1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 Failure to get a lock on a hints database can have serious
10 consequences so log it to the panic log.
12 TF/02 Log LMTP confirmation messages in the same way as SMTP,
13 controlled using the smtp_confirmation log selector.
19 PP/01 Date: & Message-Id: revert to normally being appended to a message,
20 only prepend for the Resent-* case. Fixes regression introduced in
21 Exim 4.70 by NM/22 for Bugzilla 607.
23 PP/02 Include check_rfc2047_length in configure.default because we're seeing
24 increasing numbers of administrators be bitten by this.
26 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
28 PP/03 Bugzilla 994: added openssl_options main configuration option.
30 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
32 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
34 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
36 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
37 without a peer certificate, leading to a segfault because of an
38 assumption that peers always have certificates. Be a little more
39 paranoid. Problem reported by Martin Tscholak.
41 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
42 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
43 NB: ClamAV planning to remove STREAM in "middle of 2010".
44 CL also introduces -bmalware, various -d+acl logging additions and
45 more caution in buffer sizes.
47 PP/09 Implemented reverse_ip expansion operator.
49 PP/10 Bugzilla 937: provide a "debug" ACL control.
51 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
53 PP/12 Bugzilla 973: Implement --version.
55 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
57 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
59 PP/15 Bugzilla 816: support multiple condition rules on Routers.
61 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
62 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
63 ignore trailing whitespace.
65 JJ/02 prevent non-panic DKIM error from being sent to paniclog
67 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
70 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
71 Notification from Dr Andrew Aitchison.
73 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
74 ExtendedDetectionInfo response format.
75 Notification from John Horne.
77 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
80 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
81 XSL and documented dependency on system catalogs, with examples of how
84 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
87 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
88 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
89 configuration file which is writeable by the Exim user or group.
91 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
92 of configuration files to cover files specified with the -C option if
93 they are going to be used with root privileges, not just the default
96 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
97 option (effectively making it always true).
99 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
100 files to be used while preserving root privileges.
102 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
103 that rogue child processes cannot use them.
105 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
106 run-time user, instead of root.
108 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
109 Exim run-time user without dropping privileges.
111 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
112 result string, instead of calling string_vformat() twice with the same
115 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
116 for other users. Others should always drop root privileges if they use
117 -C on the command line, even for a whitelisted configure file.
119 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
121 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
127 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
128 $data_path, and $header_path variables; fixed documentation bugs and
131 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
132 exipick to access non-standard spools, including the "frozen" queue
135 NM/01 Bugzilla 965: Support mysql stored procedures.
136 Patch from Alain Williams
138 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
140 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
141 Patch from Andreas Metzler
143 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
144 Patch from Kirill Miazine
146 NM/05 Bugzilla 671: Added umask to procmail example.
148 JJ/03 installed exipick 20100323.0, fixing doc bug
150 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
151 directory. Notification and patch from Dan Rosenberg.
153 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
155 TK/02 Improve log output when DKIM signing operation fails.
157 MH/01 Treat the transport option dkim_domain as a colon separated
158 list, not as a single string, and sign the message with each element,
159 omitting multiple occurences of the same signer.
161 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
162 Bugzilla 985, 986. Patch by Simon Arlott
164 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
165 Patch by Simon Arlott
167 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
168 MBX locking. Notification from Dan Rosenberg.
174 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
176 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
178 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
180 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
182 NM/04 Bugzilla 915: Flag broken perl installation during build.
188 TK/01 Added patch by Johannes Berg that expands the main option
189 "spamd_address" if it starts with a dollar sign.
191 TK/02 Write list of recipients to X-Envelope-Sender header when building
192 the mbox-format spool file for content scanning (suggested by Jakob
195 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
196 (http://www.dcc-servers.net/) support via dccifd. Activated by
197 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
199 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
200 by Mark Daniel Reidel <mr@df.eu>.
202 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
203 When building exim an external PCRE library is now needed -
204 PCRE is a system library on the majority of modern systems.
205 See entry on PCRE_LIBS in EDITME file.
207 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
208 conversation. Added nologin parameter to request.
209 Patch contributed by Kirill Miazine.
211 TF/01 Do not log submission mode rewrites if they do not change the address.
213 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
215 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
216 log files in place. Contributed by Roberto Lima.
218 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
220 TF/03 Bugzilla 615: When checking the local_parts router precondition
221 after a local_part_suffix or local_part_prefix option, Exim now
222 does not use the address's named list lookup cache, since this
223 contains cached lookups for the whole local part.
225 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
226 Robert Millan. Documentation is in experimental-spec.txt.
228 TF/04 Bugzilla 668: Fix parallel build (make -j).
230 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
232 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
233 Patch provided by Jan Srzednicki.
235 TF/05 Leading white space used to be stripped from $spam_report which
236 wrecked the formatting. Now it is preserved.
238 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
239 that they are available at delivery time.
241 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
243 TF/08 TLS error reporting now respects the incoming_interface and
244 incoming_port log selectors.
246 TF/09 Produce a more useful error message if an SMTP transport's hosts
247 setting expands to an empty string.
249 NM/06 Bugzilla 744: EXPN did not work under TLS.
250 Patch provided by Phil Pennock.
252 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
253 Patch provided by Richard Godbee.
255 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
256 acl_smtp_notquit, added index entry.
258 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
259 Patch provided by Eugene Bujak.
261 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
262 accept(). Patch provided by Maxim Dounin.
264 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
265 Patch provided by Phil Pennock.
267 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
269 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
270 Patch provided by Brad "anomie" Jorsch.
272 NM/14 Bugzilla 622: Added timeout setting to transport filter.
273 Patch provided by Dean Brooks.
275 TK/05 Add native DKIM support (does not depend on external libraries).
277 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
278 Patch provided by Graeme Fowler.
280 NM/16 Bugzilla 851: Documentation example syntax fix.
282 NM/17 Changed NOTICE file to remove references to embedded PCRE.
284 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
287 NM/19 Bugzilla 745: TLS version reporting.
288 Patch provided by Phil Pennock.
290 NM/20 Bugzilla 167: bool: condition support.
291 Patch provided by Phil Pennock.
293 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
294 clients. Patch provided by Phil Pennock.
296 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
297 Patch provided by Brad "anomie" Jorsch.
299 NM/23 Bugzilla 687: Fix misparses in eximstats.
300 Patch provided by Heiko Schlittermann.
302 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
303 Patch provided by Heiko Schlittermann.
305 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
306 plus update to original patch.
308 NM/26 Bugzilla 799: Documentation correction for ratelimit.
310 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
311 Patch provided by David Brownlee.
313 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
315 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
317 NM/30 Bugzilla 888: TLS documentation bugfixes.
319 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
321 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
322 Unlike the original bugzilla I have changed all shell scripts in src tree.
324 NM/33 Bugzilla 898: Transport filter timeout fix.
325 Patch by Todd Rinaldo.
327 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
328 Patch by Serge Demonchaux.
330 NM/35 Bugzilla 39: Base64 decode bug fixes.
331 Patch by Jakob Hirsch.
333 NM/36 Bugzilla 909: Correct connect() call in dcc code.
335 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
337 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
339 NM/39 Bugzilla 911: Fixed MakeLinks build script.
345 TK/01 Add preliminary DKIM support. Currently requires a forked version of
346 ALT-N's libdkim that I have put here:
347 http://duncanthrax.net/exim-experimental/
349 Note to Michael Haardt: I had to rename some vars in sieve.c. They
350 were called 'true' and it seems that C99 defines that as a reserved
351 keyword to be used with 'bool' variable types. That means you could
352 not include C99-style headers which use bools without triggering
353 build errors in sieve.c.
355 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
356 as mailq or other aliases. Changed the --help handling significantly
357 to do whats expected. exim_usage() emits usage/help information.
359 SC/01 Added the -bylocaldomain option to eximstats.
361 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
363 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
365 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
371 PH/01 Another patch from the Sieve maintainer.
373 PH/02 When an IPv6 address is converted to a string for single-key lookup
374 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
375 dots are used instead of colons so that keys in lsearch files need not
376 contain colons. This was done some time before quoting was made available
377 in lsearch files. However, iplsearch files do require colons in IPv6 keys
378 (notated using the quote facility) so as to distinguish them from IPv4
379 keys. This meant that lookups for IP addresses in host lists did not work
380 for iplsearch lookups.
382 This has been fixed by arranging for IPv6 addresses to be expressed with
383 colons if the lookup type is iplsearch. This is not incompatible, because
384 previously such lookups could never work.
386 The situation is now rather anomolous, since one *can* have colons in
387 ordinary lsearch keys. However, making the change in all cases is
388 incompatible and would probably break a number of configurations.
390 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
393 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
394 conversion specification without a maximum field width, thereby enabling
395 a rogue spamd server to cause a buffer overflow. While nobody in their
396 right mind would setup Exim to query an untrusted spamd server, an
397 attacker that gains access to a server running spamd could potentially
398 exploit this vulnerability to run arbitrary code as the Exim user.
400 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
401 $primary_hostname instead of what libspf2 thinks the hosts name is.
403 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
404 a directory entry by the name of the lookup key. Previously, if a
405 symlink pointed to a non-existing file or a file in a directory that
406 Exim lacked permissions to read, a lookup for a key matching that
407 symlink would fail. Now it is enough that a matching directory entry
408 exists, symlink or not. (Bugzilla 503.)
410 PH/03 The body_linecount and body_zerocount variables are now exported in the
413 PH/04 Added the $dnslist_matched variable.
415 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
416 This means they are set thereafter only if the connection becomes
419 PH/06 Added the client_condition to authenticators so that some can be skipped
420 by clients under certain conditions.
422 PH/07 The error message for a badly-placed control=no_multiline_responses left
423 "_responses" off the end of the name.
425 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
427 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
428 (without spaces) instead of just copying the configuration text.
430 PH/10 Added the /noupdate option to the ratelimit ACL condition.
432 PH/11 Added $max_received_linelength.
434 PH/12 Added +ignore_defer and +include_defer to host lists.
436 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
437 way in which PCRE > 7.0 is built.
439 PH/14 Implemented queue_only_load_latch.
441 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
442 MAIL command. The effect was to mangle the value on 64-bit systems.
444 PH/16 Another patch from the Sieve maintainer.
446 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
448 PH/18 If a system quota error occurred while trying to create the file for
449 a maildir delivery, the message "Mailbox is full" was not appended to the
450 bounce if the delivery eventually timed out. Change 4.67/27 below applied
451 only to a quota excession during the actual writing of the file.
453 PH/19 It seems that peer DN values may contain newlines (and other non-printing
454 characters?) which causes problems in log lines. The DN values are now
455 passed through string_printing() before being added to log lines.
457 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
458 and InterBase are left for another time.)
460 PH/21 Added message_body_newlines option.
462 PH/22 Guard against possible overflow in moan_check_errorcopy().
464 PH/23 POSIX allows open() to be a macro; guard against that.
466 PH/24 If the recipient of an error message contained an @ in the local part
467 (suitably quoted, of course), incorrect values were put in $domain and
468 $local_part during the evaluation of errors_copy.
474 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
475 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
478 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
479 issue a MAIL command.
481 PH/02 In an ACL statement such as
483 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
485 if a client was not listed at all, or was listed with a value other than
486 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
487 the condition was not true (as it should be), so access was not denied.
488 The bug was that the ! inversion was incorrectly passed on to the second
489 item. This has been fixed.
491 PH/03 Added additional dnslists conditions == and =& which are different from
492 = and & when the dns lookup returns more than one IP address.
494 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
495 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
497 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
498 FSYNC, which compiles an option called disable_fsync that allows for
499 bypassing fsync(). The documentation is heavily laced with warnings.
501 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
503 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
504 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
505 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
506 including adding "make clean"; (3) Added -fPIC when compiling the test
507 dynamically loaded module, to get rid of a warning.
509 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
510 message fails, move_frozen_messages = true and ignore_bounce_errors_after
511 = 0s. The bug is otherwise harmless.
513 PH/07 There was a bug in the dovecot authenticator such that the value of
514 $auth1 could be overwritten, and so not correctly preserved, after a
515 successful authentication. This usually meant that the value preserved by
516 the server_setid option was incorrect.
518 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
520 PH/09 Installed PCRE release 7.0.
522 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
523 run for batched SMTP input. It is now run at the start of every message
524 in the batch. While fixing this I discovered that the process information
525 (output by running exiwhat) was not always getting set for -bs and -bS
526 input. This is fixed, and it now also says "batched" for BSMTP.
528 PH/11 Added control=no_pipelining.
530 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
531 patch, slightly modified), and move the expansion of helo_data till after
532 the connection is made in the smtp transport (so it can use these
535 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
537 PH/14 Added log_selector = +pid.
539 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
541 PH/16 Add ${if forany and ${if forall.
543 PH/17 Added dsn_from option to vary the From: line in DSNs.
545 PH/18 Flush SMTP output before performing a callout, unless control =
546 no_callout_flush is set.
548 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
549 was true (the default) a successful delivery failed to delete the retry
550 item, thus causing premature timeout of the address. The bug is now
553 PH/20 Added hosts_avoid_pipelining to the smtp transport.
555 PH/21 Long custom messages for fakedefer and fakereject are now split up
556 into multiline reponses in the same way that messages for "deny" and
557 other ACL rejections are.
559 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
560 with slight modification.
562 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
563 draft, changing the syntax and factoring some duplicate code".
565 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
566 for deliveries of the second and subsequent messages over the same SMTP
569 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
570 ${reduce, with only minor "tidies".
572 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
574 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
575 expansion side effects.
577 PH/27 When a message times out after an over-quota error from an Exim-imposed
578 quota, the bounce message says "mailbox is full". This message was not
579 being given when it was a system quota that was exceeded. It now should
582 MH/03 Made $recipients available in local_scan(). local_scan() already has
583 better access to the recipient list through recipients_list[], but
584 $recipients can be useful in postmaster-provided expansion strings.
586 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
587 in the case of a MAIL command with additional options following the
588 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
589 were accidentally chopped off.
591 PH/29 SMTP synchronization checks are implemented when a command is read -
592 there is a check that no more input is waiting when there shouldn't be
593 any. However, for some commands, a delay in an ACL can mean that it is
594 some time before the response is written. In this time, more input might
595 arrive, invalidly. So now there are extra checks after an ACL has run for
596 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
597 pipelining has not been advertised.
599 PH/30 MH's patch to allow iscntrl() characters to be list separators.
601 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
602 returned in the SMTP response when smtp_return_error_details was false.
605 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
606 instead of the C I/O that was originally supplied, because problems were
609 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
610 Exim which did not show up earlier: it was assuming that a call to
611 SSL_CTX_set_info_callback() might give an error value. In fact, there is
612 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
613 was a macro that became an assignment, so it seemed to work. This has
614 changed to a proper function call with a void return, hence the compile
615 error. Exim's code has been fixed.
617 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
620 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
622 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
624 PH/37 If a message is not accepted after it has had an id assigned (e.g.
625 because it turns out to be too big or there is a timeout) there is no
626 "Completed" line in the log. When some messages of this type were
627 selected by exigrep, they were listed as "not completed". Others were
628 picked up by some special patterns. I have improved the selection
629 criteria to be more general.
631 PH/38 The host_find_failed option in the manualroute router can now be set
632 to "ignore", to completely ignore a host whose IP address cannot be
633 found. If all hosts are ignored, the behaviour is controlled by the new
634 host_all_ignored option.
636 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
637 homing or because of multiple MX records with /mx) generated more than
638 one IP address, and the following item turned out to be the local host,
639 all the secondary addresses of the first item were incorrectly removed
640 from the list, along with the local host and any following hosts (which
641 is what is supposed to happen).
643 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
644 whoever called Exim into the -H file. In the case of the daemon it was
645 behaving confusingly. When first started, it used values for whoever
646 started the daemon, but after a SIGHUP it used the Exim user (because it
647 calls itself on a restart). I have changed the code so that it now always
650 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
651 message are rejected with the same error (e.g. no authentication or bad
652 sender address), and a DATA command is nevertheless sent (as can happen
653 with PIPELINING or a stupid MUA), the error message that was given to the
654 RCPT commands is included in the rejection of the DATA command. This is
655 intended to be helpful for MUAs that show only the final error to their
658 PH/42 Another patch from the Sieve maintainer.
660 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
661 Eximstats - Fixed some broken HTML links and added missing column headers
663 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
664 columns for Rejects, Temp Rejects, Ham, and Spam rows.
666 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
668 PH/43 Yet another patch from the Sieve maintainer.
670 PH/44 I found a way to check for a TCP/IP connection going away before sending
671 the response to the final '.' that terminates a message, but only in the
672 case where the client has not sent further data following the '.'
673 (unfortunately, this is allowed). However, in many cases there won't be
674 any further data because there won't be any more messages to send. A call
675 to select() can be used: if it shows that the input is "ready", there is
676 either input waiting, or the socket has been closed. An attempt to read
677 the next input character can distinguish the two cases. Previously, Exim
678 would have sent an OK response which the client would never have see.
679 This could lead to message repetition. This fix should cure that, at
680 least in a lot of common cases.
682 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
683 advertised in response to EHLO.
689 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
690 fixed by 4.65/MH/01 (is this a record?) are fixed:
692 (i) An empty string was always treated as zero by the numeric comparison
693 operators. This behaviour has been restored.
695 (ii) It is documented that the numeric comparison operators always treat
696 their arguments as decimal numbers. This was broken in that numbers
697 starting with 0 were being interpreted as octal.
699 While fixing these problems I realized that there was another issue that
700 hadn't been noticed. Values of message_size_limit (both the global option
701 and the transport option) were treated as octal if they started with 0.
702 The documentation was vague. These values are now always treated as
703 decimal, and I will make that clear in the documentation.
709 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
710 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
713 MH/01 Don't check that the operands of numeric comparison operators are
714 integers when their expansion is in "skipping" mode (fixes bug
715 introduced by 4.64-PH/07).
717 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
718 child addresses, Exim now panics and dies. Previously, because the count
719 is held in a short int, deliveries were likely to be lost. As such a
720 large number of recipients for a single message is ridiculous
721 (performance will be very, very poor), I have chosen to impose a limit
722 rather than extend the field.
728 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
729 leftover -K file (the existence of which was triggered by #402).
730 While we were at it, introduced process PID as part of the -K
731 filename. This should rule out race conditions when creating
734 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
735 processing considerably. Previous code took too long for large mails,
736 triggering a timeout which in turn triggers #401.
738 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
739 in the DK code in transports.c. sendfile() is not really portable,
740 hence the _LINUX specificness.
742 TF/01 In the add_headers option to the mail command in an Exim filter,
743 there was a bug that Exim would claim a syntax error in any
744 header after the first one which had an odd number of characters
747 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
748 callout verification, Exim cached a "reject" for the entire domain. This
749 is correct for most verifications, but it is not correct for a recipient
750 verification with use_sender or use_postmaster set, because in that case
751 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
752 case of MAIL FROM:<> rejection from other early rejections (e.g.
753 rejection of HELO). When verifying a recipient using a non-null MAIL
754 address, the cache is ignored if it shows MAIL FROM:<> rejection.
755 Whatever the result of the callout, the value of the domain cache is
756 left unchanged (for any other kind of callout, getting as far as trying
757 RCPT means that the domain itself is ok).
759 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
762 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
763 manifest itself as EPIPE rather than ECONNECT. When tidying away a
764 session, the daemon ignores ECONNECT errors and logs others; it now
765 ignores EPIPE as well.
767 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
768 (quoted-printable decoding).
770 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
771 later the small subsequent patch to fix an introduced bug.
773 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
775 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
777 PH/08 An error is now given if message_size_limit is specified negative.
779 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
780 to be given (somewhat) arbitrary names.
782 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
785 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
786 miscellaneous code fixes
788 PH/10 Added the log_reject_target ACL modifier to specify where to log
791 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
792 hostname. This is wrong, because it relates to the incoming message (and
793 probably the interface on which it is arriving) and not to the outgoing
794 callout (which could be using a different interface). This has been
795 changed to use the value of the helo_data option from the smtp transport
796 instead - this is what is used when a message is actually being sent. If
797 there is no remote transport (possible with a router that sets up host
798 addresses), $smtp_active_hostname is used.
800 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
801 tweaks were necessary in order to get it to work (see also 21 below):
802 (a) The code assumed that strncpy() returns a negative number on buffer
803 overflow, which isn't the case. Replaced with Exim's string_format()
805 (b) There were several signed/unsigned issues. I just did the minimum
806 hacking in of casts. There is scope for a larger refactoring.
807 (c) The code used strcasecmp() which is not a standard C function.
808 Replaced with Exim's strcmpic() function.
809 (d) The code set only $1; it now sets $auth1 as well.
810 (e) A simple test gave the error "authentication client didn't specify
811 service in request". It would seem that Dovecot has changed its
812 interface. Fortunately there's a specification; I followed it and
813 changed what the client sends and it appears to be working now.
815 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
818 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
819 address A is aliased to B and C, where B exists and C does not. Without
820 -v the output is "A verified" because verification stops after a
821 successful redirection if more than one address is generated. However,
822 with -v the child addresses are also verified. Exim was outputting "A
823 failed to verify" and then showing the successful verification for C,
824 with its parentage. It now outputs "B failed to verify", showing B's
825 parentage before showing the successful verification of C.
827 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
828 look up a TXT record in a specific list after matching in a combined
831 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
832 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
833 they consult the DNS. I had assumed they would set it the way they
834 wanted; and indeed my experiments on Linux seem to show that in some
835 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
836 To be on the safe side, however, I have now made the interface to
837 host_find_byname() similar to host_find_bydns(), with an argument
838 containing the DNS resolver options. The host_find_byname() function now
839 sets these options at its start, just as host_find_bydns() does. The smtp
840 transport options dns_qualify_single and dns_search_parents are passed to
841 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
842 of host_find_byname() use the default settings of RES_DEFNAMES
843 (qualify_single) but not RES_DNSRCH (search_parents).
845 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
846 spool_read_header() do less string testing, by means of a preliminary
847 switch on the second character of optional "-foo" lines. (This is
848 overdue, caused by the large number of possibilities that now exist.
849 Originally there were few.) While I was there, I also converted the
850 str(n)cmp tests so they don't re-test the leading "-" and the first
851 character, in the hope this might squeeze out yet more improvement.
853 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
854 flag allowing group syntax was set by the header_syntax check but not
855 turned off, possible causing trouble later; (2) The flag was not being
856 set at all for the header_verify test, causing "group"-style headers to
857 be rejected. I have now set it in this case, and also caused header_
858 verify to ignore an empty address taken from a group. While doing this, I
859 came across some other cases where the code for allowing group syntax
860 while scanning a header line wasn't quite right (mostly, not resetting
861 the flag correctly in the right place). These bugs could have caused
862 trouble for malformed header lines. I hope it is now all correct.
864 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
865 with the "reply" argument non-NULL. The code, however (which originally
866 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
867 but it didn't always do it. This confused somebody who was copying the
868 code for some other use. I have removed all the tests.
870 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
871 feature that was used to support insecure browsers during the U.S. crypto
872 embargo. It requires special client support, and Exim is probably the
873 only MTA that supported it -- and would never use it because real RSA is
874 always available. This code has been removed, because it had the bad
875 effect of slowing Exim down by computing (never used) parameters for the
876 RSA_EXPORT functionality.
878 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
879 authenticator to fail if there's a tab character in the incoming data
880 (there should never be unless someone is messing about, as it's supposed
881 to be base64-encoded). Also added, on Timo's advice, the "secured" option
882 if the connection is using TLS or if the remote IP is the same as the
883 local IP, and the "valid-client-cert option" if a client certificate has
886 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
887 authenticators. This can be used for authorization after authentication
888 succeeds. (In the case of plaintext, it servers for both authentication
891 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
892 if any retry times were supplied.
894 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
895 connection, where there is no client IP address to check. In this
896 situation, the verify now always succeeds.
898 PH/25 Applied John Jetmore's -Mset patch.
900 PH/26 Added -bem to be like -Mset, but loading a message from a file.
902 PH/27 In a string expansion for a processed (not raw) header when multiple
903 headers of the same name were present, leading whitespace was being
904 removed from all of them, but trailing whitespace was being removed only
905 from the last one. Now trailing whitespace is removed from each header
906 before concatenation. Completely empty headers in a concatenation (as
909 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
910 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
912 PH/29 [Removed. This was a change that I later backed out, and forgot to
913 correct the ChangeLog entry (that I had efficiently created) before
914 committing the later change.]
916 PH/30 Exim was sometimes attempting to deliver messages that had suffered
917 address errors (4xx response to RCPT) over the same connection as other
918 messages routed to the same hosts. Such deliveries are always "forced",
919 so retry times are not inspected. This resulted in far too many retries
920 for the affected addresses. The effect occurred only when there were more
921 hosts than the hosts_max_try setting in the smtp transport when it had
922 the 4xx errors. Those hosts that it had tried were not added to the list
923 of hosts for which the message was waiting, so if all were tried, there
924 was no problem. Two fixes have been applied:
926 (i) If there are any address or message errors in an SMTP delivery, none
927 of the hosts (tried or untried) are now added to the list of hosts
928 for which the message is waiting, so the message should not be a
929 candidate for sending over the same connection that was used for a
930 successful delivery of some other message. This seems entirely
931 reasonable: after all the message is NOT "waiting for some host".
932 This is so "obvious" that I'm not sure why it wasn't done
933 previously. Hope I haven't missed anything, but it can't do any
934 harm, as the worst effect is to miss an optimization.
936 (ii) If, despite (i), such a delivery is accidentally attempted, the
937 routing retry time is respected, so at least it doesn't keep
938 hammering the server.
940 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
941 in ${readsocket because some servers need this prod.
943 PH/32 Added some extra debug output when updating a wait-xxx database.
945 PH/33 The hint "could be header name not terminated by colon", which has been
946 given for certain expansion errors for a long time, was not being given
947 for the ${if def:h_colon_omitted{... case.
949 PH/34 The spec says: "With one important exception, whenever a domain list is
950 being scanned, $domain contains the subject domain." There was at least
951 one case where this was not true.
953 PH/35 The error "getsockname() failed: connection reset by peer" was being
954 written to the panic log as well as the main log, but it isn't really
955 panic-worthy as it just means the connection died rather early on. I have
956 removed the panic log writing for the ECONNRESET error when getsockname()
959 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
960 runs only) independently of the message's sender address. This meant
961 that, if the 4xx error was in fact related to the sender, a different
962 message to the same recipient with a different sender could confuse
963 things. In particualar, this can happen when sending to a greylisting
964 server, but other circumstances could also provoke similar problems.
965 I have changed the default so that the retry time for these errors is now
966 based a combination of the sender and recipient addresses. This change
967 can be overridden by setting address_retry_include_sender=false in the
970 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
971 remote server are returned as part of bounce messages. This was not
972 happening for LMTP over a pipe (the lmtp transport), but now it is the
973 same for both kinds of LMTP.
975 PH/38 Despite being documented as not happening, Exim was rewriting addresses
976 in header lines that were in fact CNAMEs. This is no longer the case.
978 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
979 and queue runs started by the daemon processed all messages. This has
980 been fixed so that -R and -S can now usefully be given with -q<time>.
982 PH/40 Import PCRE release 6.7 (fixes some bugs).
984 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
986 PH/42 Give an error if -q is specified more than once.
988 PH/43 Renamed the variables $interface_address and $interface_port as
989 $received_ip_address and $received_port, to make it clear that these
990 values apply to message reception, and not to the outgoing interface when
991 a message is delivered. (The old names remain recognized, of course.)
993 PH/44 There was no timeout on the connect() call when using a Unix domain
994 socket in the ${readsocket expansion. There now is.
996 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
997 be meaningful with "accept".
999 SC/01 Eximstats V1.43
1000 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1002 SC/02 Eximstats V1.44
1003 Use a glob alias rather than an array ref in the generated
1004 parser. This improves both readability and performance.
1006 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1007 Collect SpamAssassin and rejection statistics.
1008 Don't display local sender or destination tables unless
1009 there is data to show.
1010 Added average volumes into the top table text output.
1012 SC/04 Eximstats V1.46
1013 Collect data on the number of addresses (recipients)
1014 as well as the number of messages.
1016 SC/05 Eximstats V1.47
1017 Added 'Message too big' to the list of mail rejection
1018 reasons (thanks to Marco Gaiarin).
1020 SC/06 Eximstats V1.48
1021 Mainlog lines which have GMT offsets and are too short to
1022 have a flag are now skipped.
1024 SC/07 Eximstats V1.49 (Alain Williams)
1025 Added the -emptyok flag.
1027 SC/08 Eximstats V1.50
1028 Fixes for obtaining the IP address from reject messages.
1030 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1031 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1032 whitesspace changes from 4.64-PH/27
1034 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1037 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1038 are found, allow negative numbers in numeric criteria)
1040 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1042 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1043 to match changes made in 4.64-PH/43
1045 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1047 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1048 "domain" is actually an IP address, and force a failure. This locks out
1049 those revolvers/nameservers that support "A-for-A" lookups, in
1050 contravention of the specifications.
1052 PH/48 When a host name was looked up from an IP address, and the subsequent
1053 forward lookup of the name timed out, the host name was left in
1054 $sender_host_name, contrary to the specification.
1056 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1057 restricted to single-key lookups, Exim was not diagnosing an error if
1058 * or *@ was used with a query-style lookup.
1060 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1062 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1063 long ago, but noone interested enough thought of it. Let's just say that
1064 the "1.1" means that there are some new functions that weren't there at
1065 some point in the past.
1067 PH/51 Error processing for expansion failure of helo_data from an smtp
1068 transport during callout processing was broken.
1070 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1071 tested/used via the -bh/-bhc/-bs options.
1073 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1074 bug, fixed in subsequent PCRE releases).
1076 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1077 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1079 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1085 SC/01 Use a glob alias rather than an array ref in eximstats generated
1086 parser. This improves both readability and performance.
1088 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1089 Don't display local sender or destination tables in eximstats unless
1090 there is data to show.
1091 Added average volumes into the eximstats top table text output.
1093 SC/03 Collect data on the number of addresses (recipients) as well
1094 as the number of messages in eximstats.
1096 TF/01 Correct an error in the documentation for the redirect router. Exim
1097 does (usually) call initgroups() when daemonizing.
1099 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1100 with consistent privilege compared to when running as a daemon.
1102 TF/03 Note in the spec that $authenticated_id is not set for local
1103 submissions from trusted users.
1105 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1106 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1108 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1109 by adding some example configuration directives to the default
1110 configuration file. A little bit of work is required to uncomment the
1111 directives and define how usernames and passwords are checked, but
1112 there is now a framework to start from.
1114 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1115 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1116 without this. I don't know how relevant this is to other LDAP libraries.
1118 PH/02 Add the verb name to the "unknown ACL verb" error.
1120 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1122 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1124 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1125 directory not expanded when it should be if an expanded home directory
1126 was set for the address (which is overridden by the transport).
1128 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1131 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1132 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1133 because it is too late at that time, and has no effect.
1135 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1136 security issue with \' (bugzilla #107). I could not use the
1137 PQescapeStringConn() function, because it needs a PGconn value as one of
1140 PH/08 When testing addresses using -bt, indicate those final addresses that
1141 are duplicates that would not cause an additional delivery. At least one
1142 person was confused, thinking that -bt output corresponded to deliveries.
1143 (Suppressing duplicates isn't a good idea as you lose the information
1144 about possibly different redirections that led to the duplicates.)
1146 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1147 systems where poll() doesn't work, in particular OS X.
1149 PH/10 Added more information to debugging output for retry time not reached.
1151 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1152 operations in malware.c.
1154 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1157 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1158 syslog_duplication set false, log lines that would normally be written
1159 both the the main log and to the reject log were not written to syslog at
1162 PH/14 In the default configuration, change the use of "message" in ACL warn
1163 statements to "add_header".
1165 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1166 not followed by a command (e.g. "seen endif").
1168 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1169 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1172 PH/17 Added extra conditions to the default value of delay_warning_condition
1176 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1177 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1178 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1181 The Auto-Submitted: and various List- headers are standardised, whereas I
1182 don't think Precedence: ever was.
1184 PH/18 Refactored debugging code in route_finduser() to show more information,
1185 in particular, the error code if getpwnam() issues one.
1187 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1188 This is apparently needed in addition to the PH/07 change above to avoid
1189 any possible encoding problems.
1191 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1192 but not after initializing Perl.
1194 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1195 output them only if debugging. By default they are written stderr,
1196 apparently, which is not desirable.
1198 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1201 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1204 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1206 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1207 authenticated or an ident call has been made. Suppress the default
1208 values for $authenticated_id and $authenticated_sender (but permit -oMai
1209 and -oMas) when testing with -bh.
1211 PH/24 Re-jigged the order of the tests in the default configuration so that the
1212 tests for valid domains and recipients precede the DNS black list and CSA
1213 tests, on the grounds that those ones are more expensive.
1215 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1216 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1217 This bug exists in every version of Exim that I still have, right back to
1220 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1221 However, an attempt to turn on case-sensitivity in a regex key by
1222 including (?-i) didn't work because the subject string was already
1223 lowercased, and the effects were non-intuitive. It turns out that a
1224 one-line patch can be used to allow (?-i) to work as expected.
1230 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1231 other effects) broke the use of negated acl sub-conditions.
1233 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1236 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1237 "Deny" causes Exim to reject the incoming connection with a 554 error.
1238 Unfortunately, if there is a major crisis, such as a disk failure,
1239 tcp-wrappers gives "deny", whereas what one would like would be some
1240 kind of temporary error. A kludge has been added to help with this.
1241 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1242 554 error is used if errno is still zero or contains ENOENT (which occurs
1243 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1246 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1248 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1249 errors. Otherwise a message that provokes a temporary error (when other
1250 messages do not) can cause a whole host to time out.
1252 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1253 addresses were routed directly to files or pipes from a redirect router.
1254 File deliveries just didn't batch; pipe deliveries might have suffered
1257 PH/06 A failure to get a lock for a hints database would erroneously always say
1258 "Failed to get write lock", even when it was really a read lock.
1260 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1261 of 0600. This has been changed to use the value of the lockfile_mode
1262 option (which defaults to 0600).
1264 PH/08 Applied small patch from the Sieve maintainer.
1266 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1267 folder from quota calculations, a direct delivery into this folder messed
1268 up the contents of the maildirsize file. This was because the regex was
1269 used only to exclude .Trash (or whatever) when the size of the mailbox
1270 was calculated. There was no check that a delivery was happening into an
1271 excluded directory. This bug has been fixed by ignoring all quota
1272 processing for deliveries into excluded directories.
1274 PH/10 Added the maildirfolder_create_regex option to appendfile.
1280 PH/01 The code for finding all the local interface addresses on a FreeBSD
1281 system running IPv6 was broken. This may well have applied to all BSD
1282 systems, as well as to others that have similar system calls. The broken
1283 code found IPv4 interfaces correctly, but gave incorrect values for the
1284 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1285 that it would not match correctly against @[] and not recognize the IPv6
1288 PH/02 The ipliteral router was not recognizing addresses of the form user@
1289 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1291 PH/03 Added disable_ipv6.
1293 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1294 decoded form, because it is most often used to construct To: headers
1295 lines in autoreplies, and the decoded form may well be syntactically
1296 invalid. However, $reply_address has leading white space removed, and all
1297 newlines turned into spaces so that the autoreply transport does not
1300 PH/05 If group was specified without a user on a router, and no group or user
1301 was specified on a transport, the group from the router was ignored.
1303 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1304 for visible compile-time settings that can be used to change these
1305 numbers, for those that want even more. Backwards compatibility with old
1306 spool files has been maintained. However, going back to a previous Exim
1307 release will lost any variables that are in spool files.
1309 PH/07 Two small changes when running in the test harness: increase delay when
1310 passing a TCP/IP connection to a new process, in case the original
1311 process has to generate a bounce, and remove special handling of
1312 127.0.0.2 (sic), which is no longer necessary.
1314 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1315 be the same on different OS.
1317 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1320 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1321 whether --show-vars was specified or not
1323 JJ/02 exipick: Added support for new ACL variable spool format introduced
1326 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1327 syntactically invalid From: or Reply-to: line, and a filter used this to
1328 generate an autoreply, and therefore failed to obtain an address for the
1329 autoreply, Exim could try to deliver to a non-existent relative file
1330 name, causing unrelated and misleading errors. What now happens is that
1331 it logs this as a hard delivery error, but does not attempt to create a
1334 PH/11 The exinext utility has a -C option for testing purposes, but although
1335 the given file was scanned by exinext itself; it wasn't being passed on
1336 when Exim was called.
1338 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1339 an end-of-file indication when reading a command response.
1341 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1342 compiled. In many other places in Exim, IPv6 addresses are always
1343 recognized, so I have changed this. It also means that IPv4 domain
1344 literals of the form [IPV4:n.n.n.n] are now always recognized.
1346 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1347 used if the router is not running as root, for example, when verifying at
1348 ACL time, or when using -bh. The debugging output from this situation was
1349 non-existent - all you got was a failure to exec. I have made two
1352 (a) Failures to set uid/gid, the current directory, or a process leader
1353 in a subprocess such as that created by queryprogram now generate
1354 suitable debugging ouput when -d is set.
1356 (b) The queryprogram router detects when it is not running as root,
1357 outputs suitable debugging information if -d is set, and then runs
1358 the subprocess without attempting to change uid/gid.
1360 PH/15 Minor change to Makefile for building test_host (undocumented testing
1363 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1364 additional section of a DNS packet that returns MX or SRV records.
1365 Instead, it always explicitly searches for A/AAAA records. This avoids
1366 major problems that occur when a DNS server includes only records of one
1367 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1368 fixed another bug: if SRV records were looked up and the corresponding
1369 address records were *not* found in the additional section, the port
1370 values from the SRV records were lost.
1372 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1373 using the correct key (the original address) when searching the retry
1374 rules in order to find which one to use for generating the retry hint.
1376 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1377 adding the default one. Similarly, if it contains a Reply-To: header, the
1378 errors_reply_to option, if set, is not used.
1380 PH/19 When calculating a retry time, Exim used to measure the "time since
1381 failure" by looking at the "first failed" field in the retry record. Now
1382 it does not use this if it is later than than the arrival time of the
1383 message. Instead it uses the arrival time. This makes for better
1384 behaviour in cases where some deliveries succeed, thus re-setting the
1385 "first failed" field. An example is a quota failure for a huge message
1386 when small messages continue to be delivered. Without this change, the
1387 "time since failure" will always be short, possible causing more frequent
1388 delivery attempts for the huge message than are intended.
1389 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1391 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1392 $1, $2, $3) because the numerical variables can be reset during some
1393 expansion items (e.g. "match"), thereby losing the authentication data.
1395 PH/21 Make -bV show the size of off_t variables so that the test suite can
1396 decide whether to run tests for quotas > 2G.
1398 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1399 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1400 filecount value is greater than 2G or if a quota value is greater than 2G
1401 on a system where the size of off_t is not greater than 4, a panic error
1404 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1405 never match. The debug and -bh output now contains an explicit error
1406 message indicating a malformed IPv4 address or mask.
1408 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1409 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1410 PH/23 above applies.
1412 PH/25 Do not write to syslog when running in the test harness. The only
1413 occasion when this arises is a failure to open the main or panic logs
1414 (for which there is an explicit test).
1416 PH/26 Added the /no_tell option to "control=freeze".
1418 PH/27 If a host name lookup failed very early in a connection, for example, if
1419 the IP address matched host_lookup and the reverse lookup yielded a name
1420 that did not have a forward lookup, an error message of the form "no IP
1421 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1422 could be logged. Now it outputs the IP address instead of "NULL".
1424 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1425 allows the sender and the authenticated sender to be set when
1426 submitting a message from within Exim. Since child_open_exim() is
1427 documented for local_scan(), the new function should be too.
1429 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1430 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1431 results in an empty string is now treated as unset.
1433 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1435 PH/31 Added sender_verify_fail logging option.
1437 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1438 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1439 tidied the source and removed it altogether.
1441 PH/33 When a queue run was abandoned because the load average was too high, a
1442 log line was always written; now it is written only if the queue_run log
1443 selector is set. In addition, the log line for abandonment now contains
1444 information about the queue run such as the pid. This is always present
1445 in "start" and "stop" lines but was omitted from the "abandon" line.
1447 PH/34 Omit spaces between a header name and the colon in the error message that
1448 is given when verify = headers_syntax fails (if there are lots of them,
1449 the message gets confusing).
1451 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1452 names, as there are now some PTR records that contain slashes. This check
1453 is only to protect against broken name servers that fall over on strange
1454 characters, so the fact that it applies to all lookups doesn't matter.
1456 PH/36 Now that the new test suite is complete, we can remove some of the
1457 special code in Exim that was needed for the old test suite. For example,
1458 sorting DNS records because real resolvers return them in an arbitrary
1459 order. The new test suite's fake resolver always returns records in the
1462 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1463 bounces) except when queue_only is set, to avoid logging races between
1464 the different processes.
1466 PH/38 Panic-die if .include specifies a non-absolute path.
1468 PH/39 A tweak to the "H" retry rule from its user.
1470 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1471 a label. They prevented compilation on older perls.
1473 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1474 a warning to be raised on newish perls.
1476 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1477 on queue. Changes to match documented behaviour of showing count of
1478 messages matching specified criteria.
1480 PH/40 Changed the default ident timeout from 30s to 5s.
1482 PH/41 Added support for the use of login_cap features, on those BSD systems
1483 that have them, for controlling the resources used by pipe deliveries.
1485 PH/42 The content-scanning code uses fopen() to create files in which to put
1486 message data. Previously it was not paying any attention to the mode of
1487 the files. Exim runs with umask(0) because the rest of the code creates
1488 files with open(), and sets the required mode explicitly. Thus, these
1489 files were ending up world-writeable. This was not a big issue, because,
1490 being within the spool directory, they were not world-accessible. I have
1491 created a function called modefopen, which takes an additional mode
1492 argument. It sets umask(777), creates the file, chmods it to the required
1493 mode, then resets the umask. All the relevant calls to fopen() in the
1494 content scanning code have been changed to use this function.
1496 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1497 to 24 hours. This avoids potential overflow problems when processing G
1498 and H retry rules. I suspect nobody ever tinkers with this value.
1500 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1502 PH/45 When the plaintext authenticator is running as a client, the server's
1503 challenges are checked to ensure they are valid base64 strings. By
1504 default, the authentication attempt is cancelled if an invalid string is
1505 received. Setting client_ignore_invalid_base64 true ignores these errors.
1506 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1507 they are received. Thus, the responses can be made to depend on the
1508 challenges. If an invalid string is ignored, an empty string is placed in
1511 PH/46 Messages that are created by the autoreply transport now contains a
1512 References: header, in accordance with RFCs 2822 and 3834.
1514 PH/47 Added authenticated_sender_force to the smtp transport.
1516 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1518 PH/49 Installed latest patch from the Sieve maintainer.
1520 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1521 was also set, the appendfile transport was unnecessarily scanning a
1522 directory of message files (e.g. for maildir delivery) to find the count
1523 of files (along with the size), even though it did not need this
1524 information. It now does the scan only if it needs to find either the
1525 size of the count of files.
1527 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1529 PH/52 Two bugs concerned with error handling when the smtp transport is
1532 (i) Exim was not creating retry information for temporary errors given
1533 for individual recipients after the DATA command when the smtp transport
1534 was used in LMTP mode. This meant that they could be retried too
1535 frequently, and not timed out correctly.
1537 (ii) Exim was setting the flag that allows error details to be returned
1538 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1539 recipients that were returned after the DATA command.
1541 PH/53 This is related to PH/52, but is more general: for any failing address,
1542 when detailed error information was permitted to be returned to the
1543 sender, but the error was temporary, then after the final timeout, only
1544 "retry timeout exceeded" was returned. Now it returns the full error as
1545 well as "retry timeout exceeded".
1547 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1548 do this, and (what is worse) MTAs that accept it.
1550 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1551 will now be deprecated.
1553 PH/56 New os.c-cygwin from the Cygwin maintainer.
1555 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1556 formats (previously only available in exim formats via -bpr, -bpru,
1557 and -bpra. Now also available in native and exiqgrep formats)
1559 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1560 with very large, slow to parse queues
1562 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1564 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1566 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1567 responses to authentication challenges, though it was showing the
1568 challenges; (ii) I've removed the CR characters from the debug output for
1571 PH/58 Allow for the insertion of a newline as well as a space when a string
1572 is turned into more than one encoded-word during RFC 2047 encoding. The
1573 Sieve code now uses this.
1575 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1576 data_4xx, lost_connection, tls_required.
1578 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1579 message was being sent as an SMTP response.
1581 PH/61 Add -l and -k options to exicyclog.
1583 PH/62 When verifying, if an address was redirected to one new address, so that
1584 verification continued, and the new address failed or deferred after
1585 having set something in $address_data, the value of $address_data was not
1586 passed back to the ACL. This was different to the case when no
1587 redirection occurred. The value is now passed back in both cases.
1589 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1590 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1591 use login_cap.h, so on its own it isn't the distinguishing feature. The
1592 new name refers directly to the setclassresources() function.
1594 PH/65 Added configuration files for NetBSD3.
1596 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1598 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1599 is preferred over IPv4.
1601 PH/68 The bounce_return_message and bounce_return_body options were not being
1602 honoured for bounces generated during the reception of non-SMTP messages.
1603 In particular, this applied to messages rejected by the ACL. This bug has
1604 been fixed. However, if bounce_return_message is true and bounce_return_
1605 body is false, the headers that are returned for a non-SMTP message
1606 include only those that have been read before the error was detected.
1607 (In the case of an ACL rejection, they have all been read.)
1609 PH/69 The HTML version of the specification is now built in a directory called
1610 spec_html instead of spec.html, because the latter looks like a path with
1611 a MIME-type, and this confuses some software.
1613 PH/70 Catch two compiler warnings in sieve.c.
1615 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1616 function verify_get_ident() calls ip_connect() to connect a socket, but
1617 if the "connect()" function timed out, ip_connect() used to close the
1618 socket. However, verify_get_ident() also closes the socket later, and in
1619 between Exim writes to the log, which may get opened at this point. When
1620 the socket was closed in ip_connect(), the log could get the same file
1621 descriptor number as the socket. This naturally causes chaos. The fix is
1622 not to close the socket in ip_connect(); the socket should be closed by
1623 the function that creates it. There was only one place in the code where
1624 this was missing, in the iplookup router, which I don't think anybody now
1625 uses, but I've fixed it anyway.
1627 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1628 well as to direct DNS lookups. Otherwise the handling of names in host
1629 lists is inconsistent and therefore confusing.
1635 PH/01 Two changes to the default runtime configuration:
1637 (1) Move the checks for relay_from_hosts and authenticated clients from
1638 after to before the (commented out) DNS black list checks.
1640 (2) Add control=submission to the relay_from_hosts and authenticated
1641 clients checks, on the grounds that messages accepted by these
1642 statements are most likely to be submissions.
1644 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1646 (1) Generate an error if the third argument for the ${prvs expansion is
1649 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1652 (3) Reset the variables that are obtained from the first argument of
1653 ${prvscheck and used in the second argument before leaving the code,
1654 because their memory is reclaimed, so using them afterwards may do
1657 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1658 one (it's much easier than Tom thought :-).
1660 (5) Because of (4), we can now allow for the use of $prvscheck_result
1661 inside the third argument.
1663 PH/03 For some reason, the default setting of PATH when running a command from
1664 a pipe transport was just "/usr/bin". I have changed it to
1667 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1668 anything to be listed in the output from -bV.
1670 PH/05 When a filter generated an autoreply, the entire To: header line was
1671 quoted in the delivery log line, like this:
1673 => >A.N.Other <ano@some.domain> <original@ddress> ...
1675 This has been changed so that it extracts the operative address. There
1676 may be more than one such address. If so, they are comma-separated, like
1679 => >ano@some.domain,ona@other.domain <original@ddress> ...
1681 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1682 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1683 not being looked up in the rDNS to get a host name, Exim was showing the
1684 IP address twice in Received: lines, even though the IP addresses were
1685 identical. For example:
1687 Received: from [1.2.3.4] (helo=[1.2.3.4])
1689 However, if the real host name was known, it was omitting the HELO data
1690 if it matched the actual IP address. This has been tidied up so that it
1691 doesn't show the same IP address twice.
1693 PH/07 When both +timestamp and +memory debugging was on, the value given by
1694 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1695 called by the debug printing, thereby overwriting the timestamp buffer.
1696 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1698 PH/08 When the original message was included in an autoreply transport, it
1699 always said "this is a copy of the message, including all the headers",
1700 even if body_only or headers_only was set. It now gives an appropriate
1703 PH/09 Applied a patch from the Sieve maintainer which:
1705 o fixes some comments
1706 o adds the (disabled) notify extension core
1707 o adds some debug output for the result of if/elsif tests
1708 o points to the current vacation draft in the documentation
1709 and documents the missing references header update
1713 o fixes a bug in processing the envelope test (when testing
1714 multiple envelope elements, the last element determinted the
1717 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1718 Electronic Mail") by including:
1720 Auto-submitted: auto-generated
1722 in the messages that it generates (bounce messages and others, such as
1723 warnings). In the case of bounce messages for non-SMTP mesages, there was
1724 also a typo: it was using "Auto_submitted" (underscore instead of
1725 hyphen). Since every message generated by Exim is necessarily in response
1726 to another message, thes have all been changed to:
1728 Auto-Submitted: auto-replied
1730 in accordance with these statements in the RFC:
1732 The auto-replied keyword:
1734 - SHOULD be used on messages sent in direct response to another
1735 message by an automatic process,
1737 - MUST NOT be used on manually-generated messages,
1739 - MAY be used on Delivery Status Notifications (DSNs) and Message
1740 Disposition Notifications (MDNs),
1742 - MUST NOT be used on messages generated by automatic or periodic
1743 processes, except for messages which are automatic responses to
1746 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1747 to the default Received: header definition.
1749 PH/12 Added log selector acl_warn_skipped (default on).
1751 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1752 variables because (a) they are in the wrong storage pool and (b) even if
1753 they were copied, it wouldn't work properly because of the caching.
1755 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1756 checking when decoding. Apparently there are clients that generate
1757 overlong encoded strings. Why am I not surprised?
1759 PH/15 If the first argument of "${if match_address" was not empty, but did not
1760 contain an "@" character, Exim crashed. Now it writes a panic log message
1761 and treats the condition as false.
1763 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1765 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1766 extension "envelope-auth". The code is finished and in agreement with
1767 other implementations, but there is no documentation so far and in fact,
1768 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1769 not changing the active code.
1771 Print executed "if" and "elsif" statements when debugging is used. This
1772 helps a great deal to understand what a filter does.
1774 Document more things not specified clearly in RFC3028. I had all this
1775 sorted out, when out of a sudden new issues came to my mind. Oops."
1777 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1780 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1781 canonical form (as documented). However, after a host name lookup from
1782 the IP address, check_host() was doing a simple string comparison with
1783 addresses acquired from the DNS when checking that the found name did
1784 have the original IP as one of its addresses. Since any found IPv6
1785 addresses are likely to be in abbreviated form, the comparison could
1786 fail. Luckily, there already exists a function for doing the comparison
1787 by converting both addresses to binary, so now that is used instead of
1788 the text comparison.
1790 PH/20 There was another similar case to PH/19, when a complete host name was
1791 given in a host list; looking up its IP address could give an abbreviated
1792 form, whereas the current host's name might or might not be abbreviated.
1793 The same fix has been applied.
1799 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1800 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1803 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1804 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1806 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1808 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1809 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1810 Also, for "Auto-Submitted", treat anything other than "no" as
1811 non-personal, in accordance with RFC 3834. (Previously it treated
1812 anything starting "auto-" as non-personal.)
1814 TF/01 The control=submission/name=... option had a problem with syntax
1815 errors if the name included a slash character. The /name= option
1816 now slurps the rest of the string, so it can include any characters
1817 but it must come last in the list of options (after /sender_retain
1820 PH/05 Some modifications to the interface to the fake nameserver for the new
1828 TK/01 Added the "success_on_redirect" address verification option. See
1829 NewStuff for rationale and an example.
1831 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1833 PH/02 Patch to exigrep to allow it to work on syslog lines.
1835 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1836 fread() to skip over the body file's header line, because in Cygwin the
1837 header line is locked and is inaccessible.
1839 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1840 co-exist for some time) to make it clear that it is the Exim ID that is
1841 referenced, not the Message-ID: header line.
1843 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1844 string_format() function, because snprintf() does not exist on all
1847 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1848 ${stat: expansion item.
1850 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1851 protocol synchronization error", to keep the pedants happy.
1853 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1854 well as for IRIX systems, when gcc is being used. See the host.c source
1857 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1859 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1862 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1863 transport and to the smtp transport in LMTP mode.
1865 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1867 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1868 run a filter in a subprocess. This could lead to confusion in subsequent
1869 lookups in the parent process. There should also be a search_tidyup() at
1870 the end of the subprocess.
1872 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1873 only if the host matched helo_try_verify_hosts, which caused the
1874 verification to occur when the EHLO/HELO command was issued. The ACL just
1875 tested the remembered result. Now, if a previous verification attempt has
1876 not happened, "verify = helo" does it there and then.
1878 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1880 TK/03 Fix log output including CR from clamd.
1882 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1883 exist provoked a memory error which could cause a segfault.
1885 PH/15 Installed PCRE 6.2
1887 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1889 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1890 of the problem. Specifically, suggested +O2 rather than +O1 for the
1893 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1895 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1896 record, the port was not correctly logged when the outgoing_port log
1897 selector was set (it logged the transort's default port).
1899 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1900 fallback_hosts, and "hosts" in the smtp transport.
1902 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1903 host errors such as "Connection refused".
1905 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1906 authentication with radiusclient 0.4.9:
1908 - Error returned from rc_read_config was caught wrongly
1909 - Username/password not passed on to radius server due to wrong length.
1911 The presumption is that some radiusclient API changes for 4.51/PH/17
1912 were not taken care of correctly. The code is still untested by me (my
1913 Linux distribution still has 0.3.2 of radiusclient), but it was
1914 contributed by a Radius user.
1916 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1917 expanding the "port" option of the smtp transport.
1919 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1920 while reading a MIME header. Thanks to Tom Hughes for a patch.
1922 PH/24 Include config.h inside local_scan.h so that configuration settings are
1925 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1926 that in an ACL for RCPT (for example), you can examine exactly what was
1929 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1930 commands, but it was not correctly comparing the address with the actual
1931 client host address. Thus, it would show the EHLO address in Received:
1932 header lines when this was not necessary.
1934 PH/27 Added the % operator to ${eval:}.
1936 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1937 it should be ignoring failures (because with -C, for example, it has lost
1938 privilege). It wasn't ignoring creation failures other than "already
1941 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1944 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1945 because an input line was too long, either on its own, or by virtue of
1946 too many continuations, the temporary file was not being removed, and the
1947 return code was incorrect.
1949 PH/31 Missing "BOOL" in function definition in filtertest.c.
1951 PH/32 Applied Sieve patches from the maintainer.
1953 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1955 PH/33 Added "verify = not_blind".
1957 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1958 Local/Makefile (with some defaults set). These are used in built scripts
1959 such as exicyclog, but they have never been used in the exim_install
1960 script (though there are many overriding facilities there). I have
1961 arranged that the exim_install script now takes note of these two
1964 PH/35 Installed configuration files for Dragonfly.
1966 PH/36 When a locally submitted message by a trusted user did not contain a
1967 From: header, and the sender address was obtained from -f or from an SMTP
1968 MAIL command, and the trusted user did not use -F to supply a sender
1969 name, $originator_name was incorrectly used when constructing a From:
1970 header. Furthermore, $originator_name was used for submission mode
1971 messages from external hosts without From: headers in a similar way,
1972 which is clearly wrong.
1974 PH/37 Added control=suppress_local_fixups.
1976 PH/38 When log_selector = +received_sender was set, and the addition of the
1977 sender made the log line's construction buffer exactly full, or one byte
1978 less than full, an overflow happened when the terminating "\n" was
1981 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1982 when the result of a list match is failure because a DNS lookup failed.
1984 PH/40 RM_COMMAND is now used in the building process.
1986 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1987 the "build-* directories that it finds.
1989 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1990 address in a domain literal was a prefix of an interface address.
1992 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1993 when verifying a sender address, unless rewrite_headers is false.
1995 PH/44 Wrote a long comment about why errors_to addresses are verified as
1996 recipients, not senders.
1998 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1999 the ratelimit ACL was added.
2001 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2003 PH/46 Added extra information about PostgreSQL errors to the error string.
2005 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2006 avoiding the need to install special zones in a real server. This is
2007 backwards compatible; if it can't find the fake resolver, it drops back.
2008 Thus, both old and new test suites can be run.
2010 TF/02 Added util/ratelimit.pl
2012 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2013 clock is set back in time.
2015 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2016 Candler <B.Candler@pobox.com>.
2018 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2019 OK for addresses that are the result of redirections.
2021 PH/48 A number of further additions for the benefit of the new test suite,
2022 including a fake gethostbyname() that interfaces to the fake DNS resolver
2025 TF/06 The fix for widen_domains has also been applied to qualify_single and
2026 search_parents which are the other dnslookup options that can cause
2029 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2032 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2034 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2035 service" if it encountered an MX record with an empty target hostname.
2036 The message is now "an MX or SRV record indicated no SMTP service".
2038 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2039 if the DNS of the sending site is misconfigured. This is quite a
2040 common situation. This change restores the behaviour of treating a
2041 helo verification defer as a failure.
2043 PH/51 If self=fail was set on a router, the bounce message did not include the
2044 actual error message.
2050 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2052 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2053 command itself ended in error, the underlying message about the transport
2054 filter timeout was being overwritten with the pipe command error. Now the
2055 underlying error message should be appended to the second error message.
2057 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2059 PH/02 Reduce the amount of output that "make" produces by default. Full output
2060 can still be requested.
2062 PH/03 The warning log line about a condition test deferring for a "warn" verb
2063 was being output only once per connection, rather than after each
2064 occurrence (because it was using the same function as for successful
2065 "warn" verbs). This seems wrong, so I have changed it.
2067 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2068 it should not have, which might have caused a crash in the right
2069 circumstances, but probably never did.
2071 PH/04 Installed a modified version of Tony Finch's patch to make submission
2072 mode fix the return path as well as the Sender: header line, and to
2073 add a /name= option so that you can make the user's friendly name appear
2076 TF/03 Added the control = fakedefer ACL modifier.
2078 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2079 Mark Lowes for thorough testing.
2081 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2083 TK/03 Merged latest SRS patch from Miles Wilton.
2085 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2086 with the definition in sysexits.h (which is #included earlier).
2087 Fortunately, Exim does not actually use EX_OK. The code used to try to
2088 preserve the sysexits.h value, by assumimg that macro definitions were
2089 scanned for macro replacements. I have been disabused of this notion,
2090 so now the code just undefines EX_OK before #including unistd.h.
2092 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2093 in the smtp transport. When a block could not be written in a single
2094 write() function, the timeout was being re-applied to each part-write.
2095 This seems wrong - if the receiver was accepting one byte at a time it
2096 would take for ever. The timeout is now adjusted when this happens. It
2097 doesn't have to be particularly precise.
2099 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2100 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2102 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2103 without a domain if the check to <postmaster@domain> fails.
2105 SC/01 Eximstats: added -xls and the ability to specify output files
2106 (patch written by Frank Heydlauf).
2108 SC/02 Eximstats: use FileHandles for outputing results.
2110 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2112 SC/04 Eximstats: fixed display of large numbers with -nvr option
2114 SC/05 Eximstats: fixed merging of reports with empty tables.
2116 SC/06 Eximstats: added the -include_original_destination flag
2118 SC/07 Eximstats: removed tabs and trailing whitespace.
2120 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2122 TK/06 MBOX spool code: Add real "From " MBOX separator line
2123 so the .eml file is really in mbox format (even though
2124 most programs do not really care). Patch from Alex Miller.
2126 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2127 The latter is generated from $received_to and is only set if the
2128 message has one envelope recipient. SA can use these headers,
2129 obviously out-of-the-box. Patch from Alex Miller.
2131 PH/08 The ${def test on a variable was returning false if the variable's
2132 value was "0", contrary to what the specification has always said!
2133 The result should be true unless the variable is empty.
2135 PH/09 The syntax error of a character other than { following "${if
2136 def:variable_name" (after optional whitespace) was not being diagnosed.
2137 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2138 accidental colon was present, for example, could give incorrect results.
2140 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2141 result is used (not including appendfile, where other changes are about
2144 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2145 This involved changing a lot of size variables from int to off_t. It
2146 should work with maildirs and everything.
2148 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2149 spamd dying while we are connected to it.
2151 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2154 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2155 function for BDB 4.3.
2157 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2159 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2160 the sources for PCRE have been re-arranged and more files are now
2163 PH/15 The code I had for printing potentially long long variables in PH/11
2164 above was not the best (it lost precision). The length of off_t variables
2165 is now inspected at build time, and an appropriate printing format (%ld
2166 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2167 to be "long long int" or "long int". This is needed for the internal
2168 formatting function string_vformat().
2170 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2171 the configuration file to be ":syslog", then the script "guesses" where
2172 the logs files are, rather than using the compiled in default. In our
2173 case the guess is not the same as the compiled default, so the script
2174 suddenly stopped working when I started to use syslog. The patch checks
2175 to see if log_file_path is "". If so, it attempts to read it from exim
2176 with no configuration file to get the compiled in version, before it
2177 falls back to the previous guessing code."
2179 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2180 implementing BATV in an Exim configuration. See NewStuff for the gory
2183 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2184 Makefile that are specific to HP-UX.
2186 PH/18 If the "use_postmaster" option was set for a recipient callout together
2187 with the "random" option, the postmaster address was used as the MAIL
2188 FROM address for the random test, but not for the subsequent recipient
2189 test. It is now used for both.
2191 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2192 patch removes a few documentation additions to RFC 3028, because the
2193 latest draft now contains them. It adds the new en;ascii-case comparator
2194 and a new error check for 8bit text in MIME parts. Comparator and
2195 require names are now matched exactly. I enabled the subaddress
2196 extension, but it is not well tested yet (read: it works for me)."
2198 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2199 rework some of the code of TK/09 above to avoid the hardwired use of
2200 "%lld" and "long long". Replaced the call to snprintf() with a call to
2203 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2204 records point to non-existent hosts", "retry timeout exceeded", and
2205 "retry time not reached for any host after a long failure period".
2207 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2208 experimental DomainKeys support:
2210 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2211 (2) On an error such as an illegally used "control", the wrong name for
2212 the control was given.
2214 These problems did NOT occur unless DomainKeys support was compiled.
2216 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2218 PH/24 Added ${if match_ip condition.
2220 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2221 databases so that it will be absolutely obvious if a crash occurs in the
2222 DB library. This is a regular occurrence (often caused by mis-matched
2225 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2226 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2227 picked up on a user's system that detects such things. There doesn't seem
2228 to be a gcc warning option for this - only an attribute that has to be
2229 put on the function's prototype. It seems that in Fedora Core 4 they have
2230 set this on a number of new functions. No doubt there will be more in due
2233 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2234 specify a transport. However, if an address that was verified by such a
2235 router was the subject of a callout, Exim crashed because it tried to
2236 read the rcpt_include_affixes from the non-existent transport. Now it
2237 just assumes that the setting of that option is false. This bug was
2238 introduced by 4.51/PH/31.
2240 PH/28 Changed -d+all to exclude +memory, because that information is very
2241 rarely of interest, but it makes the output a lot bigger. People tend to
2242 do -d+all out of habit.
2244 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2245 code in os-type was giving problems when libc.so lives in lib64, like on
2248 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2249 aren't the modern standard, and it seems that some systems' include files
2250 don't always have them. Exim was already checking for some of the newer
2251 ones like T_AAAA, and defining it itself. I've added checks for all the
2252 record types that Exim uses.
2254 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2255 not automatically generating a new one, as it is supposed to. This
2256 prevented TLS from working. If the file did exist, but contained invalid
2257 data, a new version was generated, as expected. It was only the case of a
2258 non-existent file that was broken.
2260 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2261 with a change in libdomainkeys > 0.64.
2263 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2264 from DNS. If the selector record carries the flag, it now has
2265 precedence over the domain-wide flag.
2267 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2269 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2270 the use of a port name that isn't defined in /etc/services) occurred, the
2271 message was deferred as in a normal delivery, and thus remained on the
2272 spool, instead of being failed because of the mua_wrapper setting. This
2273 is now fixed, and I tidied up some of the mua_wrapper messages at the
2276 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2277 the messages in a hash of arrays rather than using individual hashes.
2278 This is a bit cleaner and results in dramatic memory savings, albeit
2279 at a slight CPU cost.
2281 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2282 as requested by Marc Sherman.
2284 SC/10 Eximstats: added histograms for user specified patterns as requested
2287 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2289 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2290 fopen() in the content-scanning modules that did not already have it.
2296 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2297 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2299 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2301 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2303 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2304 to test Sieve filters that use "vacation".
2306 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2307 that changes the way the GnuTLS parameters are stored in the cache file.
2308 The new format can be generated externally. For backward compatibility,
2309 if the data in the cache doesn't make sense, Exim assumes it has read an
2310 old-format file, and it generates new data and writes a new file. This
2311 means that you can't go back to an older release without removing the
2314 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2315 work if there are any delivery delays because "one_time" forces the
2316 parent to be marked "delivered", so its unseen clone is never tried
2317 again. For this reason, Exim now forbids the simultaneous setting of
2320 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2321 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2322 entry if you want to know the details. The fix, however, neglected to
2323 consider the case when local delivery batching is involved. The test for
2324 "previously delivered" was not happening when checking to see if an
2325 address could be batched with a previous (undelivered) one; under
2326 certain circumstances this could lead to multiple deliveries to the same
2329 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2330 in its include files, and this causes problems building Exim.
2332 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2333 header_syntax) but Exim was just ignoring anything given after a slash.
2334 In particular, this caused confusion with an attempt to use "verify =
2335 reverse_host_lookup/defer_ok". An error is now given when options are
2336 supplied for verify items that do not have them. (Maybe reverse_host_
2337 lookup should have a defer_ok option, but that's a different point.)
2339 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2340 defined by RFC 821) to 2048, because there were problems with some AUTH
2341 commands, and RFC 1869 says the size should be increased for extended
2342 SMTP commands that take arguments.
2344 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2347 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2348 "unknown" error; now it says that the functionality isn't in the binary.
2350 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2351 an address' error message when a string expansion fails (syntax or
2352 whatever). Otherwise the password may appear in the log. Following change
2353 PH/42 below, there is no longer a chance of it appearing in a bounce
2356 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2358 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2359 of its addresses was ever tried. (Bugzilla bug #2.)
2361 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2362 the result incorrectly in the debug output. (It correctly added a newline
2363 to what was transported.)
2365 TF/01 Added $received_time.
2367 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2368 commented out examples of how to interface to a virus scanner and to
2369 SpamAssassin. Also added commented examples of av_scanner and
2370 spamd_address settings.
2372 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2373 and controls are allowed in which ACLs. There were a couple of minor
2374 errors. Some of the entries in the conditions table (which is a table of
2375 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2376 negation of where the condition IS allowed.
2378 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2380 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2381 header file does not have a version number, so I've had to invent a new
2382 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2383 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2384 radiusclient), but it was contributed by a Radius user.
2386 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2387 files or directories, for OpenSSL.
2389 PH/19 When an Exim process that is running as root has to create an Exim log
2390 file, it does so in a subprocess that runs as exim:exim so as to get the
2391 ownership right at creation (otherwise, other Exim processes might see
2392 the file with the wrong ownership). There was no test for failure of this
2393 fork() call, which would lead to the process getting stuck as it waited
2394 for a non-existent subprocess. Forks do occasionally fail when resources
2395 run out. I reviewed all the other calls to fork(); they all seem to check
2398 PH/20 When checking for unexpected SMTP input at connect time (before writing
2399 the banner), Exim was not dealing correctly with a non-positive return
2400 from the read() function. If the client had disconnected by this time,
2401 the result was a log entry for a synchronization error with an empty
2402 string after "input=" when read() returned zero. If read() returned -1
2403 (an event I could not check), uninitialized data bytes were printed.
2404 There were reports of junk text (parts of files, etc) appearing after
2407 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2409 PH/22 Added support for macro redefinition, and (re)definition in between
2410 driver and ACL definitions.
2412 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2413 forgetting to use the resulting value; it was using the unexpanded value.
2415 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2416 hadn't been configured. The fix is from Juergen Kreileder, who
2417 understands it better than I do:
2419 "Here's what I see happening with three configured cyrus_sasl
2420 authenticators configured (plain, login, cram-md5):
2422 On startup auth_cyrus_sasl_init() gets called for each of these.
2423 This means three calls to sasl_listmech() without a specified mech_list.
2424 => SASL tests which mechs of all available mechs actually work
2425 => three warnings about OTP not working
2426 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2428 With the patch, sasl_listmech() also gets called three times. But now
2429 SASL's mech_list option is set to the server_mech specified in the the
2430 authenticator. Or in other words, the answer from sasl_listmech()
2431 gets limited to just the mech you're testing for (which is different
2433 => the return list contains just 'plain' or 'login', 'cram-md5' or
2434 nothing depending on the value of ob->server_mech.
2436 I've just tested the patch: Authentication still works fine,
2437 unavailable mechs specified in the exim configuration are still
2438 caught, and the auth.log warnings about OTP are gone."
2440 PH/25 When debugging is enabled, the contents of the command line are added
2441 to the debugging output, even when log_selector=+arguments is not
2444 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2445 answer is "GNU", and only if the return is "GNU/something" is the answer
2448 PH/27 $acl_verify_message is now set immediately after the failure of a
2449 verification in an ACL, and so is available in subsequent modifiers. In
2450 particular, the message can be preserved by coding like this:
2452 warn !verify = sender
2453 set acl_m0 = $acl_verify_message
2455 Previously, $acl_verify_message was set only while expanding "message"
2456 and "log_message" when a very denied access.
2458 PH/28 Modified OS/os.c-Linux with
2460 -#ifndef OS_LOAD_AVERAGE
2461 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2463 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2464 nomenclature these days.)
2466 PH/29 Installed patch from the Sieve maintainer that adds the options
2467 sieve_useraddress and sieve_subaddress to the redirect router.
2469 PH/30 In these circumstances:
2470 . Two addresses routed to the same list of hosts;
2471 . First host does not offer TLS;
2472 . First host accepts first address;
2473 . First host gives temporary error to second address;
2474 . Second host offers TLS and a TLS session is established;
2475 . Second host accepts second address.
2476 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2477 and peerdn, if requested) that were in fact used only for the second
2480 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2481 attention to any local part prefix or suffix that was matched by the
2482 router that accepted the address. It now behaves in the same way as it
2483 does for delivery: the affixes are removed from the local part unless
2484 rcpt_include_affixes is set on the transport.
2486 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2487 timeout during the DATA phase of an incoming message.
2489 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2490 applied a patch sanctioned by the Sieve maintainer.
2492 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2493 the uid or gid is negative. A case of a negative gid caused this to be
2494 noticed. The fix allows for either to be negative.
2496 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2497 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2500 PH/36 The change PH/12 above was broken. Fixed it.
2502 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2503 the grounds that routing the same address twice would always produce the
2504 same answer. This might have been true once, but it is certainly no
2505 longer true now. Routing a child address may depend on the previous
2506 routing that produced that child. Some complicated redirection strategies
2507 went wrong when messages had multiple recipients, and made Exim's
2508 behaviour dependent on the order in which the addresses were given.
2510 I have moved the duplicate checking until after the routing is complete.
2511 Exim scans the addresses that are assigned to local and remote
2512 transports, and removes any duplicates. This means that more work will be
2513 done, as duplicates will always all be routed, but duplicates are
2514 presumably rare, so I don't expect this is of any significance.
2516 For deliveries to pipes, files, and autoreplies, the duplicate checking
2517 still happens during the routing process, since they are not going to be
2520 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2521 It corrects a timeout issue with spamd. This is Ian's comment: "The
2522 background is that sometimes spamd either never reads data from a
2523 connection it has accepted, or it never writes response data. The exiscan
2524 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2525 blindly assumes that writes won't block so it may never time out."
2527 PH/39 Allow G after quota size as well as K and M.
2529 PH/40 The value set for $authenticated_id in an authenticator may not contain
2530 binary zeroes or newlines because the value is written to log lines and
2531 to spool files. There was no check on this. Now the value is run through
2532 the string_printing() function so that such characters are converted to
2533 printable escape sequences.
2535 PH/41 $message_linecount is a new variable that contains the total number of
2536 lines in the message. Compare $body_linecount, which is the count for the
2539 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2540 bounce and delay warning messages, except in certain special cases, which
2543 (a) An SMTP error message from a remote host;
2544 (b) A message specified in a :fail: redirection;
2545 (c) A message specified in a "fail" command in a system filter;
2546 (d) A message specified in a FAIL return from the queryprogram router;
2547 (e) A message specified by the cannot_route_message router option.
2549 In these cases only, Exim does include the error details in bounce and
2550 warning messages. There are also a few cases where bland messages such
2551 as "unrouteable address" or "local delivery error" are given.
2553 PH/43 $value is now also set for the "else" part of a ${run expansion.
2555 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2556 being worked on, but at least Exim now implements the latest version to
2559 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2560 process to complete was treated as a delivery failure, a timeout while
2561 writing the message to the pipe was logged, but erroneously treated as a
2562 successful delivery. Such timeouts include transport filter timeouts. For
2563 consistency with the overall process timeout, these timeouts are now
2564 treated as errors, giving rise to delivery failures by default. However,
2565 there is now a new Boolean option for the pipe transport called
2566 timeout_defer, which, if set TRUE, converts the failures into defers for
2567 both kinds of timeout. A transport filter timeout is now identified in
2570 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2571 systems where "make" and "gmake" are different, calling "gmake" at top
2572 level broke things. I've arranged for the value of $(MAKE) to be passed
2573 from the Makefile to this script so that it can call the same version of
2577 A note about Exim versions 4.44 and 4.50
2578 ----------------------------------------
2580 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2581 changes of various kinds. As a consequence, a big documentation update was
2582 needed. This delayed the release for rather longer than seemed good, especially
2583 in the light of a couple of (minor) security issues. Therefore, the changes
2584 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2585 release. So 4.44 and 4.50 are in effect two different branches that both start
2588 I have left the 4.50 change log unchanged; it contains all the changes since
2589 4.43. The change log for 4.44 is below; many of its items are identical to
2590 those for 4.50. This seems to be the most sensible way to preserve the
2591 historical information.
2597 1. Minor wording change to the doc/README.SIEVE file.
2599 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2600 computation of the current number of files was incorrect.
2602 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2603 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2606 4. Give more explanation in the error message when the command for a transport
2607 filter fails to execute.
2609 5. There are several places where Exim runs a non-Exim command in a
2610 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2611 was being done only for the command run by the queryprogram router. It is
2612 now done for all such subprocesses. The other cases are: ${run, transport
2613 filters, and the commands run by the lmtp and pipe transports.
2615 6. Added CONFIGURE_GROUP build-time option.
2617 7. Some older OS have a limit of 256 on the maximum number of file
2618 descriptors. Exim was using setrlimit() to set 1000 as a large value
2619 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2620 systems. I've change it so that if it can't get 1000, it tries for 256.
2622 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2623 was an oversight, and furthermore, ever since the addition of extra
2624 controls (e.g. 4.43/32), the checks on when to allow different forms of
2625 "control" were broken. There should now be diagnostics for all cases when a
2626 control that does not make sense is encountered.
2628 9. Added the /retain_sender option to "control=submission".
2630 10. $recipients is now available in the predata ACL (oversight).
2632 11. Tidy the search cache before the fork to do a delivery from a message
2633 received from the command line. Otherwise the child will trigger a lookup
2634 failure and thereby defer the delivery if it tries to use (for example) a
2635 cached ldap connection that the parent has called unbind on.
2637 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2638 of $address_data from the recipient verification was clobbered by the
2639 sender verification.
2641 13. The value of address_data from a sender verification is now available in
2642 $sender_address_data in subsequent conditions in the ACL statement.
2644 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2646 15. Added a new option "connect=<time>" to callout options, to set a different
2649 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2650 was its contents. (It was OK if the option was not defined at all.)
2652 17. A "Completed" log line is now written for messages that are removed from
2653 the spool by the -Mrm option.
2655 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2656 information about exactly what failed.
2658 19. Added -dd to debug only the daemon process.
2660 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2661 handles timeouts, both on the server side and network timeouts. Renamed the
2662 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2664 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2665 It is now set to "smtps".
2667 22. $host_address is now set to the target address during the checking of
2668 ignore_target_hosts.
2670 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2671 being passed; this would have caused $sender_host_name to have been used if
2672 matching the list had actually called for a host name (not very likely,
2673 since this list is usually IP addresses). A host name is now passed as
2676 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2677 code that specifies a non-restarting handler (typically sigaction() in
2678 modern systems) in an attempt to fix a rare and obscure crash bug.
2680 25. Narrowed the window for a race in the daemon that could cause it to ignore
2681 SIGCHLD signals. This is not a major problem, because they are used only to
2682 wake it up if nothing else does.
2684 26. A malformed maildirsize file could cause Exim to calculate negative values
2685 for the mailbox size or file count. Odd effects could occur as a result.
2686 The maildirsize information is now recalculated if the size or filecount
2689 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2690 support for a long time. Removed HAVE_SYS_VFS_H.
2692 28. Installed the latest version of exipick from John Jetmore.
2694 29. In an address list, if the pattern was not a regular expression, an empty
2695 subject address (from a bounce message) matched only if the pattern was an
2696 empty string. Non-empty patterns were not even tested. This was the wrong
2697 because it is perfectly reasonable to use an empty address as part of a
2698 database query. An empty address is now tested by patterns that are
2699 lookups. However, all the other forms of pattern expect the subject to
2700 contain a local part and a domain, and therefore, for them, an empty
2701 address still always fails if the pattern is not itself empty.
2703 30. Exim went into a mad DNS loop when attempting to do a callout where the
2704 host was specified on an smtp transport, and looking it up yielded more
2705 than one IP address.
2707 31. Re-factored the code for checking spool and log partition space into a
2708 function that finds that data and another that does the check. The former
2709 is then used to implement four new variables: $spool_space, $log_space,
2710 $spool_inodes, and $log_inodes.
2712 32. The RFC2047 encoding function was originally intended for short strings
2713 such as real names; it was not keeping to the 75-character limit for
2714 encoded words that the RFC imposes. It now respects the limit, and
2715 generates multiple encoded words if necessary. To be on the safe side, I
2716 have increased the buffer size for the ${rfc2047: expansion operator from
2719 33. It is now permitted to omit both strings after an "if" condition; if the
2720 condition is true, the result is "true". As before, when the second string
2721 is omitted, a false condition yields an empty string. This makes it less
2722 cumbersome to write custom ACL and router conditions.
2724 34. Failure to deliver a bounce message always caused it to be frozen, even if
2725 there was an errors_to setting on the router. The errors_to setting is now
2728 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2729 canonical form (fully expanded) before being placed in
2730 $sender_host_address.
2732 36. The table in the code that translates DNS record types into text (T_A to
2733 "A" for instance) was missing entries for NS and CNAME. It is just possible
2734 that this could have caused confusion if both these types were looked up
2735 for the same domain, because the text type is used as part of Exim's
2736 per-process caching. But the chance of anyone hitting this buglet seems
2739 37. The dnsdb lookup has been extended in a number of ways.
2741 (1) There is a new type, "zns", which walks up the domain tree until it
2742 finds some nameserver records. It should be used with care.
2744 (2) There is a new type, "mxh", which is like "mx" except that it returns
2745 just the host names, not the priorities.
2747 (3) It is now possible to give a list of domains (or IP addresses) to be
2748 looked up. The behaviour when one of the lookups defers can be
2749 controlled by a keyword.
2751 (4) It is now possible to specify the separator character for use when
2752 multiple records are returned.
2754 38. The dnslists ACL condition has been extended: it is now possible to supply
2755 a list of IP addresses and/or domains to be looked up in a particular DNS
2758 39. Added log_selector=+queue_time_overall.
2760 40. When running the queue in the test harness, wait just a tad after forking a
2761 delivery process, to get repeatability of debugging output.
2763 41. Include certificate and key file names in error message when GnuTLS fails
2764 to set them up, because the GnuTLS error message doesn't include the name
2765 of the failing file when there is a problem reading it.
2767 42. Allow both -bf and -bF in the same test run.
2769 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2771 44. The "Exiscan patch" is now merged into the mainline Exim source.
2773 45. Sometimes the final signoff response after QUIT could fail to get
2774 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2775 < 0 before doing a fflush(). This bug looks as though it goes back to the
2776 introduction of TLS in release 3.20, but "sometimes" must have been rare
2777 because the tests only now provoked it.
2779 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2780 (this can affect the format of dates).
2782 47. exim_tidydb, when checking for the continued existence of a message for
2783 which it has found a message-specific retry record, was not finding
2784 messages that were in split spool directories. Consequently, it was
2785 deleting retry records that should have stayed in existence.
2787 48. Steve fixed some bugs in eximstats.
2789 49. The SPA authentication driver was not abandoning authentication and moving
2790 on to the next authenticator when an expansion was forced to fail,
2791 contradicting the general specification for all authenticators. Instead it
2792 was generating a temporary error. It now behaves as specified.
2794 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2795 (the order specifies the preference for clients). The order is now AES256,
2796 AES128, 3DES, ARCFOUR128.
2798 51. Small patch to Sieve code - explicitly set From: when generating an
2801 52. Exim crashed if a remote delivery caused a very long error message to be
2802 recorded - for instance if somebody sent an entire SpamAssassin report back
2803 as a large number of 550 error lines. This bug was coincidentally fixed by
2804 increasing the size of one of Exim's internal buffers (big_buffer) that
2805 happened as part of the Exiscan merge. However, to be on the safe side, I
2806 have made the code more robust (and fixed the comments that describe what
2809 53. Now that there can be additional text after "Completed" in log lines (if
2810 the queue_time_overall log selector is set), a one-byte patch to exigrep
2811 was needed to allow it to recognize "Completed" as not the last thing in
2814 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2815 patch that reportedly fixes this has been added. I am not expert enough to
2816 create a test for it. This is what the patch creator wrote:
2818 "I found a little strange behaviour of ldap code when working with
2819 Windows 2003 AD Domain, where users was placed in more than one
2820 Organization Units. When I tried to give exim partial DN, the exit code
2821 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2822 But simultaneously result of request was absolutely normal ldap result,
2823 so I produce this patch..."
2825 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2826 REFERENCE, so I have modified the code to exclude the patch when that macro
2829 55. Some experimental protocols are using DNS PTR records for new purposes. The
2830 keys for these records are domain names, not reversed IP addresses. The
2831 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2832 leaves it alone. Component reversal etc. now happens only for IP addresses.
2835 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2837 57. Double the size of the debug message buffer (to 2048) so that more of very
2838 long debug lines gets shown.
2840 58. The exicyclog utility now does better if the number of log files to keep
2841 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2843 59. Two changes related to the smtp_active_hostname option:
2845 (1) $smtp_active_hostname is now available as a variable.
2846 (2) The default for smtp_banner uses $smtp_active_hostname instead
2847 of $primary_hostname.
2849 60. The host_aton() function is supposed to be passed a string that is known
2850 to be a valid IP address. However, in the case of IPv6 addresses, it was
2851 not checking this. This is a hostage to fortune. Exim now panics and dies
2852 if the condition is not met. A case was found where this could be provoked
2853 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2854 components; fortuitously, this particular loophole had already been fixed
2855 by change 4.50/55 above.
2857 If there are any other similar loopholes, the new check in host_aton()
2858 itself should stop them being exploited. The report I received stated that
2859 data on the command line could provoke the exploit when Exim was running as
2860 exim, but did not say which command line option was involved. All I could
2861 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2862 running as the user.
2865 61. There was a buffer overflow vulnerability in the SPA authentication code
2866 (which came originally from the Samba project). I have added a test to the
2867 spa_base64_to_bits() function which I hope fixes it.
2870 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2871 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2873 63. The daemon start-up calls getloadavg() while still root for those OS that
2874 need the first call to be done as root, but it missed one case: when
2875 deliver_queue_load_max is set with deliver_drop_privilege. This is
2876 necessary for the benefit of the queue runner, because there is no re-exec
2877 when deliver_drop_privilege is set.
2879 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2880 This has been fixed.
2882 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2883 was in use, was not putting the data itself into the right store pool;
2884 consequently, it could be overwritten for a subsequent message in the same
2885 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2888 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2890 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2891 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2892 the function were treating the return as a boolean value, which happened to
2893 work because 0=false and not-0=true, but is not correct code.
2895 68. The host_aton() function was not handling scoped IPv6 addresses (those
2896 with, for example, "%eth0" on the end) correctly.
2898 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2899 negated items (that is, ~something) in unsigned ints. Some compilers
2900 apparently mutter when there is no cast.
2902 70. If an address verification called from an ACL failed, and did not produce a
2903 user-specific message (i.e. there was only a "system" message), nothing was
2904 put in $acl_verify_message. In this situation, it now puts the system
2907 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2908 session; change 4.31/43 added the unwanted input to the log line - except
2909 that it did not do this in the start of session case. It now does.
2911 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2912 This is wrong and can cause the other end to generate a synchronization
2913 error if it is another Exim or anything else that does the synchronization
2914 check. A QUIT command is no longer sent after a timeout.
2916 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2917 during host lookups.
2919 74. The defer_ok option of callout verification was not working if it was used
2920 when verifying addresses in header lines, that is, for this case:
2922 verify = header_sender/callout=defer_ok
2924 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2925 those file descriptors could be used for SMTP connections. If anything
2926 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2927 could be sent to the SMTP client, causing chaos. The daemon now opens
2928 stdin, stdout, and stderr to /dev/null when it puts itself into the
2931 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2932 log by default. The user can override this with suitable Perl magic.
2934 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2935 the log message when discard triggers, was not working for the DATA ACL or
2936 for the non-SMTP ACL.
2938 78. Error message wording change in sieve.c.
2940 79. If smtp_accept_max_per_host was set, the number of connections could be
2941 restricted to fewer than expected, because the daemon was trying to set up
2942 a new connection before checking whether the processes handling previous
2943 connections had finished. The check for completed processes is now done
2944 earlier. On busy systems, this bug wouldn't be noticed because something
2945 else would have woken the daemon, and it would have reaped the completed
2948 80. If a message was submitted locally by a user whose login name contained one
2949 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2950 It caused a spool format error. I have fixed the spool reading code. A
2951 related problem was that the "from" clause in the Received: line became
2952 illegal because of the space(s). It is now covered by ${quote_local_part.
2954 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2957 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2959 83. Patch from Sieve maintainer to fix unterminated string problem in
2960 "vacation" handling.
2962 84. Some minor changes to the Linux configuration files to help with other
2963 OS variants using glibc.
2965 85. One more patch for Sieve to update vacation handling to latest spec.
2968 ----------------------------------------------------
2969 See the note above about the 4.44 and 4.50 releases.
2970 ----------------------------------------------------
2976 1. Change 4.43/35 introduced a bug that caused file counts to be
2977 incorrectly computed when quota_filecount was set in an appendfile
2980 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2981 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2984 3. Give more explanation in the error message when the command for a transport
2985 filter fails to execute.
2987 4. There are several places where Exim runs a non-Exim command in a
2988 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2989 was being done only for the command run by the queryprogram router. It is
2990 now done for all such subprocesses. The other cases are: ${run, transport
2991 filters, and the commands run by the lmtp and pipe transports.
2993 5. Some older OS have a limit of 256 on the maximum number of file
2994 descriptors. Exim was using setrlimit() to set 1000 as a large value
2995 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2996 systems. I've change it so that if it can't get 1000, it tries for 256.
2998 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2999 was an oversight, and furthermore, ever since the addition of extra
3000 controls (e.g. 4.43/32), the checks on when to allow different forms of
3001 "control" were broken. There should now be diagnostics for all cases when a
3002 control that does not make sense is encountered.
3004 7. $recipients is now available in the predata ACL (oversight).
3006 8. Tidy the search cache before the fork to do a delivery from a message
3007 received from the command line. Otherwise the child will trigger a lookup
3008 failure and thereby defer the delivery if it tries to use (for example) a
3009 cached ldap connection that the parent has called unbind on.
3011 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3012 of $address_data from the recipient verification was clobbered by the
3013 sender verification.
3015 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3016 was its contents. (It was OK if the option was not defined at all.)
3018 11. A "Completed" log line is now written for messages that are removed from
3019 the spool by the -Mrm option.
3021 12. $host_address is now set to the target address during the checking of
3022 ignore_target_hosts.
3024 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3025 being passed; this would have caused $sender_host_name to have been used if
3026 matching the list had actually called for a host name (not very likely,
3027 since this list is usually IP addresses). A host name is now passed as
3030 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3031 code that specifies a non-restarting handler (typically sigaction() in
3032 modern systems) in an attempt to fix a rare and obscure crash bug.
3034 15. Narrowed the window for a race in the daemon that could cause it to ignore
3035 SIGCHLD signals. This is not a major problem, because they are used only to
3036 wake it up if nothing else does.
3038 16. A malformed maildirsize file could cause Exim to calculate negative values
3039 for the mailbox size or file count. Odd effects could occur as a result.
3040 The maildirsize information is now recalculated if the size or filecount
3043 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3044 support for a long time. Removed HAVE_SYS_VFS_H.
3046 18. Updated exipick to current release from John Jetmore.
3048 19. Allow an empty sender to be matched against a lookup in an address list.
3049 Previously the only cases considered were a regular expression, or an
3052 20. Exim went into a mad DNS lookup loop when doing a callout where the
3053 host was specified on the transport, if the DNS lookup yielded more than
3056 21. The RFC2047 encoding function was originally intended for short strings
3057 such as real names; it was not keeping to the 75-character limit for
3058 encoded words that the RFC imposes. It now respects the limit, and
3059 generates multiple encoded words if necessary. To be on the safe side, I
3060 have increased the buffer size for the ${rfc2047: expansion operator from
3063 22. Failure to deliver a bounce message always caused it to be frozen, even if
3064 there was an errors_to setting on the router. The errors_to setting is now
3067 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3068 canonical form (fully expanded) before being placed in
3069 $sender_host_address.
3071 24. Updated eximstats to version 1.33
3073 25. Include certificate and key file names in error message when GnuTLS fails
3074 to set them up, because the GnuTLS error message doesn't include the name
3075 of the failing file when there is a problem reading it.
3077 26. Expand error message when OpenSSL has problems setting up cert/key files.
3080 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3081 (this can affect the format of dates).
3083 28. exim_tidydb, when checking for the continued existence of a message for
3084 which it has found a message-specific retry record, was not finding
3085 messages that were in split spool directories. Consequently, it was
3086 deleting retry records that should have stayed in existence.
3088 29. eximstats updated to version 1.35
3089 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3090 1.35 - bugfix such that pie charts by volume are generated correctly
3092 30. The SPA authentication driver was not abandoning authentication and moving
3093 on to the next authenticator when an expansion was forced to fail,
3094 contradicting the general specification for all authenticators. Instead it
3095 was generating a temporary error. It now behaves as specified.
3097 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3098 (the order specifies the preference for clients). The order is now AES256,
3099 AES128, 3DES, ARCFOUR128.
3101 31. Small patch to Sieve code - explicitly set From: when generating an
3104 32. Exim crashed if a remote delivery caused a very long error message to be
3105 recorded - for instance if somebody sent an entire SpamAssassin report back
3106 as a large number of 550 error lines. This bug was coincidentally fixed by
3107 increasing the size of one of Exim's internal buffers (big_buffer) that
3108 happened as part of the Exiscan merge. However, to be on the safe side, I
3109 have made the code more robust (and fixed the comments that describe what
3112 33. Some experimental protocols are using DNS PTR records for new purposes. The
3113 keys for these records are domain names, not reversed IP addresses. The
3114 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3115 leaves it alone. Component reversal etc. now happens only for IP addresses.
3118 34. The host_aton() function is supposed to be passed a string that is known
3119 to be a valid IP address. However, in the case of IPv6 addresses, it was
3120 not checking this. This is a hostage to fortune. Exim now panics and dies
3121 if the condition is not met. A case was found where this could be provoked
3122 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3123 components; fortuitously, this particular loophole had already been fixed
3124 by change 4.50/55 or 4.44/33 above.
3126 If there are any other similar loopholes, the new check in host_aton()
3127 itself should stop them being exploited. The report I received stated that
3128 data on the command line could provoke the exploit when Exim was running as
3129 exim, but did not say which command line option was involved. All I could
3130 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3131 running as the user.
3134 35. There was a buffer overflow vulnerability in the SPA authentication code
3135 (which came originally from the Samba project). I have added a test to the
3136 spa_base64_to_bits() function which I hope fixes it.
3139 36. The daemon start-up calls getloadavg() while still root for those OS that
3140 need the first call to be done as root, but it missed one case: when
3141 deliver_queue_load_max is set with deliver_drop_privilege. This is
3142 necessary for the benefit of the queue runner, because there is no re-exec
3143 when deliver_drop_privilege is set.
3145 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3146 was in use, was not putting the data itself into the right store pool;
3147 consequently, it could be overwritten for a subsequent message in the same
3148 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3151 38. Sometimes the final signoff response after QUIT could fail to get
3152 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3153 < 0 before doing a fflush(). This bug looks as though it goes back to the
3154 introduction of TLS in release 3.20, but "sometimes" must have been rare
3155 because the tests only now provoked it.
3161 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3162 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3163 Now it returns an integer. A number of calls were still expecting a T/F
3164 return. Fortuitously, in all cases, the tests worked in OK situations,
3165 which is the norm. However, things would have gone wrong on any write
3166 failures on the smtp file descriptor. This function is used when sending
3167 messages over SMTP and also when doing verify callouts.
3169 2. When Exim is called to do synchronous delivery of a locally submitted
3170 message (the -odf or -odi options), it no longer closes stderr before doing
3173 3. Implemented the mua_wrapper option.
3175 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3177 5. Implemented the functions header_remove(), header_testname(),
3178 header_add_at_position(), and receive_remove_recipient(), and exported them
3181 6. If an ACL "warn" statement specified the addition of headers, Exim already
3182 inserted X-ACL-Warn: at the start if there was no header name. However, it
3183 was not making this test for the second and subsequent header lines if
3184 there were newlines in the string. This meant that an invalid header could
3185 be inserted if Exim was badly configured.
3187 7. Allow an ACL "warn" statement to add header lines at the start or after all
3188 the Received: headers, as well as at the end.
3190 8. Added the rcpt_4xx retry error code.
3192 9. Added postmaster_mailfrom=xxx to callout verification option.
3194 10. Added mailfrom=xxxx to the callout verification option, for verify=
3197 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3198 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3200 12. Inserted some casts to stop certain compilers warning when using pointer
3201 differences as field lengths or precisions in printf-type calls (mostly
3202 affecting debugging statements).
3204 13. Added optional readline() support for -be (dynamically loaded).
3206 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3207 same clock tick as a message's arrival, so that its received time was the
3208 same as the "first fail" time on the retry record, and that message
3209 remained on the queue past the ultimate address timeout, every queue runner
3210 would try a delivery (because it was past the ultimate address timeout) but
3211 after another failure, the ultimate address timeout, which should have then
3212 bounced the address, did not kick in. This was a "< instead of <=" error;
3213 in most cases the first failure would have been in the next clock tick
3214 after the received time, and all would be well.
3216 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3217 being recognized when the domain list was tested by the match_domain
3218 condition in an expansion string.
3220 16. Added the ${str2b64: operator.
3222 17. Exim was always calling setrlimit() to set a large limit for the number of
3223 processes, without checking whether the existing limit was already
3224 adequate. (It did check for the limit on file descriptors.) Furthermore,
3225 errors from getrlimit() and setrlimit() were being ignored. Now they are
3226 logged to the main and panic logs, but Exim does carry on, to try to do its
3227 job under whatever limits there are.
3229 18. Imported PCRE 5.0.
3231 19. Trivial typo in log message " temporarily refused connection" (the leading
3234 20. If the log selector return_path_on_delivery was set and an address was
3235 redirected to /dev/null, the delivery process crashed because it assumed
3236 that a return path would always be set for a "successful" delivery. In this
3237 case, the whole delivery is bypassed as an optimization, and therefore no
3240 21. Internal re-arrangement: the function for sending a challenge and reading
3241 a response while authentication was assuming a zero-terminated challenge
3242 string. It's now changed to take a pointer and a length, to allow for
3243 binary data in such strings.
3245 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3247 23. Exim was not respecting finduser_retries when seeking the login of the
3248 uid under which it was called; it was always trying 10 times. (The default
3249 setting of finduser_retries is zero.) Also, it was sleeping after the final
3250 failure, which is pointless.
3252 24. Implemented tls_on_connect_ports.
3254 25. Implemented acl_smtp_predata.
3256 26. If the domain in control=submission is set empty, Exim assumes that the
3257 authenticated id is a complete email address when it generates From: or
3258 Sender: header lines.
3260 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3261 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3262 chown and chgrp in /bin and hostname in /usr/bin.
3264 28. Exim was keeping the "process log" file open after each use, just as it
3265 does for the main log. This opens the possibility of it remaining open for
3266 long periods when the USR1 signal hits a daemon. Occasional processlog
3267 errors were reported, that could have been caused by this. Anyway, it seems
3268 much more sensible not to leave this file open at all, so that is what now
3271 29. The long-running daemon process does not normally write to the log once it
3272 has entered its main loop, and it closes the log before doing so. This is
3273 so that log files can straightforwardly be renamed and moved. However,
3274 there are a couple of unusual error situations where the daemon does write
3275 log entries, and I had neglected to close the log afterwards.
3277 30. The text of an SMTP error response that was received during a remote
3278 delivery was being truncated at 512 bytes. This is too short for some of
3279 the long messages that one sometimes sees. I've increased the limit to
3282 31. It is now possible to make retry rules that apply only when a message has a
3283 specific sender, in particular, an empty sender.
3285 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3286 it possible to be selective about when SMTP synchronization is enforced.
3288 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3290 32. Implemented hosts_connection_nolog.
3292 33. Added an ACL for QUIT.
3294 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3297 35. Added mailbox_size and mailbox_filecount to appendfile.
3299 36. Added control = no_multiline_responses to ACLs.
3301 37. There was a bug in the logic of the code that waits for the clock to tick
3302 in the case where the clock went backwards by a substantial amount such
3303 that the microsecond fraction of "now" was more than the microsecond
3304 fraction of "then" (but the whole seconds number was less).
3306 38. Added support for the libradius Radius client library this is found on
3307 FreeBSD (previously only the radiusclient library was supported).
3313 1. When certain lookups returned multiple values in the form name=value, the
3314 quoting of the values was not always being done properly. Specifically:
3315 (a) If the value started with a double quote, but contained no whitespace,
3317 (b) If the value contained whitespace other than a space character (i.e.
3318 tabs or newlines or carriage returns) it was not quoted.
3319 This fix has been applied to the mysql and pgsql lookups by writing a
3320 separate quoting function and calling it from the lookup code. The fix
3321 should probably also be applied to nisplus, ibase and oracle lookups, but
3322 since I cannot test any of those, I have not disturbed their existing code.
3324 2. A hit in the callout cache for a specific address caused a log line with no
3325 reason for rejecting RCPT. Now it says "Previous (cached) callout
3326 verification failure".
3328 3. There was an off-by-one bug in the queryprogram router. An over-long
3329 return line was truncated at 256 instead of 255 characters, thereby
3330 overflowing its buffer with the terminating zero. As well as fixing this, I
3331 have increased the buffer size to 1024 (and made a note to document this).
3333 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3334 when Exim is waiting for an SMTP response from a remote server, Exim
3335 restarts its select() call on the socket, thereby resetting its timeout.
3336 This is not a problem when such interrupts are rare. Somebody set up a cron
3337 job to run exiwhat every 2 minutes, which is less than the normal select()
3338 timeout (5 or 10 minutes). This meant that the select() timeout never
3339 kicked in because it was always reset. I have fixed this by comparing the
3340 time when an interrupt arrives with the time at the start of the first call
3341 to select(). If more time than the timeout has elapsed, the interrupt is
3342 treated as a timeout.
3344 5. Some internal re-factoring in preparation for the addition of Sieve
3345 extensions (by MH). In particular, the "personal" test is moved to a
3346 separate function, and given an option for scanning Cc: and Bcc: (which is
3347 not set for Exim filters).
3349 6. When Exim created an email address using the login of the caller as the
3350 local part (e.g. when creating a From: or Sender: header line), it was not
3351 quoting the local part when it contained special characters such as @.
3353 7. Installed new OpenBSD configuration files.
3355 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3356 try to make them clearer.
3358 9. Callout options, other than the timeout value, were being ignored when
3359 verifying sender addresses in header lines. For example, when using
3361 verify = header_sender/callout=no_cache
3363 the cache was (incorrectly) being used.
3365 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3366 only the Cygwin environment.
3368 11. When return_path_on_delivery was set as a log selector, if different remote
3369 addresses in the same message used different return paths and parallel
3370 remote delivery occurred, the wrong values would sometimes be logged.
3371 (Whenever a remote delivery process finished, the return path value from
3372 the most recently started remote delivery process was logged.)
3374 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3375 lines when AUTH and/or TLS are in use. This is the "received protocol"
3376 field. Exim used to use "asmtp" for authenticated SMTP, without any
3377 indication (in the protocol name) for TLS use. Now it follows the RFC and
3378 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3379 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3380 names appear in log lines as well as in Received: header lines.
3382 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3383 extensions, and comparison tests, and to fix some bugs.
3385 14. Changes to the "personal" filter test:
3387 (1) The test was buggy in that it was just doing the equivalent of
3388 "contains" tests on header lines. For example, if a user's address was
3389 anne@some.where, the "personal" test would incorrectly be true for
3391 To: susanne@some.where
3393 This test is now done by extracting each address from the header in turn,
3394 and checking the entire address. Other tests that are part of "personal"
3395 are now done using regular expressions (for example, to check local parts
3396 of addresses in From: header lines).
3398 (2) The list of non-personal local parts in From: addresses has been
3399 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3400 taken from the Sieve specification recommendations.
3402 (3) If the message contains any header line starting with "List-" it is
3403 treated as non-personal.
3405 (4) The test for "circular" in the Subject: header line has been removed
3406 because it now seems ill-conceived.
3408 15. Minor typos in src/EDITME comments corrected.
3410 16. Installed latest exipick from John Jetmore.
3412 17. If headers_add on a router specified a text string that was too long for
3413 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3414 of string_sprintf() is now avoided.
3416 18. $message_body_size was not set (it was always zero) when running the DATA
3417 ACL and the local_scan() function.
3419 19. For the "mail" command in an Exim filter, no default was being set for
3420 the once_repeat time, causing a random time value to be used if "once" was
3421 specified. (If the value happened to be <= 0, no repeat happened.) The
3422 default is now 0s, meaning "never repeat". The "vacation" command was OK
3423 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3424 (I found it when inspecting the code).
3426 20. There is now an overall timeout for performing a callout verification. It
3427 defaults to 4 times the callout timeout, which applies to individual SMTP
3428 commands during the callout. The overall timeout applies when there is more
3429 than one host that can be tried. The timeout is checked before trying the
3430 next host. This prevents very long delays if there are a large number of
3431 hosts and all are timing out (e.g. when the network connections are timing
3432 out). The value of the overall timeout can be changed by specifying an
3433 additional sub-option for "callout", called "maxwait". For example:
3435 verify = sender/callout=5s,maxwait=20s
3437 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3438 to the end before writing, but this should make it even safer).
3440 22. Exim was forgetting that it had advertised PIPELINING for the second and
3441 subsequent messages on an SMTP connection. It was also not resetting its
3442 memory on STARTTLS and an internal HELO.
3444 23. When Exim logs an SMTP synchronization error within a session, it now
3445 records whether PIPELINING has been advertised or not.
3447 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3448 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3449 rather than long int.
3451 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3453 26. Added the never_mail option to autoreply.
3459 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3460 crash if the getsockname() call failed; this can happen if a connection is
3461 closed very soon after it is established. The problem was simply in the
3462 order in which certain operations were done, causing Exim to try to write
3463 to the SMTP stream before it had set up the file descriptor. The bug has
3464 been fixed by making things happen in the correct order.
3470 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3471 before the connection was closed, thus losing the rejection response.
3473 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3474 some early Solaris releases, but causes trouble in current releases where
3475 socklen_t is defined.
3477 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3480 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3483 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3484 that releasing the top of it at the end releases what was used for sub-
3485 expansions (unless the block got too big). However, discard this block if
3486 the first thing is a variable or header, so that we can use its block when
3487 it is dynamic (useful for very large $message_headers, for example).
3489 6. Lookups now cache *every* query, not just the most recent. A new, separate
3490 store pool is used for this. It can be recovered when all lookup caches are
3491 flushed. Lookups now release memory at the end of their result strings.
3492 This has involved some general refactoring of the lookup sources.
3494 7. Some code has been added to the store_xxx() functions to reduce the amount
3495 of flapping under certain conditions.
3497 8. log_incoming_interface used to affect only the <= reception log lines. Now
3498 it causes the local interface and port to be added to several more SMTP log
3499 lines, for example "SMTP connection from", and rejection lines.
3501 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3503 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3505 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3506 could be overwritten at the end of the current message (or the start of a
3507 new message if it was set in a HELO ACL). The value is now preserved for
3508 the duration of the SMTP connection.
3510 12. If a transport had a headers_rewrite setting, and a matching header line
3511 contained an unqualified address, that address was qualified, even if it
3512 did not match any rewriting rules. The underlying bug was that the values
3513 of the flags that permit the existence of unqualified sender and recipient
3514 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3515 non-local messages, and by -bnq for local messages) were not being
3516 preserved with the message after it was received.
3518 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3519 "next input=" as part of the text comprising the host identity instead of
3520 the correct text. The code was using the same buffer for two different
3521 strings. However, depending on which order the printing function evaluated
3522 its arguments, the bug did not always show up. Under Linux, for example, my
3523 test suite worked just fine.
3525 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3526 doesn't work with some older versions of Perl. It has been changed to "my",
3527 which in any case is probably the better facility to use.
3529 15. A really picky compiler found some instances of statements for creating
3530 error messages that either had too many or two few arguments for the format
3533 16. The size of the buffer for calls to the DNS resolver has been increased
3534 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3535 for addresses that have a lot of PTR records. This alleviates a problem; it
3536 does not fully solve it.
3538 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3539 buffer now truncates the list and logs the incident, which is the same
3540 action as happens when Exim is looking up a host name and its aliases.
3541 Previously in this situation something unpredictable would happen;
3542 sometimes it was "internal error: store_reset failed".
3544 18. If a server dropped the connection unexpectedly when an Exim client was
3545 using GnuTLS and trying to read a response, the client delivery process
3546 crashed while trying to generate an error log message.
3548 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3549 string, for example:
3551 warn message = H1: something\nH2: something
3553 the text was added as a single header line from Exim's point of view
3554 though it ended up OK in the delivered message. However, searching for the
3555 second and subsequent header lines using $h_h2: did not work. This has been
3556 fixed. Similarly, if a system filter added multiple headers in this way,
3557 the routers could not see them.
3559 20. Expanded the error message when iplsearch is called with an invalid key to
3560 suggest using net-iplsearch in a host list.
3562 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3563 ACLs are no longer actually imposed (and a message to that effect is
3566 22. If a "gecos" field in a passwd entry contained escaped characters, in
3567 particular, if it contained a \" sequence, Exim got it wrong when building
3568 a From: or a Sender: header from that name. A second bug also caused
3569 incorrect handling when an unquoted " was present following a character
3570 that needed quoting.
3572 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3573 was not being matched caselessly.
3575 24. Arranged for all hyphens in the exim.8 source to be escaped with
3578 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3579 properly. Recipient callout cache records were still being keyed to include
3580 the sender, even when use_sender was set false. This led to far more
3581 callouts that were necessary. The sender is no longer included in the key
3582 when use_sender is false.
3584 26. Added "control = submission" modifier to ACLs.
3586 27. Added the ${base62d: operator to decode base 62 numbers.
3588 28. dnsdb lookups can now access SRV records.
3590 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3591 the configuration file.
3593 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3594 (-v) mode. This makes the output for a verbose queue run more intelligible.
3596 31. Added a use_postmaster feature to recipient callouts.
3598 32. Added the $body_zerocount variable, containing the number of binary zero
3599 bytes in the message body.
3601 33. The time of last modification of the "new" subdirectory is now used as the
3602 "mailbox time last read" when there is a quota error for a maildir
3605 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3607 35. Added +ignore_unknown as a special item in host lists.
3609 36. Code for decoding IPv6 addresses in host lists is now included, even if
3610 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3611 address was recognized as an IP address, but was then not correctly decoded
3612 into binary, causing unexpected and incorrect effects when compared with
3619 1. Very minor rewording of debugging text in manualroute to say "list of
3620 hosts" instead of "hostlist".
3622 2. If verify=header_syntax was set, and a header line with an unqualified
3623 address (no domain) and a large number of spaces between the end of the
3624 name and the colon was received, the reception process suffered a buffer
3625 overflow, and (when I tested it) crashed. This was caused by some obsolete
3626 code that should have been removed. The fix is to remove it!
3628 3. When running in the test harness, delay a bit after writing a bounce
3629 message to get a bit more predictability in the log output.
3631 4. Added a call to search_tidyup() just before forking a reception process. In
3632 theory, someone could use a lookup in the expansion of smtp_accept_max_
3633 per_host which, without the tidyup, could leave open a database connection.
3635 5. Added the variables $recipient_data and $sender_data which get set from a
3636 lookup success in an ACL "recipients" or "senders" condition, or a router
3637 "senders" option, similar to $domain_data and $local_part_data.
3639 6. Moved the writing of debug_print from before to after the "senders" test
3642 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3643 problems for message scanning, either using a data ACL, or using
3644 local_scan() because the Received: header was not generated till after they
3645 were called (in order to set the time as the time of reception completion).
3646 I have revised the way this works. The header is now generated after the
3647 body is received, but before the ACL or local_scan() are called. After they
3648 are run, the timestamp in the header is updated.
3654 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3655 before starting a queue runner without re-exec. This happened only when
3656 deliver_drop_privilege was set or when the Exim user was set to root. The
3657 effect of the bug was that timeouts during subsequent deliveries caused
3658 crashes instead of being properly handled. The handler is now left at its
3659 default (and expected) setting.
3661 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3662 message, again when deliver_drop_privilege is set or Exim is run as root.
3663 The bug described in (1) was not present in this case, but the tidying up
3664 of the other signals was missing. I have made the two cases consistent.
3666 3. The ignore_target_hosts setting on a manualroute router was being ignored
3667 for hosts that were looked up using the /MX notation.
3669 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3672 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3673 operated on the sender address. After changing the $sender_address to <>
3674 for the sender address verify, Exim was re-instated it as the original
3675 (before rewriting) address, but remembering that it had rewritten it, so it
3676 wasn't rewriting it again. This bug also had the effect of breaking the
3677 sender address verification caching when the sender address was rewritten.
3679 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3680 This has been changed so that if the ip literal address matches
3681 ignore_target_hosts, the router declines.
3683 7. Added expansion conditions match_domain, match_address, and match_local_
3684 part (NOT match_host).
3686 8. The placeholder for the Received: header didn't have a length field set.
3688 9. Added code to Exim itself and to exim_lock to test for a specific race
3689 condition that could lead to file corruption when using MBX delivery. The
3690 issue is with the lockfile that is created in /tmp. If this file is removed
3691 after a process has opened it but before that process has acquired a lock,
3692 there is the potential for a second process to recreate the file and also
3693 acquire a lock. This could lead to two Exim processes writing to the file
3694 at the same time. The added code performs the same test as UW imapd; it
3695 checks after acquiring the lock that its file descriptor still refers to
3696 the same named file.
3698 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3699 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3702 11. Added the smtp_active_hostname option. If used, this will typically be made
3703 to depend on the incoming interface address. Because $interface_address is
3704 not set up until the daemon has forked a reception process, error responses
3705 that can happen earlier (such as "too many connections") no longer contain
3708 12. If an expansion in a condition on a "warn" statement fails because a lookup
3709 defers, the "warn" statement is abandoned, and the next ACL statement is
3710 processed. Previously this caused the whole ACL to be aborted.
3712 13. Added the iplsearch lookup type.
3714 14. Added ident_timeout as a log selector.
3716 15. Added tls_certificate_verified as a log selector.
3718 16. Added a global option tls_require_ciphers (compare the smtp transport
3719 option of the same name). This controls incoming TLS connections.
3721 17. I finally figured out how to make tls_require_ciphers do a similar thing
3722 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3723 before starting the TLS session.
3725 18. Tabs are now shown as \t in -bP output.
3727 19. If the log selector return_path_on_delivery was set, Exim crashed when
3728 bouncing a message because it had too many Received: header lines.
3730 20. If two routers both had headers_remove settings, and the first one included
3731 a superfluous trailing colon, the final name in the first list and the
3732 first name in the second list were incorrectly joined into one item (with a
3733 colon in the middle).
3739 1. Added -C and -D options to the exinext utility, mainly to make it easier
3740 to include in the automated testing, but these could be helpful when
3741 multiple configurations are in use.
3743 2. The exinext utility was not formatting the output nicely when there was
3744 an alternate port involved in the retry record key, nor when there was a
3745 message id as well (for retries that were specific to a specific message
3746 and a specific host). It was also confused by IPv6 addresses, because of
3747 the additional colons they contain. I have fixed the IPv4 problem, and
3748 patched it up to do a reasonable job for IPv6.
3750 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3751 delivery, the log line now contains "pipelined" if PIPELINING was used.
3753 4. An SMTP transport process used to panic and die if the bind() call to set
3754 an explicit outgoing interface failed. This has been changed; it is now
3755 treated in the same way as a connect() failure.
3757 5. A reference to $sender_host_name in the part of a conditional expansion
3758 that was being skipped was still causing a DNS lookup. This no longer
3761 6. The def: expansion condition was not recognizing references to header lines
3762 that used bh_ and bheader_.
3764 7. Added the _cache feature to named lists.
3766 8. The code for checking quota_filecount in the appendfile transport was
3767 allowing one more file than it should have been.
3769 9. For compatibility with Sendmail, the command line option
3777 and sets the incoming protocol and host name (for trusted callers). The
3778 host name and its colon can be omitted when only the protocol is to be set.
3779 Note the Exim already has two private options, -pd and -ps, that refer to
3780 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3781 "s", but I don't think that's a major issue.
3783 10. A number of refactoring changes to the code, none of which should affect
3786 (a) The number of logging options was getting close to filling up the
3787 32-bit word that was used as a bit map. I have split them into two classes:
3788 those that are passed in the argument to log_write(), and those that are
3789 only ever tested independently outside of that function. These are now in
3790 separate 32-bit words, so there is plenty of room for expansion again.
3791 There is no change in the user interface or the logging behaviour.
3793 (b) When building, for example, log lines, the code previously used a
3794 macro that called string_cat() twice, in order to add two strings. This is
3795 not really sufficiently general. Furthermore, there was one instance where
3796 it was actually wrong because one of the argument was used twice, and in
3797 one call a function was used. (As it happened, calling the function twice
3798 did not affect the overall behaviour.) The macro has been replaced by a
3799 function that can join an arbitrary number of extra strings onto a growing
3802 (c) The code for expansion conditions now uses a table and a binary chop
3803 instead of a serial search (which was left over from when there were very
3804 few conditions). Also, it now recognizes conditions like "pam" even when
3805 the relevant support is not compiled in: a suitably worded error message is
3806 given if an attempt is made to use such a condition.
3808 11. Added ${time_interval:xxxxx}.
3810 12. A bug was causing one of the ddress fields not to be passed back correctly
3811 from remote delivery subprocesses. The field in question was not being
3812 subsequently used, so this caused to problems in practice.
3814 13. Added new log selectors queue_time and deliver_time.
3816 14. Might have fixed a bug in maildirsizefile handling that threw up
3817 "unexpected character" debug warnings, and recalculated the data
3818 unnecessarily. In any case, I expanded the warning message to give more
3821 15. Added the message "Restricted characters in address" to the statements in
3822 the default ACL that block characters like @ and % in local parts.
3824 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3825 Three changes have been made:
3827 (a) There was a serious bug; a negative response to MAIL caused the whole
3828 recipient domain to be cached as invalid, thereby blocking all messages
3829 to all local parts at the same domain, from all senders. This bug has
3830 been fixed. The domain is no longer cached after a negative response to
3831 MAIL if the sender used is not empty.
3833 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3836 (c) A new callout option, "use_sender" has been added for people who want
3837 the modified behaviour.
3843 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3846 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3847 indeed breaks things for older releases.
3849 3. Added additional logging to the case where there is a problem reading data
3850 from a filter that is running in a subprocess using a pipe, in order to
3851 try to track down a specific problem.
3853 4. Testing facility fudge: when running in the test harness and attempting
3854 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3855 getting "No route to host". Convert this to a timeout.
3857 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3860 6. Some OS don't have socklen_t but use size_t instead. This affects the
3861 fifth argument of getsockopt() amongst other things. This is now
3862 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3863 can be set for individual OS. I have set it for SunOS5, OSF1, and
3864 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3865 some earlier ones do not.
3867 7. Change 4.30/15 was not doing the test caselessly.
3869 8. The standard form for an IPv6 address literal was being rejected by address
3870 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3871 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3872 this, as well as the form without the "IPv6" on the front (but only when
3873 address literals are enabled, of course).
3875 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3877 10. Exim crashed if a message with an empty sender address specified by -f
3878 encountered a router with an errors_to setting. This could be provoked only
3879 by a command such as
3883 where an empty string was supplied; "<>" did not hit this bug.
3885 11. Installed PCRE release 4.5.
3887 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3888 remained set. It is now erased.
3890 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3891 times from message ids (which are base 36 rather than the normal 62).
3893 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3894 were being counted as actual protocol errors, and logged if the log
3895 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3896 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3897 and DATA following a set of rejected RCPTs do not count as protocol errors.
3898 In other words, Exim assumes they were pipelined, though this may not
3899 actually be the case. Of course, in all cases the client gets an
3900 appropriate error code.
3902 15. If a lookup fails in an ACL condition, a message about the failure may
3903 be available; it is used if testing the ACL cannot continue, because most
3904 such messages specify what the cause of the deferral is. However, some
3905 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3906 that caused an old message to be retained and used if a later statement
3907 caused a defer, replacing the real cause of the deferral.
3909 16. If an IP address had so many PTR records that the DNS lookup buffer
3910 was not large enough to hold them, Exim could crash while trying to process
3911 the truncated data. It now detects and logs this case.
3913 17. Further to 4.21/58, another change has been made: if (and only if) the
3914 first line of a message (the first header line) ends with CRLF, a bare LF
3915 in a subsequent header line has a space inserted after it, so as not to
3916 terminate the header.
3918 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3919 unnecessarily, used atoi() instead of strtol(), and didn't check the
3920 termination when getting file sizes from file names by regex.
3922 19. Completely re-implemented the support for maildirsize files, in the light
3923 of a number of problems with the previous contributed implementation
3924 (4.30/29). In particular:
3926 . If the quota is zero, the maildirsize file is maintained, but no quota is
3929 . If the maildir directory does not exist, it is created before any attempt
3930 to write a maildirsize file.
3932 . The quota value in the file is just a cache; if the quota is changed in
3933 the transport, the new value overrides.
3935 . A regular expression is available for excluding directories from the
3938 20. The autoreply transport checks the characters in options that define the
3939 message's headers; it allows continued headers, but it was checking with
3940 isspace() after an embedded newline instead of explicitly looking for a
3943 21. If all the "regular" hosts to which an address was routed had passed their
3944 expiry times, and had not reached their retry times, the address was
3945 bounced, even if fallback hosts were defined. Now Exim should go on to try
3948 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3949 equivalent code in the SMTP transport. Some hosts send humungous responses
3950 to HELO/EHLO, more than 1024 it seems.
3952 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3953 gives compiler warnings in some environments. I've now done it "properly",
3956 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3957 (because of problems with the built-in one) was declared to return uschar *
3958 instead of char *, causing compiler failure.
3960 25. Fixed a file descriptor leak when processing alias/forward files.
3962 26. Fixed a minor format string issue in dbfn.c.
3964 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3966 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3967 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3970 29. When a sender address is verified, it is cached, to save repeating the test
3971 when there is more than one recipient in a message. However, when the
3972 verification involves a callout, it is possible for different callout
3973 options to be set for different recipients. It is too complicated to keep
3974 track of this in the cache, so now Exim always runs a verification when a
3975 callout is required, relying on the callout cache for the optimization.
3976 The overhead is duplication of the address routing, but this should not be
3979 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3980 to be verified with callout=postmaster, and the main callout worked but the
3981 postmaster check failed, the verification correctly failed. However, if a
3982 subsequent RCPT command asked for sender verification *without* the
3983 postmaster check, incorrect caching caused this verification also to fail,
3986 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3987 it was not caching the DNS options (qualify_single, search_parents) that
3988 were used when the lookup failed. A subsequent lookup with different
3989 options therefore always gave the same answer, though there were cases
3990 where it should not have. (Example: a "domains = !$mx_any" option on a
3991 dnslookup router: the "domains" option is always processed without any
3992 widening, but the router might have qualify_single set.) Now Exim uses the
3993 cached value only when the same options are set.
3995 32. Added John Jetmore's "exipick" utility to the distribution.
3997 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3998 than a timeout (e.g. a certificate is required, and is not provided), an
3999 Exim server now closes the connection immediately. Previously it waited for
4000 the client to close - but if the client is SSL, it seems that they each
4001 wait for each other, leading to a delay before one of them times out.
4003 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4004 maintained 0.8.x compatibility because I don't think many are using it, and
4005 it is clearly obsolete.
4007 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4010 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4011 client certificate was expired. A simple patch fixes this, though I don't
4012 understand the full logic of why the verify callback is called multiple
4015 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4016 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4017 which causes problems with some clients (such as the Certicom SSL Plus
4018 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4019 disables the coutermeasure allowing Eudora to connect."
4021 38. Exim was not checking that a write() to a log file succeeded. This could
4022 lead to Bad Things if a log got too big, in particular if it hit a file
4023 size limit. Exim now panics and dies if it cannot write to a log file, just
4024 as it does if it cannot open a log file.
4026 39. Modified OS/Makefile-Linux so that it now contains
4028 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4030 The two -D definitions ensure that Exim is compiled with large file
4031 support, which makes it possible to handle log files that are bigger than
4034 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4035 instance) a domain was checked against a named list that involved a lookup,
4036 causing $domain_data to be set, then another domain was checked against the
4037 same list, then the first domain was re-checked, the value of $domain_data
4038 after the final check could be wrong. In particular, if the second check
4039 failed, it could be set empty. This bug probably also applied to
4042 41. The strip_trailing_dot option was not being applied to the address given
4043 with the -f command-line option.
4045 42. The code for reading a message's header from the spool was incrementing
4046 $received_count, but never initializing it. This meant that the value was
4047 incorrect (doubled) while delivering a message in the same process in which
4048 it was received. In the most common configuration of Exim, this never
4049 happens - a fresh exec is done - but it can happen when
4050 deliver_drop_privilege is set.
4052 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4053 it now includes up to 150 characters of the unexpected data in the log
4056 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4057 and building data strings. The size of both of these buffers was 10 000
4058 bytes - far larger than anybody would *ever* want, thought I. Needless to
4059 say, somebody hit the limit. I have increased the maximum line length to
4060 20 000 and the maximum data length of concatenated lines to 100 000. I have
4061 also fixed two bugs, because there was no checking on these buffers. Tsk,
4062 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4063 buffer is too small.
4065 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4066 lsearch lookups. Now it does.
4068 46. When parsing a route_list item in a manualroute router, a fixed-length
4069 buffer was used for the list of hosts. I made this 1024 bytes long,
4070 thinking that nobody would ever have a list of hosts that long. Wrong.
4071 Somebody had a whole pile of complicated expansion conditions, and the
4072 string was silently truncated, leading to an expansion error. It turns out
4073 that it is easier to change to an unlimited length (owing to other changes
4074 that have happened since this code was originally written) than to build
4075 structure for giving a limitation error. The length of the item that
4076 expands into the list of hosts is now unlimited.
4078 47. The lsearch lookup could not handle data where the length of text line was
4079 more than 4095 characters. Such lines were truncated, leading to shortened
4080 data being returned. It should now handle lines of any length.
4082 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4083 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4086 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4087 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4088 no longer happens in comment lines. A list of replacements is now placed
4089 at the head of all of the source files, except those whose only change is
4090 to replace PERL_COMMAND in the very first #! line.
4092 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4093 messages on the queue, with a bottom-up merge sort, using code contributed
4094 by Michael Haardt. This should make operations like -bp somewhat faster on
4095 large queues. It won't affect queue runners, except when queue_run_in_order
4098 51. Installed eximstats 1.31 in the distribution.
4100 52. Added support for SRV lookups to the dnslookup router.
4102 53. If an ACL referred to $message_body or $message_body_end, the value was not
4103 reset for any messages that followed in the same SMTP session.
4105 54. The store-handling optimization for building very long strings was not
4106 differentiating between the different store pools. I don't think this
4107 actually made any difference in practice, but I've tidied it.
4109 55. While running the routers to verify a sender address, $sender_address
4110 was still set to the sender address. This is wrong, because when routing to
4111 send a bounce to the sender, it would be empty. Therefore, I have changed
4112 it so that, while verifying a sender address, $sender_address is set to <>.
4113 (There is no change to what happens when verifying a recipient address.)
4115 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4116 target A or AAAA records (if not already returned) without resetting the
4117 qualify_single or search_parents options of the DNS resolver. These are
4118 inappropriate in this case because the targets of MX and SRV records must
4119 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4120 target that, when qualified, matched something in the local domain. These
4121 two options are now turned off when doing these lookups.
4123 57. It seems that at least some releases of Reiserfs (which does not have the
4124 concept of a fixed number of inodes) returns zero and not -1 for the
4125 number of available inodes. This interacted badly with check_spool_inodes,
4126 which assumed that -1 was the "no such thing" setting. What I have done is
4127 to check that the total number of inodes is greater than zero before doing
4128 the test of how many are available.
4130 58. When a "warn" ACL statement has a log_message modifier, the message is
4131 remembered, and not repeated. This is to avoid a lot of repetition when a
4132 message has many recipients that cause the same warning to be written.
4133 Howewer, Exim was preserving the list of already written lines for an
4134 entire SMTP session, which doesn't seem right. The memory is now reset if a
4135 new message is started.
4137 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4138 debugging output unless log_rewrite was also set.
4140 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4141 of (int)(handle) when we know that handle contains (void *)(-1).
4143 61. The Exim daemon panic-logs an error return when it closes the incoming
4144 connection. However "connection reset by peer" seems to be common, and
4145 isn't really an error worthy of noting specially, so that particular error
4148 62. When Exim is trying to find all the local interfaces, it used to panic and
4149 die if the ioctl to get the interface flags failed. However, it seems that
4150 on at least one OS (Solaris 9) it is possible to have an interface that is
4151 included in the list of interfaces, but for which you get a failure error
4152 for this call. This happens when the interface is not "plumbed" into a
4153 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4154 failure of the "get flags" call assumes that the interface is down.
4156 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4157 makes life easier for people who are doing arithmetic on fields extracted
4158 from dates, where you often get leading zeros that should not be
4159 interpreted as octal.
4161 64. Added qualify_domain to the redirect router, to override the global
4164 65. If a pathologically long header line contained very many addresses (the
4165 report of this problem mentioned 10 000) and each of them was rewritten,
4166 Exim could use up a very large amount of memory. (It kept on making new
4167 copies of the header line as it rewrote, and never released the old ones.)
4168 At the expense of a bit more processing, the header rewriting function has
4169 been changed so that it no longer eats memory in this way.
4171 66. The generation of the Received: header has been moved from the time that a
4172 message starts to be received, to the time that it finishes. The timestamp
4173 in the Received: header should now be very close to that of the <= log
4174 line. There are two side-effects of this change:
4176 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4177 logged header lines no longer include the local Received: line, because
4178 it has not yet been created. The same applies to a copy of the message
4179 that is returned to a non-SMTP sender when a message is rejected.
4181 (b) When a filter file is tested using -bf, no additional Received: header
4182 is added to the test message. After some thought, I decided that this
4185 This change does not affect the value of $received_for. It is still set
4186 after address rewriting, but before local_scan() is called.
4188 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4190 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4191 gave an unhelpful panic error message, and a defer error. I have managed to
4192 change this behaviour so that it now rejects any supplied certificate,
4193 which seems right, as the list of acceptable certificates is empty.
4195 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4196 gave an unhelpful defer error. I have not managed to make this reject any
4197 supplied certificates, but the error message it gives is "no certificate
4198 supplied", which is not helpful.
4200 70. exigrep's output now also includes lines that are not associated with any
4201 message, but which match the given pattern. Implemented by a patch from
4202 Martin Sluka, which also tidied up the Perl a bit.
4204 71. Recipient callout verification, like sender verification, was using <> in
4205 the MAIL FROM command. This isn't really the right thing, since the actual
4206 sender may affect whether the remote host accepts the recipient or not. I
4207 have changed it to use the actual sender in the callout; this means that
4208 the cache record is now keyed on a recipient/sender pair, not just the
4209 recipient address. There doesn't seem to be a real danger of callout loops,
4210 since a callout by the remote host to check the sender would use <>.
4211 [SEE ABOVE: changed after hitting problems.]
4213 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4214 temporary errors. However, in the case of such a code being given after
4215 the end of a data transmission (i.e. after ".") Exim was failing to write
4216 a retry record for the message. (Yes, there was some broken host that was
4217 actually sending 8xx at this point.)
4219 73. An unknown lookup type in a host list could cause Exim to panic-die when
4220 the list was checked. (An example that provoked this was putting <; in the
4221 middle of a list instead of at the start.) If this happened during a DATA
4222 ACL check, a -D file could be left lying around. This kind of configuration
4223 error no longer causes Exim to die; instead it causes a defer errror. The
4224 incident is still logged to the main and panic logs.
4226 74. Buglet left over from Exim 3 conversion. The message "too many messages
4227 in one connection" was written to the rejectlog but not the mainlog, except
4228 when address rewriting (yes!) was being logged.
4230 75. Added write_rejectlog option.
4232 76. When a system filter was run not as root (that is, when system_filter_user
4233 was set), the values of the $n variables were not being returned to the
4234 main process; thus, they were not subsequently available in the $sn
4237 77. Added +return_path_on_delivery log selector.
4239 78. A connection timeout was being treated differently from recipients deferred
4240 when testing hosts_max_try with a message that was older than the host's
4241 retry timeout. (The host should not be counted, thus allowing all hosts to
4242 be tried at least once before bouncing.) This may have been the cause of an
4243 occasionally reported bug whereby a message would remain on the queue
4244 longer than the retry timeout, but would be bounced if a delivery was
4245 forced. I say "may" because I never totally pinned down the problem;
4246 setting up timeout/retry tests is difficult. See also the next item.
4248 79. The ultimate address timeout was not being applied to errors that involved
4249 a combination of host plus message (for example, a timeout on a MAIL
4250 command). When an address resolved to a number of possible hosts, and they
4251 were not all tried for each delivery (e.g. because of hosts_max_try), a
4252 message could remain on the queue longer than the retry timeout.
4254 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4257 81. Fixed an obscure SMTP outgoing bug which required at least the following
4258 conditions: (a) there was another message waiting for the same server;
4259 (b) the server returned 5xx to all RCPT commands in the first message so
4260 that the message was not completed; (c) the server dropped the connection
4261 or gave a negative response to the RSET that Exim sends to abort the
4262 transaction. The observed case was a dropped connection after DATA that had
4263 been sent in pipelining mode. That is, the server had advertised PIPELINING
4264 but was not implementing it correctly. The effect of the bug was incorrect
4265 behaviour, such as trying another host, and this could lead to a crash.
4271 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4272 and daemon.c were passed as pointers to ints; they should have been
4273 pointers to socklen_t variables (which are typically unsigned ints).
4275 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4278 3. Fixed a really odd bug that affected only the testing scheme; patching a
4279 certain fixed string in the binary changed the value of another string that
4280 happened to be identical to the end of the original first string.
4282 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4283 name", it returns that address as the IP address. On some operating
4284 systems (e.g. Solaris), it also passes back the IP address string as the
4285 "host name". However, on others (e.g. Linux), it passes back an empty
4286 string. Exim wasn't checking for this, and was changing the host name to an
4287 empty string, assuming it had been canonicized.
4289 5. Although rare, it is permitted to have more than one PTR record for a given
4290 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4291 all the names associated with an address, because they do in Solaris.
4292 However, it seems that they do not in Linux for data that comes from the
4293 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4294 I found this out when I moved to a new Linux workstation and tried to run
4295 the Exim test suite.
4297 To get round this problem I have changed the code so that it now does its
4298 own call to the DNS to look up PTR records when searching for a host name.
4299 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4300 addresses that are only in /etc/hosts are still found.
4302 This behaviour is, however, controlled by an option called host_lookup_
4303 order, which defaults to "bydns:byaddr". If people want to use the other
4304 order, or indeed, just use one or the other means of lookup, they can
4305 specify it in this variable.
4307 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4308 some operating systems, this comes back from gethostbyaddr() as an empty
4309 string, and this is what Exim used to test for. However, it seems that in
4310 other systems, "." is yielded. Exim now tests for this case too.
4312 7. The values of check_spool_space and check_log_space are now held internally
4313 as a number of kilobytes instead of an absolute number of bytes. If a
4314 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4315 kilobyte. This means that much larger values can be stored.
4317 8. Exim monitor: an attempt to get the action menu when not actually pointing
4318 at a message produces an empty menu entitled "No message selected". This
4319 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4320 no entries in it ("Shell widget menu has zero width and/or height"). So I
4321 have added a single, blank menu entry in this case.
4323 9. Added ${quote_local_part.
4325 10. MIME decoding is now applied to the contents of Subject: header lines when
4328 11. Now that a reference to $sender_host_address automatically causes a reverse
4329 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4330 host lookup before query-style lookups in lists that might use this
4331 variable. This has therefore been abolished, and the "net-" prefix is no
4332 longer necessary for query-style lookups.
4334 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4335 have been a typo for LFLAGS, so it has been changed.
4337 13. The install script calls Exim with "-C /dev/null" in order to find the
4338 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4339 to be output. Howeve, since Exim outputs its version number before the
4340 error, it didn't break the script. It just looked ugly. I fixed this by
4341 always allowing "-C /dev/null" if the caller is root.
4343 14. Ignore overlarge ACL variable number when reading spool file - insurance
4344 against a later release with more variables having written the file.
4346 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4347 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4348 this, as well as the form without the "IPv6" on the front.
4350 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4351 OS/Makefile-Darwin file.
4353 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4354 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4356 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4357 variable. However, some OS use ints for inodes. I've added cast to long int
4358 to get rid of the compiler warning.
4360 19. I had forgotten to lock out "/../" in configuration file names when
4361 ALT_CONFIG_PREFIX was set.
4363 20. Routers used for verification do not need to specify transports. However,
4364 if such a router generated a host list, and callout was configured, Exim
4365 crashed, because it could not find a port number from the (non-existent)
4366 transport. It now assumes port 25 in this circumstance.
4368 21. Added the -t option to exigrep.
4370 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4371 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4372 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4373 them is set, LOOKUP_LSEARCH is forced.
4375 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4377 24. Added sender and host information to the "rejected by local_scan()" log
4378 line; previously there was no indication of these.
4380 25. Added .include_if_exists.
4382 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4383 Linux. It turns out that not all file systems support this. Apparently some
4384 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4385 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4386 not supported on the file descriptor, is now ignored when Exim is trying to
4387 sync a directory. This applies only to Linux.
4389 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4391 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4392 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4395 29. Added support for maildirsize files from supplied patch (modified a bit).
4397 30. The use of :fail: followed by an empty string could lead Exim to respond to
4398 sender verification failures with (e.g.):
4400 550 Verification failed for <xxx>
4401 550 Sender verify failed
4403 where the first response line was missing the '-' that indicates it is not
4404 the final line of the response.
4406 31. The loop for finding the name of the user that called Exim had a hardwired
4407 limit of 10; it now uses the value of finduser_retries, which is used for
4408 all other user lookups.
4410 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4413 33. Exim was neglecting to zero errno before one call of strtol() when
4414 expanding a string and expecting an integer value. On some systems this
4415 resulted in spurious "integer overflow" errors. Also, it was casting the
4416 result into an int without checking.
4418 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4419 did not work. The code looks as if it has *never* worked, though it appears
4420 to have been documented since at least releast 1.62. I have made it work.
4422 35. The "timeout_DNS" error in retry rules, also documented since at least
4423 1.62, also never worked. As it isn't clear exactly what this means, and
4424 clearly it isn't a major issue, I have abolished the feature by treating it
4425 as "timeout", and writing a warning to the main and panic logs.
4427 36. The display of retry rules for -brt wasn't always showing the error code
4430 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4431 timeout_connect_A, timeout_connect_MX.
4433 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4434 to the empty sender.
4436 39. The daemon was not analysing the content of -oX till after it had closed
4437 stderr and disconnected from the controlling terminal. This meant that any
4438 syntax errors were only noted on the panic log, and the return code from
4439 the command was 0. By re-arranging the code a little, I've made the
4440 decoding happen first, so such errors now appear on stderr, and the return
4441 code is 1. However, the actual setting up of the sockets still happens in
4442 the disconnected process, so errors there are still only recorded on the
4445 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4446 connections (as happens on some IP stacks) was logged at start up time as
4447 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4448 it from "IPv6 and IPv4", which means that two separate sockets are being
4451 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4452 says whether AF_INET or AF_INET6 was passed as an argument.
4454 42. Exiwhat output was messed up when time zones were included in log
4457 43. Exiwhat now gives more information about the daemon's listening ports,
4458 and whether -tls-on-connect was used.
4460 44. The "port" option of the smtp transport is now expanded.
4462 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4463 silently ignored. Now an error message is written to the main and panic
4466 46. There's a new ACL modifier called "logwrite" which writes to a log file
4467 as soon as it is encountered.
4469 47. Added $local_user_uid and $local_user_gid at routing time.
4471 48. Exim crashed when trying to verify a sender address that was being
4474 49. Exim was recognizing only a space character after ".include". It now also
4475 recognizes a tab character.
4477 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4478 extracting the relevant information from the specification. The man page no
4479 longer contains scrambled data for the -d option, and I've added a section
4480 at the front about calling Exim under different names.
4482 51. Added "extra_headers" argument to the "mail" command in filter files.
4484 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4487 53. Installed eximstats 1.29.
4489 54. Added transport_filter_timeout as a generic transport option.
4491 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4492 Cc: header lines. This was required by RFC 822, but it not required by RFC
4495 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4496 incoming messages that did not have them. Now it does so only if the
4497 message originates locally, that is, if there is no associated remote host
4498 address. When Resent- header lines are present, this applies to the Resent-
4499 lines rather than the non-Resent- lines.
4501 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4502 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4504 58. Messages for configuration errors now include the name of the main
4505 configuration file - useful now that there may be more than one file in a
4506 list (.included file names were always shown).
4508 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4509 for those rare installations that do not start the daemon as root or run it
4510 setuid root. I've cut out the call to initgroups() if the daemon is not
4513 60. The Exim user and group can now be bound into the binary as text strings
4514 that are looked up at the start of Exim's processing.
4516 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4518 62. Added $mailstore_basename variable.
4520 63. Installed patch to sieve.c from Michael Haardt.
4522 64. When Exim failed to open the panic log after failing to open the main log,
4523 the original message it was trying to log was written to stderr and debug
4524 output, but if they were not available (the usual case in production), it
4525 was lost. Now it is written to syslog before the two lines that record the
4526 failures to open the logs.
4528 65. Users' Exim filters run in subprocesses under the user's uid. It is
4529 possible for a "deliver" command or an alias in a "personal" command to
4530 provoke an address rewrite. If logging of address rewriting is configured,
4531 this fails because the process is not running as root or exim. There may be
4532 a better way of dealing with this, but for the moment (because 4.30 needs
4533 to be released), I have disabled address rewrite logging when running a
4534 filter in a non-root, non-exim process.
4540 1. The buildconfig auxiliary program wasn't quoting the value set for
4541 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4542 not defined. This bug was masked in 4.22 by the effect that was fixed in
4545 2. Some messages that were rejected after a message id was allocated were
4546 shown as "incomplete" by exigrep. It no longer does this for messages that
4547 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4549 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4550 have allow_domain_literals set, the ID did not get logged in the <= line.
4551 Domain literals are now always recognized in Message-ID: header lines.
4553 4. The first argument for a ${extract expansion item is the key name or field
4554 number. Leading and trailing spaces in this item were not being ignored,
4555 causing some misleading effects.
4557 5. When deliver_drop_privilege was set, single queue runner processes started
4558 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4559 same command in the process it spins off) were not dropping privilege.
4561 6. When the daemon running as "exim" started a queue runner, it always
4562 re-executed Exim in the spun-off process. This is a waste of effort when
4563 deliver_drop_privilege is set. The new process now just calls the
4564 queue-runner function directly.
4570 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4573 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4574 ignored. Though the use of -oP was forcing the writing of a pid file, it
4575 was always written to the default place.
4577 3. If the message "no IP address found for host xxxx" is generated during
4578 incoming verification, it is now followed by identification of the incoming
4579 connection (so you can more easily find what provoked it).
4581 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4583 5. Added some features to "harden" Exim a bit more against certain attacks:
4585 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4586 be put in Local/Makefile. This is like the never_users runtime option,
4587 but it cannot be overridden. The default setting is "root".
4589 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4590 prefix string with which any file named in a -C command line option
4593 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4594 is retained for -C and -D only if the caller of Exim is root. Without
4595 it, the exim user may also use -C and -D and retain privilege.
4597 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4598 command line option is disabled.
4600 6. Macro names set by the -D option must start with an upper case letter, just
4601 like macro names defined in the configuration file.
4603 7. Added "dereference=" facility to LDAP.
4605 8. Two instances of the typo "uknown" in the source files are fixed.
4607 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4608 the Configure-Makefile script screwed up while processing it.
4610 10. Incorporated PCRE 4.4.
4612 11. The SMTP synchronization check was not operating right at the start of an
4613 SMTP session. For example, it could not catch a HELO sent before the client
4614 waited for the greeting. There is now a check for outstanding input at the
4615 point when the greeting is written. Because of the duplex, asynchronous
4616 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4617 way, but not yet received, when the check is performed.
4619 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4620 on TCP/IP sockets, because this apparently causes some broken clients to
4623 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4624 unchanged) from the Cygwin maintainer.
4626 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4627 was supported instead of testing for mbx. Effectively a typo.
4629 15. The spa authenticator server code was not checking that the input it
4630 received was valid base64.
4632 16. The debug output line for the "set" modifier in ACLs was not showing the
4633 name of the variable that was being set.
4635 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4637 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4638 lookup on its own. Something else was needed to trigger the lookup. For
4639 example, a match in host_lookup or the need for a host name in a host list.
4640 Now, if $sender_host_name is referenced and the host name has not yet been
4641 looked up, a lookup is performed. If the lookup fails, the variable remains
4642 empty, and $host_lookup_failed is set to "1".
4644 19. Added "eqi" as a case-independent comparison operator.
4646 20. The saslauthd authentication condition could segfault if neither service
4647 nor realm was specified.
4649 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4650 error message that was logged was misleading, and incoming SMTP
4651 connections were dropped. The message is now more accurate, and temporary
4652 errors are given to SMTP connections.
4654 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4655 (except RSET) with a 421 error, until QUIT is received. However, it was
4656 failing to send a response to QUIT.
4658 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4659 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4660 ACL is now run after the helo_try_verify_hosts code.
4662 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4663 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4664 case-independent, so other case variants are also recognized). Apparently
4665 some systems use these upper case variants.
4667 25. If more than two messages were waiting for the same host, and a transport
4668 filter was specified for the transport, Exim sent two messages over the
4669 same TCP/IP connection, and then failed with "socket operation on non-
4670 socket" when it tried to send the third.
4672 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4674 27. The extern definition of crypt16() in expand.c was not being excluded when
4675 the OS had its own crypt16() function.
4677 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4678 as a preferred synonym for return_size_limit, both as an option and as an
4681 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4683 30. Changed the default configuration ACL to relax the local part checking rule
4684 for addresses that are not in any local domains. For these addresses,
4685 slashes and pipe symbols are allowed within local parts, but the sequence
4686 /../ is explicitly forbidden.
4688 31. SPA server authentication was not clearing the challenge buffer before
4691 32. log_message in a "warn" ACL statement was writing to the reject log as
4692 well as to the main log, which contradicts the documentation and doesn't
4693 seem right (because no rejection is happening). So I have stopped it.
4695 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4696 However, I am unable to do any testing of this.
4698 34. Fixed an infelicity in the appendfile transport. When checking directories
4699 for a mailbox, to see if any needed to be created, it was accidentally
4700 using path names with one or more superfluous leading slashes; tracing
4701 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4703 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4704 contents are added to the log line that is written for every discarded
4705 recipient. (Previously a log_message setting was ignored.)
4707 36. The ${quote: operator now quotes the string if it is empty.
4709 37. The install script runs exim in order to find its version number. If for
4710 some reason other than non-existence or emptiness, which it checks, it
4711 could not run './exim', it was installing it with an empty version number,
4712 i.e. as "exim-". This error state is now caught, and the installation is
4715 38. An argument was missing from the function that creates an error message
4716 when Exim fails to connect to the socket for saslauthd authentication.
4717 This could cause Exim to crash, or give a corrupted message.
4719 39. Added isip, isip4, and isip6 to ${if conditions.
4721 40. The ACL variables $acl_xx are now saved with the message, and can be
4722 accessed later in routers, transports, and filters.
4724 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4725 strings in the file are not string-expanded.
4727 42. If a MAIL command specified a SIZE value that was too large to fit into an
4728 int variable, the check against message_size_limit failed. Such values are
4729 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4730 day this will have to be increased, but I don't think I want to be around
4731 when emails are that large.
4738 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4739 iconv() is not standard in FreeBSD.
4741 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4742 IPv6 enabled. The observed symptom was a segmentation fault on return from
4743 the function os_common_find_running_interfaces() in src/os.c.
4745 3. In the check_special_case() function in daemon.c I had used "errno" as an
4746 argument name, which causes warnings on some systems. This was basically a
4747 typo, since it was named "eno" in the comments!
4749 4. The code that waits for the clock to tick (at a resolution of some fraction
4750 of a second) so as to ensure message-id uniqueness was always waiting for
4751 at least one whole tick, when it could have waited for less. [This is
4752 almost certainly not relevant at current processor speeds, where it is
4753 unlikely to ever wait at all. But we try to future-proof.]
4755 5. The function that sleeps for a time interval that includes fractions of a
4756 second contained a race. It did not block SIGALRM between setting the
4757 timer, and suspending (a couple of lines later). If the interval was short
4758 and the sigsuspend() was delayed until after it had expired, the suspension
4759 never ended. On busy systems this could lead to processes getting stuck for
4762 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4763 process, before it forks any delivery processes. The open lookup caching
4764 mechanism meant that the open file or database connection was passed into
4765 the delivery process. The problem was that delivery processes always tidy
4766 up cached lookup data. This could cause a problem for the next delivery
4767 process started by the queue runner, because the external queue runner
4768 process does not know about the closure. So the next delivery process
4769 still has data in the lookup cache. In the case of a file lookup, there was
4770 no problem because closing a file descriptor in a subprocess doesn't affect
4771 the parent. However, if the lookup was caching a connection to a database,
4772 the connection was closed, and the second delivery process was likely to
4773 see errors such as "PGSQL: query failed: server closed the connection
4774 unexpectedly". The problem has been fixed by closing all cached lookups
4775 in a queue runner before running a delivery process.
4777 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4778 seem to have the "const" qualifier which it has on other OS. I've
4781 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4782 *of the same type* (client or server) with the same public name that an
4783 error should be diagnosed.
4785 9. When Exim looked up a host name for an IP address, but failed to find the
4786 original IP address when looking up the host name (a safety check), it
4787 output the message "<ip address> does not match any IP for NULL", which was
4788 confusing, to say the least. The bug was that the host name should have
4789 appeared instead of "NULL".
4791 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4792 user that is built into the binary, and the -C or -D options is used, root
4793 privilege is dropped before the configuration file is read. In addition,
4794 logging is switched to stderr instead of the normal log files. If the
4795 configuration then re-defines the Exim user, the unprivileged environment
4796 is probably not what is expected, so Exim logs a panic warning message (but
4799 However, if deliver_drop_privilege is set, the unprivileged state may well
4800 be exactly what is intended, so the warning has been cut out in that case,
4801 and Exim is allowed to try to write to its normal log files.
4807 1. smtp_return_error_details was not giving details for temporary sender
4808 or receiver verification errors.
4810 2. Diagnose a configuration error if two authenticators have the same public
4813 3. Exim used not to create the message log file for a message until the first
4814 delivery attempt. This could be confusing when incoming messages were held
4815 for policy or load reasons. The message log file is now created at the time
4816 the message is received, and an initial "Received" line is written to it.
4818 4. The automatically generated man page for command line options had a minor
4819 bug that caused no ill effects; however, a more serious problem was that
4820 the procedure for building the man page automatically didn't always
4821 operate. Consequently, release 4.20 contains an out-of-date version. This
4822 shouldn't happen again.
4824 5. When building Exim with embedded Perl support, the script that builds the
4825 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4826 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4828 6. The freeze_tell option was not being used for messages that were frozen on
4829 arrival, either by an ACL or by local_scan().
4831 7. Added the smtp_incomplete_transaction log selector.
4833 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4834 was accepting AUTH without a new EHLO.
4836 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4837 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4840 10. Logging of TCP/IP connections (when configured) now happens in the main
4841 daemon process instead of the child process, so that the TCP/IP connection
4842 count is more accurate (but it can never be perfect).
4844 11. The use of "drop" in a nested ACL was not being handled correctly in the
4845 outer ACL. Now, if condition failure induced by the nested "drop" causes
4846 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4847 or "require"), the connection is dropped.
4849 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4850 that yield "discard" can now be used with an "accept" or a "discard" verb,
4851 but an error is generated for any others (because I can't see a useful way
4852 to define what should happen).
4854 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4855 are now processed in the same way as lines in the Exim configuration file.
4856 In particular, continuation lines are supported.
4858 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4860 15. Added -ti meaning -t -i.
4862 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4863 domains, and warn by logging if others are found.
4865 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4866 returned from GIFCONF when Exim is trying to find the list of interfaces on
4867 a host. The code in os.c has been modified to copy each ifreq to an aligned
4868 structure in all cases.
4870 Also, in some cases, the returned ifreq's were being copied to a 'struct
4871 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4872 means the last couple of bytes of an IPv6 address could be chopped if the
4873 ifreq contained only a normal sockaddr (14 bytes storage).
4875 18. Named domain lists were not supported in the hosts_treat_as_local option.
4876 An entry such as +xxxx was not recognized, and was treated as a literal
4879 19. Ensure that header lines added by a DATA ACL are included in the reject log
4880 if the ACL subsequently rejects the message.
4882 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4883 MD5 (which is deprecated).
4885 21. When testing a filter file using -bf, Exim was writing a message when it
4886 took the sender from a "From " line in the message, but it was not doing so
4887 when it took $return_path from a Return-Path: header line. It now does.
4889 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4890 with a valid header line field name (a series of printing characters
4891 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4893 23. Changed "disc" in the source to "disk" to conform to the documentation and
4894 the book and for uniformity.
4896 24. Ignore Sendmail's -Ooption=value command line item.
4898 25. When execve() failed while trying to run a command in a pipe transport,
4899 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4900 could be confused with a return value of 69 from the command itself. This
4901 has been changed to 127, the value the shell returns if it is asked to run
4902 a non-existent command. The wording for the related log line suggests a
4903 non-existent command as the problem.
4905 26. If received_header_text expands to an empty string, do not add a Received:
4906 header line to the message. (Well, it adds a token one on the spool, but
4907 marks it "old" so that it doesn't get used or transmitted.)
4909 27. Installed eximstats 1.28 (addition of -nt option).
4911 28. There was no check for failure on the call to getsockname() in the daemon
4912 code. This can fail if there is a shortage of resources on the system, with
4913 ENOMEM, for example. A temporary error is now given on failure.
4915 29. Contrary to the C standard, it seems that in some environments, the
4916 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4917 program. Exim now does this explicitly; it affects the formatting of
4918 timestamps using strftime().
4920 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4921 complaints. I've now initialized all the variables, to avoid this.
4923 32. Header lines added by a system filter were not being "seen" during
4924 transport-time rewrites.
4926 33. The info_callback() function passed to OpenSSL is set up with type void
4927 (*)(SSL *, int, int), as described somewhere. However, when calling the
4928 function (actually a macro) that sets it up, the type void(*)() is
4929 expected. I've put in a cast to prevent warnings from picky compilers.
4931 34. If a DNS black list lookup found a CNAME record, but there were no A
4932 records associated with the domain it pointed at, Exim crashed.
4934 35. If a DNS black list lookup returned more than one A record, Exim ignored
4935 all but the first. It now scans all returned addresses if a particular IP
4936 value is being sought. In this situation, the contents of the
4937 $dnslist_value variable are a list of all the addresses, separated by a
4940 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4941 to accept a host name and all its aliases if the forward lookup for any of
4942 them yielded the IP address of the incoming connection. Now it accepts only
4943 those names whose forward lookup yields the correct IP address. Any other
4944 names are discarded. This closes a loophole whereby a rogue DNS
4945 administrator could create reverse DNS records to break through a
4946 wildcarded host restriction in an ACL.
4948 37. If a user filter or a system filter that ran in a subprocess used any of
4949 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4950 the wrong values were passed to the pipe command ($thisaddress had the
4951 value of $0, $0 had the value of $1, etc). This bug was introduced by
4952 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4954 38. Improved the line breaking for long SMTP error messages from ACLs.
4955 Previously, if there was no break point between 40 and 75 characters, Exim
4956 left the rest of the message alone. Two changes have been made: (a) I've
4957 reduced the minimum length to 35 characters; (b) if it can't find a break
4958 point between 35 and 75 characters, it looks ahead and uses the first one
4959 that it finds. This may give the occasional overlong line, but at least the
4960 remaining text gets split now.
4962 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4963 file descriptors might be low, and that setting 1000 would always raise it.
4964 It turns out that in some environments, the limit is already over 1000 and
4965 that lowering it causes trouble. So now Exim takes care not to decrease it.
4967 40. When delivering a message, the value of $return_path is set to $sender_
4968 address at the start of routing (routers may change the value). By an
4969 oversight, this default was not being set up when an address was tested by
4970 -bt or -bv, which affected the outcome if any router or filter referred to
4973 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4974 log only when "message" or "log_message" are set. However, if one of the
4975 conditions was an address verification, or a call to a nested ACL, the
4976 messages generated by the underlying test were being passed through. This
4977 no longer happens. The underlying message is available in $acl_verify_
4978 message for both "message" and "log_message" expansions, so it can be
4979 passed through if needed.
4981 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4982 new expansion $bh_ to give the encoded byte string without charset
4983 translation. Translation happens only if iconv() is available; HAVE_ICONV
4984 indicates this at build time. HEADERS_CHARSET gives the charset to
4985 translate to; headers_charset can change it in the configuration, and
4986 "headers charset" can change it in an individual filter file.
4988 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4989 that creates RFC 2047 encoded "words" labels them as that charset instead
4990 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4991 expansion operator; (ii) when Exim creates a From: line for a local
4992 message; (iii) when a header line is rewritten to include a "phrase" part.
4994 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4995 buggy, causing it to skip the first lines of messages whose message ID
4996 ended in 'D'. This would not have bitten before Exim release 4.14, because
4997 message IDs were unlikely to end in 'D' before then. The effect was to have
4998 incorrect size information for certain domains.
5000 45. #include "config.h" was missing at the start of the crypt16.c module. This
5001 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5004 46. If there was a timeout during a "random" callout check, Exim treated it as
5005 a failure of the random address, and carried on sending RSET and the real
5006 address. If the delay was just some slowness somewhere, the response to the
5007 original RCPT would be taken as a response to RSET and so on, causing
5008 mayhem of various kinds.
5010 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5011 when I implemented it. It didn't allow for the fact that some option values
5012 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5013 the right test for positive values.
5015 48. Domain names in DNS records are case-independent. Exim always looks them up
5016 in lower case. Some resolvers return domain names in exactly the case they
5017 appear in the zone file, that is, they may contain uppercase letters. Not
5018 all resolvers do this - some return always lower case. Exim was treating a
5019 change of case by a resolver as a change of domain, similar to a widening
5020 of a domain abbreviation. This triggered its re-routing code and so it was
5021 trying to route what was effectively the same domain again. This normally
5022 caused routing to fail (because the router wouldn't handle the domain
5023 twice). Now Exim checks for this case specially, and just changes the
5024 casing of the domain that it ultimately uses when it transmits the message
5027 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5030 50. If a filter generated a file delivery with a non-absolute name (possible if
5031 no home directory exists for the router), the forbid_file option was not
5034 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5035 the existing equality matching.
5037 52. Exim was using ints instead of ino_t variables in some places where it was
5038 dealing with inode numbers.
5040 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5041 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5042 called TMPDIR, and if it finds it is different, it changes its value.
5044 54. The smtp_printf() function is now made available to local_scan() so
5045 additional output lines can be written before returning. There is also an
5046 smtp_fflush() function to enable the detection of a dropped connection.
5047 The variables smtp_input and smtp_batched_input are exported to
5050 55. Changed the default runtime configuration: the message "Unknown user"
5051 has been removed from the ACL, and instead placed on the localuser router,
5052 using the cannot_route_message feature. This means that any verification
5053 failures that generate their own messages won't get overridden. Similarly,
5054 the "Unrouteable address" message that was in the ACL for unverifiable
5055 relay addresses has also been removed.
5057 56. Added hosts_avoid_esmtp to the smtp transport.
5059 57. The exicyclog script was not checking for the esoteric option
5060 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5061 will work only if exicyclog is run under the appropriate euid.
5063 58. Following a discussion on the list, the rules by which Exim recognises line
5064 endings on incoming messages have been changed. The -dropcr and drop_cr
5065 options are now no-ops, retained only for backwards compatibility. The
5066 following line terminators are recognized: LF CRLF CR. However, special
5067 processing applies to CR:
5069 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5070 nor a local message in the state where . is a terminator.
5072 (ii) If a bare CR is encountered in a header line, an extra space is added
5073 after the line terminator so as not to end the header. The reasoning
5074 behind this is that bare CRs in header lines are most likely either
5075 to be mistakes, or people trying to play silly games.
5077 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5078 was being incorrectly given as 18 bytes larger than it should have been.
5079 This is a VOB (very old bug).
5081 60. This may never have affected anything current, but just in case it has:
5082 When the local host is found other than at the start of a list of hosts,
5083 the local host, those with the same MX, and any that follow, are discarded.
5084 When the list in question was part of a longer list of hosts, the following
5085 hosts (not currently being processed) were also being discarded. This no
5086 longer happens. I'm not sure if this situation could ever has previously
5089 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5092 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5093 header. This is something that is recommended in a new Internet Draft, and
5094 is something that is documented as being done by Sendmail. There are two
5095 possible values. For messages generated by the autoreply transport, Exim
5098 Auto-Submitted: auto-replied
5100 whereas for all other generated messages (e.g. bounces) it adds
5102 Auto-Submitted: auto-generated
5104 63. The "personal" condition in filters now includes a test for the
5105 Auto-Submitted: header. If it contains the string "auto-" the message it
5106 not considered personal.
5108 64. Added rcpt_include_affixes as a generic transport option.
5110 65. Added queue_only_override (default true).
5112 66. Added the syslog_duplication option.
5114 67. If what should have been the first header line of a message consisted of
5115 a space followed by a colon, Exim was mis-interpreting it as a header line.
5116 It isn't of course - it is syntactically invalid and should therefore be
5117 treated as the start of the message body. The misbehaviour could have
5118 caused a number of strange effects, including loss of data in subsequent
5119 header lines, and spool format errors.
5121 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5122 client host had authenticated. This control can now be exercised by an ACL
5123 for more flexibility.
5125 69. By default, callouts do not happen when testing with -bh. There is now a
5126 variant, -bhc, which does actually run the callout code, including
5127 consulting and updating the callout cache.
5129 70. Added support for saslauthd authentication, courtesy of Alexander
5132 71. If statvfs() failed on the spool or log directories while checking their
5133 size for availability, Exim confusingly gave the error "space shortage".
5134 Furthermore, in debugging mode it crashed with a floating point exception.
5135 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5136 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5137 serious problem, Exim now writes to the main and panic logs when this
5138 happens, with details of the failure. It then refuses to accept the
5139 incoming message, giving the message "spool directory problem" or "log
5140 directory problem" with a 421 code for SMTP messages.
5142 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5143 0, 1, and 2 exist, because some OS complain for execs without them (see
5144 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5145 descriptors. However, the code omitted to check that the open succeeded,
5146 causing mysterious errors if for some reason the permissions on /dev/null
5147 got screwed. Now Exim writes a message to the main and panic logs, and
5148 bombs out if it can't open /dev/null.
5150 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5151 interact so that it is all more flexible. It is supposed to remain
5152 backwards compatible. Also added extra_local_interfaces.
5154 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5155 to bomb out with an assertion failure - to the client this appears as a
5156 connection drop. This problem occurs in the part of the code that was taken
5157 from the Samba project. Fortunately, the assertion is in a very simple
5158 function, so I have fixed this by reproducing the function inline in the
5159 one place where it is called, and arranging for authentication to fail
5160 instead of killing the process with assert().
5162 75. The SPA client code was not working when the server requested OEM rather
5163 than Unicode encoding.
5165 76. Added code to make require_files with a specific uid setting more usable in
5166 the case where statting the file as root fails - usually a non-root-mounted
5167 NFS file system. When this happens and the failure is EACCES, Exim now
5168 forks a subprocess and does the per-uid checking as the relevant uid.
5170 77. Added process_log_path.
5172 78. If log_file_path was not explicitly set, a setting of check_log_space or
5173 check_log_inodes was ignored.
5175 79. If a space check for the spool or log partitions fails, the incident is now
5176 logged. Of course, in the latter case the data may get lost...
5178 80. Added the %p formatting code to string_format() so that it can be used to
5179 print addresses in debug_print(). Adjusted all the address printing in the
5180 debugging in store.c to use %p rather than %d.
5182 81. There was a concern that a line of code in smtp_in.c could overflow a
5183 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5184 initially expressed, the concern was not well-founded, because trailing
5185 spaces are removed early. However, if the trailing spaces were followed by
5186 a NULL, they did not get removed, so the overflow was possible. Two fixes
5189 (a) I re-wrote the offending code in a cleaner fashion.
5190 (b) If an incoming SMTP command contains a NULL character, it is rejected
5193 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5194 runs initgroups(), so that if the Exim user is in any additional groups,
5195 they will be used during message reception.
5201 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob