Refer to the symlink's contents as "symlink data", not "symlink name".
[rsync/rsync.git] / io.c
1 /*
2  * Socket and pipe I/O utilities used in rsync.
3  *
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) 2003-2008 Wayne Davison
8  *
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.
13  *
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.
18  *
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.
21  */
22
23 /* Rsync provides its own multiplexing system, which is used to send
24  * stderr and stdout over a single socket.
25  *
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(). */
29
30 #include "rsync.h"
31 #include "ifuncs.h"
32
33 /** If no timeout is specified then use a 60 second select timeout */
34 #define SELECT_TIMEOUT 60
35
36 extern int bwlimit;
37 extern size_t bwlimit_writemax;
38 extern int io_timeout;
39 extern int allowed_lull;
40 extern int am_server;
41 extern int am_daemon;
42 extern int am_sender;
43 extern int am_generator;
44 extern int inc_recurse;
45 extern int io_error;
46 extern int eol_nulls;
47 extern int flist_eof;
48 extern int list_only;
49 extern int read_batch;
50 extern int csum_length;
51 extern int protect_args;
52 extern int checksum_seed;
53 extern int protocol_version;
54 extern int remove_source_files;
55 extern int preserve_hard_links;
56 extern struct stats stats;
57 extern struct file_list *cur_flist;
58 #ifdef ICONV_OPTION
59 extern int filesfrom_convert;
60 extern iconv_t ic_send, ic_recv;
61 #endif
62
63 const char phase_unknown[] = "unknown";
64 int ignore_timeout = 0;
65 int batch_fd = -1;
66 int msgdone_cnt = 0;
67 int check_for_io_err = 0;
68
69 /* Ignore an EOF error if non-zero. See whine_about_eof(). */
70 int kluge_around_eof = 0;
71
72 int msg_fd_in = -1;
73 int msg_fd_out = -1;
74 int sock_f_in = -1;
75 int sock_f_out = -1;
76
77 static int iobuf_f_in = -1;
78 static char *iobuf_in;
79 static size_t iobuf_in_siz;
80 static size_t iobuf_in_ndx;
81 static size_t iobuf_in_remaining;
82
83 static int iobuf_f_out = -1;
84 static char *iobuf_out;
85 static int iobuf_out_cnt;
86
87 int flist_forward_from = -1;
88
89 static int io_multiplexing_out;
90 static int io_multiplexing_in;
91 static time_t last_io_in;
92 static time_t last_io_out;
93 static int no_flush;
94
95 static int write_batch_monitor_in = -1;
96 static int write_batch_monitor_out = -1;
97
98 static int io_filesfrom_f_in = -1;
99 static int io_filesfrom_f_out = -1;
100 static xbuf ff_buf = EMPTY_XBUF;
101 static char ff_lastchar;
102 #ifdef ICONV_OPTION
103 static xbuf iconv_buf = EMPTY_XBUF;
104 #endif
105 static int defer_forwarding_messages = 0, defer_forwarding_keep = 0;
106 static int select_timeout = SELECT_TIMEOUT;
107 static int active_filecnt = 0;
108 static OFF_T active_bytecnt = 0;
109 static int first_message = 1;
110
111 static char int_byte_extra[64] = {
112         0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* (00 - 3F)/4 */
113         0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* (40 - 7F)/4 */
114         1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* (80 - BF)/4 */
115         2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 5, 6, /* (C0 - FF)/4 */
116 };
117
118 #define REMOTE_OPTION_ERROR "rsync: on remote machine: -"
119 #define REMOTE_OPTION_ERROR2 ": unknown option"
120
121 enum festatus { FES_SUCCESS, FES_REDO, FES_NO_SEND };
122
123 static void readfd(int fd, char *buffer, size_t N);
124 static void writefd(int fd, const char *buf, size_t len);
125 static void writefd_unbuffered(int fd, const char *buf, size_t len);
126 static void mplex_write(int fd, enum msgcode code, const char *buf, size_t len, int convert);
127
128 struct flist_ndx_item {
129         struct flist_ndx_item *next;
130         int ndx;
131 };
132
133 struct flist_ndx_list {
134         struct flist_ndx_item *head, *tail;
135 };
136
137 static struct flist_ndx_list redo_list, hlink_list;
138
139 struct msg_list_item {
140         struct msg_list_item *next;
141         char convert;
142         char buf[1];
143 };
144
145 struct msg_list {
146         struct msg_list_item *head, *tail;
147 };
148
149 static struct msg_list msg_queue;
150
151 static void flist_ndx_push(struct flist_ndx_list *lp, int ndx)
152 {
153         struct flist_ndx_item *item;
154
155         if (!(item = new(struct flist_ndx_item)))
156                 out_of_memory("flist_ndx_push");
157         item->next = NULL;
158         item->ndx = ndx;
159         if (lp->tail)
160                 lp->tail->next = item;
161         else
162                 lp->head = item;
163         lp->tail = item;
164 }
165
166 static int flist_ndx_pop(struct flist_ndx_list *lp)
167 {
168         struct flist_ndx_item *next;
169         int ndx;
170
171         if (!lp->head)
172                 return -1;
173
174         ndx = lp->head->ndx;
175         next = lp->head->next;
176         free(lp->head);
177         lp->head = next;
178         if (!next)
179                 lp->tail = NULL;
180
181         return ndx;
182 }
183
184 static void got_flist_entry_status(enum festatus status, const char *buf)
185 {
186         int ndx = IVAL(buf, 0);
187         struct file_list *flist = flist_for_ndx(ndx);
188
189         assert(flist != NULL);
190
191         if (remove_source_files) {
192                 active_filecnt--;
193                 active_bytecnt -= F_LENGTH(flist->files[ndx - flist->ndx_start]);
194         }
195
196         if (inc_recurse)
197                 flist->in_progress--;
198
199         switch (status) {
200         case FES_SUCCESS:
201                 if (remove_source_files)
202                         send_msg(MSG_SUCCESS, buf, 4, 0);
203                 if (preserve_hard_links) {
204                         struct file_struct *file = flist->files[ndx - flist->ndx_start];
205                         if (F_IS_HLINKED(file)) {
206                                 flist_ndx_push(&hlink_list, ndx);
207                                 flist->in_progress++;
208                         }
209                 }
210                 break;
211         case FES_REDO:
212                 if (inc_recurse)
213                         flist->to_redo++;
214                 flist_ndx_push(&redo_list, ndx);
215                 break;
216         case FES_NO_SEND:
217                 break;
218         }
219 }
220
221 static void check_timeout(void)
222 {
223         time_t t;
224
225         if (!io_timeout || ignore_timeout)
226                 return;
227
228         if (!last_io_in) {
229                 last_io_in = time(NULL);
230                 return;
231         }
232
233         t = time(NULL);
234
235         if (t - last_io_in >= io_timeout) {
236                 if (!am_server && !am_daemon) {
237                         rprintf(FERROR, "io timeout after %d seconds -- exiting\n",
238                                 (int)(t-last_io_in));
239                 }
240                 exit_cleanup(RERR_TIMEOUT);
241         }
242 }
243
244 /* Note the fds used for the main socket (which might really be a pipe
245  * for a local transfer, but we can ignore that). */
246 void io_set_sock_fds(int f_in, int f_out)
247 {
248         sock_f_in = f_in;
249         sock_f_out = f_out;
250 }
251
252 void set_io_timeout(int secs)
253 {
254         io_timeout = secs;
255
256         if (!io_timeout || io_timeout > SELECT_TIMEOUT)
257                 select_timeout = SELECT_TIMEOUT;
258         else
259                 select_timeout = io_timeout;
260
261         allowed_lull = read_batch ? 0 : (io_timeout + 1) / 2;
262 }
263
264 /* Setup the fd used to receive MSG_* messages.  Only needed during the
265  * early stages of being a local sender (up through the sending of the
266  * file list) or when we're the generator (to fetch the messages from
267  * the receiver). */
268 void set_msg_fd_in(int fd)
269 {
270         msg_fd_in = fd;
271 }
272
273 /* Setup the fd used to send our MSG_* messages.  Only needed when
274  * we're the receiver (to send our messages to the generator). */
275 void set_msg_fd_out(int fd)
276 {
277         msg_fd_out = fd;
278         set_nonblocking(msg_fd_out);
279 }
280
281 /* Add a message to the pending MSG_* list. */
282 static void msg_list_add(struct msg_list *lst, int code, const char *buf, int len, int convert)
283 {
284         struct msg_list_item *m;
285         int sz = len + 4 + sizeof m[0] - 1;
286
287         if (!(m = (struct msg_list_item *)new_array(char, sz)))
288                 out_of_memory("msg_list_add");
289         m->next = NULL;
290         m->convert = convert;
291         SIVAL(m->buf, 0, ((code+MPLEX_BASE)<<24) | len);
292         memcpy(m->buf + 4, buf, len);
293         if (lst->tail)
294                 lst->tail->next = m;
295         else
296                 lst->head = m;
297         lst->tail = m;
298 }
299
300 static void msg_flush(void)
301 {
302         if (am_generator) {
303                 while (msg_queue.head && io_multiplexing_out) {
304                         struct msg_list_item *m = msg_queue.head;
305                         int len = IVAL(m->buf, 0) & 0xFFFFFF;
306                         int tag = *((uchar*)m->buf+3) - MPLEX_BASE;
307                         if (!(msg_queue.head = m->next))
308                                 msg_queue.tail = NULL;
309                         stats.total_written += len + 4;
310                         defer_forwarding_messages++;
311                         mplex_write(sock_f_out, tag, m->buf + 4, len, m->convert);
312                         defer_forwarding_messages--;
313                         free(m);
314                 }
315         } else {
316                 while (msg_queue.head) {
317                         struct msg_list_item *m = msg_queue.head;
318                         int len = IVAL(m->buf, 0) & 0xFFFFFF;
319                         int tag = *((uchar*)m->buf+3) - MPLEX_BASE;
320                         if (!(msg_queue.head = m->next))
321                                 msg_queue.tail = NULL;
322                         defer_forwarding_messages++;
323                         mplex_write(msg_fd_out, tag, m->buf + 4, len, m->convert);
324                         defer_forwarding_messages--;
325                         free(m);
326                 }
327         }
328 }
329
330 static void check_for_d_option_error(const char *msg)
331 {
332         static char rsync263_opts[] = "BCDHIKLPRSTWabceghlnopqrtuvxz";
333         char *colon;
334         int saw_d = 0;
335
336         if (*msg != 'r'
337          || strncmp(msg, REMOTE_OPTION_ERROR, sizeof REMOTE_OPTION_ERROR - 1) != 0)
338                 return;
339
340         msg += sizeof REMOTE_OPTION_ERROR - 1;
341         if (*msg == '-' || (colon = strchr(msg, ':')) == NULL
342          || strncmp(colon, REMOTE_OPTION_ERROR2, sizeof REMOTE_OPTION_ERROR2 - 1) != 0)
343                 return;
344
345         for ( ; *msg != ':'; msg++) {
346                 if (*msg == 'd')
347                         saw_d = 1;
348                 else if (*msg == 'e')
349                         break;
350                 else if (strchr(rsync263_opts, *msg) == NULL)
351                         return;
352         }
353
354         if (saw_d) {
355                 rprintf(FWARNING,
356                     "*** Try using \"--old-d\" if remote rsync is <= 2.6.3 ***\n");
357         }
358 }
359
360 /* Read a message from the MSG_* fd and handle it.  This is called either
361  * during the early stages of being a local sender (up through the sending
362  * of the file list) or when we're the generator (to fetch the messages
363  * from the receiver). */
364 static void read_msg_fd(void)
365 {
366         char buf[2048];
367         size_t n;
368         struct file_list *flist;
369         int fd = msg_fd_in;
370         int tag, len;
371
372         /* Temporarily disable msg_fd_in.  This is needed to avoid looping back
373          * to this routine from writefd_unbuffered(). */
374         no_flush++;
375         msg_fd_in = -1;
376         defer_forwarding_messages++;
377
378         readfd(fd, buf, 4);
379         tag = IVAL(buf, 0);
380
381         len = tag & 0xFFFFFF;
382         tag = (tag >> 24) - MPLEX_BASE;
383
384         check_for_io_err = 0;
385
386         switch (tag) {
387         case MSG_DONE:
388                 if (len < 0 || len > 1 || !am_generator) {
389                   invalid_msg:
390                         rprintf(FERROR, "invalid message %d:%d [%s%s]\n",
391                                 tag, len, who_am_i(),
392                                 inc_recurse ? "/inc" : "");
393                         exit_cleanup(RERR_STREAMIO);
394                 }
395                 if (len) {
396                         readfd(fd, buf, len);
397                         stats.total_read = read_varlong(fd, 3);
398                 }
399                 msgdone_cnt++;
400                 break;
401         case MSG_REDO:
402                 if (len != 4 || !am_generator)
403                         goto invalid_msg;
404                 readfd(fd, buf, 4);
405                 got_flist_entry_status(FES_REDO, buf);
406                 break;
407         case MSG_FLIST:
408                 if (len != 4 || !am_generator || !inc_recurse)
409                         goto invalid_msg;
410                 readfd(fd, buf, 4);
411                 /* Read extra file list from receiver. */
412                 assert(iobuf_in != NULL);
413                 assert(iobuf_f_in == fd);
414                 if (verbose > 3) {
415                         rprintf(FINFO, "[%s] receiving flist for dir %d\n",
416                                 who_am_i(), IVAL(buf,0));
417                 }
418                 flist = recv_file_list(fd);
419                 flist->parent_ndx = IVAL(buf,0);
420                 /* If the sender is going to send us an MSG_IO_ERROR value, it
421                  * will always be the very next message following MSG_FLIST. */
422                 check_for_io_err = 1;
423 #ifdef SUPPORT_HARD_LINKS
424                 if (preserve_hard_links)
425                         match_hard_links(flist);
426 #endif
427                 break;
428         case MSG_FLIST_EOF:
429                 if (len != 0 || !am_generator || !inc_recurse)
430                         goto invalid_msg;
431                 flist_eof = 1;
432                 break;
433         case MSG_IO_ERROR:
434                 if (len != 4)
435                         goto invalid_msg;
436                 readfd(fd, buf, len);
437                 io_error |= IVAL(buf, 0);
438                 break;
439         case MSG_DELETED:
440                 if (len >= (int)sizeof buf || !am_generator)
441                         goto invalid_msg;
442                 readfd(fd, buf, len);
443                 send_msg(MSG_DELETED, buf, len, 1);
444                 break;
445         case MSG_SUCCESS:
446                 if (len != 4 || !am_generator)
447                         goto invalid_msg;
448                 readfd(fd, buf, 4);
449                 got_flist_entry_status(FES_SUCCESS, buf);
450                 break;
451         case MSG_NO_SEND:
452                 if (len != 4 || !am_generator)
453                         goto invalid_msg;
454                 readfd(fd, buf, 4);
455                 got_flist_entry_status(FES_NO_SEND, buf);
456                 break;
457         case MSG_ERROR_SOCKET:
458         case MSG_CLIENT:
459                 if (!am_generator)
460                         goto invalid_msg;
461                 if (tag == MSG_ERROR_SOCKET)
462                         io_end_multiplex_out();
463                 /* FALL THROUGH */
464         case MSG_INFO:
465         case MSG_ERROR:
466         case MSG_ERROR_XFER:
467         case MSG_WARNING:
468         case MSG_LOG:
469                 while (len) {
470                         n = len;
471                         if (n >= sizeof buf)
472                                 n = sizeof buf - 1;
473                         readfd(fd, buf, n);
474                         rwrite((enum logcode)tag, buf, n, !am_generator);
475                         len -= n;
476                 }
477                 break;
478         default:
479                 rprintf(FERROR, "unknown message %d:%d [%s]\n",
480                         tag, len, who_am_i());
481                 exit_cleanup(RERR_STREAMIO);
482         }
483
484         no_flush--;
485         msg_fd_in = fd;
486         if (!--defer_forwarding_messages && !no_flush)
487                 msg_flush();
488 }
489
490 /* This is used by the generator to limit how many file transfers can
491  * be active at once when --remove-source-files is specified.  Without
492  * this, sender-side deletions were mostly happening at the end. */
493 void increment_active_files(int ndx, int itemizing, enum logcode code)
494 {
495         /* TODO: tune these limits? */
496         while (active_filecnt >= (active_bytecnt >= 128*1024 ? 10 : 50)) {
497                 check_for_finished_files(itemizing, code, 0);
498                 if (iobuf_out_cnt)
499                         io_flush(NORMAL_FLUSH);
500                 else
501                         read_msg_fd();
502         }
503
504         active_filecnt++;
505         active_bytecnt += F_LENGTH(cur_flist->files[ndx - cur_flist->ndx_start]);
506 }
507
508 /* Write an message to a multiplexed stream. If this fails, rsync exits. */
509 static void mplex_write(int fd, enum msgcode code, const char *buf, size_t len, int convert)
510 {
511         char buffer[BIGPATHBUFLEN]; /* Oversized for use by iconv code. */
512         size_t n = len;
513
514 #ifdef ICONV_OPTION
515         /* We need to convert buf before doing anything else so that we
516          * can include the (converted) byte length in the message header. */
517         if (convert && ic_send != (iconv_t)-1) {
518                 xbuf outbuf, inbuf;
519
520                 INIT_XBUF(outbuf, buffer + 4, 0, sizeof buffer - 4);
521                 INIT_XBUF(inbuf, (char*)buf, len, -1);
522
523                 iconvbufs(ic_send, &inbuf, &outbuf,
524                           ICB_INCLUDE_BAD | ICB_INCLUDE_INCOMPLETE);
525                 if (inbuf.len > 0) {
526                         rprintf(FERROR, "overflowed conversion buffer in mplex_write");
527                         exit_cleanup(RERR_UNSUPPORTED);
528                 }
529
530                 n = len = outbuf.len;
531         } else
532 #endif
533         if (n > 1024 - 4) /* BIGPATHBUFLEN can handle 1024 bytes */
534                 n = 0;    /* We'd rather do 2 writes than too much memcpy(). */
535         else
536                 memcpy(buffer + 4, buf, n);
537
538         SIVAL(buffer, 0, ((MPLEX_BASE + (int)code)<<24) + len);
539
540         defer_forwarding_keep = 1; /* defer_forwarding_messages++ on return */
541         writefd_unbuffered(fd, buffer, n+4);
542         defer_forwarding_keep = 0;
543
544         if (len > n)
545                 writefd_unbuffered(fd, buf+n, len-n);
546
547         if (!--defer_forwarding_messages && !no_flush)
548                 msg_flush();
549 }
550
551 int send_msg(enum msgcode code, const char *buf, int len, int convert)
552 {
553         if (msg_fd_out < 0) {
554                 if (!defer_forwarding_messages)
555                         return io_multiplex_write(code, buf, len, convert);
556                 if (!io_multiplexing_out)
557                         return 0;
558                 msg_list_add(&msg_queue, code, buf, len, convert);
559                 return 1;
560         }
561         if (flist_forward_from >= 0)
562                 msg_list_add(&msg_queue, code, buf, len, convert);
563         else
564                 mplex_write(msg_fd_out, code, buf, len, convert);
565         return 1;
566 }
567
568 void send_msg_int(enum msgcode code, int num)
569 {
570         char numbuf[4];
571         SIVAL(numbuf, 0, num);
572         send_msg(code, numbuf, 4, 0);
573 }
574
575 void wait_for_receiver(void)
576 {
577         if (iobuf_out_cnt)
578                 io_flush(NORMAL_FLUSH);
579         else
580                 read_msg_fd();
581 }
582
583 int get_redo_num(void)
584 {
585         return flist_ndx_pop(&redo_list);
586 }
587
588 int get_hlink_num(void)
589 {
590         return flist_ndx_pop(&hlink_list);
591 }
592
593 /**
594  * When we're the receiver and we have a local --files-from list of names
595  * that needs to be sent over the socket to the sender, we have to do two
596  * things at the same time: send the sender a list of what files we're
597  * processing and read the incoming file+info list from the sender.  We do
598  * this by augmenting the read_timeout() function to copy this data.  It
599  * uses ff_buf to read a block of data from f_in (when it is ready, since
600  * it might be a pipe) and then blast it out f_out (when it is ready to
601  * receive more data).
602  */
603 void io_set_filesfrom_fds(int f_in, int f_out)
604 {
605         io_filesfrom_f_in = f_in;
606         io_filesfrom_f_out = f_out;
607         alloc_xbuf(&ff_buf, 2048);
608 #ifdef ICONV_OPTION
609         if (protect_args)
610                 alloc_xbuf(&iconv_buf, 1024);
611 #endif
612 }
613
614 /* It's almost always an error to get an EOF when we're trying to read from the
615  * network, because the protocol is (for the most part) self-terminating.
616  *
617  * There is one case for the receiver when it is at the end of the transfer
618  * (hanging around reading any keep-alive packets that might come its way): if
619  * the sender dies before the generator's kill-signal comes through, we can end
620  * up here needing to loop until the kill-signal arrives.  In this situation,
621  * kluge_around_eof will be < 0.
622  *
623  * There is another case for older protocol versions (< 24) where the module
624  * listing was not terminated, so we must ignore an EOF error in that case and
625  * exit.  In this situation, kluge_around_eof will be > 0. */
626 static void whine_about_eof(int fd)
627 {
628         if (kluge_around_eof && fd == sock_f_in) {
629                 int i;
630                 if (kluge_around_eof > 0)
631                         exit_cleanup(0);
632                 /* If we're still here after 10 seconds, exit with an error. */
633                 for (i = 10*1000/20; i--; )
634                         msleep(20);
635         }
636
637         rprintf(FERROR, RSYNC_NAME ": connection unexpectedly closed "
638                 "(%.0f bytes received so far) [%s]\n",
639                 (double)stats.total_read, who_am_i());
640
641         exit_cleanup(RERR_STREAMIO);
642 }
643
644 /**
645  * Read from a socket with I/O timeout. return the number of bytes
646  * read. If no bytes can be read then exit, never return a number <= 0.
647  *
648  * TODO: If the remote shell connection fails, then current versions
649  * actually report an "unexpected EOF" error here.  Since it's a
650  * fairly common mistake to try to use rsh when ssh is required, we
651  * should trap that: if we fail to read any data at all, we should
652  * give a better explanation.  We can tell whether the connection has
653  * started by looking e.g. at whether the remote version is known yet.
654  */
655 static int read_timeout(int fd, char *buf, size_t len)
656 {
657         int n, cnt = 0;
658
659         io_flush(FULL_FLUSH);
660
661         while (cnt == 0) {
662                 /* until we manage to read *something* */
663                 fd_set r_fds, w_fds;
664                 struct timeval tv;
665                 int maxfd = fd;
666                 int count;
667
668                 FD_ZERO(&r_fds);
669                 FD_ZERO(&w_fds);
670                 FD_SET(fd, &r_fds);
671                 if (io_filesfrom_f_out >= 0) {
672                         int new_fd;
673                         if (ff_buf.len == 0) {
674                                 if (io_filesfrom_f_in >= 0) {
675                                         FD_SET(io_filesfrom_f_in, &r_fds);
676                                         new_fd = io_filesfrom_f_in;
677                                 } else {
678                                         io_filesfrom_f_out = -1;
679                                         new_fd = -1;
680                                 }
681                         } else {
682                                 FD_SET(io_filesfrom_f_out, &w_fds);
683                                 new_fd = io_filesfrom_f_out;
684                         }
685                         if (new_fd > maxfd)
686                                 maxfd = new_fd;
687                 }
688
689                 tv.tv_sec = select_timeout;
690                 tv.tv_usec = 0;
691
692                 errno = 0;
693
694                 count = select(maxfd + 1, &r_fds, &w_fds, NULL, &tv);
695
696                 if (count <= 0) {
697                         if (errno == EBADF) {
698                                 defer_forwarding_messages = 0;
699                                 exit_cleanup(RERR_SOCKETIO);
700                         }
701                         check_timeout();
702                         continue;
703                 }
704
705                 if (io_filesfrom_f_out >= 0) {
706                         if (ff_buf.len) {
707                                 if (FD_ISSET(io_filesfrom_f_out, &w_fds)) {
708                                         int l = write(io_filesfrom_f_out,
709                                                       ff_buf.buf + ff_buf.pos,
710                                                       ff_buf.len);
711                                         if (l > 0) {
712                                                 if (!(ff_buf.len -= l))
713                                                         ff_buf.pos = 0;
714                                                 else
715                                                         ff_buf.pos += l;
716                                         } else if (errno != EINTR) {
717                                                 /* XXX should we complain? */
718                                                 io_filesfrom_f_out = -1;
719                                         }
720                                 }
721                         } else if (io_filesfrom_f_in >= 0) {
722                                 if (FD_ISSET(io_filesfrom_f_in, &r_fds)) {
723 #ifdef ICONV_OPTION
724                                         xbuf *ibuf = filesfrom_convert ? &iconv_buf : &ff_buf;
725 #else
726                                         xbuf *ibuf = &ff_buf;
727 #endif
728                                         int l = read(io_filesfrom_f_in, ibuf->buf, ibuf->size);
729                                         if (l <= 0) {
730                                                 if (l == 0 || errno != EINTR) {
731                                                         /* Send end-of-file marker */
732                                                         memcpy(ff_buf.buf, "\0\0", 2);
733                                                         ff_buf.len = ff_lastchar? 2 : 1;
734                                                         ff_buf.pos = 0;
735                                                         io_filesfrom_f_in = -1;
736                                                 }
737                                         } else {
738 #ifdef ICONV_OPTION
739                                                 if (filesfrom_convert) {
740                                                         iconv_buf.pos = 0;
741                                                         iconv_buf.len = l;
742                                                         iconvbufs(ic_send, &iconv_buf, &ff_buf,
743                                                             ICB_EXPAND_OUT|ICB_INCLUDE_BAD|ICB_INCLUDE_INCOMPLETE);
744                                                         l = ff_buf.len;
745                                                 }
746 #endif
747                                                 if (!eol_nulls) {
748                                                         char *s = ff_buf.buf + l;
749                                                         /* Transform CR and/or LF into '\0' */
750                                                         while (s-- > ff_buf.buf) {
751                                                                 if (*s == '\n' || *s == '\r')
752                                                                         *s = '\0';
753                                                         }
754                                                 }
755                                                 if (!ff_lastchar) {
756                                                         /* Last buf ended with a '\0', so don't
757                                                          * let this buf start with one. */
758                                                         while (l && ff_buf.buf[ff_buf.pos] == '\0')
759                                                                 ff_buf.pos++, l--;
760                                                 }
761                                                 if (!l)
762                                                         ff_buf.pos = 0;
763                                                 else {
764                                                         char *f = ff_buf.buf + ff_buf.pos;
765                                                         char *t = f;
766                                                         char *eob = f + l;
767                                                         /* Eliminate any multi-'\0' runs. */
768                                                         while (f != eob) {
769                                                                 if (!(*t++ = *f++)) {
770                                                                         while (f != eob && !*f)
771                                                                                 f++, l--;
772                                                                 }
773                                                         }
774                                                         ff_lastchar = f[-1];
775                                                 }
776                                                 ff_buf.len = l;
777                                         }
778                                 }
779                         }
780                 }
781
782                 if (!FD_ISSET(fd, &r_fds))
783                         continue;
784
785                 n = read(fd, buf, len);
786
787                 if (n <= 0) {
788                         if (n == 0)
789                                 whine_about_eof(fd); /* Doesn't return. */
790                         if (errno == EINTR || errno == EWOULDBLOCK
791                             || errno == EAGAIN)
792                                 continue;
793
794                         /* Don't write errors on a dead socket. */
795                         if (fd == sock_f_in) {
796                                 io_end_multiplex_out();
797                                 rsyserr(FERROR_SOCKET, errno, "read error");
798                         } else
799                                 rsyserr(FERROR, errno, "read error");
800                         exit_cleanup(RERR_STREAMIO);
801                 }
802
803                 buf += n;
804                 len -= n;
805                 cnt += n;
806
807                 if (fd == sock_f_in && io_timeout)
808                         last_io_in = time(NULL);
809         }
810
811         return cnt;
812 }
813
814 /* Read a line into the "buf" buffer. */
815 int read_line(int fd, char *buf, size_t bufsiz, int flags)
816 {
817         char ch, *s, *eob;
818         int cnt;
819
820 #ifdef ICONV_OPTION
821         if (flags & RL_CONVERT && iconv_buf.size < bufsiz)
822                 realloc_xbuf(&iconv_buf, bufsiz + 1024);
823 #endif
824
825   start:
826 #ifdef ICONV_OPTION
827         s = flags & RL_CONVERT ? iconv_buf.buf : buf;
828 #else
829         s = buf;
830 #endif
831         eob = s + bufsiz - 1;
832         while (1) {
833                 cnt = read(fd, &ch, 1);
834                 if (cnt < 0 && (errno == EWOULDBLOCK
835                   || errno == EINTR || errno == EAGAIN)) {
836                         struct timeval tv;
837                         fd_set r_fds, e_fds;
838                         FD_ZERO(&r_fds);
839                         FD_SET(fd, &r_fds);
840                         FD_ZERO(&e_fds);
841                         FD_SET(fd, &e_fds);
842                         tv.tv_sec = select_timeout;
843                         tv.tv_usec = 0;
844                         if (!select(fd+1, &r_fds, NULL, &e_fds, &tv))
845                                 check_timeout();
846                         /*if (FD_ISSET(fd, &e_fds))
847                                 rprintf(FINFO, "select exception on fd %d\n", fd); */
848                         continue;
849                 }
850                 if (cnt != 1)
851                         break;
852                 if (flags & RL_EOL_NULLS ? ch == '\0' : (ch == '\r' || ch == '\n')) {
853                         /* Skip empty lines if dumping comments. */
854                         if (flags & RL_DUMP_COMMENTS && s == buf)
855                                 continue;
856                         break;
857                 }
858                 if (s < eob)
859                         *s++ = ch;
860         }
861         *s = '\0';
862
863         if (flags & RL_DUMP_COMMENTS && (*buf == '#' || *buf == ';'))
864                 goto start;
865
866 #ifdef ICONV_OPTION
867         if (flags & RL_CONVERT) {
868                 xbuf outbuf;
869                 INIT_XBUF(outbuf, buf, 0, bufsiz);
870                 iconv_buf.pos = 0;
871                 iconv_buf.len = s - iconv_buf.buf;
872                 iconvbufs(ic_recv, &iconv_buf, &outbuf,
873                           ICB_INCLUDE_BAD | ICB_INCLUDE_INCOMPLETE);
874                 outbuf.buf[outbuf.len] = '\0';
875                 return outbuf.len;
876         }
877 #endif
878
879         return s - buf;
880 }
881
882 void read_args(int f_in, char *mod_name, char *buf, size_t bufsiz, int rl_nulls,
883                char ***argv_p, int *argc_p, char **request_p)
884 {
885         int maxargs = MAX_ARGS;
886         int dot_pos = 0;
887         int argc = 0;
888         char **argv, *p;
889         int rl_flags = (rl_nulls ? RL_EOL_NULLS : 0);
890
891 #ifdef ICONV_OPTION
892         rl_flags |= (protect_args && ic_recv != (iconv_t)-1 ? RL_CONVERT : 0);
893 #endif
894
895         if (!(argv = new_array(char *, maxargs)))
896                 out_of_memory("read_args");
897         if (mod_name && !protect_args)
898                 argv[argc++] = "rsyncd";
899
900         while (1) {
901                 if (read_line(f_in, buf, bufsiz, rl_flags) == 0)
902                         break;
903
904                 if (argc == maxargs-1) {
905                         maxargs += MAX_ARGS;
906                         if (!(argv = realloc_array(argv, char *, maxargs)))
907                                 out_of_memory("read_args");
908                 }
909
910                 if (dot_pos) {
911                         if (request_p) {
912                                 *request_p = strdup(buf);
913                                 request_p = NULL;
914                         }
915                         if (mod_name)
916                                 glob_expand_module(mod_name, buf, &argv, &argc, &maxargs);
917                         else
918                                 glob_expand(buf, &argv, &argc, &maxargs);
919                 } else {
920                         if (!(p = strdup(buf)))
921                                 out_of_memory("read_args");
922                         argv[argc++] = p;
923                         if (*p == '.' && p[1] == '\0')
924                                 dot_pos = argc;
925                 }
926         }
927         argv[argc] = NULL;
928
929         glob_expand(NULL, NULL, NULL, NULL);
930
931         *argc_p = argc;
932         *argv_p = argv;
933 }
934
935 int io_start_buffering_out(int f_out)
936 {
937         if (iobuf_out) {
938                 assert(f_out == iobuf_f_out);
939                 return 0;
940         }
941         if (!(iobuf_out = new_array(char, IO_BUFFER_SIZE)))
942                 out_of_memory("io_start_buffering_out");
943         iobuf_out_cnt = 0;
944         iobuf_f_out = f_out;
945         return 1;
946 }
947
948 int io_start_buffering_in(int f_in)
949 {
950         if (iobuf_in) {
951                 assert(f_in == iobuf_f_in);
952                 return 0;
953         }
954         iobuf_in_siz = 2 * IO_BUFFER_SIZE;
955         if (!(iobuf_in = new_array(char, iobuf_in_siz)))
956                 out_of_memory("io_start_buffering_in");
957         iobuf_f_in = f_in;
958         return 1;
959 }
960
961 void io_end_buffering_in(void)
962 {
963         if (!iobuf_in)
964                 return;
965         free(iobuf_in);
966         iobuf_in = NULL;
967         iobuf_in_ndx = 0;
968         iobuf_in_remaining = 0;
969         iobuf_f_in = -1;
970 }
971
972 void io_end_buffering_out(void)
973 {
974         if (!iobuf_out)
975                 return;
976         io_flush(FULL_FLUSH);
977         free(iobuf_out);
978         iobuf_out = NULL;
979         iobuf_f_out = -1;
980 }
981
982 void maybe_flush_socket(int important)
983 {
984         if (iobuf_out && iobuf_out_cnt
985          && (important || time(NULL) - last_io_out >= 5))
986                 io_flush(NORMAL_FLUSH);
987 }
988
989 void maybe_send_keepalive(void)
990 {
991         if (time(NULL) - last_io_out >= allowed_lull) {
992                 if (!iobuf_out || !iobuf_out_cnt) {
993                         if (protocol_version < 29)
994                                 return; /* there's nothing we can do */
995                         if (protocol_version >= 30)
996                                 send_msg(MSG_NOOP, "", 0, 0);
997                         else {
998                                 write_int(sock_f_out, cur_flist->used);
999                                 write_shortint(sock_f_out, ITEM_IS_NEW);
1000                         }
1001                 }
1002                 if (iobuf_out)
1003                         io_flush(NORMAL_FLUSH);
1004         }
1005 }
1006
1007 void start_flist_forward(int f_in)
1008 {
1009         assert(iobuf_out != NULL);
1010         assert(iobuf_f_out == msg_fd_out);
1011         flist_forward_from = f_in;
1012 }
1013
1014 void stop_flist_forward()
1015 {
1016         flist_forward_from = -1;
1017         io_flush(FULL_FLUSH);
1018 }
1019
1020 /**
1021  * Continue trying to read len bytes - don't return until len has been
1022  * read.
1023  **/
1024 static void read_loop(int fd, char *buf, size_t len)
1025 {
1026         while (len) {
1027                 int n = read_timeout(fd, buf, len);
1028
1029                 buf += n;
1030                 len -= n;
1031         }
1032 }
1033
1034 /**
1035  * Read from the file descriptor handling multiplexing - return number
1036  * of bytes read.
1037  *
1038  * Never returns <= 0.
1039  */
1040 static int readfd_unbuffered(int fd, char *buf, size_t len)
1041 {
1042         size_t msg_bytes;
1043         int tag, cnt = 0;
1044         char line[BIGPATHBUFLEN];
1045
1046         if (!iobuf_in || fd != iobuf_f_in)
1047                 return read_timeout(fd, buf, len);
1048
1049         if (!io_multiplexing_in && iobuf_in_remaining == 0) {
1050                 iobuf_in_remaining = read_timeout(fd, iobuf_in, iobuf_in_siz);
1051                 iobuf_in_ndx = 0;
1052         }
1053
1054         while (cnt == 0) {
1055                 if (iobuf_in_remaining) {
1056                         len = MIN(len, iobuf_in_remaining);
1057                         memcpy(buf, iobuf_in + iobuf_in_ndx, len);
1058                         iobuf_in_ndx += len;
1059                         iobuf_in_remaining -= len;
1060                         cnt = len;
1061                         break;
1062                 }
1063
1064                 read_loop(fd, line, 4);
1065                 tag = IVAL(line, 0);
1066
1067                 msg_bytes = tag & 0xFFFFFF;
1068                 tag = (tag >> 24) - MPLEX_BASE;
1069
1070                 check_for_io_err = 0;
1071
1072                 switch (tag) {
1073                 case MSG_DATA:
1074                         if (msg_bytes > iobuf_in_siz) {
1075                                 if (!(iobuf_in = realloc_array(iobuf_in, char,
1076                                                                msg_bytes)))
1077                                         out_of_memory("readfd_unbuffered");
1078                                 iobuf_in_siz = msg_bytes;
1079                         }
1080                         read_loop(fd, iobuf_in, msg_bytes);
1081                         iobuf_in_remaining = msg_bytes;
1082                         iobuf_in_ndx = 0;
1083                         break;
1084                 case MSG_NOOP:
1085                         if (am_sender)
1086                                 maybe_send_keepalive();
1087                         break;
1088                 case MSG_IO_ERROR:
1089                         if (msg_bytes != 4)
1090                                 goto invalid_msg;
1091                         read_loop(fd, line, msg_bytes);
1092                         send_msg_int(MSG_IO_ERROR, IVAL(line, 0));
1093                         io_error |= IVAL(line, 0);
1094                         break;
1095                 case MSG_DELETED:
1096                         if (msg_bytes >= sizeof line)
1097                                 goto overflow;
1098 #ifdef ICONV_OPTION
1099                         if (ic_recv != (iconv_t)-1) {
1100                                 xbuf outbuf, inbuf;
1101                                 char ibuf[512];
1102                                 int add_null = 0;
1103                                 int pos = 0;
1104
1105                                 INIT_CONST_XBUF(outbuf, line);
1106                                 INIT_XBUF(inbuf, ibuf, 0, -1);
1107
1108                                 while (msg_bytes) {
1109                                         inbuf.len = msg_bytes > sizeof ibuf
1110                                                   ? sizeof ibuf : msg_bytes;
1111                                         read_loop(fd, inbuf.buf, inbuf.len);
1112                                         if (!(msg_bytes -= inbuf.len)
1113                                          && !ibuf[inbuf.len-1])
1114                                                 inbuf.len--, add_null = 1;
1115                                         if (iconvbufs(ic_send, &inbuf, &outbuf,
1116                                             ICB_INCLUDE_BAD | ICB_INCLUDE_INCOMPLETE) < 0)
1117                                                 goto overflow;
1118                                         pos = -1;
1119                                 }
1120                                 if (add_null) {
1121                                         if (outbuf.len == outbuf.size)
1122                                                 goto overflow;
1123                                         outbuf.buf[outbuf.len++] = '\0';
1124                                 }
1125                                 msg_bytes = outbuf.len;
1126                         } else
1127 #endif
1128                                 read_loop(fd, line, msg_bytes);
1129                         /* A directory name was sent with the trailing null */
1130                         if (msg_bytes > 0 && !line[msg_bytes-1])
1131                                 log_delete(line, S_IFDIR);
1132                         else {
1133                                 line[msg_bytes] = '\0';
1134                                 log_delete(line, S_IFREG);
1135                         }
1136                         break;
1137                 case MSG_SUCCESS:
1138                         if (msg_bytes != 4) {
1139                           invalid_msg:
1140                                 rprintf(FERROR, "invalid multi-message %d:%ld [%s]\n",
1141                                         tag, (long)msg_bytes, who_am_i());
1142                                 exit_cleanup(RERR_STREAMIO);
1143                         }
1144                         read_loop(fd, line, msg_bytes);
1145                         successful_send(IVAL(line, 0));
1146                         break;
1147                 case MSG_NO_SEND:
1148                         if (msg_bytes != 4)
1149                                 goto invalid_msg;
1150                         read_loop(fd, line, msg_bytes);
1151                         send_msg_int(MSG_NO_SEND, IVAL(line, 0));
1152                         break;
1153                 case MSG_INFO:
1154                 case MSG_ERROR:
1155                 case MSG_ERROR_XFER:
1156                 case MSG_WARNING:
1157                         if (msg_bytes >= sizeof line) {
1158                             overflow:
1159                                 rprintf(FERROR,
1160                                         "multiplexing overflow %d:%ld [%s]\n",
1161                                         tag, (long)msg_bytes, who_am_i());
1162                                 exit_cleanup(RERR_STREAMIO);
1163                         }
1164                         read_loop(fd, line, msg_bytes);
1165                         rwrite((enum logcode)tag, line, msg_bytes, 1);
1166                         if (first_message) {
1167                                 if (list_only && !am_sender && tag == 1) {
1168                                         line[msg_bytes] = '\0';
1169                                         check_for_d_option_error(line);
1170                                 }
1171                                 first_message = 0;
1172                         }
1173                         break;
1174                 default:
1175                         rprintf(FERROR, "unexpected tag %d [%s]\n",
1176                                 tag, who_am_i());
1177                         exit_cleanup(RERR_STREAMIO);
1178                 }
1179         }
1180
1181         if (iobuf_in_remaining == 0)
1182                 io_flush(NORMAL_FLUSH);
1183
1184         return cnt;
1185 }
1186
1187 /* Do a buffered read from fd.  Don't return until all N bytes have
1188  * been read.  If all N can't be read then exit with an error. */
1189 static void readfd(int fd, char *buffer, size_t N)
1190 {
1191         int  cnt;
1192         size_t total = 0;
1193
1194         while (total < N) {
1195                 cnt = readfd_unbuffered(fd, buffer + total, N-total);
1196                 total += cnt;
1197         }
1198
1199         if (fd == write_batch_monitor_in) {
1200                 if ((size_t)write(batch_fd, buffer, total) != total)
1201                         exit_cleanup(RERR_FILEIO);
1202         }
1203
1204         if (fd == flist_forward_from)
1205                 writefd(iobuf_f_out, buffer, total);
1206
1207         if (fd == sock_f_in)
1208                 stats.total_read += total;
1209 }
1210
1211 unsigned short read_shortint(int f)
1212 {
1213         char b[2];
1214         readfd(f, b, 2);
1215         return (UVAL(b, 1) << 8) + UVAL(b, 0);
1216 }
1217
1218 int32 read_int(int f)
1219 {
1220         char b[4];
1221         int32 num;
1222
1223         readfd(f, b, 4);
1224         num = IVAL(b, 0);
1225 #if SIZEOF_INT32 > 4
1226         if (num & (int32)0x80000000)
1227                 num |= ~(int32)0xffffffff;
1228 #endif
1229         return num;
1230 }
1231
1232 int32 read_varint(int f)
1233 {
1234         union {
1235             char b[5];
1236             int32 x;
1237         } u;
1238         uchar ch;
1239         int extra;
1240
1241         u.x = 0;
1242         readfd(f, (char*)&ch, 1);
1243         extra = int_byte_extra[ch / 4];
1244         if (extra) {
1245                 uchar bit = ((uchar)1<<(8-extra));
1246                 if (extra >= (int)sizeof u.b) {
1247                         rprintf(FERROR, "Overflow in read_varint()\n");
1248                         exit_cleanup(RERR_STREAMIO);
1249                 }
1250                 readfd(f, u.b, extra);
1251                 u.b[extra] = ch & (bit-1);
1252         } else
1253                 u.b[0] = ch;
1254 #if CAREFUL_ALIGNMENT
1255         u.x = IVAL(u.b,0);
1256 #endif
1257 #if SIZEOF_INT32 > 4
1258         if (u.x & (int32)0x80000000)
1259                 u.x |= ~(int32)0xffffffff;
1260 #endif
1261         return u.x;
1262 }
1263
1264 int64 read_varlong(int f, uchar min_bytes)
1265 {
1266         union {
1267             char b[9];
1268             int64 x;
1269         } u;
1270         char b2[8];
1271         int extra;
1272
1273 #if SIZEOF_INT64 < 8
1274         memset(u.b, 0, 8);
1275 #else
1276         u.x = 0;
1277 #endif
1278         readfd(f, b2, min_bytes);
1279         memcpy(u.b, b2+1, min_bytes-1);
1280         extra = int_byte_extra[CVAL(b2, 0) / 4];
1281         if (extra) {
1282                 uchar bit = ((uchar)1<<(8-extra));
1283                 if (min_bytes + extra > (int)sizeof u.b) {
1284                         rprintf(FERROR, "Overflow in read_varlong()\n");
1285                         exit_cleanup(RERR_STREAMIO);
1286                 }
1287                 readfd(f, u.b + min_bytes - 1, extra);
1288                 u.b[min_bytes + extra - 1] = CVAL(b2, 0) & (bit-1);
1289 #if SIZEOF_INT64 < 8
1290                 if (min_bytes + extra > 5 || u.b[4] || CVAL(u.b,3) & 0x80) {
1291                         rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1292                         exit_cleanup(RERR_UNSUPPORTED);
1293                 }
1294 #endif
1295         } else
1296                 u.b[min_bytes + extra - 1] = CVAL(b2, 0);
1297 #if SIZEOF_INT64 < 8
1298         u.x = IVAL(u.b,0);
1299 #elif CAREFUL_ALIGNMENT
1300         u.x = IVAL(u.b,0) | (((int64)IVAL(u.b,4))<<32);
1301 #endif
1302         return u.x;
1303 }
1304
1305 int64 read_longint(int f)
1306 {
1307 #if SIZEOF_INT64 >= 8
1308         char b[9];
1309 #endif
1310         int32 num = read_int(f);
1311
1312         if (num != (int32)0xffffffff)
1313                 return num;
1314
1315 #if SIZEOF_INT64 < 8
1316         rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1317         exit_cleanup(RERR_UNSUPPORTED);
1318 #else
1319         readfd(f, b, 8);
1320         return IVAL(b,0) | (((int64)IVAL(b,4))<<32);
1321 #endif
1322 }
1323
1324 void read_buf(int f, char *buf, size_t len)
1325 {
1326         readfd(f,buf,len);
1327 }
1328
1329 void read_sbuf(int f, char *buf, size_t len)
1330 {
1331         readfd(f, buf, len);
1332         buf[len] = '\0';
1333 }
1334
1335 uchar read_byte(int f)
1336 {
1337         uchar c;
1338         readfd(f, (char *)&c, 1);
1339         return c;
1340 }
1341
1342 int read_vstring(int f, char *buf, int bufsize)
1343 {
1344         int len = read_byte(f);
1345
1346         if (len & 0x80)
1347                 len = (len & ~0x80) * 0x100 + read_byte(f);
1348
1349         if (len >= bufsize) {
1350                 rprintf(FERROR, "over-long vstring received (%d > %d)\n",
1351                         len, bufsize - 1);
1352                 return -1;
1353         }
1354
1355         if (len)
1356                 readfd(f, buf, len);
1357         buf[len] = '\0';
1358         return len;
1359 }
1360
1361 /* Populate a sum_struct with values from the socket.  This is
1362  * called by both the sender and the receiver. */
1363 void read_sum_head(int f, struct sum_struct *sum)
1364 {
1365         int32 max_blength = protocol_version < 30 ? OLD_MAX_BLOCK_SIZE : MAX_BLOCK_SIZE;
1366         sum->count = read_int(f);
1367         if (sum->count < 0) {
1368                 rprintf(FERROR, "Invalid checksum count %ld [%s]\n",
1369                         (long)sum->count, who_am_i());
1370                 exit_cleanup(RERR_PROTOCOL);
1371         }
1372         sum->blength = read_int(f);
1373         if (sum->blength < 0 || sum->blength > max_blength) {
1374                 rprintf(FERROR, "Invalid block length %ld [%s]\n",
1375                         (long)sum->blength, who_am_i());
1376                 exit_cleanup(RERR_PROTOCOL);
1377         }
1378         sum->s2length = protocol_version < 27 ? csum_length : (int)read_int(f);
1379         if (sum->s2length < 0 || sum->s2length > MAX_DIGEST_LEN) {
1380                 rprintf(FERROR, "Invalid checksum length %d [%s]\n",
1381                         sum->s2length, who_am_i());
1382                 exit_cleanup(RERR_PROTOCOL);
1383         }
1384         sum->remainder = read_int(f);
1385         if (sum->remainder < 0 || sum->remainder > sum->blength) {
1386                 rprintf(FERROR, "Invalid remainder length %ld [%s]\n",
1387                         (long)sum->remainder, who_am_i());
1388                 exit_cleanup(RERR_PROTOCOL);
1389         }
1390 }
1391
1392 /* Send the values from a sum_struct over the socket.  Set sum to
1393  * NULL if there are no checksums to send.  This is called by both
1394  * the generator and the sender. */
1395 void write_sum_head(int f, struct sum_struct *sum)
1396 {
1397         static struct sum_struct null_sum;
1398
1399         if (sum == NULL)
1400                 sum = &null_sum;
1401
1402         write_int(f, sum->count);
1403         write_int(f, sum->blength);
1404         if (protocol_version >= 27)
1405                 write_int(f, sum->s2length);
1406         write_int(f, sum->remainder);
1407 }
1408
1409 /**
1410  * Sleep after writing to limit I/O bandwidth usage.
1411  *
1412  * @todo Rather than sleeping after each write, it might be better to
1413  * use some kind of averaging.  The current algorithm seems to always
1414  * use a bit less bandwidth than specified, because it doesn't make up
1415  * for slow periods.  But arguably this is a feature.  In addition, we
1416  * ought to take the time used to write the data into account.
1417  *
1418  * During some phases of big transfers (file FOO is uptodate) this is
1419  * called with a small bytes_written every time.  As the kernel has to
1420  * round small waits up to guarantee that we actually wait at least the
1421  * requested number of microseconds, this can become grossly inaccurate.
1422  * We therefore keep track of the bytes we've written over time and only
1423  * sleep when the accumulated delay is at least 1 tenth of a second.
1424  **/
1425 static void sleep_for_bwlimit(int bytes_written)
1426 {
1427         static struct timeval prior_tv;
1428         static long total_written = 0;
1429         struct timeval tv, start_tv;
1430         long elapsed_usec, sleep_usec;
1431
1432 #define ONE_SEC 1000000L /* # of microseconds in a second */
1433
1434         if (!bwlimit_writemax)
1435                 return;
1436
1437         total_written += bytes_written;
1438
1439         gettimeofday(&start_tv, NULL);
1440         if (prior_tv.tv_sec) {
1441                 elapsed_usec = (start_tv.tv_sec - prior_tv.tv_sec) * ONE_SEC
1442                              + (start_tv.tv_usec - prior_tv.tv_usec);
1443                 total_written -= elapsed_usec * bwlimit / (ONE_SEC/1024);
1444                 if (total_written < 0)
1445                         total_written = 0;
1446         }
1447
1448         sleep_usec = total_written * (ONE_SEC/1024) / bwlimit;
1449         if (sleep_usec < ONE_SEC / 10) {
1450                 prior_tv = start_tv;
1451                 return;
1452         }
1453
1454         tv.tv_sec  = sleep_usec / ONE_SEC;
1455         tv.tv_usec = sleep_usec % ONE_SEC;
1456         select(0, NULL, NULL, NULL, &tv);
1457
1458         gettimeofday(&prior_tv, NULL);
1459         elapsed_usec = (prior_tv.tv_sec - start_tv.tv_sec) * ONE_SEC
1460                      + (prior_tv.tv_usec - start_tv.tv_usec);
1461         total_written = (sleep_usec - elapsed_usec) * bwlimit / (ONE_SEC/1024);
1462 }
1463
1464 /* Write len bytes to the file descriptor fd, looping as necessary to get
1465  * the job done and also (in certain circumstances) reading any data on
1466  * msg_fd_in to avoid deadlock.
1467  *
1468  * This function underlies the multiplexing system.  The body of the
1469  * application never calls this function directly. */
1470 static void writefd_unbuffered(int fd, const char *buf, size_t len)
1471 {
1472         size_t n, total = 0;
1473         fd_set w_fds, r_fds, e_fds;
1474         int maxfd, count, cnt, using_r_fds;
1475         int defer_inc = 0;
1476         struct timeval tv;
1477
1478         if (no_flush++)
1479                 defer_forwarding_messages++, defer_inc++;
1480
1481         while (total < len) {
1482                 FD_ZERO(&w_fds);
1483                 FD_SET(fd, &w_fds);
1484                 FD_ZERO(&e_fds);
1485                 FD_SET(fd, &e_fds);
1486                 maxfd = fd;
1487
1488                 if (msg_fd_in >= 0) {
1489                         FD_ZERO(&r_fds);
1490                         FD_SET(msg_fd_in, &r_fds);
1491                         if (msg_fd_in > maxfd)
1492                                 maxfd = msg_fd_in;
1493                         using_r_fds = 1;
1494                 } else
1495                         using_r_fds = 0;
1496
1497                 tv.tv_sec = select_timeout;
1498                 tv.tv_usec = 0;
1499
1500                 errno = 0;
1501                 count = select(maxfd + 1, using_r_fds ? &r_fds : NULL,
1502                                &w_fds, &e_fds, &tv);
1503
1504                 if (count <= 0) {
1505                         if (count < 0 && errno == EBADF)
1506                                 exit_cleanup(RERR_SOCKETIO);
1507                         check_timeout();
1508                         continue;
1509                 }
1510
1511                 /*if (FD_ISSET(fd, &e_fds))
1512                         rprintf(FINFO, "select exception on fd %d\n", fd); */
1513
1514                 if (using_r_fds && FD_ISSET(msg_fd_in, &r_fds))
1515                         read_msg_fd();
1516
1517                 if (!FD_ISSET(fd, &w_fds))
1518                         continue;
1519
1520                 n = len - total;
1521                 if (bwlimit_writemax && n > bwlimit_writemax)
1522                         n = bwlimit_writemax;
1523                 cnt = write(fd, buf + total, n);
1524
1525                 if (cnt <= 0) {
1526                         if (cnt < 0) {
1527                                 if (errno == EINTR)
1528                                         continue;
1529                                 if (errno == EWOULDBLOCK || errno == EAGAIN) {
1530                                         msleep(1);
1531                                         continue;
1532                                 }
1533                         }
1534
1535                         /* Don't try to write errors back across the stream. */
1536                         if (fd == sock_f_out)
1537                                 io_end_multiplex_out();
1538                         /* Don't try to write errors down a failing msg pipe. */
1539                         if (am_server && fd == msg_fd_out)
1540                                 exit_cleanup(RERR_STREAMIO);
1541                         rsyserr(FERROR, errno,
1542                                 "writefd_unbuffered failed to write %ld bytes [%s]",
1543                                 (long)len, who_am_i());
1544                         /* If the other side is sending us error messages, try
1545                          * to grab any messages they sent before they died. */
1546                         while (!am_server && fd == sock_f_out && io_multiplexing_in) {
1547                                 char buf[1024];
1548                                 set_io_timeout(30);
1549                                 ignore_timeout = 0;
1550                                 readfd_unbuffered(sock_f_in, buf, sizeof buf);
1551                         }
1552                         exit_cleanup(RERR_STREAMIO);
1553                 }
1554
1555                 total += cnt;
1556                 defer_forwarding_messages++, defer_inc++;
1557
1558                 if (fd == sock_f_out) {
1559                         if (io_timeout || am_generator)
1560                                 last_io_out = time(NULL);
1561                         sleep_for_bwlimit(cnt);
1562                 }
1563         }
1564
1565         no_flush--;
1566         defer_inc -= defer_forwarding_keep;
1567         if (!(defer_forwarding_messages -= defer_inc) && !no_flush)
1568                 msg_flush();
1569 }
1570
1571 void io_flush(int flush_it_all)
1572 {
1573         if (!iobuf_out_cnt || no_flush)
1574                 return;
1575
1576         if (io_multiplexing_out)
1577                 mplex_write(sock_f_out, MSG_DATA, iobuf_out, iobuf_out_cnt, 0);
1578         else
1579                 writefd_unbuffered(iobuf_f_out, iobuf_out, iobuf_out_cnt);
1580         iobuf_out_cnt = 0;
1581
1582         if (flush_it_all && !defer_forwarding_messages)
1583                 msg_flush();
1584 }
1585
1586 static void writefd(int fd, const char *buf, size_t len)
1587 {
1588         if (fd == sock_f_out)
1589                 stats.total_written += len;
1590
1591         if (fd == write_batch_monitor_out) {
1592                 if ((size_t)write(batch_fd, buf, len) != len)
1593                         exit_cleanup(RERR_FILEIO);
1594         }
1595
1596         if (!iobuf_out || fd != iobuf_f_out) {
1597                 writefd_unbuffered(fd, buf, len);
1598                 return;
1599         }
1600
1601         while (len) {
1602                 int n = MIN((int)len, IO_BUFFER_SIZE - iobuf_out_cnt);
1603                 if (n > 0) {
1604                         memcpy(iobuf_out+iobuf_out_cnt, buf, n);
1605                         buf += n;
1606                         len -= n;
1607                         iobuf_out_cnt += n;
1608                 }
1609
1610                 if (iobuf_out_cnt == IO_BUFFER_SIZE)
1611                         io_flush(NORMAL_FLUSH);
1612         }
1613 }
1614
1615 void write_shortint(int f, unsigned short x)
1616 {
1617         char b[2];
1618         b[0] = (char)x;
1619         b[1] = (char)(x >> 8);
1620         writefd(f, b, 2);
1621 }
1622
1623 void write_int(int f, int32 x)
1624 {
1625         char b[4];
1626         SIVAL(b, 0, x);
1627         writefd(f, b, 4);
1628 }
1629
1630 void write_varint(int f, int32 x)
1631 {
1632         char b[5];
1633         uchar bit;
1634         int cnt = 4;
1635
1636         SIVAL(b, 1, x);
1637
1638         while (cnt > 1 && b[cnt] == 0)
1639                 cnt--;
1640         bit = ((uchar)1<<(7-cnt+1));
1641         if (CVAL(b, cnt) >= bit) {
1642                 cnt++;
1643                 *b = ~(bit-1);
1644         } else if (cnt > 1)
1645                 *b = b[cnt] | ~(bit*2-1);
1646         else
1647                 *b = b[cnt];
1648
1649         writefd(f, b, cnt);
1650 }
1651
1652 void write_varlong(int f, int64 x, uchar min_bytes)
1653 {
1654         char b[9];
1655         uchar bit;
1656         int cnt = 8;
1657
1658         SIVAL(b, 1, x);
1659 #if SIZEOF_INT64 >= 8
1660         SIVAL(b, 5, x >> 32);
1661 #else
1662         if (x <= 0x7FFFFFFF && x >= 0)
1663                 memset(b + 5, 0, 4);
1664         else {
1665                 rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1666                 exit_cleanup(RERR_UNSUPPORTED);
1667         }
1668 #endif
1669
1670         while (cnt > min_bytes && b[cnt] == 0)
1671                 cnt--;
1672         bit = ((uchar)1<<(7-cnt+min_bytes));
1673         if (CVAL(b, cnt) >= bit) {
1674                 cnt++;
1675                 *b = ~(bit-1);
1676         } else if (cnt > min_bytes)
1677                 *b = b[cnt] | ~(bit*2-1);
1678         else
1679                 *b = b[cnt];
1680
1681         writefd(f, b, cnt);
1682 }
1683
1684 /*
1685  * Note: int64 may actually be a 32-bit type if ./configure couldn't find any
1686  * 64-bit types on this platform.
1687  */
1688 void write_longint(int f, int64 x)
1689 {
1690         char b[12], * const s = b+4;
1691
1692         SIVAL(s, 0, x);
1693         if (x <= 0x7FFFFFFF && x >= 0) {
1694                 writefd(f, s, 4);
1695                 return;
1696         }
1697
1698 #if SIZEOF_INT64 < 8
1699         rprintf(FERROR, "Integer overflow: attempted 64-bit offset\n");
1700         exit_cleanup(RERR_UNSUPPORTED);
1701 #else
1702         memset(b, 0xFF, 4);
1703         SIVAL(s, 4, x >> 32);
1704         writefd(f, b, 12);
1705 #endif
1706 }
1707
1708 void write_buf(int f, const char *buf, size_t len)
1709 {
1710         writefd(f,buf,len);
1711 }
1712
1713 /** Write a string to the connection */
1714 void write_sbuf(int f, const char *buf)
1715 {
1716         writefd(f, buf, strlen(buf));
1717 }
1718
1719 void write_byte(int f, uchar c)
1720 {
1721         writefd(f, (char *)&c, 1);
1722 }
1723
1724 void write_vstring(int f, const char *str, int len)
1725 {
1726         uchar lenbuf[3], *lb = lenbuf;
1727
1728         if (len > 0x7F) {
1729                 if (len > 0x7FFF) {
1730                         rprintf(FERROR,
1731                                 "attempting to send over-long vstring (%d > %d)\n",
1732                                 len, 0x7FFF);
1733                         exit_cleanup(RERR_PROTOCOL);
1734                 }
1735                 *lb++ = len / 0x100 + 0x80;
1736         }
1737         *lb = len;
1738
1739         writefd(f, (char*)lenbuf, lb - lenbuf + 1);
1740         if (len)
1741                 writefd(f, str, len);
1742 }
1743
1744 /* Send a file-list index using a byte-reduction method. */
1745 void write_ndx(int f, int32 ndx)
1746 {
1747         static int32 prev_positive = -1, prev_negative = 1;
1748         int32 diff, cnt = 0;
1749         char b[6];
1750
1751         if (protocol_version < 30 || read_batch) {
1752                 write_int(f, ndx);
1753                 return;
1754         }
1755
1756         /* Send NDX_DONE as a single-byte 0 with no side effects.  Send
1757          * negative nums as a positive after sending a leading 0xFF. */
1758         if (ndx >= 0) {
1759                 diff = ndx - prev_positive;
1760                 prev_positive = ndx;
1761         } else if (ndx == NDX_DONE) {
1762                 *b = 0;
1763                 writefd(f, b, 1);
1764                 return;
1765         } else {
1766                 b[cnt++] = (char)0xFF;
1767                 ndx = -ndx;
1768                 diff = ndx - prev_negative;
1769                 prev_negative = ndx;
1770         }
1771
1772         /* A diff of 1 - 253 is sent as a one-byte diff; a diff of 254 - 32767
1773          * or 0 is sent as a 0xFE + a two-byte diff; otherwise we send 0xFE
1774          * & all 4 bytes of the (non-negative) num with the high-bit set. */
1775         if (diff < 0xFE && diff > 0)
1776                 b[cnt++] = (char)diff;
1777         else if (diff < 0 || diff > 0x7FFF) {
1778                 b[cnt++] = (char)0xFE;
1779                 b[cnt++] = (char)((ndx >> 24) | 0x80);
1780                 b[cnt++] = (char)ndx;
1781                 b[cnt++] = (char)(ndx >> 8);
1782                 b[cnt++] = (char)(ndx >> 16);
1783         } else {
1784                 b[cnt++] = (char)0xFE;
1785                 b[cnt++] = (char)(diff >> 8);
1786                 b[cnt++] = (char)diff;
1787         }
1788         writefd(f, b, cnt);
1789 }
1790
1791 /* Receive a file-list index using a byte-reduction method. */
1792 int32 read_ndx(int f)
1793 {
1794         static int32 prev_positive = -1, prev_negative = 1;
1795         int32 *prev_ptr, num;
1796         char b[4];
1797
1798         if (protocol_version < 30)
1799                 return read_int(f);
1800
1801         readfd(f, b, 1);
1802         if (CVAL(b, 0) == 0xFF) {
1803                 readfd(f, b, 1);
1804                 prev_ptr = &prev_negative;
1805         } else if (CVAL(b, 0) == 0)
1806                 return NDX_DONE;
1807         else
1808                 prev_ptr = &prev_positive;
1809         if (CVAL(b, 0) == 0xFE) {
1810                 readfd(f, b, 2);
1811                 if (CVAL(b, 0) & 0x80) {
1812                         b[3] = CVAL(b, 0) & ~0x80;
1813                         b[0] = b[1];
1814                         readfd(f, b+1, 2);
1815                         num = IVAL(b, 0);
1816                 } else
1817                         num = (UVAL(b,0)<<8) + UVAL(b,1) + *prev_ptr;
1818         } else
1819                 num = UVAL(b, 0) + *prev_ptr;
1820         *prev_ptr = num;
1821         if (prev_ptr == &prev_negative)
1822                 num = -num;
1823         return num;
1824 }
1825
1826 /* Read a line of up to bufsiz-1 characters into buf.  Strips
1827  * the (required) trailing newline and all carriage returns.
1828  * Returns 1 for success; 0 for I/O error or truncation. */
1829 int read_line_old(int f, char *buf, size_t bufsiz)
1830 {
1831         bufsiz--; /* leave room for the null */
1832         while (bufsiz > 0) {
1833                 buf[0] = 0;
1834                 read_buf(f, buf, 1);
1835                 if (buf[0] == 0)
1836                         return 0;
1837                 if (buf[0] == '\n')
1838                         break;
1839                 if (buf[0] != '\r') {
1840                         buf++;
1841                         bufsiz--;
1842                 }
1843         }
1844         *buf = '\0';
1845         return bufsiz > 0;
1846 }
1847
1848 void io_printf(int fd, const char *format, ...)
1849 {
1850         va_list ap;
1851         char buf[BIGPATHBUFLEN];
1852         int len;
1853
1854         va_start(ap, format);
1855         len = vsnprintf(buf, sizeof buf, format, ap);
1856         va_end(ap);
1857
1858         if (len < 0)
1859                 exit_cleanup(RERR_STREAMIO);
1860
1861         if (len > (int)sizeof buf) {
1862                 rprintf(FERROR, "io_printf() was too long for the buffer.\n");
1863                 exit_cleanup(RERR_STREAMIO);
1864         }
1865
1866         write_sbuf(fd, buf);
1867 }
1868
1869 /** Setup for multiplexing a MSG_* stream with the data stream. */
1870 void io_start_multiplex_out(void)
1871 {
1872         io_flush(NORMAL_FLUSH);
1873         io_start_buffering_out(sock_f_out);
1874         io_multiplexing_out = 1;
1875 }
1876
1877 /** Setup for multiplexing a MSG_* stream with the data stream. */
1878 void io_start_multiplex_in(void)
1879 {
1880         io_flush(NORMAL_FLUSH);
1881         io_start_buffering_in(sock_f_in);
1882         io_multiplexing_in = 1;
1883 }
1884
1885 /** Write an message to the multiplexed data stream. */
1886 int io_multiplex_write(enum msgcode code, const char *buf, size_t len, int convert)
1887 {
1888         if (!io_multiplexing_out)
1889                 return 0;
1890         io_flush(NORMAL_FLUSH);
1891         stats.total_written += (len+4);
1892         mplex_write(sock_f_out, code, buf, len, convert);
1893         return 1;
1894 }
1895
1896 void io_end_multiplex_in(void)
1897 {
1898         io_multiplexing_in = 0;
1899         io_end_buffering_in();
1900 }
1901
1902 /** Stop output multiplexing. */
1903 void io_end_multiplex_out(void)
1904 {
1905         io_multiplexing_out = 0;
1906         io_end_buffering_out();
1907 }
1908
1909 void start_write_batch(int fd)
1910 {
1911         /* Some communication has already taken place, but we don't
1912          * enable batch writing until here so that we can write a
1913          * canonical record of the communication even though the
1914          * actual communication so far depends on whether a daemon
1915          * is involved. */
1916         write_int(batch_fd, protocol_version);
1917         if (protocol_version >= 30)
1918                 write_byte(batch_fd, inc_recurse);
1919         write_int(batch_fd, checksum_seed);
1920
1921         if (am_sender)
1922                 write_batch_monitor_out = fd;
1923         else
1924                 write_batch_monitor_in = fd;
1925 }
1926
1927 void stop_write_batch(void)
1928 {
1929         write_batch_monitor_out = -1;
1930         write_batch_monitor_in = -1;
1931 }