2 * Socket and pipe I/O utilities used in rsync.
4 * Copyright (C) 1996-2001 Andrew Tridgell
5 * Copyright (C) 1996 Paul Mackerras
6 * Copyright (C) 2001, 2002 Martin Pool <mbp@samba.org>
7 * Copyright (C) 2003-2007 Wayne Davison
9 * This program is free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License as published by
11 * the Free Software Foundation; either version 3 of the License, or
12 * (at your option) any later version.
14 * This program is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 * GNU General Public License for more details.
19 * You should have received a copy of the GNU General Public License along
20 * with this program; if not, visit the http://fsf.org website.
23 /* Rsync provides its own multiplexing system, which is used to send
24 * stderr and stdout over a single socket.
26 * For historical reasons this is off during the start of the
27 * connection, but it's switched on quite early using
28 * io_start_multiplex_out() and io_start_multiplex_in(). */
32 /** If no timeout is specified then use a 60 second select timeout */
33 #define SELECT_TIMEOUT 60
36 extern size_t bwlimit_writemax;
37 extern int io_timeout;
38 extern int allowed_lull;
42 extern int am_generator;
43 extern int inc_recurse;
47 extern int read_batch;
48 extern int csum_length;
49 extern int protect_args;
50 extern int checksum_seed;
51 extern int protocol_version;
52 extern int remove_source_files;
53 extern int preserve_hard_links;
54 extern struct stats stats;
55 extern struct file_list *cur_flist, *first_flist;
57 extern int filesfrom_convert;
58 extern iconv_t ic_send, ic_recv;
61 const char phase_unknown[] = "unknown";
62 int ignore_timeout = 0;
66 /* Ignore an EOF error if non-zero. See whine_about_eof(). */
67 int kluge_around_eof = 0;
74 static int iobuf_f_in = -1;
75 static char *iobuf_in;
76 static size_t iobuf_in_siz;
77 static size_t iobuf_in_ndx;
78 static size_t iobuf_in_remaining;
80 static int iobuf_f_out = -1;
81 static char *iobuf_out;
82 static int iobuf_out_cnt;
84 int flist_forward_from = -1;
86 static int io_multiplexing_out;
87 static int io_multiplexing_in;
88 static time_t last_io_in;
89 static time_t last_io_out;
92 static int write_batch_monitor_in = -1;
93 static int write_batch_monitor_out = -1;
95 static int io_filesfrom_f_in = -1;
96 static int io_filesfrom_f_out = -1;
97 static xbuf ff_buf = EMPTY_XBUF;
98 static char ff_lastchar;
100 static xbuf iconv_buf = EMPTY_XBUF;
102 static int defer_forwarding_messages = 0;
103 static int select_timeout = SELECT_TIMEOUT;
104 static int active_filecnt = 0;
105 static OFF_T active_bytecnt = 0;
107 static char int_byte_extra[64] = {
108 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* (00 - 3F)/4 */
109 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* (40 - 7F)/4 */
110 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* (80 - BF)/4 */
111 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 5, 6, /* (C0 - FF)/4 */
114 enum festatus { FES_SUCCESS, FES_REDO, FES_NO_SEND };
116 static void readfd(int fd, char *buffer, size_t N);
117 static void writefd(int fd, const char *buf, size_t len);
118 static void writefd_unbuffered(int fd, const char *buf, size_t len);
119 static void mplex_write(int fd, enum msgcode code, const char *buf, size_t len, int convert);
121 struct flist_ndx_item {
122 struct flist_ndx_item *next;
126 struct flist_ndx_list {
127 struct flist_ndx_item *head, *tail;
130 static struct flist_ndx_list redo_list, hlink_list;
132 struct msg_list_item {
133 struct msg_list_item *next;
139 struct msg_list_item *head, *tail;
142 static struct msg_list msg_queue;
144 static void flist_ndx_push(struct flist_ndx_list *lp, int ndx)
146 struct flist_ndx_item *item;
148 if (!(item = new(struct flist_ndx_item)))
149 out_of_memory("flist_ndx_push");
153 lp->tail->next = item;
159 static int flist_ndx_pop(struct flist_ndx_list *lp)
161 struct flist_ndx_item *next;
168 next = lp->head->next;
177 static void got_flist_entry_status(enum festatus status, const char *buf)
179 int ndx = IVAL(buf, 0);
180 struct file_list *flist = flist_for_ndx(ndx);
182 assert(flist != NULL);
183 assert(ndx >= flist->ndx_start);
185 if (remove_source_files) {
187 active_bytecnt -= F_LENGTH(flist->files[ndx - flist->ndx_start]);
191 flist->in_progress--;
195 if (remove_source_files)
196 send_msg(MSG_SUCCESS, buf, 4, 0);
197 if (preserve_hard_links) {
198 struct file_struct *file = flist->files[ndx - flist->ndx_start];
199 if (F_IS_HLINKED(file))
200 flist_ndx_push(&hlink_list, ndx);
206 flist_ndx_push(&redo_list, ndx);
213 static void check_timeout(void)
217 if (!io_timeout || ignore_timeout)
221 last_io_in = time(NULL);
227 if (t - last_io_in >= io_timeout) {
228 if (!am_server && !am_daemon) {
229 rprintf(FERROR, "io timeout after %d seconds -- exiting\n",
230 (int)(t-last_io_in));
232 exit_cleanup(RERR_TIMEOUT);
236 /* Note the fds used for the main socket (which might really be a pipe
237 * for a local transfer, but we can ignore that). */
238 void io_set_sock_fds(int f_in, int f_out)
244 void set_io_timeout(int secs)
248 if (!io_timeout || io_timeout > SELECT_TIMEOUT)
249 select_timeout = SELECT_TIMEOUT;
251 select_timeout = io_timeout;
253 allowed_lull = read_batch ? 0 : (io_timeout + 1) / 2;
256 /* Setup the fd used to receive MSG_* messages. Only needed during the
257 * early stages of being a local sender (up through the sending of the
258 * file list) or when we're the generator (to fetch the messages from
260 void set_msg_fd_in(int fd)
265 /* Setup the fd used to send our MSG_* messages. Only needed when
266 * we're the receiver (to send our messages to the generator). */
267 void set_msg_fd_out(int fd)
270 set_nonblocking(msg_fd_out);
273 /* Add a message to the pending MSG_* list. */
274 static void msg_list_add(struct msg_list *lst, int code, const char *buf, int len, int convert)
276 struct msg_list_item *m;
277 int sz = len + 4 + sizeof m[0] - 1;
279 if (!(m = (struct msg_list_item *)new_array(char, sz)))
280 out_of_memory("msg_list_add");
282 m->convert = convert;
283 SIVAL(m->buf, 0, ((code+MPLEX_BASE)<<24) | len);
284 memcpy(m->buf + 4, buf, len);
292 static void msg_flush(void)
295 while (msg_queue.head && io_multiplexing_out) {
296 struct msg_list_item *m = msg_queue.head;
297 int len = IVAL(m->buf, 0) & 0xFFFFFF;
298 int tag = *((uchar*)m->buf+3) - MPLEX_BASE;
299 if (!(msg_queue.head = m->next))
300 msg_queue.tail = NULL;
301 stats.total_written += len + 4;
302 defer_forwarding_messages++;
303 mplex_write(sock_f_out, tag, m->buf + 4, len, m->convert);
304 defer_forwarding_messages--;
308 while (msg_queue.head) {
309 struct msg_list_item *m = msg_queue.head;
310 int len = IVAL(m->buf, 0) & 0xFFFFFF;
311 int tag = *((uchar*)m->buf+3) - MPLEX_BASE;
312 if (!(msg_queue.head = m->next))
313 msg_queue.tail = NULL;
314 defer_forwarding_messages++;
315 mplex_write(msg_fd_out, tag, m->buf + 4, len, m->convert);
316 defer_forwarding_messages--;
322 /* Read a message from the MSG_* fd and handle it. This is called either
323 * during the early stages of being a local sender (up through the sending
324 * of the file list) or when we're the generator (to fetch the messages
325 * from the receiver). */
326 static void read_msg_fd(void)
330 struct file_list *flist;
334 /* Temporarily disable msg_fd_in. This is needed to avoid looping back
335 * to this routine from writefd_unbuffered(). */
338 defer_forwarding_messages++;
343 len = tag & 0xFFFFFF;
344 tag = (tag >> 24) - MPLEX_BASE;
348 if (len < 0 || len > 1 || !am_generator) {
350 rprintf(FERROR, "invalid message %d:%d [%s%s]\n",
351 tag, len, who_am_i(),
352 inc_recurse ? "/inc" : "");
353 exit_cleanup(RERR_STREAMIO);
356 readfd(fd, buf, len);
357 stats.total_read = read_varlong(fd, 3);
362 if (len != 4 || !am_generator)
365 got_flist_entry_status(FES_REDO, buf);
368 if (len != 4 || !am_generator || !inc_recurse)
371 /* Read extra file list from receiver. */
372 assert(iobuf_in != NULL);
373 assert(iobuf_f_in == fd);
375 rprintf(FINFO, "[%s] receiving flist for dir %d\n",
376 who_am_i(), IVAL(buf,0));
378 flist = recv_file_list(fd);
379 flist->parent_ndx = IVAL(buf,0);
380 #ifdef SUPPORT_HARD_LINKS
381 if (preserve_hard_links)
382 match_hard_links(flist);
386 if (len != 0 || !am_generator || !inc_recurse)
391 if (len >= (int)sizeof buf || !am_generator)
393 readfd(fd, buf, len);
394 send_msg(MSG_DELETED, buf, len, 1);
397 if (len != 4 || !am_generator)
400 got_flist_entry_status(FES_SUCCESS, buf);
403 if (len != 4 || !am_generator)
406 got_flist_entry_status(FES_NO_SEND, buf);
412 if (tag == MSG_SOCKERR)
413 io_end_multiplex_out();
423 rwrite((enum logcode)tag, buf, n, !am_generator);
428 rprintf(FERROR, "unknown message %d:%d [%s]\n",
429 tag, len, who_am_i());
430 exit_cleanup(RERR_STREAMIO);
435 if (!--defer_forwarding_messages)
439 /* This is used by the generator to limit how many file transfers can
440 * be active at once when --remove-source-files is specified. Without
441 * this, sender-side deletions were mostly happening at the end. */
442 void increment_active_files(int ndx, int itemizing, enum logcode code)
444 /* TODO: tune these limits? */
445 while (active_filecnt >= (active_bytecnt >= 128*1024 ? 10 : 50)) {
446 check_for_finished_files(itemizing, code, 0);
448 io_flush(NORMAL_FLUSH);
454 active_bytecnt += F_LENGTH(cur_flist->files[ndx - cur_flist->ndx_start]);
457 /* Write an message to a multiplexed stream. If this fails, rsync exits. */
458 static void mplex_write(int fd, enum msgcode code, const char *buf, size_t len, int convert)
460 char buffer[BIGPATHBUFLEN]; /* Oversized for use by iconv code. */
463 SIVAL(buffer, 0, ((MPLEX_BASE + (int)code)<<24) + len);
466 if (convert && ic_send == (iconv_t)-1)
470 if (convert || n > 1024 - 4) /* BIGPATHBUFLEN can handle 1024 bytes */
473 memcpy(buffer + 4, buf, n);
475 writefd_unbuffered(fd, buffer, n+4);
484 INIT_CONST_XBUF(outbuf, buffer);
485 INIT_XBUF(inbuf, (char*)buf, len, -1);
487 defer_forwarding_messages++;
489 iconvbufs(ic_send, &inbuf, &outbuf,
490 ICB_INCLUDE_BAD | ICB_INCLUDE_INCOMPLETE);
491 writefd_unbuffered(fd, outbuf.buf, outbuf.len);
493 if (!--defer_forwarding_messages)
498 defer_forwarding_messages++;
499 writefd_unbuffered(fd, buf, len);
500 if (!--defer_forwarding_messages)
505 int send_msg(enum msgcode code, const char *buf, int len, int convert)
507 if (msg_fd_out < 0) {
508 if (!defer_forwarding_messages)
509 return io_multiplex_write(code, buf, len, convert);
510 if (!io_multiplexing_out)
512 msg_list_add(&msg_queue, code, buf, len, convert);
515 if (flist_forward_from >= 0)
516 msg_list_add(&msg_queue, code, buf, len, convert);
518 mplex_write(msg_fd_out, code, buf, len, convert);
522 void send_msg_int(enum msgcode code, int num)
525 SIVAL(numbuf, 0, num);
526 send_msg(code, numbuf, 4, 0);
529 void wait_for_receiver(void)
532 io_flush(NORMAL_FLUSH);
537 int get_redo_num(void)
539 return flist_ndx_pop(&redo_list);
542 int get_hlink_num(void)
544 return flist_ndx_pop(&hlink_list);
548 * When we're the receiver and we have a local --files-from list of names
549 * that needs to be sent over the socket to the sender, we have to do two
550 * things at the same time: send the sender a list of what files we're
551 * processing and read the incoming file+info list from the sender. We do
552 * this by augmenting the read_timeout() function to copy this data. It
553 * uses ff_buf to read a block of data from f_in (when it is ready, since
554 * it might be a pipe) and then blast it out f_out (when it is ready to
555 * receive more data).
557 void io_set_filesfrom_fds(int f_in, int f_out)
559 io_filesfrom_f_in = f_in;
560 io_filesfrom_f_out = f_out;
561 alloc_xbuf(&ff_buf, 2048);
564 alloc_xbuf(&iconv_buf, 1024);
568 /* It's almost always an error to get an EOF when we're trying to read from the
569 * network, because the protocol is (for the most part) self-terminating.
571 * There is one case for the receiver when it is at the end of the transfer
572 * (hanging around reading any keep-alive packets that might come its way): if
573 * the sender dies before the generator's kill-signal comes through, we can end
574 * up here needing to loop until the kill-signal arrives. In this situation,
575 * kluge_around_eof will be < 0.
577 * There is another case for older protocol versions (< 24) where the module
578 * listing was not terminated, so we must ignore an EOF error in that case and
579 * exit. In this situation, kluge_around_eof will be > 0. */
580 static void whine_about_eof(int fd)
582 if (kluge_around_eof && fd == sock_f_in) {
584 if (kluge_around_eof > 0)
586 /* If we're still here after 10 seconds, exit with an error. */
587 for (i = 10*1000/20; i--; )
591 rprintf(FERROR, RSYNC_NAME ": connection unexpectedly closed "
592 "(%.0f bytes received so far) [%s]\n",
593 (double)stats.total_read, who_am_i());
595 exit_cleanup(RERR_STREAMIO);
599 * Read from a socket with I/O timeout. return the number of bytes
600 * read. If no bytes can be read then exit, never return a number <= 0.
602 * TODO: If the remote shell connection fails, then current versions
603 * actually report an "unexpected EOF" error here. Since it's a
604 * fairly common mistake to try to use rsh when ssh is required, we
605 * should trap that: if we fail to read any data at all, we should
606 * give a better explanation. We can tell whether the connection has
607 * started by looking e.g. at whether the remote version is known yet.
609 static int read_timeout(int fd, char *buf, size_t len)
613 io_flush(FULL_FLUSH);
616 /* until we manage to read *something* */
625 if (io_filesfrom_f_out >= 0) {
627 if (ff_buf.len == 0) {
628 if (io_filesfrom_f_in >= 0) {
629 FD_SET(io_filesfrom_f_in, &r_fds);
630 new_fd = io_filesfrom_f_in;
632 io_filesfrom_f_out = -1;
636 FD_SET(io_filesfrom_f_out, &w_fds);
637 new_fd = io_filesfrom_f_out;
643 tv.tv_sec = select_timeout;
648 count = select(maxfd + 1, &r_fds, &w_fds, NULL, &tv);
651 if (errno == EBADF) {
652 defer_forwarding_messages = 0;
653 exit_cleanup(RERR_SOCKETIO);
659 if (io_filesfrom_f_out >= 0) {
661 if (FD_ISSET(io_filesfrom_f_out, &w_fds)) {
662 int l = write(io_filesfrom_f_out,
663 ff_buf.buf + ff_buf.pos,
666 if (!(ff_buf.len -= l))
670 } else if (errno != EINTR) {
671 /* XXX should we complain? */
672 io_filesfrom_f_out = -1;
675 } else if (io_filesfrom_f_in >= 0) {
676 if (FD_ISSET(io_filesfrom_f_in, &r_fds)) {
678 xbuf *ibuf = filesfrom_convert ? &iconv_buf : &ff_buf;
680 xbuf *ibuf = &ff_buf;
682 int l = read(io_filesfrom_f_in, ibuf->buf, ibuf->size);
684 if (l == 0 || errno != EINTR) {
685 /* Send end-of-file marker */
686 memcpy(ff_buf.buf, "\0\0", 2);
687 ff_buf.len = ff_lastchar? 2 : 1;
689 io_filesfrom_f_in = -1;
693 if (filesfrom_convert) {
696 iconvbufs(ic_send, &iconv_buf, &ff_buf,
697 ICB_EXPAND_OUT|ICB_INCLUDE_BAD|ICB_INCLUDE_INCOMPLETE);
702 char *s = ff_buf.buf + l;
703 /* Transform CR and/or LF into '\0' */
704 while (s-- > ff_buf.buf) {
705 if (*s == '\n' || *s == '\r')
710 /* Last buf ended with a '\0', so don't
711 * let this buf start with one. */
712 while (l && ff_buf.buf[ff_buf.pos] == '\0')
718 char *f = ff_buf.buf + ff_buf.pos;
721 /* Eliminate any multi-'\0' runs. */
723 if (!(*t++ = *f++)) {
724 while (f != eob && !*f)
736 if (!FD_ISSET(fd, &r_fds))
739 n = read(fd, buf, len);
743 whine_about_eof(fd); /* Doesn't return. */
744 if (errno == EINTR || errno == EWOULDBLOCK
748 /* Don't write errors on a dead socket. */
749 if (fd == sock_f_in) {
750 io_end_multiplex_out();
751 rsyserr(FSOCKERR, errno, "read error");
753 rsyserr(FERROR, errno, "read error");
754 exit_cleanup(RERR_STREAMIO);
761 if (fd == sock_f_in && io_timeout)
762 last_io_in = time(NULL);
768 /* Read a line into the "buf" buffer. */
769 int read_line(int fd, char *buf, size_t bufsiz, int flags)
775 if (flags & RL_CONVERT && iconv_buf.size < bufsiz)
776 realloc_xbuf(&iconv_buf, bufsiz + 1024);
781 s = flags & RL_CONVERT ? iconv_buf.buf : buf;
785 eob = s + bufsiz - 1;
787 cnt = read(fd, &ch, 1);
788 if (cnt < 0 && (errno == EWOULDBLOCK
789 || errno == EINTR || errno == EAGAIN)) {
796 tv.tv_sec = select_timeout;
798 if (!select(fd+1, &r_fds, NULL, &e_fds, &tv))
800 /*if (FD_ISSET(fd, &e_fds))
801 rprintf(FINFO, "select exception on fd %d\n", fd); */
806 if (flags & RL_EOL_NULLS ? ch == '\0' : (ch == '\r' || ch == '\n')) {
807 /* Skip empty lines if dumping comments. */
808 if (flags & RL_DUMP_COMMENTS && s == buf)
817 if (flags & RL_DUMP_COMMENTS && (*buf == '#' || *buf == ';'))
821 if (flags & RL_CONVERT) {
823 INIT_XBUF(outbuf, buf, 0, bufsiz);
825 iconv_buf.len = s - iconv_buf.buf;
826 iconvbufs(ic_recv, &iconv_buf, &outbuf,
827 ICB_INCLUDE_BAD | ICB_INCLUDE_INCOMPLETE);
828 outbuf.buf[outbuf.len] = '\0';
836 int read_args(int f_in, char *mod_name, char *buf, size_t bufsiz, int rl_nulls,
837 char ***argv_p, int *argc_p, char **request_p)
839 int maxargs = MAX_ARGS;
843 int rl_flags = (rl_nulls ? RL_EOL_NULLS : 0);
846 rl_flags |= (protect_args && ic_recv != (iconv_t)-1 ? RL_CONVERT : 0);
849 if (!(argv = new_array(char *, maxargs)))
850 out_of_memory("read_args");
852 argv[argc++] = "rsyncd";
855 if (read_line(f_in, buf, bufsiz, rl_flags) == 0)
858 if (argc == maxargs) {
860 if (!(argv = realloc_array(argv, char *, maxargs)))
861 out_of_memory("read_args");
866 *request_p = strdup(buf);
870 glob_expand_module(mod_name, buf, &argv, &argc, &maxargs);
872 glob_expand(buf, &argv, &argc, &maxargs);
874 if (!(p = strdup(buf)))
875 out_of_memory("read_args");
877 if (*p == '.' && p[1] == '\0')
885 return dot_pos ? dot_pos : argc;
888 int io_start_buffering_out(int f_out)
891 assert(f_out == iobuf_f_out);
894 if (!(iobuf_out = new_array(char, IO_BUFFER_SIZE)))
895 out_of_memory("io_start_buffering_out");
901 int io_start_buffering_in(int f_in)
904 assert(f_in == iobuf_f_in);
907 iobuf_in_siz = 2 * IO_BUFFER_SIZE;
908 if (!(iobuf_in = new_array(char, iobuf_in_siz)))
909 out_of_memory("io_start_buffering_in");
914 void io_end_buffering_in(void)
921 iobuf_in_remaining = 0;
925 void io_end_buffering_out(void)
929 io_flush(FULL_FLUSH);
935 void maybe_flush_socket(int important)
937 if (iobuf_out && iobuf_out_cnt
938 && (important || time(NULL) - last_io_out >= 5))
939 io_flush(NORMAL_FLUSH);
942 void maybe_send_keepalive(void)
944 if (time(NULL) - last_io_out >= allowed_lull) {
945 if (!iobuf_out || !iobuf_out_cnt) {
946 if (protocol_version < 29)
947 return; /* there's nothing we can do */
948 if (protocol_version >= 30)
949 send_msg(MSG_NOOP, "", 0, 0);
951 write_int(sock_f_out, cur_flist->used);
952 write_shortint(sock_f_out, ITEM_IS_NEW);
956 io_flush(NORMAL_FLUSH);
960 void start_flist_forward(int f_in)
962 assert(iobuf_out != NULL);
963 assert(iobuf_f_out == msg_fd_out);
964 flist_forward_from = f_in;
967 void stop_flist_forward()
969 flist_forward_from = -1;
970 io_flush(FULL_FLUSH);
974 * Continue trying to read len bytes - don't return until len has been
977 static void read_loop(int fd, char *buf, size_t len)
980 int n = read_timeout(fd, buf, len);
988 * Read from the file descriptor handling multiplexing - return number
991 * Never returns <= 0.
993 static int readfd_unbuffered(int fd, char *buf, size_t len)
997 char line[BIGPATHBUFLEN];
999 if (!iobuf_in || fd != iobuf_f_in)
1000 return read_timeout(fd, buf, len);
1002 if (!io_multiplexing_in && iobuf_in_remaining == 0) {
1003 iobuf_in_remaining = read_timeout(fd, iobuf_in, iobuf_in_siz);
1008 if (iobuf_in_remaining) {
1009 len = MIN(len, iobuf_in_remaining);
1010 memcpy(buf, iobuf_in + iobuf_in_ndx, len);
1011 iobuf_in_ndx += len;
1012 iobuf_in_remaining -= len;
1017 read_loop(fd, line, 4);
1018 tag = IVAL(line, 0);
1020 msg_bytes = tag & 0xFFFFFF;
1021 tag = (tag >> 24) - MPLEX_BASE;
1025 if (msg_bytes > iobuf_in_siz) {
1026 if (!(iobuf_in = realloc_array(iobuf_in, char,
1028 out_of_memory("readfd_unbuffered");
1029 iobuf_in_siz = msg_bytes;
1031 read_loop(fd, iobuf_in, msg_bytes);
1032 iobuf_in_remaining = msg_bytes;
1037 maybe_send_keepalive();
1042 read_loop(fd, line, msg_bytes);
1043 io_error |= IVAL(line, 0);
1046 if (msg_bytes >= sizeof line)
1049 if (ic_recv != (iconv_t)-1) {
1055 INIT_CONST_XBUF(outbuf, line);
1059 inbuf.len = msg_bytes > sizeof ibuf
1060 ? sizeof ibuf : msg_bytes;
1061 read_loop(fd, inbuf.buf, inbuf.len);
1062 if (!(msg_bytes -= inbuf.len)
1063 && !ibuf[inbuf.len-1])
1064 inbuf.len--, add_null = 1;
1065 if (iconvbufs(ic_send, &inbuf, &outbuf,
1066 ICB_INCLUDE_BAD | ICB_INCLUDE_INCOMPLETE) < 0)
1071 if (outbuf.len == outbuf.size)
1073 outbuf.buf[outbuf.len++] = '\0';
1075 msg_bytes = outbuf.len;
1078 read_loop(fd, line, msg_bytes);
1079 /* A directory name was sent with the trailing null */
1080 if (msg_bytes > 0 && !line[msg_bytes-1])
1081 log_delete(line, S_IFDIR);
1083 line[msg_bytes] = '\0';
1084 log_delete(line, S_IFREG);
1088 if (msg_bytes != 4) {
1090 rprintf(FERROR, "invalid multi-message %d:%ld [%s]\n",
1091 tag, (long)msg_bytes, who_am_i());
1092 exit_cleanup(RERR_STREAMIO);
1094 read_loop(fd, line, msg_bytes);
1095 successful_send(IVAL(line, 0));
1100 read_loop(fd, line, msg_bytes);
1101 send_msg_int(MSG_NO_SEND, IVAL(line, 0));
1105 if (msg_bytes >= sizeof line) {
1108 "multiplexing overflow %d:%ld [%s]\n",
1109 tag, (long)msg_bytes, who_am_i());
1110 exit_cleanup(RERR_STREAMIO);
1112 read_loop(fd, line, msg_bytes);
1113 rwrite((enum logcode)tag, line, msg_bytes, 1);
1116 rprintf(FERROR, "unexpected tag %d [%s]\n",
1118 exit_cleanup(RERR_STREAMIO);
1122 if (iobuf_in_remaining == 0)
1123 io_flush(NORMAL_FLUSH);
1128 /* Do a buffered read from fd. Don't return until all N bytes have
1129 * been read. If all N can't be read then exit with an error. */
1130 static void readfd(int fd, char *buffer, size_t N)
1136 cnt = readfd_unbuffered(fd, buffer + total, N-total);
1140 if (fd == write_batch_monitor_in) {
1141 if ((size_t)write(batch_fd, buffer, total) != total)
1142 exit_cleanup(RERR_FILEIO);
1145 if (fd == flist_forward_from)
1146 writefd(iobuf_f_out, buffer, total);
1148 if (fd == sock_f_in)
1149 stats.total_read += total;
1152 unsigned short read_shortint(int f)
1156 return (UVAL(b, 1) << 8) + UVAL(b, 0);
1159 int32 read_int(int f)
1166 #if SIZEOF_INT32 > 4
1167 if (num & (int32)0x80000000)
1168 num |= ~(int32)0xffffffff;
1173 int32 read_varint(int f)
1183 readfd(f, (char*)&ch, 1);
1184 extra = int_byte_extra[ch / 4];
1186 uchar bit = ((uchar)1<<(8-extra));
1187 if (extra >= (int)sizeof u.b) {
1188 rprintf(FERROR, "Overflow in read_varint()\n");
1189 exit_cleanup(RERR_STREAMIO);
1191 readfd(f, u.b, extra);
1192 u.b[extra] = ch & (bit-1);
1195 #if CAREFUL_ALIGNMENT
1198 #if SIZEOF_INT32 > 4
1199 if (u.x & (int32)0x80000000)
1200 u.x |= ~(int32)0xffffffff;
1205 int64 read_varlong(int f, uchar min_bytes)
1214 #if SIZEOF_INT64 < 8
1219 readfd(f, b2, min_bytes);
1220 memcpy(u.b, b2+1, min_bytes-1);
1221 extra = int_byte_extra[CVAL(b2, 0) / 4];
1223 uchar bit = ((uchar)1<<(8-extra));
1224 if (min_bytes + extra > (int)sizeof u.b) {
1225 rprintf(FERROR, "Overflow in read_varlong()\n");
1226 exit_cleanup(RERR_STREAMIO);
1228 readfd(f, u.b + min_bytes - 1, extra);
1229 u.b[min_bytes + extra - 1] = CVAL(b2, 0) & (bit-1);
1230 #if SIZEOF_INT64 < 8
1231 if (min_bytes + extra > 5 || u.b[4] || CVAL(u.b,3) & 0x80) {
1232 rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1233 exit_cleanup(RERR_UNSUPPORTED);
1237 u.b[min_bytes + extra - 1] = CVAL(b2, 0);
1238 #if SIZEOF_INT64 < 8
1240 #elif CAREFUL_ALIGNMENT
1241 u.x = IVAL(u.b,0) | (((int64)IVAL(u.b,4))<<32);
1246 int64 read_longint(int f)
1248 #if SIZEOF_INT64 >= 8
1251 int32 num = read_int(f);
1253 if (num != (int32)0xffffffff)
1256 #if SIZEOF_INT64 < 8
1257 rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1258 exit_cleanup(RERR_UNSUPPORTED);
1261 return IVAL(b,0) | (((int64)IVAL(b,4))<<32);
1265 void read_buf(int f, char *buf, size_t len)
1270 void read_sbuf(int f, char *buf, size_t len)
1272 readfd(f, buf, len);
1276 uchar read_byte(int f)
1279 readfd(f, (char *)&c, 1);
1283 int read_vstring(int f, char *buf, int bufsize)
1285 int len = read_byte(f);
1288 len = (len & ~0x80) * 0x100 + read_byte(f);
1290 if (len >= bufsize) {
1291 rprintf(FERROR, "over-long vstring received (%d > %d)\n",
1297 readfd(f, buf, len);
1302 /* Populate a sum_struct with values from the socket. This is
1303 * called by both the sender and the receiver. */
1304 void read_sum_head(int f, struct sum_struct *sum)
1306 sum->count = read_int(f);
1307 if (sum->count < 0) {
1308 rprintf(FERROR, "Invalid checksum count %ld [%s]\n",
1309 (long)sum->count, who_am_i());
1310 exit_cleanup(RERR_PROTOCOL);
1312 sum->blength = read_int(f);
1313 if (sum->blength < 0 || sum->blength > MAX_BLOCK_SIZE) {
1314 rprintf(FERROR, "Invalid block length %ld [%s]\n",
1315 (long)sum->blength, who_am_i());
1316 exit_cleanup(RERR_PROTOCOL);
1318 sum->s2length = protocol_version < 27 ? csum_length : (int)read_int(f);
1319 if (sum->s2length < 0 || sum->s2length > MAX_DIGEST_LEN) {
1320 rprintf(FERROR, "Invalid checksum length %d [%s]\n",
1321 sum->s2length, who_am_i());
1322 exit_cleanup(RERR_PROTOCOL);
1324 sum->remainder = read_int(f);
1325 if (sum->remainder < 0 || sum->remainder > sum->blength) {
1326 rprintf(FERROR, "Invalid remainder length %ld [%s]\n",
1327 (long)sum->remainder, who_am_i());
1328 exit_cleanup(RERR_PROTOCOL);
1332 /* Send the values from a sum_struct over the socket. Set sum to
1333 * NULL if there are no checksums to send. This is called by both
1334 * the generator and the sender. */
1335 void write_sum_head(int f, struct sum_struct *sum)
1337 static struct sum_struct null_sum;
1342 write_int(f, sum->count);
1343 write_int(f, sum->blength);
1344 if (protocol_version >= 27)
1345 write_int(f, sum->s2length);
1346 write_int(f, sum->remainder);
1350 * Sleep after writing to limit I/O bandwidth usage.
1352 * @todo Rather than sleeping after each write, it might be better to
1353 * use some kind of averaging. The current algorithm seems to always
1354 * use a bit less bandwidth than specified, because it doesn't make up
1355 * for slow periods. But arguably this is a feature. In addition, we
1356 * ought to take the time used to write the data into account.
1358 * During some phases of big transfers (file FOO is uptodate) this is
1359 * called with a small bytes_written every time. As the kernel has to
1360 * round small waits up to guarantee that we actually wait at least the
1361 * requested number of microseconds, this can become grossly inaccurate.
1362 * We therefore keep track of the bytes we've written over time and only
1363 * sleep when the accumulated delay is at least 1 tenth of a second.
1365 static void sleep_for_bwlimit(int bytes_written)
1367 static struct timeval prior_tv;
1368 static long total_written = 0;
1369 struct timeval tv, start_tv;
1370 long elapsed_usec, sleep_usec;
1372 #define ONE_SEC 1000000L /* # of microseconds in a second */
1374 if (!bwlimit_writemax)
1377 total_written += bytes_written;
1379 gettimeofday(&start_tv, NULL);
1380 if (prior_tv.tv_sec) {
1381 elapsed_usec = (start_tv.tv_sec - prior_tv.tv_sec) * ONE_SEC
1382 + (start_tv.tv_usec - prior_tv.tv_usec);
1383 total_written -= elapsed_usec * bwlimit / (ONE_SEC/1024);
1384 if (total_written < 0)
1388 sleep_usec = total_written * (ONE_SEC/1024) / bwlimit;
1389 if (sleep_usec < ONE_SEC / 10) {
1390 prior_tv = start_tv;
1394 tv.tv_sec = sleep_usec / ONE_SEC;
1395 tv.tv_usec = sleep_usec % ONE_SEC;
1396 select(0, NULL, NULL, NULL, &tv);
1398 gettimeofday(&prior_tv, NULL);
1399 elapsed_usec = (prior_tv.tv_sec - start_tv.tv_sec) * ONE_SEC
1400 + (prior_tv.tv_usec - start_tv.tv_usec);
1401 total_written = (sleep_usec - elapsed_usec) * bwlimit / (ONE_SEC/1024);
1404 /* Write len bytes to the file descriptor fd, looping as necessary to get
1405 * the job done and also (in certain circumstances) reading any data on
1406 * msg_fd_in to avoid deadlock.
1408 * This function underlies the multiplexing system. The body of the
1409 * application never calls this function directly. */
1410 static void writefd_unbuffered(int fd, const char *buf, size_t len)
1412 size_t n, total = 0;
1413 fd_set w_fds, r_fds, e_fds;
1414 int maxfd, count, cnt, using_r_fds;
1419 defer_forwarding_messages++, defer_inc++;
1421 while (total < len) {
1428 if (msg_fd_in >= 0) {
1430 FD_SET(msg_fd_in, &r_fds);
1431 if (msg_fd_in > maxfd)
1437 tv.tv_sec = select_timeout;
1441 count = select(maxfd + 1, using_r_fds ? &r_fds : NULL,
1442 &w_fds, &e_fds, &tv);
1445 if (count < 0 && errno == EBADF)
1446 exit_cleanup(RERR_SOCKETIO);
1451 /*if (FD_ISSET(fd, &e_fds))
1452 rprintf(FINFO, "select exception on fd %d\n", fd); */
1454 if (using_r_fds && FD_ISSET(msg_fd_in, &r_fds))
1457 if (!FD_ISSET(fd, &w_fds))
1461 if (bwlimit_writemax && n > bwlimit_writemax)
1462 n = bwlimit_writemax;
1463 cnt = write(fd, buf + total, n);
1469 if (errno == EWOULDBLOCK || errno == EAGAIN) {
1475 /* Don't try to write errors back across the stream. */
1476 if (fd == sock_f_out)
1477 io_end_multiplex_out();
1478 /* Don't try to write errors down a failing msg pipe. */
1479 if (am_server && fd == msg_fd_out)
1480 exit_cleanup(RERR_STREAMIO);
1481 rsyserr(FERROR, errno,
1482 "writefd_unbuffered failed to write %ld bytes [%s]",
1483 (long)len, who_am_i());
1484 /* If the other side is sending us error messages, try
1485 * to grab any messages they sent before they died. */
1486 while (!am_server && fd == sock_f_out && io_multiplexing_in) {
1490 readfd_unbuffered(sock_f_in, buf, sizeof buf);
1492 exit_cleanup(RERR_STREAMIO);
1496 defer_forwarding_messages++, defer_inc++;
1498 if (fd == sock_f_out) {
1499 if (io_timeout || am_generator)
1500 last_io_out = time(NULL);
1501 sleep_for_bwlimit(cnt);
1506 if (!(defer_forwarding_messages -= defer_inc))
1510 void io_flush(int flush_it_all)
1512 if (!iobuf_out_cnt || no_flush)
1515 if (io_multiplexing_out)
1516 mplex_write(sock_f_out, MSG_DATA, iobuf_out, iobuf_out_cnt, 0);
1518 writefd_unbuffered(iobuf_f_out, iobuf_out, iobuf_out_cnt);
1521 if (flush_it_all && !defer_forwarding_messages)
1525 static void writefd(int fd, const char *buf, size_t len)
1527 if (fd == sock_f_out)
1528 stats.total_written += len;
1530 if (fd == write_batch_monitor_out) {
1531 if ((size_t)write(batch_fd, buf, len) != len)
1532 exit_cleanup(RERR_FILEIO);
1535 if (!iobuf_out || fd != iobuf_f_out) {
1536 writefd_unbuffered(fd, buf, len);
1541 int n = MIN((int)len, IO_BUFFER_SIZE - iobuf_out_cnt);
1543 memcpy(iobuf_out+iobuf_out_cnt, buf, n);
1549 if (iobuf_out_cnt == IO_BUFFER_SIZE)
1550 io_flush(NORMAL_FLUSH);
1554 void write_shortint(int f, unsigned short x)
1558 b[1] = (char)(x >> 8);
1562 void write_int(int f, int32 x)
1569 void write_varint(int f, int32 x)
1577 while (cnt > 1 && b[cnt] == 0)
1579 bit = ((uchar)1<<(7-cnt+1));
1580 if (CVAL(b, cnt) >= bit) {
1584 *b = b[cnt] | ~(bit*2-1);
1591 void write_varlong(int f, int64 x, uchar min_bytes)
1598 #if SIZEOF_INT64 >= 8
1599 SIVAL(b, 5, x >> 32);
1601 if (x <= 0x7FFFFFFF && x >= 0)
1602 memset(b + 5, 0, 4);
1604 rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1605 exit_cleanup(RERR_UNSUPPORTED);
1609 while (cnt > min_bytes && b[cnt] == 0)
1611 bit = ((uchar)1<<(7-cnt+min_bytes));
1612 if (CVAL(b, cnt) >= bit) {
1615 } else if (cnt > min_bytes)
1616 *b = b[cnt] | ~(bit*2-1);
1624 * Note: int64 may actually be a 32-bit type if ./configure couldn't find any
1625 * 64-bit types on this platform.
1627 void write_longint(int f, int64 x)
1629 char b[12], * const s = b+4;
1632 if (x <= 0x7FFFFFFF && x >= 0) {
1637 #if SIZEOF_INT64 < 8
1638 rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1639 exit_cleanup(RERR_UNSUPPORTED);
1642 SIVAL(s, 4, x >> 32);
1647 void write_buf(int f, const char *buf, size_t len)
1652 /** Write a string to the connection */
1653 void write_sbuf(int f, const char *buf)
1655 writefd(f, buf, strlen(buf));
1658 void write_byte(int f, uchar c)
1660 writefd(f, (char *)&c, 1);
1663 void write_vstring(int f, const char *str, int len)
1665 uchar lenbuf[3], *lb = lenbuf;
1670 "attempting to send over-long vstring (%d > %d)\n",
1672 exit_cleanup(RERR_PROTOCOL);
1674 *lb++ = len / 0x100 + 0x80;
1678 writefd(f, (char*)lenbuf, lb - lenbuf + 1);
1680 writefd(f, str, len);
1683 /* Send a file-list index using a byte-reduction method. */
1684 void write_ndx(int f, int32 ndx)
1686 static int32 prev_positive = -1, prev_negative = 1;
1687 int32 diff, cnt = 0;
1690 if (protocol_version < 30 || read_batch) {
1695 /* Send NDX_DONE as a single-byte 0 with no side effects. Send
1696 * negative nums as a positive after sending a leading 0xFF. */
1698 diff = ndx - prev_positive;
1699 prev_positive = ndx;
1700 } else if (ndx == NDX_DONE) {
1705 b[cnt++] = (char)0xFF;
1707 diff = ndx - prev_negative;
1708 prev_negative = ndx;
1711 /* A diff of 1 - 253 is sent as a one-byte diff; a diff of 254 - 32767
1712 * or 0 is sent as a 0xFE + a two-byte diff; otherwise we send 0xFE
1713 * & all 4 bytes of the (non-negative) num with the high-bit set. */
1714 if (diff < 0xFE && diff > 0)
1715 b[cnt++] = (char)diff;
1716 else if (diff < 0 || diff > 0x7FFF) {
1717 b[cnt++] = (char)0xFE;
1718 b[cnt++] = (char)((ndx >> 24) | 0x80);
1719 b[cnt++] = (char)ndx;
1720 b[cnt++] = (char)(ndx >> 8);
1721 b[cnt++] = (char)(ndx >> 16);
1723 b[cnt++] = (char)0xFE;
1724 b[cnt++] = (char)(diff >> 8);
1725 b[cnt++] = (char)diff;
1730 /* Receive a file-list index using a byte-reduction method. */
1731 int32 read_ndx(int f)
1733 static int32 prev_positive = -1, prev_negative = 1;
1734 int32 *prev_ptr, num;
1737 if (protocol_version < 30)
1741 if (CVAL(b, 0) == 0xFF) {
1743 prev_ptr = &prev_negative;
1744 } else if (CVAL(b, 0) == 0)
1747 prev_ptr = &prev_positive;
1748 if (CVAL(b, 0) == 0xFE) {
1750 if (CVAL(b, 0) & 0x80) {
1751 b[3] = CVAL(b, 0) & ~0x80;
1756 num = (UVAL(b,0)<<8) + UVAL(b,1) + *prev_ptr;
1758 num = UVAL(b, 0) + *prev_ptr;
1760 if (prev_ptr == &prev_negative)
1765 /* Read a line of up to bufsiz-1 characters into buf. Strips
1766 * the (required) trailing newline and all carriage returns.
1767 * Returns 1 for success; 0 for I/O error or truncation. */
1768 int read_line_old(int f, char *buf, size_t bufsiz)
1770 bufsiz--; /* leave room for the null */
1771 while (bufsiz > 0) {
1773 read_buf(f, buf, 1);
1778 if (buf[0] != '\r') {
1787 void io_printf(int fd, const char *format, ...)
1790 char buf[BIGPATHBUFLEN];
1793 va_start(ap, format);
1794 len = vsnprintf(buf, sizeof buf, format, ap);
1798 exit_cleanup(RERR_STREAMIO);
1800 if (len > (int)sizeof buf) {
1801 rprintf(FERROR, "io_printf() was too long for the buffer.\n");
1802 exit_cleanup(RERR_STREAMIO);
1805 write_sbuf(fd, buf);
1808 /** Setup for multiplexing a MSG_* stream with the data stream. */
1809 void io_start_multiplex_out(void)
1811 io_flush(NORMAL_FLUSH);
1812 io_start_buffering_out(sock_f_out);
1813 io_multiplexing_out = 1;
1816 /** Setup for multiplexing a MSG_* stream with the data stream. */
1817 void io_start_multiplex_in(void)
1819 io_flush(NORMAL_FLUSH);
1820 io_start_buffering_in(sock_f_in);
1821 io_multiplexing_in = 1;
1824 /** Write an message to the multiplexed data stream. */
1825 int io_multiplex_write(enum msgcode code, const char *buf, size_t len, int convert)
1827 if (!io_multiplexing_out)
1829 io_flush(NORMAL_FLUSH);
1830 stats.total_written += (len+4);
1831 mplex_write(sock_f_out, code, buf, len, convert);
1835 void io_end_multiplex_in(void)
1837 io_multiplexing_in = 0;
1838 io_end_buffering_in();
1841 /** Stop output multiplexing. */
1842 void io_end_multiplex_out(void)
1844 io_multiplexing_out = 0;
1845 io_end_buffering_out();
1848 void start_write_batch(int fd)
1850 /* Some communication has already taken place, but we don't
1851 * enable batch writing until here so that we can write a
1852 * canonical record of the communication even though the
1853 * actual communication so far depends on whether a daemon
1855 write_int(batch_fd, protocol_version);
1856 write_int(batch_fd, checksum_seed);
1859 write_batch_monitor_out = fd;
1861 write_batch_monitor_in = fd;
1864 void stop_write_batch(void)
1866 write_batch_monitor_out = -1;
1867 write_batch_monitor_in = -1;