high; i++) { if (F_IS_ACTIVE(flist->sorted[i])) break; } flist->low = i; for (i = flist->high; i >= flist->low; i--) { if (F_IS_ACTIVE(flist->sorted[i])) break; } flist->high = i; } } static void output_flist(struct file_list *flist) { char uidbuf[16], gidbuf[16], depthbuf[16]; struct file_struct *file; const char *root, *dir, *slash, *name, *trail; const char *who = who_am_i(); int i; rprintf(FINFO, "[%s] flist start=%d, used=%d, low=%d, high=%d\n", who, flist->ndx_start, flist->used, flist->low, flist->high); for (i = 0; i < flist->used; i++) { file = flist->files[i]; if ((am_root || am_sender) && uid_ndx) { snprintf(uidbuf, sizeof uidbuf, " uid=%u", F_OWNER(file)); } else *uidbuf = '\0'; if (gid_ndx) { static char parens[] = "(\0)\0\0\0"; char *pp = parens + (file->flags & FLAG_SKIP_GROUP ? 0 : 3); snprintf(gidbuf, sizeof gidbuf, " gid=%s%u%s", pp, F_GROUP(file), pp + 2); } else *gidbuf = '\0'; if (!am_sender) snprintf(depthbuf, sizeof depthbuf, "%d", F_DEPTH(file)); if (F_IS_ACTIVE(file)) { root = am_sender ? NS(F_PATHNAME(file)) : depthbuf; if ((dir = file->dirname) == NULL) dir = slash = ""; else slash = "/"; name = file->basename; trail = S_ISDIR(file->mode) ? "/" : ""; } else root = dir = slash = name = trail = ""; rprintf(FINFO, "[%s] i=%d %s %s%s%s%s mode=0%o len=%s%s%s flags=%x\n", who, i + flist->ndx_start, root, dir, slash, name, trail, (int)file->mode, comma_num(F_LENGTH(file)), uidbuf, gidbuf, file->flags); } } enum fnc_state { s_DIR, s_SLASH, s_BASE, s_TRAILING }; enum fnc_type { t_PATH, t_ITEM }; static int found_prefix; /* Compare the names of two file_struct entities, similar to how strcmp() * would do if it were operating on the joined strings. * * Some differences beginning with protocol_version 29: (1) directory names * are compared with an assumed trailing slash so that they compare in a * way that would cause them to sort immediately prior to any content they * may have; (2) a directory of any name compares after a non-directory of * any name at the same depth; (3) a directory with name "." compares prior * to anything else. These changes mean that a directory and a non-dir * with the same name will not compare as equal (protocol_version >= 29). * * The dirname component can be an empty string, but the basename component * cannot (and never is in the current codebase). The basename component * may be NULL (for a removed item), in which case it is considered to be * after any existing item. */ int f_name_cmp(const struct file_struct *f1, const struct file_struct *f2) { int dif; const uchar *c1, *c2; enum fnc_state state1, state2; enum fnc_type type1, type2; enum fnc_type t_path = protocol_version >= 29 ? t_PATH : t_ITEM; if (!f1 || !F_IS_ACTIVE(f1)) { if (!f2 || !F_IS_ACTIVE(f2)) return 0; return -1; } if (!f2 || !F_IS_ACTIVE(f2)) return 1; c1 = (uchar*)f1->dirname; c2 = (uchar*)f2->dirname; if (c1 == c2) c1 = c2 = NULL; if (!c1) { type1 = S_ISDIR(f1->mode) ? t_path : t_ITEM; c1 = (const uchar*)f1->basename; if (type1 == t_PATH && *c1 == '.' && !c1[1]) { type1 = t_ITEM; state1 = s_TRAILING; c1 = (uchar*)""; } else state1 = s_BASE; } else { type1 = t_path; state1 = s_DIR; } if (!c2) { type2 = S_ISDIR(f2->mode) ? t_path : t_ITEM; c2 = (const uchar*)f2->basename; if (type2 == t_PATH && *c2 == '.' && !c2[1]) { type2 = t_ITEM; state2 = s_TRAILING; c2 = (uchar*)""; } else state2 = s_BASE; } else { type2 = t_path; state2 = s_DIR; } if (type1 != type2) return type1 == t_PATH ? 1 : -1; do { if (!*c1) { switch (state1) { case s_DIR: state1 = s_SLASH; c1 = (uchar*)"/"; break; case s_SLASH: type1 = S_ISDIR(f1->mode) ? t_path : t_ITEM; c1 = (const uchar*)f1->basename; if (type1 == t_PATH && *c1 == '.' && !c1[1]) { type1 = t_ITEM; state1 = s_TRAILING; c1 = (uchar*)""; } else state1 = s_BASE; break; case s_BASE: state1 = s_TRAILING; if (type1 == t_PATH) { c1 = (uchar*)"/"; break; } /* FALL THROUGH */ case s_TRAILING: type1 = t_ITEM; break; } if (*c2 && type1 != type2) return type1 == t_PATH ? 1 : -1; } if (!*c2) { switch (state2) { case s_DIR: state2 = s_SLASH; c2 = (uchar*)"/"; break; case s_SLASH: type2 = S_ISDIR(f2->mode) ? t_path : t_ITEM; c2 = (const uchar*)f2->basename; if (type2 == t_PATH && *c2 == '.' && !c2[1]) { type2 = t_ITEM; state2 = s_TRAILING; c2 = (uchar*)""; } else state2 = s_BASE; break; case s_BASE: state2 = s_TRAILING; if (type2 == t_PATH) { c2 = (uchar*)"/"; break; } /* FALL THROUGH */ case s_TRAILING: found_prefix = 1; if (!*c1) return 0; type2 = t_ITEM; break; } if (type1 != type2) return type1 == t_PATH ? 1 : -1; } } while ((dif = (int)*c1++ - (int)*c2++) == 0); return dif; } /* Returns 1 if f1's filename has all of f2's filename as a prefix. This does * not match if f2's basename is not an exact match of a path element in f1. * E.g. /path/foo is not a prefix of /path/foobar/baz, but /path/foobar is. */ int f_name_has_prefix(const struct file_struct *f1, const struct file_struct *f2) { found_prefix = 0; f_name_cmp(f1, f2); return found_prefix; } char *f_name_buf(void) { static char names[5][MAXPATHLEN]; static unsigned int n; n = (n + 1) % (sizeof names / sizeof names[0]); return names[n]; } /* Return a copy of the full filename of a flist entry, using the indicated * buffer or one of 5 static buffers if fbuf is NULL. No size-checking is * done because we checked the size when creating the file_struct entry. */ char *f_name(const struct file_struct *f, char *fbuf) { if (!f || !F_IS_ACTIVE(f)) return NULL; if (!fbuf) fbuf = f_name_buf(); if (f->dirname) { int len = strlen(f->dirname); memcpy(fbuf, f->dirname, len); fbuf[len] = '/'; strlcpy(fbuf + len + 1, f->basename, MAXPATHLEN - (len + 1)); } else strlcpy(fbuf, f->basename, MAXPATHLEN); return fbuf; } /* Do a non-recursive scan of the named directory, possibly ignoring all * exclude rules except for the daemon's. If "dlen" is >=0, it is the length * of the dirname string, and also indicates that "dirname" is a MAXPATHLEN * buffer (the functions we call will append names onto the end, but the old * dir value will be restored on exit). */ struct file_list *get_dirlist(char *dirname, int dlen, int flags) { struct file_list *dirlist; char dirbuf[MAXPATHLEN]; int save_recurse = recurse; int save_xfer_dirs = xfer_dirs; int save_prune_empty_dirs = prune_empty_dirs; int senddir_fd = flags & GDL_IGNORE_FILTER_RULES ? -2 : -1; if (dlen < 0) { dlen = strlcpy(dirbuf, dirname, MAXPATHLEN); if (dlen >= MAXPATHLEN) return NULL; dirname = dirbuf; } dirlist = flist_new(FLIST_TEMP, "get_dirlist"); recurse = 0; xfer_dirs = 1; send_directory(senddir_fd, dirlist, dirname, dlen, FLAG_CONTENT_DIR); xfer_dirs = save_xfer_dirs; recurse = save_recurse; if (INFO_GTE(PROGRESS, 1)) flist_count_offset += dirlist->used; prune_empty_dirs = 0; dirlist->sorted = dirlist->files; flist_sort_and_clean(dirlist, 0); prune_empty_dirs = save_prune_empty_dirs; if (DEBUG_GTE(FLIST, 3)) output_flist(dirlist); return dirlist; }