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.
15 TF/03 Include the error message when we fail to unlink a spool file.
17 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
18 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
19 for maintaining out-of-tree patches for some time.
21 PP/01 Bugzilla 139: Documentation and portability issues.
22 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
23 Handle per-OS dynamic-module compilation flags.
25 PP/02 Let /dev/null have normal permissions.
26 The 4.73 fixes were a little too stringent and complained about the
27 permissions on /dev/null. Exempt it from some checks.
28 Reported by Andreas M. Kirchwitz.
30 PP/03 Report version information for many libraries, including
31 Exim version information for dynamically loaded libraries. Created
32 version.h, now support a version extension string for distributors
33 who patch heavily. Dynamic module ABI change.
35 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
36 privilege escalation vulnerability whereby the Exim run-time user
37 can cause root to append content of the attacker's choosing to
40 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
47 PP/01 Date: & Message-Id: revert to normally being appended to a message,
48 only prepend for the Resent-* case. Fixes regression introduced in
49 Exim 4.70 by NM/22 for Bugzilla 607.
51 PP/02 Include check_rfc2047_length in configure.default because we're seeing
52 increasing numbers of administrators be bitten by this.
54 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
56 PP/03 Bugzilla 994: added openssl_options main configuration option.
58 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
60 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
62 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
64 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
65 without a peer certificate, leading to a segfault because of an
66 assumption that peers always have certificates. Be a little more
67 paranoid. Problem reported by Martin Tscholak.
69 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
70 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
71 NB: ClamAV planning to remove STREAM in "middle of 2010".
72 CL also introduces -bmalware, various -d+acl logging additions and
73 more caution in buffer sizes.
75 PP/09 Implemented reverse_ip expansion operator.
77 PP/10 Bugzilla 937: provide a "debug" ACL control.
79 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
81 PP/12 Bugzilla 973: Implement --version.
83 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
85 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
87 PP/15 Bugzilla 816: support multiple condition rules on Routers.
89 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
90 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
91 ignore trailing whitespace.
93 JJ/02 prevent non-panic DKIM error from being sent to paniclog
95 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
98 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
99 Notification from Dr Andrew Aitchison.
101 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
102 ExtendedDetectionInfo response format.
103 Notification from John Horne.
105 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
108 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
109 XSL and documented dependency on system catalogs, with examples of how
112 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
115 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
116 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
117 configuration file which is writeable by the Exim user or group.
119 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
120 of configuration files to cover files specified with the -C option if
121 they are going to be used with root privileges, not just the default
124 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
125 option (effectively making it always true).
127 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
128 files to be used while preserving root privileges.
130 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
131 that rogue child processes cannot use them.
133 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
134 run-time user, instead of root.
136 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
137 Exim run-time user without dropping privileges.
139 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
140 result string, instead of calling string_vformat() twice with the same
143 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
144 for other users. Others should always drop root privileges if they use
145 -C on the command line, even for a whitelisted configure file.
147 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
149 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
155 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
156 $data_path, and $header_path variables; fixed documentation bugs and
159 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
160 exipick to access non-standard spools, including the "frozen" queue
163 NM/01 Bugzilla 965: Support mysql stored procedures.
164 Patch from Alain Williams
166 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
168 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
169 Patch from Andreas Metzler
171 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
172 Patch from Kirill Miazine
174 NM/05 Bugzilla 671: Added umask to procmail example.
176 JJ/03 installed exipick 20100323.0, fixing doc bug
178 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
179 directory. Notification and patch from Dan Rosenberg.
181 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
183 TK/02 Improve log output when DKIM signing operation fails.
185 MH/01 Treat the transport option dkim_domain as a colon separated
186 list, not as a single string, and sign the message with each element,
187 omitting multiple occurences of the same signer.
189 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
190 Bugzilla 985, 986. Patch by Simon Arlott
192 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
193 Patch by Simon Arlott
195 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
196 MBX locking. Notification from Dan Rosenberg.
202 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
204 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
206 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
208 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
210 NM/04 Bugzilla 915: Flag broken perl installation during build.
216 TK/01 Added patch by Johannes Berg that expands the main option
217 "spamd_address" if it starts with a dollar sign.
219 TK/02 Write list of recipients to X-Envelope-Sender header when building
220 the mbox-format spool file for content scanning (suggested by Jakob
223 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
224 (http://www.dcc-servers.net/) support via dccifd. Activated by
225 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
227 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
228 by Mark Daniel Reidel <mr@df.eu>.
230 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
231 When building exim an external PCRE library is now needed -
232 PCRE is a system library on the majority of modern systems.
233 See entry on PCRE_LIBS in EDITME file.
235 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
236 conversation. Added nologin parameter to request.
237 Patch contributed by Kirill Miazine.
239 TF/01 Do not log submission mode rewrites if they do not change the address.
241 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
243 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
244 log files in place. Contributed by Roberto Lima.
246 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
248 TF/03 Bugzilla 615: When checking the local_parts router precondition
249 after a local_part_suffix or local_part_prefix option, Exim now
250 does not use the address's named list lookup cache, since this
251 contains cached lookups for the whole local part.
253 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
254 Robert Millan. Documentation is in experimental-spec.txt.
256 TF/04 Bugzilla 668: Fix parallel build (make -j).
258 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
260 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
261 Patch provided by Jan Srzednicki.
263 TF/05 Leading white space used to be stripped from $spam_report which
264 wrecked the formatting. Now it is preserved.
266 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
267 that they are available at delivery time.
269 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
271 TF/08 TLS error reporting now respects the incoming_interface and
272 incoming_port log selectors.
274 TF/09 Produce a more useful error message if an SMTP transport's hosts
275 setting expands to an empty string.
277 NM/06 Bugzilla 744: EXPN did not work under TLS.
278 Patch provided by Phil Pennock.
280 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
281 Patch provided by Richard Godbee.
283 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
284 acl_smtp_notquit, added index entry.
286 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
287 Patch provided by Eugene Bujak.
289 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
290 accept(). Patch provided by Maxim Dounin.
292 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
293 Patch provided by Phil Pennock.
295 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
297 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
298 Patch provided by Brad "anomie" Jorsch.
300 NM/14 Bugzilla 622: Added timeout setting to transport filter.
301 Patch provided by Dean Brooks.
303 TK/05 Add native DKIM support (does not depend on external libraries).
305 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
306 Patch provided by Graeme Fowler.
308 NM/16 Bugzilla 851: Documentation example syntax fix.
310 NM/17 Changed NOTICE file to remove references to embedded PCRE.
312 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
315 NM/19 Bugzilla 745: TLS version reporting.
316 Patch provided by Phil Pennock.
318 NM/20 Bugzilla 167: bool: condition support.
319 Patch provided by Phil Pennock.
321 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
322 clients. Patch provided by Phil Pennock.
324 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
325 Patch provided by Brad "anomie" Jorsch.
327 NM/23 Bugzilla 687: Fix misparses in eximstats.
328 Patch provided by Heiko Schlittermann.
330 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
331 Patch provided by Heiko Schlittermann.
333 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
334 plus update to original patch.
336 NM/26 Bugzilla 799: Documentation correction for ratelimit.
338 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
339 Patch provided by David Brownlee.
341 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
343 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
345 NM/30 Bugzilla 888: TLS documentation bugfixes.
347 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
349 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
350 Unlike the original bugzilla I have changed all shell scripts in src tree.
352 NM/33 Bugzilla 898: Transport filter timeout fix.
353 Patch by Todd Rinaldo.
355 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
356 Patch by Serge Demonchaux.
358 NM/35 Bugzilla 39: Base64 decode bug fixes.
359 Patch by Jakob Hirsch.
361 NM/36 Bugzilla 909: Correct connect() call in dcc code.
363 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
365 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
367 NM/39 Bugzilla 911: Fixed MakeLinks build script.
373 TK/01 Add preliminary DKIM support. Currently requires a forked version of
374 ALT-N's libdkim that I have put here:
375 http://duncanthrax.net/exim-experimental/
377 Note to Michael Haardt: I had to rename some vars in sieve.c. They
378 were called 'true' and it seems that C99 defines that as a reserved
379 keyword to be used with 'bool' variable types. That means you could
380 not include C99-style headers which use bools without triggering
381 build errors in sieve.c.
383 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
384 as mailq or other aliases. Changed the --help handling significantly
385 to do whats expected. exim_usage() emits usage/help information.
387 SC/01 Added the -bylocaldomain option to eximstats.
389 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
391 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
393 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
399 PH/01 Another patch from the Sieve maintainer.
401 PH/02 When an IPv6 address is converted to a string for single-key lookup
402 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
403 dots are used instead of colons so that keys in lsearch files need not
404 contain colons. This was done some time before quoting was made available
405 in lsearch files. However, iplsearch files do require colons in IPv6 keys
406 (notated using the quote facility) so as to distinguish them from IPv4
407 keys. This meant that lookups for IP addresses in host lists did not work
408 for iplsearch lookups.
410 This has been fixed by arranging for IPv6 addresses to be expressed with
411 colons if the lookup type is iplsearch. This is not incompatible, because
412 previously such lookups could never work.
414 The situation is now rather anomolous, since one *can* have colons in
415 ordinary lsearch keys. However, making the change in all cases is
416 incompatible and would probably break a number of configurations.
418 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
421 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
422 conversion specification without a maximum field width, thereby enabling
423 a rogue spamd server to cause a buffer overflow. While nobody in their
424 right mind would setup Exim to query an untrusted spamd server, an
425 attacker that gains access to a server running spamd could potentially
426 exploit this vulnerability to run arbitrary code as the Exim user.
428 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
429 $primary_hostname instead of what libspf2 thinks the hosts name is.
431 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
432 a directory entry by the name of the lookup key. Previously, if a
433 symlink pointed to a non-existing file or a file in a directory that
434 Exim lacked permissions to read, a lookup for a key matching that
435 symlink would fail. Now it is enough that a matching directory entry
436 exists, symlink or not. (Bugzilla 503.)
438 PH/03 The body_linecount and body_zerocount variables are now exported in the
441 PH/04 Added the $dnslist_matched variable.
443 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
444 This means they are set thereafter only if the connection becomes
447 PH/06 Added the client_condition to authenticators so that some can be skipped
448 by clients under certain conditions.
450 PH/07 The error message for a badly-placed control=no_multiline_responses left
451 "_responses" off the end of the name.
453 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
455 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
456 (without spaces) instead of just copying the configuration text.
458 PH/10 Added the /noupdate option to the ratelimit ACL condition.
460 PH/11 Added $max_received_linelength.
462 PH/12 Added +ignore_defer and +include_defer to host lists.
464 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
465 way in which PCRE > 7.0 is built.
467 PH/14 Implemented queue_only_load_latch.
469 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
470 MAIL command. The effect was to mangle the value on 64-bit systems.
472 PH/16 Another patch from the Sieve maintainer.
474 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
476 PH/18 If a system quota error occurred while trying to create the file for
477 a maildir delivery, the message "Mailbox is full" was not appended to the
478 bounce if the delivery eventually timed out. Change 4.67/27 below applied
479 only to a quota excession during the actual writing of the file.
481 PH/19 It seems that peer DN values may contain newlines (and other non-printing
482 characters?) which causes problems in log lines. The DN values are now
483 passed through string_printing() before being added to log lines.
485 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
486 and InterBase are left for another time.)
488 PH/21 Added message_body_newlines option.
490 PH/22 Guard against possible overflow in moan_check_errorcopy().
492 PH/23 POSIX allows open() to be a macro; guard against that.
494 PH/24 If the recipient of an error message contained an @ in the local part
495 (suitably quoted, of course), incorrect values were put in $domain and
496 $local_part during the evaluation of errors_copy.
502 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
503 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
506 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
507 issue a MAIL command.
509 PH/02 In an ACL statement such as
511 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
513 if a client was not listed at all, or was listed with a value other than
514 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
515 the condition was not true (as it should be), so access was not denied.
516 The bug was that the ! inversion was incorrectly passed on to the second
517 item. This has been fixed.
519 PH/03 Added additional dnslists conditions == and =& which are different from
520 = and & when the dns lookup returns more than one IP address.
522 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
523 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
525 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
526 FSYNC, which compiles an option called disable_fsync that allows for
527 bypassing fsync(). The documentation is heavily laced with warnings.
529 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
531 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
532 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
533 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
534 including adding "make clean"; (3) Added -fPIC when compiling the test
535 dynamically loaded module, to get rid of a warning.
537 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
538 message fails, move_frozen_messages = true and ignore_bounce_errors_after
539 = 0s. The bug is otherwise harmless.
541 PH/07 There was a bug in the dovecot authenticator such that the value of
542 $auth1 could be overwritten, and so not correctly preserved, after a
543 successful authentication. This usually meant that the value preserved by
544 the server_setid option was incorrect.
546 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
548 PH/09 Installed PCRE release 7.0.
550 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
551 run for batched SMTP input. It is now run at the start of every message
552 in the batch. While fixing this I discovered that the process information
553 (output by running exiwhat) was not always getting set for -bs and -bS
554 input. This is fixed, and it now also says "batched" for BSMTP.
556 PH/11 Added control=no_pipelining.
558 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
559 patch, slightly modified), and move the expansion of helo_data till after
560 the connection is made in the smtp transport (so it can use these
563 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
565 PH/14 Added log_selector = +pid.
567 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
569 PH/16 Add ${if forany and ${if forall.
571 PH/17 Added dsn_from option to vary the From: line in DSNs.
573 PH/18 Flush SMTP output before performing a callout, unless control =
574 no_callout_flush is set.
576 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
577 was true (the default) a successful delivery failed to delete the retry
578 item, thus causing premature timeout of the address. The bug is now
581 PH/20 Added hosts_avoid_pipelining to the smtp transport.
583 PH/21 Long custom messages for fakedefer and fakereject are now split up
584 into multiline reponses in the same way that messages for "deny" and
585 other ACL rejections are.
587 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
588 with slight modification.
590 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
591 draft, changing the syntax and factoring some duplicate code".
593 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
594 for deliveries of the second and subsequent messages over the same SMTP
597 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
598 ${reduce, with only minor "tidies".
600 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
602 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
603 expansion side effects.
605 PH/27 When a message times out after an over-quota error from an Exim-imposed
606 quota, the bounce message says "mailbox is full". This message was not
607 being given when it was a system quota that was exceeded. It now should
610 MH/03 Made $recipients available in local_scan(). local_scan() already has
611 better access to the recipient list through recipients_list[], but
612 $recipients can be useful in postmaster-provided expansion strings.
614 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
615 in the case of a MAIL command with additional options following the
616 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
617 were accidentally chopped off.
619 PH/29 SMTP synchronization checks are implemented when a command is read -
620 there is a check that no more input is waiting when there shouldn't be
621 any. However, for some commands, a delay in an ACL can mean that it is
622 some time before the response is written. In this time, more input might
623 arrive, invalidly. So now there are extra checks after an ACL has run for
624 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
625 pipelining has not been advertised.
627 PH/30 MH's patch to allow iscntrl() characters to be list separators.
629 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
630 returned in the SMTP response when smtp_return_error_details was false.
633 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
634 instead of the C I/O that was originally supplied, because problems were
637 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
638 Exim which did not show up earlier: it was assuming that a call to
639 SSL_CTX_set_info_callback() might give an error value. In fact, there is
640 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
641 was a macro that became an assignment, so it seemed to work. This has
642 changed to a proper function call with a void return, hence the compile
643 error. Exim's code has been fixed.
645 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
648 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
650 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
652 PH/37 If a message is not accepted after it has had an id assigned (e.g.
653 because it turns out to be too big or there is a timeout) there is no
654 "Completed" line in the log. When some messages of this type were
655 selected by exigrep, they were listed as "not completed". Others were
656 picked up by some special patterns. I have improved the selection
657 criteria to be more general.
659 PH/38 The host_find_failed option in the manualroute router can now be set
660 to "ignore", to completely ignore a host whose IP address cannot be
661 found. If all hosts are ignored, the behaviour is controlled by the new
662 host_all_ignored option.
664 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
665 homing or because of multiple MX records with /mx) generated more than
666 one IP address, and the following item turned out to be the local host,
667 all the secondary addresses of the first item were incorrectly removed
668 from the list, along with the local host and any following hosts (which
669 is what is supposed to happen).
671 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
672 whoever called Exim into the -H file. In the case of the daemon it was
673 behaving confusingly. When first started, it used values for whoever
674 started the daemon, but after a SIGHUP it used the Exim user (because it
675 calls itself on a restart). I have changed the code so that it now always
678 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
679 message are rejected with the same error (e.g. no authentication or bad
680 sender address), and a DATA command is nevertheless sent (as can happen
681 with PIPELINING or a stupid MUA), the error message that was given to the
682 RCPT commands is included in the rejection of the DATA command. This is
683 intended to be helpful for MUAs that show only the final error to their
686 PH/42 Another patch from the Sieve maintainer.
688 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
689 Eximstats - Fixed some broken HTML links and added missing column headers
691 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
692 columns for Rejects, Temp Rejects, Ham, and Spam rows.
694 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
696 PH/43 Yet another patch from the Sieve maintainer.
698 PH/44 I found a way to check for a TCP/IP connection going away before sending
699 the response to the final '.' that terminates a message, but only in the
700 case where the client has not sent further data following the '.'
701 (unfortunately, this is allowed). However, in many cases there won't be
702 any further data because there won't be any more messages to send. A call
703 to select() can be used: if it shows that the input is "ready", there is
704 either input waiting, or the socket has been closed. An attempt to read
705 the next input character can distinguish the two cases. Previously, Exim
706 would have sent an OK response which the client would never have see.
707 This could lead to message repetition. This fix should cure that, at
708 least in a lot of common cases.
710 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
711 advertised in response to EHLO.
717 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
718 fixed by 4.65/MH/01 (is this a record?) are fixed:
720 (i) An empty string was always treated as zero by the numeric comparison
721 operators. This behaviour has been restored.
723 (ii) It is documented that the numeric comparison operators always treat
724 their arguments as decimal numbers. This was broken in that numbers
725 starting with 0 were being interpreted as octal.
727 While fixing these problems I realized that there was another issue that
728 hadn't been noticed. Values of message_size_limit (both the global option
729 and the transport option) were treated as octal if they started with 0.
730 The documentation was vague. These values are now always treated as
731 decimal, and I will make that clear in the documentation.
737 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
738 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
741 MH/01 Don't check that the operands of numeric comparison operators are
742 integers when their expansion is in "skipping" mode (fixes bug
743 introduced by 4.64-PH/07).
745 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
746 child addresses, Exim now panics and dies. Previously, because the count
747 is held in a short int, deliveries were likely to be lost. As such a
748 large number of recipients for a single message is ridiculous
749 (performance will be very, very poor), I have chosen to impose a limit
750 rather than extend the field.
756 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
757 leftover -K file (the existence of which was triggered by #402).
758 While we were at it, introduced process PID as part of the -K
759 filename. This should rule out race conditions when creating
762 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
763 processing considerably. Previous code took too long for large mails,
764 triggering a timeout which in turn triggers #401.
766 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
767 in the DK code in transports.c. sendfile() is not really portable,
768 hence the _LINUX specificness.
770 TF/01 In the add_headers option to the mail command in an Exim filter,
771 there was a bug that Exim would claim a syntax error in any
772 header after the first one which had an odd number of characters
775 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
776 callout verification, Exim cached a "reject" for the entire domain. This
777 is correct for most verifications, but it is not correct for a recipient
778 verification with use_sender or use_postmaster set, because in that case
779 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
780 case of MAIL FROM:<> rejection from other early rejections (e.g.
781 rejection of HELO). When verifying a recipient using a non-null MAIL
782 address, the cache is ignored if it shows MAIL FROM:<> rejection.
783 Whatever the result of the callout, the value of the domain cache is
784 left unchanged (for any other kind of callout, getting as far as trying
785 RCPT means that the domain itself is ok).
787 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
790 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
791 manifest itself as EPIPE rather than ECONNECT. When tidying away a
792 session, the daemon ignores ECONNECT errors and logs others; it now
793 ignores EPIPE as well.
795 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
796 (quoted-printable decoding).
798 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
799 later the small subsequent patch to fix an introduced bug.
801 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
803 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
805 PH/08 An error is now given if message_size_limit is specified negative.
807 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
808 to be given (somewhat) arbitrary names.
810 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
813 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
814 miscellaneous code fixes
816 PH/10 Added the log_reject_target ACL modifier to specify where to log
819 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
820 hostname. This is wrong, because it relates to the incoming message (and
821 probably the interface on which it is arriving) and not to the outgoing
822 callout (which could be using a different interface). This has been
823 changed to use the value of the helo_data option from the smtp transport
824 instead - this is what is used when a message is actually being sent. If
825 there is no remote transport (possible with a router that sets up host
826 addresses), $smtp_active_hostname is used.
828 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
829 tweaks were necessary in order to get it to work (see also 21 below):
830 (a) The code assumed that strncpy() returns a negative number on buffer
831 overflow, which isn't the case. Replaced with Exim's string_format()
833 (b) There were several signed/unsigned issues. I just did the minimum
834 hacking in of casts. There is scope for a larger refactoring.
835 (c) The code used strcasecmp() which is not a standard C function.
836 Replaced with Exim's strcmpic() function.
837 (d) The code set only $1; it now sets $auth1 as well.
838 (e) A simple test gave the error "authentication client didn't specify
839 service in request". It would seem that Dovecot has changed its
840 interface. Fortunately there's a specification; I followed it and
841 changed what the client sends and it appears to be working now.
843 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
846 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
847 address A is aliased to B and C, where B exists and C does not. Without
848 -v the output is "A verified" because verification stops after a
849 successful redirection if more than one address is generated. However,
850 with -v the child addresses are also verified. Exim was outputting "A
851 failed to verify" and then showing the successful verification for C,
852 with its parentage. It now outputs "B failed to verify", showing B's
853 parentage before showing the successful verification of C.
855 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
856 look up a TXT record in a specific list after matching in a combined
859 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
860 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
861 they consult the DNS. I had assumed they would set it the way they
862 wanted; and indeed my experiments on Linux seem to show that in some
863 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
864 To be on the safe side, however, I have now made the interface to
865 host_find_byname() similar to host_find_bydns(), with an argument
866 containing the DNS resolver options. The host_find_byname() function now
867 sets these options at its start, just as host_find_bydns() does. The smtp
868 transport options dns_qualify_single and dns_search_parents are passed to
869 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
870 of host_find_byname() use the default settings of RES_DEFNAMES
871 (qualify_single) but not RES_DNSRCH (search_parents).
873 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
874 spool_read_header() do less string testing, by means of a preliminary
875 switch on the second character of optional "-foo" lines. (This is
876 overdue, caused by the large number of possibilities that now exist.
877 Originally there were few.) While I was there, I also converted the
878 str(n)cmp tests so they don't re-test the leading "-" and the first
879 character, in the hope this might squeeze out yet more improvement.
881 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
882 flag allowing group syntax was set by the header_syntax check but not
883 turned off, possible causing trouble later; (2) The flag was not being
884 set at all for the header_verify test, causing "group"-style headers to
885 be rejected. I have now set it in this case, and also caused header_
886 verify to ignore an empty address taken from a group. While doing this, I
887 came across some other cases where the code for allowing group syntax
888 while scanning a header line wasn't quite right (mostly, not resetting
889 the flag correctly in the right place). These bugs could have caused
890 trouble for malformed header lines. I hope it is now all correct.
892 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
893 with the "reply" argument non-NULL. The code, however (which originally
894 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
895 but it didn't always do it. This confused somebody who was copying the
896 code for some other use. I have removed all the tests.
898 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
899 feature that was used to support insecure browsers during the U.S. crypto
900 embargo. It requires special client support, and Exim is probably the
901 only MTA that supported it -- and would never use it because real RSA is
902 always available. This code has been removed, because it had the bad
903 effect of slowing Exim down by computing (never used) parameters for the
904 RSA_EXPORT functionality.
906 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
907 authenticator to fail if there's a tab character in the incoming data
908 (there should never be unless someone is messing about, as it's supposed
909 to be base64-encoded). Also added, on Timo's advice, the "secured" option
910 if the connection is using TLS or if the remote IP is the same as the
911 local IP, and the "valid-client-cert option" if a client certificate has
914 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
915 authenticators. This can be used for authorization after authentication
916 succeeds. (In the case of plaintext, it servers for both authentication
919 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
920 if any retry times were supplied.
922 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
923 connection, where there is no client IP address to check. In this
924 situation, the verify now always succeeds.
926 PH/25 Applied John Jetmore's -Mset patch.
928 PH/26 Added -bem to be like -Mset, but loading a message from a file.
930 PH/27 In a string expansion for a processed (not raw) header when multiple
931 headers of the same name were present, leading whitespace was being
932 removed from all of them, but trailing whitespace was being removed only
933 from the last one. Now trailing whitespace is removed from each header
934 before concatenation. Completely empty headers in a concatenation (as
937 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
938 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
940 PH/29 [Removed. This was a change that I later backed out, and forgot to
941 correct the ChangeLog entry (that I had efficiently created) before
942 committing the later change.]
944 PH/30 Exim was sometimes attempting to deliver messages that had suffered
945 address errors (4xx response to RCPT) over the same connection as other
946 messages routed to the same hosts. Such deliveries are always "forced",
947 so retry times are not inspected. This resulted in far too many retries
948 for the affected addresses. The effect occurred only when there were more
949 hosts than the hosts_max_try setting in the smtp transport when it had
950 the 4xx errors. Those hosts that it had tried were not added to the list
951 of hosts for which the message was waiting, so if all were tried, there
952 was no problem. Two fixes have been applied:
954 (i) If there are any address or message errors in an SMTP delivery, none
955 of the hosts (tried or untried) are now added to the list of hosts
956 for which the message is waiting, so the message should not be a
957 candidate for sending over the same connection that was used for a
958 successful delivery of some other message. This seems entirely
959 reasonable: after all the message is NOT "waiting for some host".
960 This is so "obvious" that I'm not sure why it wasn't done
961 previously. Hope I haven't missed anything, but it can't do any
962 harm, as the worst effect is to miss an optimization.
964 (ii) If, despite (i), such a delivery is accidentally attempted, the
965 routing retry time is respected, so at least it doesn't keep
966 hammering the server.
968 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
969 in ${readsocket because some servers need this prod.
971 PH/32 Added some extra debug output when updating a wait-xxx database.
973 PH/33 The hint "could be header name not terminated by colon", which has been
974 given for certain expansion errors for a long time, was not being given
975 for the ${if def:h_colon_omitted{... case.
977 PH/34 The spec says: "With one important exception, whenever a domain list is
978 being scanned, $domain contains the subject domain." There was at least
979 one case where this was not true.
981 PH/35 The error "getsockname() failed: connection reset by peer" was being
982 written to the panic log as well as the main log, but it isn't really
983 panic-worthy as it just means the connection died rather early on. I have
984 removed the panic log writing for the ECONNRESET error when getsockname()
987 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
988 runs only) independently of the message's sender address. This meant
989 that, if the 4xx error was in fact related to the sender, a different
990 message to the same recipient with a different sender could confuse
991 things. In particualar, this can happen when sending to a greylisting
992 server, but other circumstances could also provoke similar problems.
993 I have changed the default so that the retry time for these errors is now
994 based a combination of the sender and recipient addresses. This change
995 can be overridden by setting address_retry_include_sender=false in the
998 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
999 remote server are returned as part of bounce messages. This was not
1000 happening for LMTP over a pipe (the lmtp transport), but now it is the
1001 same for both kinds of LMTP.
1003 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1004 in header lines that were in fact CNAMEs. This is no longer the case.
1006 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1007 and queue runs started by the daemon processed all messages. This has
1008 been fixed so that -R and -S can now usefully be given with -q<time>.
1010 PH/40 Import PCRE release 6.7 (fixes some bugs).
1012 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1014 PH/42 Give an error if -q is specified more than once.
1016 PH/43 Renamed the variables $interface_address and $interface_port as
1017 $received_ip_address and $received_port, to make it clear that these
1018 values apply to message reception, and not to the outgoing interface when
1019 a message is delivered. (The old names remain recognized, of course.)
1021 PH/44 There was no timeout on the connect() call when using a Unix domain
1022 socket in the ${readsocket expansion. There now is.
1024 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1025 be meaningful with "accept".
1027 SC/01 Eximstats V1.43
1028 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1030 SC/02 Eximstats V1.44
1031 Use a glob alias rather than an array ref in the generated
1032 parser. This improves both readability and performance.
1034 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1035 Collect SpamAssassin and rejection statistics.
1036 Don't display local sender or destination tables unless
1037 there is data to show.
1038 Added average volumes into the top table text output.
1040 SC/04 Eximstats V1.46
1041 Collect data on the number of addresses (recipients)
1042 as well as the number of messages.
1044 SC/05 Eximstats V1.47
1045 Added 'Message too big' to the list of mail rejection
1046 reasons (thanks to Marco Gaiarin).
1048 SC/06 Eximstats V1.48
1049 Mainlog lines which have GMT offsets and are too short to
1050 have a flag are now skipped.
1052 SC/07 Eximstats V1.49 (Alain Williams)
1053 Added the -emptyok flag.
1055 SC/08 Eximstats V1.50
1056 Fixes for obtaining the IP address from reject messages.
1058 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1059 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1060 whitesspace changes from 4.64-PH/27
1062 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1065 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1066 are found, allow negative numbers in numeric criteria)
1068 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1070 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1071 to match changes made in 4.64-PH/43
1073 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1075 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1076 "domain" is actually an IP address, and force a failure. This locks out
1077 those revolvers/nameservers that support "A-for-A" lookups, in
1078 contravention of the specifications.
1080 PH/48 When a host name was looked up from an IP address, and the subsequent
1081 forward lookup of the name timed out, the host name was left in
1082 $sender_host_name, contrary to the specification.
1084 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1085 restricted to single-key lookups, Exim was not diagnosing an error if
1086 * or *@ was used with a query-style lookup.
1088 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1090 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1091 long ago, but noone interested enough thought of it. Let's just say that
1092 the "1.1" means that there are some new functions that weren't there at
1093 some point in the past.
1095 PH/51 Error processing for expansion failure of helo_data from an smtp
1096 transport during callout processing was broken.
1098 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1099 tested/used via the -bh/-bhc/-bs options.
1101 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1102 bug, fixed in subsequent PCRE releases).
1104 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1105 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1107 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1113 SC/01 Use a glob alias rather than an array ref in eximstats generated
1114 parser. This improves both readability and performance.
1116 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1117 Don't display local sender or destination tables in eximstats unless
1118 there is data to show.
1119 Added average volumes into the eximstats top table text output.
1121 SC/03 Collect data on the number of addresses (recipients) as well
1122 as the number of messages in eximstats.
1124 TF/01 Correct an error in the documentation for the redirect router. Exim
1125 does (usually) call initgroups() when daemonizing.
1127 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1128 with consistent privilege compared to when running as a daemon.
1130 TF/03 Note in the spec that $authenticated_id is not set for local
1131 submissions from trusted users.
1133 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1134 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1136 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1137 by adding some example configuration directives to the default
1138 configuration file. A little bit of work is required to uncomment the
1139 directives and define how usernames and passwords are checked, but
1140 there is now a framework to start from.
1142 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1143 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1144 without this. I don't know how relevant this is to other LDAP libraries.
1146 PH/02 Add the verb name to the "unknown ACL verb" error.
1148 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1150 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1152 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1153 directory not expanded when it should be if an expanded home directory
1154 was set for the address (which is overridden by the transport).
1156 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1159 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1160 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1161 because it is too late at that time, and has no effect.
1163 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1164 security issue with \' (bugzilla #107). I could not use the
1165 PQescapeStringConn() function, because it needs a PGconn value as one of
1168 PH/08 When testing addresses using -bt, indicate those final addresses that
1169 are duplicates that would not cause an additional delivery. At least one
1170 person was confused, thinking that -bt output corresponded to deliveries.
1171 (Suppressing duplicates isn't a good idea as you lose the information
1172 about possibly different redirections that led to the duplicates.)
1174 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1175 systems where poll() doesn't work, in particular OS X.
1177 PH/10 Added more information to debugging output for retry time not reached.
1179 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1180 operations in malware.c.
1182 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1185 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1186 syslog_duplication set false, log lines that would normally be written
1187 both the the main log and to the reject log were not written to syslog at
1190 PH/14 In the default configuration, change the use of "message" in ACL warn
1191 statements to "add_header".
1193 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1194 not followed by a command (e.g. "seen endif").
1196 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1197 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1200 PH/17 Added extra conditions to the default value of delay_warning_condition
1204 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1205 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1206 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1209 The Auto-Submitted: and various List- headers are standardised, whereas I
1210 don't think Precedence: ever was.
1212 PH/18 Refactored debugging code in route_finduser() to show more information,
1213 in particular, the error code if getpwnam() issues one.
1215 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1216 This is apparently needed in addition to the PH/07 change above to avoid
1217 any possible encoding problems.
1219 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1220 but not after initializing Perl.
1222 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1223 output them only if debugging. By default they are written stderr,
1224 apparently, which is not desirable.
1226 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1229 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1232 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1234 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1235 authenticated or an ident call has been made. Suppress the default
1236 values for $authenticated_id and $authenticated_sender (but permit -oMai
1237 and -oMas) when testing with -bh.
1239 PH/24 Re-jigged the order of the tests in the default configuration so that the
1240 tests for valid domains and recipients precede the DNS black list and CSA
1241 tests, on the grounds that those ones are more expensive.
1243 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1244 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1245 This bug exists in every version of Exim that I still have, right back to
1248 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1249 However, an attempt to turn on case-sensitivity in a regex key by
1250 including (?-i) didn't work because the subject string was already
1251 lowercased, and the effects were non-intuitive. It turns out that a
1252 one-line patch can be used to allow (?-i) to work as expected.
1258 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1259 other effects) broke the use of negated acl sub-conditions.
1261 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1264 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1265 "Deny" causes Exim to reject the incoming connection with a 554 error.
1266 Unfortunately, if there is a major crisis, such as a disk failure,
1267 tcp-wrappers gives "deny", whereas what one would like would be some
1268 kind of temporary error. A kludge has been added to help with this.
1269 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1270 554 error is used if errno is still zero or contains ENOENT (which occurs
1271 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1274 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1276 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1277 errors. Otherwise a message that provokes a temporary error (when other
1278 messages do not) can cause a whole host to time out.
1280 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1281 addresses were routed directly to files or pipes from a redirect router.
1282 File deliveries just didn't batch; pipe deliveries might have suffered
1285 PH/06 A failure to get a lock for a hints database would erroneously always say
1286 "Failed to get write lock", even when it was really a read lock.
1288 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1289 of 0600. This has been changed to use the value of the lockfile_mode
1290 option (which defaults to 0600).
1292 PH/08 Applied small patch from the Sieve maintainer.
1294 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1295 folder from quota calculations, a direct delivery into this folder messed
1296 up the contents of the maildirsize file. This was because the regex was
1297 used only to exclude .Trash (or whatever) when the size of the mailbox
1298 was calculated. There was no check that a delivery was happening into an
1299 excluded directory. This bug has been fixed by ignoring all quota
1300 processing for deliveries into excluded directories.
1302 PH/10 Added the maildirfolder_create_regex option to appendfile.
1308 PH/01 The code for finding all the local interface addresses on a FreeBSD
1309 system running IPv6 was broken. This may well have applied to all BSD
1310 systems, as well as to others that have similar system calls. The broken
1311 code found IPv4 interfaces correctly, but gave incorrect values for the
1312 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1313 that it would not match correctly against @[] and not recognize the IPv6
1316 PH/02 The ipliteral router was not recognizing addresses of the form user@
1317 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1319 PH/03 Added disable_ipv6.
1321 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1322 decoded form, because it is most often used to construct To: headers
1323 lines in autoreplies, and the decoded form may well be syntactically
1324 invalid. However, $reply_address has leading white space removed, and all
1325 newlines turned into spaces so that the autoreply transport does not
1328 PH/05 If group was specified without a user on a router, and no group or user
1329 was specified on a transport, the group from the router was ignored.
1331 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1332 for visible compile-time settings that can be used to change these
1333 numbers, for those that want even more. Backwards compatibility with old
1334 spool files has been maintained. However, going back to a previous Exim
1335 release will lost any variables that are in spool files.
1337 PH/07 Two small changes when running in the test harness: increase delay when
1338 passing a TCP/IP connection to a new process, in case the original
1339 process has to generate a bounce, and remove special handling of
1340 127.0.0.2 (sic), which is no longer necessary.
1342 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1343 be the same on different OS.
1345 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1348 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1349 whether --show-vars was specified or not
1351 JJ/02 exipick: Added support for new ACL variable spool format introduced
1354 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1355 syntactically invalid From: or Reply-to: line, and a filter used this to
1356 generate an autoreply, and therefore failed to obtain an address for the
1357 autoreply, Exim could try to deliver to a non-existent relative file
1358 name, causing unrelated and misleading errors. What now happens is that
1359 it logs this as a hard delivery error, but does not attempt to create a
1362 PH/11 The exinext utility has a -C option for testing purposes, but although
1363 the given file was scanned by exinext itself; it wasn't being passed on
1364 when Exim was called.
1366 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1367 an end-of-file indication when reading a command response.
1369 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1370 compiled. In many other places in Exim, IPv6 addresses are always
1371 recognized, so I have changed this. It also means that IPv4 domain
1372 literals of the form [IPV4:n.n.n.n] are now always recognized.
1374 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1375 used if the router is not running as root, for example, when verifying at
1376 ACL time, or when using -bh. The debugging output from this situation was
1377 non-existent - all you got was a failure to exec. I have made two
1380 (a) Failures to set uid/gid, the current directory, or a process leader
1381 in a subprocess such as that created by queryprogram now generate
1382 suitable debugging ouput when -d is set.
1384 (b) The queryprogram router detects when it is not running as root,
1385 outputs suitable debugging information if -d is set, and then runs
1386 the subprocess without attempting to change uid/gid.
1388 PH/15 Minor change to Makefile for building test_host (undocumented testing
1391 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1392 additional section of a DNS packet that returns MX or SRV records.
1393 Instead, it always explicitly searches for A/AAAA records. This avoids
1394 major problems that occur when a DNS server includes only records of one
1395 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1396 fixed another bug: if SRV records were looked up and the corresponding
1397 address records were *not* found in the additional section, the port
1398 values from the SRV records were lost.
1400 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1401 using the correct key (the original address) when searching the retry
1402 rules in order to find which one to use for generating the retry hint.
1404 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1405 adding the default one. Similarly, if it contains a Reply-To: header, the
1406 errors_reply_to option, if set, is not used.
1408 PH/19 When calculating a retry time, Exim used to measure the "time since
1409 failure" by looking at the "first failed" field in the retry record. Now
1410 it does not use this if it is later than than the arrival time of the
1411 message. Instead it uses the arrival time. This makes for better
1412 behaviour in cases where some deliveries succeed, thus re-setting the
1413 "first failed" field. An example is a quota failure for a huge message
1414 when small messages continue to be delivered. Without this change, the
1415 "time since failure" will always be short, possible causing more frequent
1416 delivery attempts for the huge message than are intended.
1417 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1419 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1420 $1, $2, $3) because the numerical variables can be reset during some
1421 expansion items (e.g. "match"), thereby losing the authentication data.
1423 PH/21 Make -bV show the size of off_t variables so that the test suite can
1424 decide whether to run tests for quotas > 2G.
1426 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1427 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1428 filecount value is greater than 2G or if a quota value is greater than 2G
1429 on a system where the size of off_t is not greater than 4, a panic error
1432 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1433 never match. The debug and -bh output now contains an explicit error
1434 message indicating a malformed IPv4 address or mask.
1436 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1437 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1438 PH/23 above applies.
1440 PH/25 Do not write to syslog when running in the test harness. The only
1441 occasion when this arises is a failure to open the main or panic logs
1442 (for which there is an explicit test).
1444 PH/26 Added the /no_tell option to "control=freeze".
1446 PH/27 If a host name lookup failed very early in a connection, for example, if
1447 the IP address matched host_lookup and the reverse lookup yielded a name
1448 that did not have a forward lookup, an error message of the form "no IP
1449 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1450 could be logged. Now it outputs the IP address instead of "NULL".
1452 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1453 allows the sender and the authenticated sender to be set when
1454 submitting a message from within Exim. Since child_open_exim() is
1455 documented for local_scan(), the new function should be too.
1457 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1458 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1459 results in an empty string is now treated as unset.
1461 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1463 PH/31 Added sender_verify_fail logging option.
1465 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1466 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1467 tidied the source and removed it altogether.
1469 PH/33 When a queue run was abandoned because the load average was too high, a
1470 log line was always written; now it is written only if the queue_run log
1471 selector is set. In addition, the log line for abandonment now contains
1472 information about the queue run such as the pid. This is always present
1473 in "start" and "stop" lines but was omitted from the "abandon" line.
1475 PH/34 Omit spaces between a header name and the colon in the error message that
1476 is given when verify = headers_syntax fails (if there are lots of them,
1477 the message gets confusing).
1479 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1480 names, as there are now some PTR records that contain slashes. This check
1481 is only to protect against broken name servers that fall over on strange
1482 characters, so the fact that it applies to all lookups doesn't matter.
1484 PH/36 Now that the new test suite is complete, we can remove some of the
1485 special code in Exim that was needed for the old test suite. For example,
1486 sorting DNS records because real resolvers return them in an arbitrary
1487 order. The new test suite's fake resolver always returns records in the
1490 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1491 bounces) except when queue_only is set, to avoid logging races between
1492 the different processes.
1494 PH/38 Panic-die if .include specifies a non-absolute path.
1496 PH/39 A tweak to the "H" retry rule from its user.
1498 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1499 a label. They prevented compilation on older perls.
1501 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1502 a warning to be raised on newish perls.
1504 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1505 on queue. Changes to match documented behaviour of showing count of
1506 messages matching specified criteria.
1508 PH/40 Changed the default ident timeout from 30s to 5s.
1510 PH/41 Added support for the use of login_cap features, on those BSD systems
1511 that have them, for controlling the resources used by pipe deliveries.
1513 PH/42 The content-scanning code uses fopen() to create files in which to put
1514 message data. Previously it was not paying any attention to the mode of
1515 the files. Exim runs with umask(0) because the rest of the code creates
1516 files with open(), and sets the required mode explicitly. Thus, these
1517 files were ending up world-writeable. This was not a big issue, because,
1518 being within the spool directory, they were not world-accessible. I have
1519 created a function called modefopen, which takes an additional mode
1520 argument. It sets umask(777), creates the file, chmods it to the required
1521 mode, then resets the umask. All the relevant calls to fopen() in the
1522 content scanning code have been changed to use this function.
1524 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1525 to 24 hours. This avoids potential overflow problems when processing G
1526 and H retry rules. I suspect nobody ever tinkers with this value.
1528 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1530 PH/45 When the plaintext authenticator is running as a client, the server's
1531 challenges are checked to ensure they are valid base64 strings. By
1532 default, the authentication attempt is cancelled if an invalid string is
1533 received. Setting client_ignore_invalid_base64 true ignores these errors.
1534 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1535 they are received. Thus, the responses can be made to depend on the
1536 challenges. If an invalid string is ignored, an empty string is placed in
1539 PH/46 Messages that are created by the autoreply transport now contains a
1540 References: header, in accordance with RFCs 2822 and 3834.
1542 PH/47 Added authenticated_sender_force to the smtp transport.
1544 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1546 PH/49 Installed latest patch from the Sieve maintainer.
1548 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1549 was also set, the appendfile transport was unnecessarily scanning a
1550 directory of message files (e.g. for maildir delivery) to find the count
1551 of files (along with the size), even though it did not need this
1552 information. It now does the scan only if it needs to find either the
1553 size of the count of files.
1555 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1557 PH/52 Two bugs concerned with error handling when the smtp transport is
1560 (i) Exim was not creating retry information for temporary errors given
1561 for individual recipients after the DATA command when the smtp transport
1562 was used in LMTP mode. This meant that they could be retried too
1563 frequently, and not timed out correctly.
1565 (ii) Exim was setting the flag that allows error details to be returned
1566 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1567 recipients that were returned after the DATA command.
1569 PH/53 This is related to PH/52, but is more general: for any failing address,
1570 when detailed error information was permitted to be returned to the
1571 sender, but the error was temporary, then after the final timeout, only
1572 "retry timeout exceeded" was returned. Now it returns the full error as
1573 well as "retry timeout exceeded".
1575 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1576 do this, and (what is worse) MTAs that accept it.
1578 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1579 will now be deprecated.
1581 PH/56 New os.c-cygwin from the Cygwin maintainer.
1583 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1584 formats (previously only available in exim formats via -bpr, -bpru,
1585 and -bpra. Now also available in native and exiqgrep formats)
1587 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1588 with very large, slow to parse queues
1590 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1592 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1594 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1595 responses to authentication challenges, though it was showing the
1596 challenges; (ii) I've removed the CR characters from the debug output for
1599 PH/58 Allow for the insertion of a newline as well as a space when a string
1600 is turned into more than one encoded-word during RFC 2047 encoding. The
1601 Sieve code now uses this.
1603 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1604 data_4xx, lost_connection, tls_required.
1606 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1607 message was being sent as an SMTP response.
1609 PH/61 Add -l and -k options to exicyclog.
1611 PH/62 When verifying, if an address was redirected to one new address, so that
1612 verification continued, and the new address failed or deferred after
1613 having set something in $address_data, the value of $address_data was not
1614 passed back to the ACL. This was different to the case when no
1615 redirection occurred. The value is now passed back in both cases.
1617 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1618 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1619 use login_cap.h, so on its own it isn't the distinguishing feature. The
1620 new name refers directly to the setclassresources() function.
1622 PH/65 Added configuration files for NetBSD3.
1624 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1626 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1627 is preferred over IPv4.
1629 PH/68 The bounce_return_message and bounce_return_body options were not being
1630 honoured for bounces generated during the reception of non-SMTP messages.
1631 In particular, this applied to messages rejected by the ACL. This bug has
1632 been fixed. However, if bounce_return_message is true and bounce_return_
1633 body is false, the headers that are returned for a non-SMTP message
1634 include only those that have been read before the error was detected.
1635 (In the case of an ACL rejection, they have all been read.)
1637 PH/69 The HTML version of the specification is now built in a directory called
1638 spec_html instead of spec.html, because the latter looks like a path with
1639 a MIME-type, and this confuses some software.
1641 PH/70 Catch two compiler warnings in sieve.c.
1643 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1644 function verify_get_ident() calls ip_connect() to connect a socket, but
1645 if the "connect()" function timed out, ip_connect() used to close the
1646 socket. However, verify_get_ident() also closes the socket later, and in
1647 between Exim writes to the log, which may get opened at this point. When
1648 the socket was closed in ip_connect(), the log could get the same file
1649 descriptor number as the socket. This naturally causes chaos. The fix is
1650 not to close the socket in ip_connect(); the socket should be closed by
1651 the function that creates it. There was only one place in the code where
1652 this was missing, in the iplookup router, which I don't think anybody now
1653 uses, but I've fixed it anyway.
1655 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1656 well as to direct DNS lookups. Otherwise the handling of names in host
1657 lists is inconsistent and therefore confusing.
1663 PH/01 Two changes to the default runtime configuration:
1665 (1) Move the checks for relay_from_hosts and authenticated clients from
1666 after to before the (commented out) DNS black list checks.
1668 (2) Add control=submission to the relay_from_hosts and authenticated
1669 clients checks, on the grounds that messages accepted by these
1670 statements are most likely to be submissions.
1672 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1674 (1) Generate an error if the third argument for the ${prvs expansion is
1677 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1680 (3) Reset the variables that are obtained from the first argument of
1681 ${prvscheck and used in the second argument before leaving the code,
1682 because their memory is reclaimed, so using them afterwards may do
1685 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1686 one (it's much easier than Tom thought :-).
1688 (5) Because of (4), we can now allow for the use of $prvscheck_result
1689 inside the third argument.
1691 PH/03 For some reason, the default setting of PATH when running a command from
1692 a pipe transport was just "/usr/bin". I have changed it to
1695 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1696 anything to be listed in the output from -bV.
1698 PH/05 When a filter generated an autoreply, the entire To: header line was
1699 quoted in the delivery log line, like this:
1701 => >A.N.Other <ano@some.domain> <original@ddress> ...
1703 This has been changed so that it extracts the operative address. There
1704 may be more than one such address. If so, they are comma-separated, like
1707 => >ano@some.domain,ona@other.domain <original@ddress> ...
1709 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1710 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1711 not being looked up in the rDNS to get a host name, Exim was showing the
1712 IP address twice in Received: lines, even though the IP addresses were
1713 identical. For example:
1715 Received: from [1.2.3.4] (helo=[1.2.3.4])
1717 However, if the real host name was known, it was omitting the HELO data
1718 if it matched the actual IP address. This has been tidied up so that it
1719 doesn't show the same IP address twice.
1721 PH/07 When both +timestamp and +memory debugging was on, the value given by
1722 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1723 called by the debug printing, thereby overwriting the timestamp buffer.
1724 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1726 PH/08 When the original message was included in an autoreply transport, it
1727 always said "this is a copy of the message, including all the headers",
1728 even if body_only or headers_only was set. It now gives an appropriate
1731 PH/09 Applied a patch from the Sieve maintainer which:
1733 o fixes some comments
1734 o adds the (disabled) notify extension core
1735 o adds some debug output for the result of if/elsif tests
1736 o points to the current vacation draft in the documentation
1737 and documents the missing references header update
1741 o fixes a bug in processing the envelope test (when testing
1742 multiple envelope elements, the last element determinted the
1745 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1746 Electronic Mail") by including:
1748 Auto-submitted: auto-generated
1750 in the messages that it generates (bounce messages and others, such as
1751 warnings). In the case of bounce messages for non-SMTP mesages, there was
1752 also a typo: it was using "Auto_submitted" (underscore instead of
1753 hyphen). Since every message generated by Exim is necessarily in response
1754 to another message, thes have all been changed to:
1756 Auto-Submitted: auto-replied
1758 in accordance with these statements in the RFC:
1760 The auto-replied keyword:
1762 - SHOULD be used on messages sent in direct response to another
1763 message by an automatic process,
1765 - MUST NOT be used on manually-generated messages,
1767 - MAY be used on Delivery Status Notifications (DSNs) and Message
1768 Disposition Notifications (MDNs),
1770 - MUST NOT be used on messages generated by automatic or periodic
1771 processes, except for messages which are automatic responses to
1774 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1775 to the default Received: header definition.
1777 PH/12 Added log selector acl_warn_skipped (default on).
1779 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1780 variables because (a) they are in the wrong storage pool and (b) even if
1781 they were copied, it wouldn't work properly because of the caching.
1783 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1784 checking when decoding. Apparently there are clients that generate
1785 overlong encoded strings. Why am I not surprised?
1787 PH/15 If the first argument of "${if match_address" was not empty, but did not
1788 contain an "@" character, Exim crashed. Now it writes a panic log message
1789 and treats the condition as false.
1791 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1793 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1794 extension "envelope-auth". The code is finished and in agreement with
1795 other implementations, but there is no documentation so far and in fact,
1796 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1797 not changing the active code.
1799 Print executed "if" and "elsif" statements when debugging is used. This
1800 helps a great deal to understand what a filter does.
1802 Document more things not specified clearly in RFC3028. I had all this
1803 sorted out, when out of a sudden new issues came to my mind. Oops."
1805 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1808 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1809 canonical form (as documented). However, after a host name lookup from
1810 the IP address, check_host() was doing a simple string comparison with
1811 addresses acquired from the DNS when checking that the found name did
1812 have the original IP as one of its addresses. Since any found IPv6
1813 addresses are likely to be in abbreviated form, the comparison could
1814 fail. Luckily, there already exists a function for doing the comparison
1815 by converting both addresses to binary, so now that is used instead of
1816 the text comparison.
1818 PH/20 There was another similar case to PH/19, when a complete host name was
1819 given in a host list; looking up its IP address could give an abbreviated
1820 form, whereas the current host's name might or might not be abbreviated.
1821 The same fix has been applied.
1827 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1828 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1831 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1832 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1834 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1836 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1837 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1838 Also, for "Auto-Submitted", treat anything other than "no" as
1839 non-personal, in accordance with RFC 3834. (Previously it treated
1840 anything starting "auto-" as non-personal.)
1842 TF/01 The control=submission/name=... option had a problem with syntax
1843 errors if the name included a slash character. The /name= option
1844 now slurps the rest of the string, so it can include any characters
1845 but it must come last in the list of options (after /sender_retain
1848 PH/05 Some modifications to the interface to the fake nameserver for the new
1856 TK/01 Added the "success_on_redirect" address verification option. See
1857 NewStuff for rationale and an example.
1859 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1861 PH/02 Patch to exigrep to allow it to work on syslog lines.
1863 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1864 fread() to skip over the body file's header line, because in Cygwin the
1865 header line is locked and is inaccessible.
1867 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1868 co-exist for some time) to make it clear that it is the Exim ID that is
1869 referenced, not the Message-ID: header line.
1871 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1872 string_format() function, because snprintf() does not exist on all
1875 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1876 ${stat: expansion item.
1878 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1879 protocol synchronization error", to keep the pedants happy.
1881 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1882 well as for IRIX systems, when gcc is being used. See the host.c source
1885 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1887 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1890 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1891 transport and to the smtp transport in LMTP mode.
1893 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1895 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1896 run a filter in a subprocess. This could lead to confusion in subsequent
1897 lookups in the parent process. There should also be a search_tidyup() at
1898 the end of the subprocess.
1900 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1901 only if the host matched helo_try_verify_hosts, which caused the
1902 verification to occur when the EHLO/HELO command was issued. The ACL just
1903 tested the remembered result. Now, if a previous verification attempt has
1904 not happened, "verify = helo" does it there and then.
1906 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1908 TK/03 Fix log output including CR from clamd.
1910 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1911 exist provoked a memory error which could cause a segfault.
1913 PH/15 Installed PCRE 6.2
1915 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1917 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1918 of the problem. Specifically, suggested +O2 rather than +O1 for the
1921 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1923 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1924 record, the port was not correctly logged when the outgoing_port log
1925 selector was set (it logged the transort's default port).
1927 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1928 fallback_hosts, and "hosts" in the smtp transport.
1930 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1931 host errors such as "Connection refused".
1933 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1934 authentication with radiusclient 0.4.9:
1936 - Error returned from rc_read_config was caught wrongly
1937 - Username/password not passed on to radius server due to wrong length.
1939 The presumption is that some radiusclient API changes for 4.51/PH/17
1940 were not taken care of correctly. The code is still untested by me (my
1941 Linux distribution still has 0.3.2 of radiusclient), but it was
1942 contributed by a Radius user.
1944 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1945 expanding the "port" option of the smtp transport.
1947 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1948 while reading a MIME header. Thanks to Tom Hughes for a patch.
1950 PH/24 Include config.h inside local_scan.h so that configuration settings are
1953 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1954 that in an ACL for RCPT (for example), you can examine exactly what was
1957 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1958 commands, but it was not correctly comparing the address with the actual
1959 client host address. Thus, it would show the EHLO address in Received:
1960 header lines when this was not necessary.
1962 PH/27 Added the % operator to ${eval:}.
1964 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1965 it should be ignoring failures (because with -C, for example, it has lost
1966 privilege). It wasn't ignoring creation failures other than "already
1969 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1972 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1973 because an input line was too long, either on its own, or by virtue of
1974 too many continuations, the temporary file was not being removed, and the
1975 return code was incorrect.
1977 PH/31 Missing "BOOL" in function definition in filtertest.c.
1979 PH/32 Applied Sieve patches from the maintainer.
1981 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1983 PH/33 Added "verify = not_blind".
1985 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1986 Local/Makefile (with some defaults set). These are used in built scripts
1987 such as exicyclog, but they have never been used in the exim_install
1988 script (though there are many overriding facilities there). I have
1989 arranged that the exim_install script now takes note of these two
1992 PH/35 Installed configuration files for Dragonfly.
1994 PH/36 When a locally submitted message by a trusted user did not contain a
1995 From: header, and the sender address was obtained from -f or from an SMTP
1996 MAIL command, and the trusted user did not use -F to supply a sender
1997 name, $originator_name was incorrectly used when constructing a From:
1998 header. Furthermore, $originator_name was used for submission mode
1999 messages from external hosts without From: headers in a similar way,
2000 which is clearly wrong.
2002 PH/37 Added control=suppress_local_fixups.
2004 PH/38 When log_selector = +received_sender was set, and the addition of the
2005 sender made the log line's construction buffer exactly full, or one byte
2006 less than full, an overflow happened when the terminating "\n" was
2009 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2010 when the result of a list match is failure because a DNS lookup failed.
2012 PH/40 RM_COMMAND is now used in the building process.
2014 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2015 the "build-* directories that it finds.
2017 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2018 address in a domain literal was a prefix of an interface address.
2020 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2021 when verifying a sender address, unless rewrite_headers is false.
2023 PH/44 Wrote a long comment about why errors_to addresses are verified as
2024 recipients, not senders.
2026 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2027 the ratelimit ACL was added.
2029 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2031 PH/46 Added extra information about PostgreSQL errors to the error string.
2033 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2034 avoiding the need to install special zones in a real server. This is
2035 backwards compatible; if it can't find the fake resolver, it drops back.
2036 Thus, both old and new test suites can be run.
2038 TF/02 Added util/ratelimit.pl
2040 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2041 clock is set back in time.
2043 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2044 Candler <B.Candler@pobox.com>.
2046 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2047 OK for addresses that are the result of redirections.
2049 PH/48 A number of further additions for the benefit of the new test suite,
2050 including a fake gethostbyname() that interfaces to the fake DNS resolver
2053 TF/06 The fix for widen_domains has also been applied to qualify_single and
2054 search_parents which are the other dnslookup options that can cause
2057 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2060 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2062 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2063 service" if it encountered an MX record with an empty target hostname.
2064 The message is now "an MX or SRV record indicated no SMTP service".
2066 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2067 if the DNS of the sending site is misconfigured. This is quite a
2068 common situation. This change restores the behaviour of treating a
2069 helo verification defer as a failure.
2071 PH/51 If self=fail was set on a router, the bounce message did not include the
2072 actual error message.
2078 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2080 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2081 command itself ended in error, the underlying message about the transport
2082 filter timeout was being overwritten with the pipe command error. Now the
2083 underlying error message should be appended to the second error message.
2085 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2087 PH/02 Reduce the amount of output that "make" produces by default. Full output
2088 can still be requested.
2090 PH/03 The warning log line about a condition test deferring for a "warn" verb
2091 was being output only once per connection, rather than after each
2092 occurrence (because it was using the same function as for successful
2093 "warn" verbs). This seems wrong, so I have changed it.
2095 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2096 it should not have, which might have caused a crash in the right
2097 circumstances, but probably never did.
2099 PH/04 Installed a modified version of Tony Finch's patch to make submission
2100 mode fix the return path as well as the Sender: header line, and to
2101 add a /name= option so that you can make the user's friendly name appear
2104 TF/03 Added the control = fakedefer ACL modifier.
2106 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2107 Mark Lowes for thorough testing.
2109 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2111 TK/03 Merged latest SRS patch from Miles Wilton.
2113 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2114 with the definition in sysexits.h (which is #included earlier).
2115 Fortunately, Exim does not actually use EX_OK. The code used to try to
2116 preserve the sysexits.h value, by assumimg that macro definitions were
2117 scanned for macro replacements. I have been disabused of this notion,
2118 so now the code just undefines EX_OK before #including unistd.h.
2120 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2121 in the smtp transport. When a block could not be written in a single
2122 write() function, the timeout was being re-applied to each part-write.
2123 This seems wrong - if the receiver was accepting one byte at a time it
2124 would take for ever. The timeout is now adjusted when this happens. It
2125 doesn't have to be particularly precise.
2127 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2128 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2130 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2131 without a domain if the check to <postmaster@domain> fails.
2133 SC/01 Eximstats: added -xls and the ability to specify output files
2134 (patch written by Frank Heydlauf).
2136 SC/02 Eximstats: use FileHandles for outputing results.
2138 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2140 SC/04 Eximstats: fixed display of large numbers with -nvr option
2142 SC/05 Eximstats: fixed merging of reports with empty tables.
2144 SC/06 Eximstats: added the -include_original_destination flag
2146 SC/07 Eximstats: removed tabs and trailing whitespace.
2148 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2150 TK/06 MBOX spool code: Add real "From " MBOX separator line
2151 so the .eml file is really in mbox format (even though
2152 most programs do not really care). Patch from Alex Miller.
2154 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2155 The latter is generated from $received_to and is only set if the
2156 message has one envelope recipient. SA can use these headers,
2157 obviously out-of-the-box. Patch from Alex Miller.
2159 PH/08 The ${def test on a variable was returning false if the variable's
2160 value was "0", contrary to what the specification has always said!
2161 The result should be true unless the variable is empty.
2163 PH/09 The syntax error of a character other than { following "${if
2164 def:variable_name" (after optional whitespace) was not being diagnosed.
2165 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2166 accidental colon was present, for example, could give incorrect results.
2168 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2169 result is used (not including appendfile, where other changes are about
2172 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2173 This involved changing a lot of size variables from int to off_t. It
2174 should work with maildirs and everything.
2176 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2177 spamd dying while we are connected to it.
2179 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2182 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2183 function for BDB 4.3.
2185 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2187 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2188 the sources for PCRE have been re-arranged and more files are now
2191 PH/15 The code I had for printing potentially long long variables in PH/11
2192 above was not the best (it lost precision). The length of off_t variables
2193 is now inspected at build time, and an appropriate printing format (%ld
2194 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2195 to be "long long int" or "long int". This is needed for the internal
2196 formatting function string_vformat().
2198 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2199 the configuration file to be ":syslog", then the script "guesses" where
2200 the logs files are, rather than using the compiled in default. In our
2201 case the guess is not the same as the compiled default, so the script
2202 suddenly stopped working when I started to use syslog. The patch checks
2203 to see if log_file_path is "". If so, it attempts to read it from exim
2204 with no configuration file to get the compiled in version, before it
2205 falls back to the previous guessing code."
2207 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2208 implementing BATV in an Exim configuration. See NewStuff for the gory
2211 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2212 Makefile that are specific to HP-UX.
2214 PH/18 If the "use_postmaster" option was set for a recipient callout together
2215 with the "random" option, the postmaster address was used as the MAIL
2216 FROM address for the random test, but not for the subsequent recipient
2217 test. It is now used for both.
2219 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2220 patch removes a few documentation additions to RFC 3028, because the
2221 latest draft now contains them. It adds the new en;ascii-case comparator
2222 and a new error check for 8bit text in MIME parts. Comparator and
2223 require names are now matched exactly. I enabled the subaddress
2224 extension, but it is not well tested yet (read: it works for me)."
2226 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2227 rework some of the code of TK/09 above to avoid the hardwired use of
2228 "%lld" and "long long". Replaced the call to snprintf() with a call to
2231 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2232 records point to non-existent hosts", "retry timeout exceeded", and
2233 "retry time not reached for any host after a long failure period".
2235 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2236 experimental DomainKeys support:
2238 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2239 (2) On an error such as an illegally used "control", the wrong name for
2240 the control was given.
2242 These problems did NOT occur unless DomainKeys support was compiled.
2244 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2246 PH/24 Added ${if match_ip condition.
2248 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2249 databases so that it will be absolutely obvious if a crash occurs in the
2250 DB library. This is a regular occurrence (often caused by mis-matched
2253 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2254 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2255 picked up on a user's system that detects such things. There doesn't seem
2256 to be a gcc warning option for this - only an attribute that has to be
2257 put on the function's prototype. It seems that in Fedora Core 4 they have
2258 set this on a number of new functions. No doubt there will be more in due
2261 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2262 specify a transport. However, if an address that was verified by such a
2263 router was the subject of a callout, Exim crashed because it tried to
2264 read the rcpt_include_affixes from the non-existent transport. Now it
2265 just assumes that the setting of that option is false. This bug was
2266 introduced by 4.51/PH/31.
2268 PH/28 Changed -d+all to exclude +memory, because that information is very
2269 rarely of interest, but it makes the output a lot bigger. People tend to
2270 do -d+all out of habit.
2272 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2273 code in os-type was giving problems when libc.so lives in lib64, like on
2276 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2277 aren't the modern standard, and it seems that some systems' include files
2278 don't always have them. Exim was already checking for some of the newer
2279 ones like T_AAAA, and defining it itself. I've added checks for all the
2280 record types that Exim uses.
2282 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2283 not automatically generating a new one, as it is supposed to. This
2284 prevented TLS from working. If the file did exist, but contained invalid
2285 data, a new version was generated, as expected. It was only the case of a
2286 non-existent file that was broken.
2288 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2289 with a change in libdomainkeys > 0.64.
2291 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2292 from DNS. If the selector record carries the flag, it now has
2293 precedence over the domain-wide flag.
2295 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2297 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2298 the use of a port name that isn't defined in /etc/services) occurred, the
2299 message was deferred as in a normal delivery, and thus remained on the
2300 spool, instead of being failed because of the mua_wrapper setting. This
2301 is now fixed, and I tidied up some of the mua_wrapper messages at the
2304 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2305 the messages in a hash of arrays rather than using individual hashes.
2306 This is a bit cleaner and results in dramatic memory savings, albeit
2307 at a slight CPU cost.
2309 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2310 as requested by Marc Sherman.
2312 SC/10 Eximstats: added histograms for user specified patterns as requested
2315 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2317 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2318 fopen() in the content-scanning modules that did not already have it.
2324 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2325 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2327 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2329 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2331 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2332 to test Sieve filters that use "vacation".
2334 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2335 that changes the way the GnuTLS parameters are stored in the cache file.
2336 The new format can be generated externally. For backward compatibility,
2337 if the data in the cache doesn't make sense, Exim assumes it has read an
2338 old-format file, and it generates new data and writes a new file. This
2339 means that you can't go back to an older release without removing the
2342 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2343 work if there are any delivery delays because "one_time" forces the
2344 parent to be marked "delivered", so its unseen clone is never tried
2345 again. For this reason, Exim now forbids the simultaneous setting of
2348 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2349 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2350 entry if you want to know the details. The fix, however, neglected to
2351 consider the case when local delivery batching is involved. The test for
2352 "previously delivered" was not happening when checking to see if an
2353 address could be batched with a previous (undelivered) one; under
2354 certain circumstances this could lead to multiple deliveries to the same
2357 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2358 in its include files, and this causes problems building Exim.
2360 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2361 header_syntax) but Exim was just ignoring anything given after a slash.
2362 In particular, this caused confusion with an attempt to use "verify =
2363 reverse_host_lookup/defer_ok". An error is now given when options are
2364 supplied for verify items that do not have them. (Maybe reverse_host_
2365 lookup should have a defer_ok option, but that's a different point.)
2367 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2368 defined by RFC 821) to 2048, because there were problems with some AUTH
2369 commands, and RFC 1869 says the size should be increased for extended
2370 SMTP commands that take arguments.
2372 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2375 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2376 "unknown" error; now it says that the functionality isn't in the binary.
2378 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2379 an address' error message when a string expansion fails (syntax or
2380 whatever). Otherwise the password may appear in the log. Following change
2381 PH/42 below, there is no longer a chance of it appearing in a bounce
2384 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2386 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2387 of its addresses was ever tried. (Bugzilla bug #2.)
2389 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2390 the result incorrectly in the debug output. (It correctly added a newline
2391 to what was transported.)
2393 TF/01 Added $received_time.
2395 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2396 commented out examples of how to interface to a virus scanner and to
2397 SpamAssassin. Also added commented examples of av_scanner and
2398 spamd_address settings.
2400 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2401 and controls are allowed in which ACLs. There were a couple of minor
2402 errors. Some of the entries in the conditions table (which is a table of
2403 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2404 negation of where the condition IS allowed.
2406 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2408 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2409 header file does not have a version number, so I've had to invent a new
2410 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2411 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2412 radiusclient), but it was contributed by a Radius user.
2414 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2415 files or directories, for OpenSSL.
2417 PH/19 When an Exim process that is running as root has to create an Exim log
2418 file, it does so in a subprocess that runs as exim:exim so as to get the
2419 ownership right at creation (otherwise, other Exim processes might see
2420 the file with the wrong ownership). There was no test for failure of this
2421 fork() call, which would lead to the process getting stuck as it waited
2422 for a non-existent subprocess. Forks do occasionally fail when resources
2423 run out. I reviewed all the other calls to fork(); they all seem to check
2426 PH/20 When checking for unexpected SMTP input at connect time (before writing
2427 the banner), Exim was not dealing correctly with a non-positive return
2428 from the read() function. If the client had disconnected by this time,
2429 the result was a log entry for a synchronization error with an empty
2430 string after "input=" when read() returned zero. If read() returned -1
2431 (an event I could not check), uninitialized data bytes were printed.
2432 There were reports of junk text (parts of files, etc) appearing after
2435 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2437 PH/22 Added support for macro redefinition, and (re)definition in between
2438 driver and ACL definitions.
2440 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2441 forgetting to use the resulting value; it was using the unexpanded value.
2443 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2444 hadn't been configured. The fix is from Juergen Kreileder, who
2445 understands it better than I do:
2447 "Here's what I see happening with three configured cyrus_sasl
2448 authenticators configured (plain, login, cram-md5):
2450 On startup auth_cyrus_sasl_init() gets called for each of these.
2451 This means three calls to sasl_listmech() without a specified mech_list.
2452 => SASL tests which mechs of all available mechs actually work
2453 => three warnings about OTP not working
2454 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2456 With the patch, sasl_listmech() also gets called three times. But now
2457 SASL's mech_list option is set to the server_mech specified in the the
2458 authenticator. Or in other words, the answer from sasl_listmech()
2459 gets limited to just the mech you're testing for (which is different
2461 => the return list contains just 'plain' or 'login', 'cram-md5' or
2462 nothing depending on the value of ob->server_mech.
2464 I've just tested the patch: Authentication still works fine,
2465 unavailable mechs specified in the exim configuration are still
2466 caught, and the auth.log warnings about OTP are gone."
2468 PH/25 When debugging is enabled, the contents of the command line are added
2469 to the debugging output, even when log_selector=+arguments is not
2472 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2473 answer is "GNU", and only if the return is "GNU/something" is the answer
2476 PH/27 $acl_verify_message is now set immediately after the failure of a
2477 verification in an ACL, and so is available in subsequent modifiers. In
2478 particular, the message can be preserved by coding like this:
2480 warn !verify = sender
2481 set acl_m0 = $acl_verify_message
2483 Previously, $acl_verify_message was set only while expanding "message"
2484 and "log_message" when a very denied access.
2486 PH/28 Modified OS/os.c-Linux with
2488 -#ifndef OS_LOAD_AVERAGE
2489 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2491 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2492 nomenclature these days.)
2494 PH/29 Installed patch from the Sieve maintainer that adds the options
2495 sieve_useraddress and sieve_subaddress to the redirect router.
2497 PH/30 In these circumstances:
2498 . Two addresses routed to the same list of hosts;
2499 . First host does not offer TLS;
2500 . First host accepts first address;
2501 . First host gives temporary error to second address;
2502 . Second host offers TLS and a TLS session is established;
2503 . Second host accepts second address.
2504 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2505 and peerdn, if requested) that were in fact used only for the second
2508 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2509 attention to any local part prefix or suffix that was matched by the
2510 router that accepted the address. It now behaves in the same way as it
2511 does for delivery: the affixes are removed from the local part unless
2512 rcpt_include_affixes is set on the transport.
2514 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2515 timeout during the DATA phase of an incoming message.
2517 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2518 applied a patch sanctioned by the Sieve maintainer.
2520 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2521 the uid or gid is negative. A case of a negative gid caused this to be
2522 noticed. The fix allows for either to be negative.
2524 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2525 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2528 PH/36 The change PH/12 above was broken. Fixed it.
2530 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2531 the grounds that routing the same address twice would always produce the
2532 same answer. This might have been true once, but it is certainly no
2533 longer true now. Routing a child address may depend on the previous
2534 routing that produced that child. Some complicated redirection strategies
2535 went wrong when messages had multiple recipients, and made Exim's
2536 behaviour dependent on the order in which the addresses were given.
2538 I have moved the duplicate checking until after the routing is complete.
2539 Exim scans the addresses that are assigned to local and remote
2540 transports, and removes any duplicates. This means that more work will be
2541 done, as duplicates will always all be routed, but duplicates are
2542 presumably rare, so I don't expect this is of any significance.
2544 For deliveries to pipes, files, and autoreplies, the duplicate checking
2545 still happens during the routing process, since they are not going to be
2548 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2549 It corrects a timeout issue with spamd. This is Ian's comment: "The
2550 background is that sometimes spamd either never reads data from a
2551 connection it has accepted, or it never writes response data. The exiscan
2552 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2553 blindly assumes that writes won't block so it may never time out."
2555 PH/39 Allow G after quota size as well as K and M.
2557 PH/40 The value set for $authenticated_id in an authenticator may not contain
2558 binary zeroes or newlines because the value is written to log lines and
2559 to spool files. There was no check on this. Now the value is run through
2560 the string_printing() function so that such characters are converted to
2561 printable escape sequences.
2563 PH/41 $message_linecount is a new variable that contains the total number of
2564 lines in the message. Compare $body_linecount, which is the count for the
2567 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2568 bounce and delay warning messages, except in certain special cases, which
2571 (a) An SMTP error message from a remote host;
2572 (b) A message specified in a :fail: redirection;
2573 (c) A message specified in a "fail" command in a system filter;
2574 (d) A message specified in a FAIL return from the queryprogram router;
2575 (e) A message specified by the cannot_route_message router option.
2577 In these cases only, Exim does include the error details in bounce and
2578 warning messages. There are also a few cases where bland messages such
2579 as "unrouteable address" or "local delivery error" are given.
2581 PH/43 $value is now also set for the "else" part of a ${run expansion.
2583 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2584 being worked on, but at least Exim now implements the latest version to
2587 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2588 process to complete was treated as a delivery failure, a timeout while
2589 writing the message to the pipe was logged, but erroneously treated as a
2590 successful delivery. Such timeouts include transport filter timeouts. For
2591 consistency with the overall process timeout, these timeouts are now
2592 treated as errors, giving rise to delivery failures by default. However,
2593 there is now a new Boolean option for the pipe transport called
2594 timeout_defer, which, if set TRUE, converts the failures into defers for
2595 both kinds of timeout. A transport filter timeout is now identified in
2598 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2599 systems where "make" and "gmake" are different, calling "gmake" at top
2600 level broke things. I've arranged for the value of $(MAKE) to be passed
2601 from the Makefile to this script so that it can call the same version of
2605 A note about Exim versions 4.44 and 4.50
2606 ----------------------------------------
2608 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2609 changes of various kinds. As a consequence, a big documentation update was
2610 needed. This delayed the release for rather longer than seemed good, especially
2611 in the light of a couple of (minor) security issues. Therefore, the changes
2612 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2613 release. So 4.44 and 4.50 are in effect two different branches that both start
2616 I have left the 4.50 change log unchanged; it contains all the changes since
2617 4.43. The change log for 4.44 is below; many of its items are identical to
2618 those for 4.50. This seems to be the most sensible way to preserve the
2619 historical information.
2625 1. Minor wording change to the doc/README.SIEVE file.
2627 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2628 computation of the current number of files was incorrect.
2630 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2631 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2634 4. Give more explanation in the error message when the command for a transport
2635 filter fails to execute.
2637 5. There are several places where Exim runs a non-Exim command in a
2638 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2639 was being done only for the command run by the queryprogram router. It is
2640 now done for all such subprocesses. The other cases are: ${run, transport
2641 filters, and the commands run by the lmtp and pipe transports.
2643 6. Added CONFIGURE_GROUP build-time option.
2645 7. Some older OS have a limit of 256 on the maximum number of file
2646 descriptors. Exim was using setrlimit() to set 1000 as a large value
2647 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2648 systems. I've change it so that if it can't get 1000, it tries for 256.
2650 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2651 was an oversight, and furthermore, ever since the addition of extra
2652 controls (e.g. 4.43/32), the checks on when to allow different forms of
2653 "control" were broken. There should now be diagnostics for all cases when a
2654 control that does not make sense is encountered.
2656 9. Added the /retain_sender option to "control=submission".
2658 10. $recipients is now available in the predata ACL (oversight).
2660 11. Tidy the search cache before the fork to do a delivery from a message
2661 received from the command line. Otherwise the child will trigger a lookup
2662 failure and thereby defer the delivery if it tries to use (for example) a
2663 cached ldap connection that the parent has called unbind on.
2665 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2666 of $address_data from the recipient verification was clobbered by the
2667 sender verification.
2669 13. The value of address_data from a sender verification is now available in
2670 $sender_address_data in subsequent conditions in the ACL statement.
2672 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2674 15. Added a new option "connect=<time>" to callout options, to set a different
2677 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2678 was its contents. (It was OK if the option was not defined at all.)
2680 17. A "Completed" log line is now written for messages that are removed from
2681 the spool by the -Mrm option.
2683 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2684 information about exactly what failed.
2686 19. Added -dd to debug only the daemon process.
2688 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2689 handles timeouts, both on the server side and network timeouts. Renamed the
2690 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2692 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2693 It is now set to "smtps".
2695 22. $host_address is now set to the target address during the checking of
2696 ignore_target_hosts.
2698 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2699 being passed; this would have caused $sender_host_name to have been used if
2700 matching the list had actually called for a host name (not very likely,
2701 since this list is usually IP addresses). A host name is now passed as
2704 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2705 code that specifies a non-restarting handler (typically sigaction() in
2706 modern systems) in an attempt to fix a rare and obscure crash bug.
2708 25. Narrowed the window for a race in the daemon that could cause it to ignore
2709 SIGCHLD signals. This is not a major problem, because they are used only to
2710 wake it up if nothing else does.
2712 26. A malformed maildirsize file could cause Exim to calculate negative values
2713 for the mailbox size or file count. Odd effects could occur as a result.
2714 The maildirsize information is now recalculated if the size or filecount
2717 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2718 support for a long time. Removed HAVE_SYS_VFS_H.
2720 28. Installed the latest version of exipick from John Jetmore.
2722 29. In an address list, if the pattern was not a regular expression, an empty
2723 subject address (from a bounce message) matched only if the pattern was an
2724 empty string. Non-empty patterns were not even tested. This was the wrong
2725 because it is perfectly reasonable to use an empty address as part of a
2726 database query. An empty address is now tested by patterns that are
2727 lookups. However, all the other forms of pattern expect the subject to
2728 contain a local part and a domain, and therefore, for them, an empty
2729 address still always fails if the pattern is not itself empty.
2731 30. Exim went into a mad DNS loop when attempting to do a callout where the
2732 host was specified on an smtp transport, and looking it up yielded more
2733 than one IP address.
2735 31. Re-factored the code for checking spool and log partition space into a
2736 function that finds that data and another that does the check. The former
2737 is then used to implement four new variables: $spool_space, $log_space,
2738 $spool_inodes, and $log_inodes.
2740 32. The RFC2047 encoding function was originally intended for short strings
2741 such as real names; it was not keeping to the 75-character limit for
2742 encoded words that the RFC imposes. It now respects the limit, and
2743 generates multiple encoded words if necessary. To be on the safe side, I
2744 have increased the buffer size for the ${rfc2047: expansion operator from
2747 33. It is now permitted to omit both strings after an "if" condition; if the
2748 condition is true, the result is "true". As before, when the second string
2749 is omitted, a false condition yields an empty string. This makes it less
2750 cumbersome to write custom ACL and router conditions.
2752 34. Failure to deliver a bounce message always caused it to be frozen, even if
2753 there was an errors_to setting on the router. The errors_to setting is now
2756 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2757 canonical form (fully expanded) before being placed in
2758 $sender_host_address.
2760 36. The table in the code that translates DNS record types into text (T_A to
2761 "A" for instance) was missing entries for NS and CNAME. It is just possible
2762 that this could have caused confusion if both these types were looked up
2763 for the same domain, because the text type is used as part of Exim's
2764 per-process caching. But the chance of anyone hitting this buglet seems
2767 37. The dnsdb lookup has been extended in a number of ways.
2769 (1) There is a new type, "zns", which walks up the domain tree until it
2770 finds some nameserver records. It should be used with care.
2772 (2) There is a new type, "mxh", which is like "mx" except that it returns
2773 just the host names, not the priorities.
2775 (3) It is now possible to give a list of domains (or IP addresses) to be
2776 looked up. The behaviour when one of the lookups defers can be
2777 controlled by a keyword.
2779 (4) It is now possible to specify the separator character for use when
2780 multiple records are returned.
2782 38. The dnslists ACL condition has been extended: it is now possible to supply
2783 a list of IP addresses and/or domains to be looked up in a particular DNS
2786 39. Added log_selector=+queue_time_overall.
2788 40. When running the queue in the test harness, wait just a tad after forking a
2789 delivery process, to get repeatability of debugging output.
2791 41. Include certificate and key file names in error message when GnuTLS fails
2792 to set them up, because the GnuTLS error message doesn't include the name
2793 of the failing file when there is a problem reading it.
2795 42. Allow both -bf and -bF in the same test run.
2797 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2799 44. The "Exiscan patch" is now merged into the mainline Exim source.
2801 45. Sometimes the final signoff response after QUIT could fail to get
2802 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2803 < 0 before doing a fflush(). This bug looks as though it goes back to the
2804 introduction of TLS in release 3.20, but "sometimes" must have been rare
2805 because the tests only now provoked it.
2807 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2808 (this can affect the format of dates).
2810 47. exim_tidydb, when checking for the continued existence of a message for
2811 which it has found a message-specific retry record, was not finding
2812 messages that were in split spool directories. Consequently, it was
2813 deleting retry records that should have stayed in existence.
2815 48. Steve fixed some bugs in eximstats.
2817 49. The SPA authentication driver was not abandoning authentication and moving
2818 on to the next authenticator when an expansion was forced to fail,
2819 contradicting the general specification for all authenticators. Instead it
2820 was generating a temporary error. It now behaves as specified.
2822 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2823 (the order specifies the preference for clients). The order is now AES256,
2824 AES128, 3DES, ARCFOUR128.
2826 51. Small patch to Sieve code - explicitly set From: when generating an
2829 52. Exim crashed if a remote delivery caused a very long error message to be
2830 recorded - for instance if somebody sent an entire SpamAssassin report back
2831 as a large number of 550 error lines. This bug was coincidentally fixed by
2832 increasing the size of one of Exim's internal buffers (big_buffer) that
2833 happened as part of the Exiscan merge. However, to be on the safe side, I
2834 have made the code more robust (and fixed the comments that describe what
2837 53. Now that there can be additional text after "Completed" in log lines (if
2838 the queue_time_overall log selector is set), a one-byte patch to exigrep
2839 was needed to allow it to recognize "Completed" as not the last thing in
2842 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2843 patch that reportedly fixes this has been added. I am not expert enough to
2844 create a test for it. This is what the patch creator wrote:
2846 "I found a little strange behaviour of ldap code when working with
2847 Windows 2003 AD Domain, where users was placed in more than one
2848 Organization Units. When I tried to give exim partial DN, the exit code
2849 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2850 But simultaneously result of request was absolutely normal ldap result,
2851 so I produce this patch..."
2853 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2854 REFERENCE, so I have modified the code to exclude the patch when that macro
2857 55. Some experimental protocols are using DNS PTR records for new purposes. The
2858 keys for these records are domain names, not reversed IP addresses. The
2859 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2860 leaves it alone. Component reversal etc. now happens only for IP addresses.
2863 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2865 57. Double the size of the debug message buffer (to 2048) so that more of very
2866 long debug lines gets shown.
2868 58. The exicyclog utility now does better if the number of log files to keep
2869 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2871 59. Two changes related to the smtp_active_hostname option:
2873 (1) $smtp_active_hostname is now available as a variable.
2874 (2) The default for smtp_banner uses $smtp_active_hostname instead
2875 of $primary_hostname.
2877 60. The host_aton() function is supposed to be passed a string that is known
2878 to be a valid IP address. However, in the case of IPv6 addresses, it was
2879 not checking this. This is a hostage to fortune. Exim now panics and dies
2880 if the condition is not met. A case was found where this could be provoked
2881 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2882 components; fortuitously, this particular loophole had already been fixed
2883 by change 4.50/55 above.
2885 If there are any other similar loopholes, the new check in host_aton()
2886 itself should stop them being exploited. The report I received stated that
2887 data on the command line could provoke the exploit when Exim was running as
2888 exim, but did not say which command line option was involved. All I could
2889 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2890 running as the user.
2893 61. There was a buffer overflow vulnerability in the SPA authentication code
2894 (which came originally from the Samba project). I have added a test to the
2895 spa_base64_to_bits() function which I hope fixes it.
2898 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2899 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2901 63. The daemon start-up calls getloadavg() while still root for those OS that
2902 need the first call to be done as root, but it missed one case: when
2903 deliver_queue_load_max is set with deliver_drop_privilege. This is
2904 necessary for the benefit of the queue runner, because there is no re-exec
2905 when deliver_drop_privilege is set.
2907 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2908 This has been fixed.
2910 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2911 was in use, was not putting the data itself into the right store pool;
2912 consequently, it could be overwritten for a subsequent message in the same
2913 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2916 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2918 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2919 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2920 the function were treating the return as a boolean value, which happened to
2921 work because 0=false and not-0=true, but is not correct code.
2923 68. The host_aton() function was not handling scoped IPv6 addresses (those
2924 with, for example, "%eth0" on the end) correctly.
2926 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2927 negated items (that is, ~something) in unsigned ints. Some compilers
2928 apparently mutter when there is no cast.
2930 70. If an address verification called from an ACL failed, and did not produce a
2931 user-specific message (i.e. there was only a "system" message), nothing was
2932 put in $acl_verify_message. In this situation, it now puts the system
2935 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2936 session; change 4.31/43 added the unwanted input to the log line - except
2937 that it did not do this in the start of session case. It now does.
2939 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2940 This is wrong and can cause the other end to generate a synchronization
2941 error if it is another Exim or anything else that does the synchronization
2942 check. A QUIT command is no longer sent after a timeout.
2944 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2945 during host lookups.
2947 74. The defer_ok option of callout verification was not working if it was used
2948 when verifying addresses in header lines, that is, for this case:
2950 verify = header_sender/callout=defer_ok
2952 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2953 those file descriptors could be used for SMTP connections. If anything
2954 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2955 could be sent to the SMTP client, causing chaos. The daemon now opens
2956 stdin, stdout, and stderr to /dev/null when it puts itself into the
2959 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2960 log by default. The user can override this with suitable Perl magic.
2962 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2963 the log message when discard triggers, was not working for the DATA ACL or
2964 for the non-SMTP ACL.
2966 78. Error message wording change in sieve.c.
2968 79. If smtp_accept_max_per_host was set, the number of connections could be
2969 restricted to fewer than expected, because the daemon was trying to set up
2970 a new connection before checking whether the processes handling previous
2971 connections had finished. The check for completed processes is now done
2972 earlier. On busy systems, this bug wouldn't be noticed because something
2973 else would have woken the daemon, and it would have reaped the completed
2976 80. If a message was submitted locally by a user whose login name contained one
2977 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2978 It caused a spool format error. I have fixed the spool reading code. A
2979 related problem was that the "from" clause in the Received: line became
2980 illegal because of the space(s). It is now covered by ${quote_local_part.
2982 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2985 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2987 83. Patch from Sieve maintainer to fix unterminated string problem in
2988 "vacation" handling.
2990 84. Some minor changes to the Linux configuration files to help with other
2991 OS variants using glibc.
2993 85. One more patch for Sieve to update vacation handling to latest spec.
2996 ----------------------------------------------------
2997 See the note above about the 4.44 and 4.50 releases.
2998 ----------------------------------------------------
3004 1. Change 4.43/35 introduced a bug that caused file counts to be
3005 incorrectly computed when quota_filecount was set in an appendfile
3008 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3009 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3012 3. Give more explanation in the error message when the command for a transport
3013 filter fails to execute.
3015 4. There are several places where Exim runs a non-Exim command in a
3016 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3017 was being done only for the command run by the queryprogram router. It is
3018 now done for all such subprocesses. The other cases are: ${run, transport
3019 filters, and the commands run by the lmtp and pipe transports.
3021 5. Some older OS have a limit of 256 on the maximum number of file
3022 descriptors. Exim was using setrlimit() to set 1000 as a large value
3023 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3024 systems. I've change it so that if it can't get 1000, it tries for 256.
3026 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3027 was an oversight, and furthermore, ever since the addition of extra
3028 controls (e.g. 4.43/32), the checks on when to allow different forms of
3029 "control" were broken. There should now be diagnostics for all cases when a
3030 control that does not make sense is encountered.
3032 7. $recipients is now available in the predata ACL (oversight).
3034 8. Tidy the search cache before the fork to do a delivery from a message
3035 received from the command line. Otherwise the child will trigger a lookup
3036 failure and thereby defer the delivery if it tries to use (for example) a
3037 cached ldap connection that the parent has called unbind on.
3039 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3040 of $address_data from the recipient verification was clobbered by the
3041 sender verification.
3043 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3044 was its contents. (It was OK if the option was not defined at all.)
3046 11. A "Completed" log line is now written for messages that are removed from
3047 the spool by the -Mrm option.
3049 12. $host_address is now set to the target address during the checking of
3050 ignore_target_hosts.
3052 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3053 being passed; this would have caused $sender_host_name to have been used if
3054 matching the list had actually called for a host name (not very likely,
3055 since this list is usually IP addresses). A host name is now passed as
3058 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3059 code that specifies a non-restarting handler (typically sigaction() in
3060 modern systems) in an attempt to fix a rare and obscure crash bug.
3062 15. Narrowed the window for a race in the daemon that could cause it to ignore
3063 SIGCHLD signals. This is not a major problem, because they are used only to
3064 wake it up if nothing else does.
3066 16. A malformed maildirsize file could cause Exim to calculate negative values
3067 for the mailbox size or file count. Odd effects could occur as a result.
3068 The maildirsize information is now recalculated if the size or filecount
3071 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3072 support for a long time. Removed HAVE_SYS_VFS_H.
3074 18. Updated exipick to current release from John Jetmore.
3076 19. Allow an empty sender to be matched against a lookup in an address list.
3077 Previously the only cases considered were a regular expression, or an
3080 20. Exim went into a mad DNS lookup loop when doing a callout where the
3081 host was specified on the transport, if the DNS lookup yielded more than
3084 21. The RFC2047 encoding function was originally intended for short strings
3085 such as real names; it was not keeping to the 75-character limit for
3086 encoded words that the RFC imposes. It now respects the limit, and
3087 generates multiple encoded words if necessary. To be on the safe side, I
3088 have increased the buffer size for the ${rfc2047: expansion operator from
3091 22. Failure to deliver a bounce message always caused it to be frozen, even if
3092 there was an errors_to setting on the router. The errors_to setting is now
3095 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3096 canonical form (fully expanded) before being placed in
3097 $sender_host_address.
3099 24. Updated eximstats to version 1.33
3101 25. Include certificate and key file names in error message when GnuTLS fails
3102 to set them up, because the GnuTLS error message doesn't include the name
3103 of the failing file when there is a problem reading it.
3105 26. Expand error message when OpenSSL has problems setting up cert/key files.
3108 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3109 (this can affect the format of dates).
3111 28. exim_tidydb, when checking for the continued existence of a message for
3112 which it has found a message-specific retry record, was not finding
3113 messages that were in split spool directories. Consequently, it was
3114 deleting retry records that should have stayed in existence.
3116 29. eximstats updated to version 1.35
3117 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3118 1.35 - bugfix such that pie charts by volume are generated correctly
3120 30. The SPA authentication driver was not abandoning authentication and moving
3121 on to the next authenticator when an expansion was forced to fail,
3122 contradicting the general specification for all authenticators. Instead it
3123 was generating a temporary error. It now behaves as specified.
3125 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3126 (the order specifies the preference for clients). The order is now AES256,
3127 AES128, 3DES, ARCFOUR128.
3129 31. Small patch to Sieve code - explicitly set From: when generating an
3132 32. Exim crashed if a remote delivery caused a very long error message to be
3133 recorded - for instance if somebody sent an entire SpamAssassin report back
3134 as a large number of 550 error lines. This bug was coincidentally fixed by
3135 increasing the size of one of Exim's internal buffers (big_buffer) that
3136 happened as part of the Exiscan merge. However, to be on the safe side, I
3137 have made the code more robust (and fixed the comments that describe what
3140 33. Some experimental protocols are using DNS PTR records for new purposes. The
3141 keys for these records are domain names, not reversed IP addresses. The
3142 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3143 leaves it alone. Component reversal etc. now happens only for IP addresses.
3146 34. The host_aton() function is supposed to be passed a string that is known
3147 to be a valid IP address. However, in the case of IPv6 addresses, it was
3148 not checking this. This is a hostage to fortune. Exim now panics and dies
3149 if the condition is not met. A case was found where this could be provoked
3150 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3151 components; fortuitously, this particular loophole had already been fixed
3152 by change 4.50/55 or 4.44/33 above.
3154 If there are any other similar loopholes, the new check in host_aton()
3155 itself should stop them being exploited. The report I received stated that
3156 data on the command line could provoke the exploit when Exim was running as
3157 exim, but did not say which command line option was involved. All I could
3158 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3159 running as the user.
3162 35. There was a buffer overflow vulnerability in the SPA authentication code
3163 (which came originally from the Samba project). I have added a test to the
3164 spa_base64_to_bits() function which I hope fixes it.
3167 36. The daemon start-up calls getloadavg() while still root for those OS that
3168 need the first call to be done as root, but it missed one case: when
3169 deliver_queue_load_max is set with deliver_drop_privilege. This is
3170 necessary for the benefit of the queue runner, because there is no re-exec
3171 when deliver_drop_privilege is set.
3173 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3174 was in use, was not putting the data itself into the right store pool;
3175 consequently, it could be overwritten for a subsequent message in the same
3176 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3179 38. Sometimes the final signoff response after QUIT could fail to get
3180 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3181 < 0 before doing a fflush(). This bug looks as though it goes back to the
3182 introduction of TLS in release 3.20, but "sometimes" must have been rare
3183 because the tests only now provoked it.
3189 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3190 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3191 Now it returns an integer. A number of calls were still expecting a T/F
3192 return. Fortuitously, in all cases, the tests worked in OK situations,
3193 which is the norm. However, things would have gone wrong on any write
3194 failures on the smtp file descriptor. This function is used when sending
3195 messages over SMTP and also when doing verify callouts.
3197 2. When Exim is called to do synchronous delivery of a locally submitted
3198 message (the -odf or -odi options), it no longer closes stderr before doing
3201 3. Implemented the mua_wrapper option.
3203 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3205 5. Implemented the functions header_remove(), header_testname(),
3206 header_add_at_position(), and receive_remove_recipient(), and exported them
3209 6. If an ACL "warn" statement specified the addition of headers, Exim already
3210 inserted X-ACL-Warn: at the start if there was no header name. However, it
3211 was not making this test for the second and subsequent header lines if
3212 there were newlines in the string. This meant that an invalid header could
3213 be inserted if Exim was badly configured.
3215 7. Allow an ACL "warn" statement to add header lines at the start or after all
3216 the Received: headers, as well as at the end.
3218 8. Added the rcpt_4xx retry error code.
3220 9. Added postmaster_mailfrom=xxx to callout verification option.
3222 10. Added mailfrom=xxxx to the callout verification option, for verify=
3225 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3226 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3228 12. Inserted some casts to stop certain compilers warning when using pointer
3229 differences as field lengths or precisions in printf-type calls (mostly
3230 affecting debugging statements).
3232 13. Added optional readline() support for -be (dynamically loaded).
3234 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3235 same clock tick as a message's arrival, so that its received time was the
3236 same as the "first fail" time on the retry record, and that message
3237 remained on the queue past the ultimate address timeout, every queue runner
3238 would try a delivery (because it was past the ultimate address timeout) but
3239 after another failure, the ultimate address timeout, which should have then
3240 bounced the address, did not kick in. This was a "< instead of <=" error;
3241 in most cases the first failure would have been in the next clock tick
3242 after the received time, and all would be well.
3244 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3245 being recognized when the domain list was tested by the match_domain
3246 condition in an expansion string.
3248 16. Added the ${str2b64: operator.
3250 17. Exim was always calling setrlimit() to set a large limit for the number of
3251 processes, without checking whether the existing limit was already
3252 adequate. (It did check for the limit on file descriptors.) Furthermore,
3253 errors from getrlimit() and setrlimit() were being ignored. Now they are
3254 logged to the main and panic logs, but Exim does carry on, to try to do its
3255 job under whatever limits there are.
3257 18. Imported PCRE 5.0.
3259 19. Trivial typo in log message " temporarily refused connection" (the leading
3262 20. If the log selector return_path_on_delivery was set and an address was
3263 redirected to /dev/null, the delivery process crashed because it assumed
3264 that a return path would always be set for a "successful" delivery. In this
3265 case, the whole delivery is bypassed as an optimization, and therefore no
3268 21. Internal re-arrangement: the function for sending a challenge and reading
3269 a response while authentication was assuming a zero-terminated challenge
3270 string. It's now changed to take a pointer and a length, to allow for
3271 binary data in such strings.
3273 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3275 23. Exim was not respecting finduser_retries when seeking the login of the
3276 uid under which it was called; it was always trying 10 times. (The default
3277 setting of finduser_retries is zero.) Also, it was sleeping after the final
3278 failure, which is pointless.
3280 24. Implemented tls_on_connect_ports.
3282 25. Implemented acl_smtp_predata.
3284 26. If the domain in control=submission is set empty, Exim assumes that the
3285 authenticated id is a complete email address when it generates From: or
3286 Sender: header lines.
3288 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3289 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3290 chown and chgrp in /bin and hostname in /usr/bin.
3292 28. Exim was keeping the "process log" file open after each use, just as it
3293 does for the main log. This opens the possibility of it remaining open for
3294 long periods when the USR1 signal hits a daemon. Occasional processlog
3295 errors were reported, that could have been caused by this. Anyway, it seems
3296 much more sensible not to leave this file open at all, so that is what now
3299 29. The long-running daemon process does not normally write to the log once it
3300 has entered its main loop, and it closes the log before doing so. This is
3301 so that log files can straightforwardly be renamed and moved. However,
3302 there are a couple of unusual error situations where the daemon does write
3303 log entries, and I had neglected to close the log afterwards.
3305 30. The text of an SMTP error response that was received during a remote
3306 delivery was being truncated at 512 bytes. This is too short for some of
3307 the long messages that one sometimes sees. I've increased the limit to
3310 31. It is now possible to make retry rules that apply only when a message has a
3311 specific sender, in particular, an empty sender.
3313 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3314 it possible to be selective about when SMTP synchronization is enforced.
3316 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3318 32. Implemented hosts_connection_nolog.
3320 33. Added an ACL for QUIT.
3322 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3325 35. Added mailbox_size and mailbox_filecount to appendfile.
3327 36. Added control = no_multiline_responses to ACLs.
3329 37. There was a bug in the logic of the code that waits for the clock to tick
3330 in the case where the clock went backwards by a substantial amount such
3331 that the microsecond fraction of "now" was more than the microsecond
3332 fraction of "then" (but the whole seconds number was less).
3334 38. Added support for the libradius Radius client library this is found on
3335 FreeBSD (previously only the radiusclient library was supported).
3341 1. When certain lookups returned multiple values in the form name=value, the
3342 quoting of the values was not always being done properly. Specifically:
3343 (a) If the value started with a double quote, but contained no whitespace,
3345 (b) If the value contained whitespace other than a space character (i.e.
3346 tabs or newlines or carriage returns) it was not quoted.
3347 This fix has been applied to the mysql and pgsql lookups by writing a
3348 separate quoting function and calling it from the lookup code. The fix
3349 should probably also be applied to nisplus, ibase and oracle lookups, but
3350 since I cannot test any of those, I have not disturbed their existing code.
3352 2. A hit in the callout cache for a specific address caused a log line with no
3353 reason for rejecting RCPT. Now it says "Previous (cached) callout
3354 verification failure".
3356 3. There was an off-by-one bug in the queryprogram router. An over-long
3357 return line was truncated at 256 instead of 255 characters, thereby
3358 overflowing its buffer with the terminating zero. As well as fixing this, I
3359 have increased the buffer size to 1024 (and made a note to document this).
3361 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3362 when Exim is waiting for an SMTP response from a remote server, Exim
3363 restarts its select() call on the socket, thereby resetting its timeout.
3364 This is not a problem when such interrupts are rare. Somebody set up a cron
3365 job to run exiwhat every 2 minutes, which is less than the normal select()
3366 timeout (5 or 10 minutes). This meant that the select() timeout never
3367 kicked in because it was always reset. I have fixed this by comparing the
3368 time when an interrupt arrives with the time at the start of the first call
3369 to select(). If more time than the timeout has elapsed, the interrupt is
3370 treated as a timeout.
3372 5. Some internal re-factoring in preparation for the addition of Sieve
3373 extensions (by MH). In particular, the "personal" test is moved to a
3374 separate function, and given an option for scanning Cc: and Bcc: (which is
3375 not set for Exim filters).
3377 6. When Exim created an email address using the login of the caller as the
3378 local part (e.g. when creating a From: or Sender: header line), it was not
3379 quoting the local part when it contained special characters such as @.
3381 7. Installed new OpenBSD configuration files.
3383 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3384 try to make them clearer.
3386 9. Callout options, other than the timeout value, were being ignored when
3387 verifying sender addresses in header lines. For example, when using
3389 verify = header_sender/callout=no_cache
3391 the cache was (incorrectly) being used.
3393 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3394 only the Cygwin environment.
3396 11. When return_path_on_delivery was set as a log selector, if different remote
3397 addresses in the same message used different return paths and parallel
3398 remote delivery occurred, the wrong values would sometimes be logged.
3399 (Whenever a remote delivery process finished, the return path value from
3400 the most recently started remote delivery process was logged.)
3402 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3403 lines when AUTH and/or TLS are in use. This is the "received protocol"
3404 field. Exim used to use "asmtp" for authenticated SMTP, without any
3405 indication (in the protocol name) for TLS use. Now it follows the RFC and
3406 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3407 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3408 names appear in log lines as well as in Received: header lines.
3410 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3411 extensions, and comparison tests, and to fix some bugs.
3413 14. Changes to the "personal" filter test:
3415 (1) The test was buggy in that it was just doing the equivalent of
3416 "contains" tests on header lines. For example, if a user's address was
3417 anne@some.where, the "personal" test would incorrectly be true for
3419 To: susanne@some.where
3421 This test is now done by extracting each address from the header in turn,
3422 and checking the entire address. Other tests that are part of "personal"
3423 are now done using regular expressions (for example, to check local parts
3424 of addresses in From: header lines).
3426 (2) The list of non-personal local parts in From: addresses has been
3427 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3428 taken from the Sieve specification recommendations.
3430 (3) If the message contains any header line starting with "List-" it is
3431 treated as non-personal.
3433 (4) The test for "circular" in the Subject: header line has been removed
3434 because it now seems ill-conceived.
3436 15. Minor typos in src/EDITME comments corrected.
3438 16. Installed latest exipick from John Jetmore.
3440 17. If headers_add on a router specified a text string that was too long for
3441 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3442 of string_sprintf() is now avoided.
3444 18. $message_body_size was not set (it was always zero) when running the DATA
3445 ACL and the local_scan() function.
3447 19. For the "mail" command in an Exim filter, no default was being set for
3448 the once_repeat time, causing a random time value to be used if "once" was
3449 specified. (If the value happened to be <= 0, no repeat happened.) The
3450 default is now 0s, meaning "never repeat". The "vacation" command was OK
3451 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3452 (I found it when inspecting the code).
3454 20. There is now an overall timeout for performing a callout verification. It
3455 defaults to 4 times the callout timeout, which applies to individual SMTP
3456 commands during the callout. The overall timeout applies when there is more
3457 than one host that can be tried. The timeout is checked before trying the
3458 next host. This prevents very long delays if there are a large number of
3459 hosts and all are timing out (e.g. when the network connections are timing
3460 out). The value of the overall timeout can be changed by specifying an
3461 additional sub-option for "callout", called "maxwait". For example:
3463 verify = sender/callout=5s,maxwait=20s
3465 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3466 to the end before writing, but this should make it even safer).
3468 22. Exim was forgetting that it had advertised PIPELINING for the second and
3469 subsequent messages on an SMTP connection. It was also not resetting its
3470 memory on STARTTLS and an internal HELO.
3472 23. When Exim logs an SMTP synchronization error within a session, it now
3473 records whether PIPELINING has been advertised or not.
3475 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3476 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3477 rather than long int.
3479 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3481 26. Added the never_mail option to autoreply.
3487 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3488 crash if the getsockname() call failed; this can happen if a connection is
3489 closed very soon after it is established. The problem was simply in the
3490 order in which certain operations were done, causing Exim to try to write
3491 to the SMTP stream before it had set up the file descriptor. The bug has
3492 been fixed by making things happen in the correct order.
3498 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3499 before the connection was closed, thus losing the rejection response.
3501 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3502 some early Solaris releases, but causes trouble in current releases where
3503 socklen_t is defined.
3505 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3508 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3511 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3512 that releasing the top of it at the end releases what was used for sub-
3513 expansions (unless the block got too big). However, discard this block if
3514 the first thing is a variable or header, so that we can use its block when
3515 it is dynamic (useful for very large $message_headers, for example).
3517 6. Lookups now cache *every* query, not just the most recent. A new, separate
3518 store pool is used for this. It can be recovered when all lookup caches are
3519 flushed. Lookups now release memory at the end of their result strings.
3520 This has involved some general refactoring of the lookup sources.
3522 7. Some code has been added to the store_xxx() functions to reduce the amount
3523 of flapping under certain conditions.
3525 8. log_incoming_interface used to affect only the <= reception log lines. Now
3526 it causes the local interface and port to be added to several more SMTP log
3527 lines, for example "SMTP connection from", and rejection lines.
3529 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3531 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3533 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3534 could be overwritten at the end of the current message (or the start of a
3535 new message if it was set in a HELO ACL). The value is now preserved for
3536 the duration of the SMTP connection.
3538 12. If a transport had a headers_rewrite setting, and a matching header line
3539 contained an unqualified address, that address was qualified, even if it
3540 did not match any rewriting rules. The underlying bug was that the values
3541 of the flags that permit the existence of unqualified sender and recipient
3542 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3543 non-local messages, and by -bnq for local messages) were not being
3544 preserved with the message after it was received.
3546 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3547 "next input=" as part of the text comprising the host identity instead of
3548 the correct text. The code was using the same buffer for two different
3549 strings. However, depending on which order the printing function evaluated
3550 its arguments, the bug did not always show up. Under Linux, for example, my
3551 test suite worked just fine.
3553 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3554 doesn't work with some older versions of Perl. It has been changed to "my",
3555 which in any case is probably the better facility to use.
3557 15. A really picky compiler found some instances of statements for creating
3558 error messages that either had too many or two few arguments for the format
3561 16. The size of the buffer for calls to the DNS resolver has been increased
3562 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3563 for addresses that have a lot of PTR records. This alleviates a problem; it
3564 does not fully solve it.
3566 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3567 buffer now truncates the list and logs the incident, which is the same
3568 action as happens when Exim is looking up a host name and its aliases.
3569 Previously in this situation something unpredictable would happen;
3570 sometimes it was "internal error: store_reset failed".
3572 18. If a server dropped the connection unexpectedly when an Exim client was
3573 using GnuTLS and trying to read a response, the client delivery process
3574 crashed while trying to generate an error log message.
3576 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3577 string, for example:
3579 warn message = H1: something\nH2: something
3581 the text was added as a single header line from Exim's point of view
3582 though it ended up OK in the delivered message. However, searching for the
3583 second and subsequent header lines using $h_h2: did not work. This has been
3584 fixed. Similarly, if a system filter added multiple headers in this way,
3585 the routers could not see them.
3587 20. Expanded the error message when iplsearch is called with an invalid key to
3588 suggest using net-iplsearch in a host list.
3590 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3591 ACLs are no longer actually imposed (and a message to that effect is
3594 22. If a "gecos" field in a passwd entry contained escaped characters, in
3595 particular, if it contained a \" sequence, Exim got it wrong when building
3596 a From: or a Sender: header from that name. A second bug also caused
3597 incorrect handling when an unquoted " was present following a character
3598 that needed quoting.
3600 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3601 was not being matched caselessly.
3603 24. Arranged for all hyphens in the exim.8 source to be escaped with
3606 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3607 properly. Recipient callout cache records were still being keyed to include
3608 the sender, even when use_sender was set false. This led to far more
3609 callouts that were necessary. The sender is no longer included in the key
3610 when use_sender is false.
3612 26. Added "control = submission" modifier to ACLs.
3614 27. Added the ${base62d: operator to decode base 62 numbers.
3616 28. dnsdb lookups can now access SRV records.
3618 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3619 the configuration file.
3621 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3622 (-v) mode. This makes the output for a verbose queue run more intelligible.
3624 31. Added a use_postmaster feature to recipient callouts.
3626 32. Added the $body_zerocount variable, containing the number of binary zero
3627 bytes in the message body.
3629 33. The time of last modification of the "new" subdirectory is now used as the
3630 "mailbox time last read" when there is a quota error for a maildir
3633 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3635 35. Added +ignore_unknown as a special item in host lists.
3637 36. Code for decoding IPv6 addresses in host lists is now included, even if
3638 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3639 address was recognized as an IP address, but was then not correctly decoded
3640 into binary, causing unexpected and incorrect effects when compared with
3647 1. Very minor rewording of debugging text in manualroute to say "list of
3648 hosts" instead of "hostlist".
3650 2. If verify=header_syntax was set, and a header line with an unqualified
3651 address (no domain) and a large number of spaces between the end of the
3652 name and the colon was received, the reception process suffered a buffer
3653 overflow, and (when I tested it) crashed. This was caused by some obsolete
3654 code that should have been removed. The fix is to remove it!
3656 3. When running in the test harness, delay a bit after writing a bounce
3657 message to get a bit more predictability in the log output.
3659 4. Added a call to search_tidyup() just before forking a reception process. In
3660 theory, someone could use a lookup in the expansion of smtp_accept_max_
3661 per_host which, without the tidyup, could leave open a database connection.
3663 5. Added the variables $recipient_data and $sender_data which get set from a
3664 lookup success in an ACL "recipients" or "senders" condition, or a router
3665 "senders" option, similar to $domain_data and $local_part_data.
3667 6. Moved the writing of debug_print from before to after the "senders" test
3670 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3671 problems for message scanning, either using a data ACL, or using
3672 local_scan() because the Received: header was not generated till after they
3673 were called (in order to set the time as the time of reception completion).
3674 I have revised the way this works. The header is now generated after the
3675 body is received, but before the ACL or local_scan() are called. After they
3676 are run, the timestamp in the header is updated.
3682 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3683 before starting a queue runner without re-exec. This happened only when
3684 deliver_drop_privilege was set or when the Exim user was set to root. The
3685 effect of the bug was that timeouts during subsequent deliveries caused
3686 crashes instead of being properly handled. The handler is now left at its
3687 default (and expected) setting.
3689 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3690 message, again when deliver_drop_privilege is set or Exim is run as root.
3691 The bug described in (1) was not present in this case, but the tidying up
3692 of the other signals was missing. I have made the two cases consistent.
3694 3. The ignore_target_hosts setting on a manualroute router was being ignored
3695 for hosts that were looked up using the /MX notation.
3697 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3700 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3701 operated on the sender address. After changing the $sender_address to <>
3702 for the sender address verify, Exim was re-instated it as the original
3703 (before rewriting) address, but remembering that it had rewritten it, so it
3704 wasn't rewriting it again. This bug also had the effect of breaking the
3705 sender address verification caching when the sender address was rewritten.
3707 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3708 This has been changed so that if the ip literal address matches
3709 ignore_target_hosts, the router declines.
3711 7. Added expansion conditions match_domain, match_address, and match_local_
3712 part (NOT match_host).
3714 8. The placeholder for the Received: header didn't have a length field set.
3716 9. Added code to Exim itself and to exim_lock to test for a specific race
3717 condition that could lead to file corruption when using MBX delivery. The
3718 issue is with the lockfile that is created in /tmp. If this file is removed
3719 after a process has opened it but before that process has acquired a lock,
3720 there is the potential for a second process to recreate the file and also
3721 acquire a lock. This could lead to two Exim processes writing to the file
3722 at the same time. The added code performs the same test as UW imapd; it
3723 checks after acquiring the lock that its file descriptor still refers to
3724 the same named file.
3726 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3727 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3730 11. Added the smtp_active_hostname option. If used, this will typically be made
3731 to depend on the incoming interface address. Because $interface_address is
3732 not set up until the daemon has forked a reception process, error responses
3733 that can happen earlier (such as "too many connections") no longer contain
3736 12. If an expansion in a condition on a "warn" statement fails because a lookup
3737 defers, the "warn" statement is abandoned, and the next ACL statement is
3738 processed. Previously this caused the whole ACL to be aborted.
3740 13. Added the iplsearch lookup type.
3742 14. Added ident_timeout as a log selector.
3744 15. Added tls_certificate_verified as a log selector.
3746 16. Added a global option tls_require_ciphers (compare the smtp transport
3747 option of the same name). This controls incoming TLS connections.
3749 17. I finally figured out how to make tls_require_ciphers do a similar thing
3750 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3751 before starting the TLS session.
3753 18. Tabs are now shown as \t in -bP output.
3755 19. If the log selector return_path_on_delivery was set, Exim crashed when
3756 bouncing a message because it had too many Received: header lines.
3758 20. If two routers both had headers_remove settings, and the first one included
3759 a superfluous trailing colon, the final name in the first list and the
3760 first name in the second list were incorrectly joined into one item (with a
3761 colon in the middle).
3767 1. Added -C and -D options to the exinext utility, mainly to make it easier
3768 to include in the automated testing, but these could be helpful when
3769 multiple configurations are in use.
3771 2. The exinext utility was not formatting the output nicely when there was
3772 an alternate port involved in the retry record key, nor when there was a
3773 message id as well (for retries that were specific to a specific message
3774 and a specific host). It was also confused by IPv6 addresses, because of
3775 the additional colons they contain. I have fixed the IPv4 problem, and
3776 patched it up to do a reasonable job for IPv6.
3778 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3779 delivery, the log line now contains "pipelined" if PIPELINING was used.
3781 4. An SMTP transport process used to panic and die if the bind() call to set
3782 an explicit outgoing interface failed. This has been changed; it is now
3783 treated in the same way as a connect() failure.
3785 5. A reference to $sender_host_name in the part of a conditional expansion
3786 that was being skipped was still causing a DNS lookup. This no longer
3789 6. The def: expansion condition was not recognizing references to header lines
3790 that used bh_ and bheader_.
3792 7. Added the _cache feature to named lists.
3794 8. The code for checking quota_filecount in the appendfile transport was
3795 allowing one more file than it should have been.
3797 9. For compatibility with Sendmail, the command line option
3805 and sets the incoming protocol and host name (for trusted callers). The
3806 host name and its colon can be omitted when only the protocol is to be set.
3807 Note the Exim already has two private options, -pd and -ps, that refer to
3808 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3809 "s", but I don't think that's a major issue.
3811 10. A number of refactoring changes to the code, none of which should affect
3814 (a) The number of logging options was getting close to filling up the
3815 32-bit word that was used as a bit map. I have split them into two classes:
3816 those that are passed in the argument to log_write(), and those that are
3817 only ever tested independently outside of that function. These are now in
3818 separate 32-bit words, so there is plenty of room for expansion again.
3819 There is no change in the user interface or the logging behaviour.
3821 (b) When building, for example, log lines, the code previously used a
3822 macro that called string_cat() twice, in order to add two strings. This is
3823 not really sufficiently general. Furthermore, there was one instance where
3824 it was actually wrong because one of the argument was used twice, and in
3825 one call a function was used. (As it happened, calling the function twice
3826 did not affect the overall behaviour.) The macro has been replaced by a
3827 function that can join an arbitrary number of extra strings onto a growing
3830 (c) The code for expansion conditions now uses a table and a binary chop
3831 instead of a serial search (which was left over from when there were very
3832 few conditions). Also, it now recognizes conditions like "pam" even when
3833 the relevant support is not compiled in: a suitably worded error message is
3834 given if an attempt is made to use such a condition.
3836 11. Added ${time_interval:xxxxx}.
3838 12. A bug was causing one of the ddress fields not to be passed back correctly
3839 from remote delivery subprocesses. The field in question was not being
3840 subsequently used, so this caused to problems in practice.
3842 13. Added new log selectors queue_time and deliver_time.
3844 14. Might have fixed a bug in maildirsizefile handling that threw up
3845 "unexpected character" debug warnings, and recalculated the data
3846 unnecessarily. In any case, I expanded the warning message to give more
3849 15. Added the message "Restricted characters in address" to the statements in
3850 the default ACL that block characters like @ and % in local parts.
3852 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3853 Three changes have been made:
3855 (a) There was a serious bug; a negative response to MAIL caused the whole
3856 recipient domain to be cached as invalid, thereby blocking all messages
3857 to all local parts at the same domain, from all senders. This bug has
3858 been fixed. The domain is no longer cached after a negative response to
3859 MAIL if the sender used is not empty.
3861 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3864 (c) A new callout option, "use_sender" has been added for people who want
3865 the modified behaviour.
3871 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3874 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3875 indeed breaks things for older releases.
3877 3. Added additional logging to the case where there is a problem reading data
3878 from a filter that is running in a subprocess using a pipe, in order to
3879 try to track down a specific problem.
3881 4. Testing facility fudge: when running in the test harness and attempting
3882 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3883 getting "No route to host". Convert this to a timeout.
3885 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3888 6. Some OS don't have socklen_t but use size_t instead. This affects the
3889 fifth argument of getsockopt() amongst other things. This is now
3890 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3891 can be set for individual OS. I have set it for SunOS5, OSF1, and
3892 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3893 some earlier ones do not.
3895 7. Change 4.30/15 was not doing the test caselessly.
3897 8. The standard form for an IPv6 address literal was being rejected by address
3898 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3899 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3900 this, as well as the form without the "IPv6" on the front (but only when
3901 address literals are enabled, of course).
3903 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3905 10. Exim crashed if a message with an empty sender address specified by -f
3906 encountered a router with an errors_to setting. This could be provoked only
3907 by a command such as
3911 where an empty string was supplied; "<>" did not hit this bug.
3913 11. Installed PCRE release 4.5.
3915 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3916 remained set. It is now erased.
3918 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3919 times from message ids (which are base 36 rather than the normal 62).
3921 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3922 were being counted as actual protocol errors, and logged if the log
3923 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3924 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3925 and DATA following a set of rejected RCPTs do not count as protocol errors.
3926 In other words, Exim assumes they were pipelined, though this may not
3927 actually be the case. Of course, in all cases the client gets an
3928 appropriate error code.
3930 15. If a lookup fails in an ACL condition, a message about the failure may
3931 be available; it is used if testing the ACL cannot continue, because most
3932 such messages specify what the cause of the deferral is. However, some
3933 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3934 that caused an old message to be retained and used if a later statement
3935 caused a defer, replacing the real cause of the deferral.
3937 16. If an IP address had so many PTR records that the DNS lookup buffer
3938 was not large enough to hold them, Exim could crash while trying to process
3939 the truncated data. It now detects and logs this case.
3941 17. Further to 4.21/58, another change has been made: if (and only if) the
3942 first line of a message (the first header line) ends with CRLF, a bare LF
3943 in a subsequent header line has a space inserted after it, so as not to
3944 terminate the header.
3946 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3947 unnecessarily, used atoi() instead of strtol(), and didn't check the
3948 termination when getting file sizes from file names by regex.
3950 19. Completely re-implemented the support for maildirsize files, in the light
3951 of a number of problems with the previous contributed implementation
3952 (4.30/29). In particular:
3954 . If the quota is zero, the maildirsize file is maintained, but no quota is
3957 . If the maildir directory does not exist, it is created before any attempt
3958 to write a maildirsize file.
3960 . The quota value in the file is just a cache; if the quota is changed in
3961 the transport, the new value overrides.
3963 . A regular expression is available for excluding directories from the
3966 20. The autoreply transport checks the characters in options that define the
3967 message's headers; it allows continued headers, but it was checking with
3968 isspace() after an embedded newline instead of explicitly looking for a
3971 21. If all the "regular" hosts to which an address was routed had passed their
3972 expiry times, and had not reached their retry times, the address was
3973 bounced, even if fallback hosts were defined. Now Exim should go on to try
3976 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3977 equivalent code in the SMTP transport. Some hosts send humungous responses
3978 to HELO/EHLO, more than 1024 it seems.
3980 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3981 gives compiler warnings in some environments. I've now done it "properly",
3984 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3985 (because of problems with the built-in one) was declared to return uschar *
3986 instead of char *, causing compiler failure.
3988 25. Fixed a file descriptor leak when processing alias/forward files.
3990 26. Fixed a minor format string issue in dbfn.c.
3992 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3994 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3995 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3998 29. When a sender address is verified, it is cached, to save repeating the test
3999 when there is more than one recipient in a message. However, when the
4000 verification involves a callout, it is possible for different callout
4001 options to be set for different recipients. It is too complicated to keep
4002 track of this in the cache, so now Exim always runs a verification when a
4003 callout is required, relying on the callout cache for the optimization.
4004 The overhead is duplication of the address routing, but this should not be
4007 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4008 to be verified with callout=postmaster, and the main callout worked but the
4009 postmaster check failed, the verification correctly failed. However, if a
4010 subsequent RCPT command asked for sender verification *without* the
4011 postmaster check, incorrect caching caused this verification also to fail,
4014 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4015 it was not caching the DNS options (qualify_single, search_parents) that
4016 were used when the lookup failed. A subsequent lookup with different
4017 options therefore always gave the same answer, though there were cases
4018 where it should not have. (Example: a "domains = !$mx_any" option on a
4019 dnslookup router: the "domains" option is always processed without any
4020 widening, but the router might have qualify_single set.) Now Exim uses the
4021 cached value only when the same options are set.
4023 32. Added John Jetmore's "exipick" utility to the distribution.
4025 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4026 than a timeout (e.g. a certificate is required, and is not provided), an
4027 Exim server now closes the connection immediately. Previously it waited for
4028 the client to close - but if the client is SSL, it seems that they each
4029 wait for each other, leading to a delay before one of them times out.
4031 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4032 maintained 0.8.x compatibility because I don't think many are using it, and
4033 it is clearly obsolete.
4035 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4038 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4039 client certificate was expired. A simple patch fixes this, though I don't
4040 understand the full logic of why the verify callback is called multiple
4043 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4044 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4045 which causes problems with some clients (such as the Certicom SSL Plus
4046 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4047 disables the coutermeasure allowing Eudora to connect."
4049 38. Exim was not checking that a write() to a log file succeeded. This could
4050 lead to Bad Things if a log got too big, in particular if it hit a file
4051 size limit. Exim now panics and dies if it cannot write to a log file, just
4052 as it does if it cannot open a log file.
4054 39. Modified OS/Makefile-Linux so that it now contains
4056 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4058 The two -D definitions ensure that Exim is compiled with large file
4059 support, which makes it possible to handle log files that are bigger than
4062 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4063 instance) a domain was checked against a named list that involved a lookup,
4064 causing $domain_data to be set, then another domain was checked against the
4065 same list, then the first domain was re-checked, the value of $domain_data
4066 after the final check could be wrong. In particular, if the second check
4067 failed, it could be set empty. This bug probably also applied to
4070 41. The strip_trailing_dot option was not being applied to the address given
4071 with the -f command-line option.
4073 42. The code for reading a message's header from the spool was incrementing
4074 $received_count, but never initializing it. This meant that the value was
4075 incorrect (doubled) while delivering a message in the same process in which
4076 it was received. In the most common configuration of Exim, this never
4077 happens - a fresh exec is done - but it can happen when
4078 deliver_drop_privilege is set.
4080 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4081 it now includes up to 150 characters of the unexpected data in the log
4084 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4085 and building data strings. The size of both of these buffers was 10 000
4086 bytes - far larger than anybody would *ever* want, thought I. Needless to
4087 say, somebody hit the limit. I have increased the maximum line length to
4088 20 000 and the maximum data length of concatenated lines to 100 000. I have
4089 also fixed two bugs, because there was no checking on these buffers. Tsk,
4090 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4091 buffer is too small.
4093 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4094 lsearch lookups. Now it does.
4096 46. When parsing a route_list item in a manualroute router, a fixed-length
4097 buffer was used for the list of hosts. I made this 1024 bytes long,
4098 thinking that nobody would ever have a list of hosts that long. Wrong.
4099 Somebody had a whole pile of complicated expansion conditions, and the
4100 string was silently truncated, leading to an expansion error. It turns out
4101 that it is easier to change to an unlimited length (owing to other changes
4102 that have happened since this code was originally written) than to build
4103 structure for giving a limitation error. The length of the item that
4104 expands into the list of hosts is now unlimited.
4106 47. The lsearch lookup could not handle data where the length of text line was
4107 more than 4095 characters. Such lines were truncated, leading to shortened
4108 data being returned. It should now handle lines of any length.
4110 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4111 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4114 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4115 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4116 no longer happens in comment lines. A list of replacements is now placed
4117 at the head of all of the source files, except those whose only change is
4118 to replace PERL_COMMAND in the very first #! line.
4120 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4121 messages on the queue, with a bottom-up merge sort, using code contributed
4122 by Michael Haardt. This should make operations like -bp somewhat faster on
4123 large queues. It won't affect queue runners, except when queue_run_in_order
4126 51. Installed eximstats 1.31 in the distribution.
4128 52. Added support for SRV lookups to the dnslookup router.
4130 53. If an ACL referred to $message_body or $message_body_end, the value was not
4131 reset for any messages that followed in the same SMTP session.
4133 54. The store-handling optimization for building very long strings was not
4134 differentiating between the different store pools. I don't think this
4135 actually made any difference in practice, but I've tidied it.
4137 55. While running the routers to verify a sender address, $sender_address
4138 was still set to the sender address. This is wrong, because when routing to
4139 send a bounce to the sender, it would be empty. Therefore, I have changed
4140 it so that, while verifying a sender address, $sender_address is set to <>.
4141 (There is no change to what happens when verifying a recipient address.)
4143 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4144 target A or AAAA records (if not already returned) without resetting the
4145 qualify_single or search_parents options of the DNS resolver. These are
4146 inappropriate in this case because the targets of MX and SRV records must
4147 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4148 target that, when qualified, matched something in the local domain. These
4149 two options are now turned off when doing these lookups.
4151 57. It seems that at least some releases of Reiserfs (which does not have the
4152 concept of a fixed number of inodes) returns zero and not -1 for the
4153 number of available inodes. This interacted badly with check_spool_inodes,
4154 which assumed that -1 was the "no such thing" setting. What I have done is
4155 to check that the total number of inodes is greater than zero before doing
4156 the test of how many are available.
4158 58. When a "warn" ACL statement has a log_message modifier, the message is
4159 remembered, and not repeated. This is to avoid a lot of repetition when a
4160 message has many recipients that cause the same warning to be written.
4161 Howewer, Exim was preserving the list of already written lines for an
4162 entire SMTP session, which doesn't seem right. The memory is now reset if a
4163 new message is started.
4165 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4166 debugging output unless log_rewrite was also set.
4168 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4169 of (int)(handle) when we know that handle contains (void *)(-1).
4171 61. The Exim daemon panic-logs an error return when it closes the incoming
4172 connection. However "connection reset by peer" seems to be common, and
4173 isn't really an error worthy of noting specially, so that particular error
4176 62. When Exim is trying to find all the local interfaces, it used to panic and
4177 die if the ioctl to get the interface flags failed. However, it seems that
4178 on at least one OS (Solaris 9) it is possible to have an interface that is
4179 included in the list of interfaces, but for which you get a failure error
4180 for this call. This happens when the interface is not "plumbed" into a
4181 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4182 failure of the "get flags" call assumes that the interface is down.
4184 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4185 makes life easier for people who are doing arithmetic on fields extracted
4186 from dates, where you often get leading zeros that should not be
4187 interpreted as octal.
4189 64. Added qualify_domain to the redirect router, to override the global
4192 65. If a pathologically long header line contained very many addresses (the
4193 report of this problem mentioned 10 000) and each of them was rewritten,
4194 Exim could use up a very large amount of memory. (It kept on making new
4195 copies of the header line as it rewrote, and never released the old ones.)
4196 At the expense of a bit more processing, the header rewriting function has
4197 been changed so that it no longer eats memory in this way.
4199 66. The generation of the Received: header has been moved from the time that a
4200 message starts to be received, to the time that it finishes. The timestamp
4201 in the Received: header should now be very close to that of the <= log
4202 line. There are two side-effects of this change:
4204 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4205 logged header lines no longer include the local Received: line, because
4206 it has not yet been created. The same applies to a copy of the message
4207 that is returned to a non-SMTP sender when a message is rejected.
4209 (b) When a filter file is tested using -bf, no additional Received: header
4210 is added to the test message. After some thought, I decided that this
4213 This change does not affect the value of $received_for. It is still set
4214 after address rewriting, but before local_scan() is called.
4216 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4218 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4219 gave an unhelpful panic error message, and a defer error. I have managed to
4220 change this behaviour so that it now rejects any supplied certificate,
4221 which seems right, as the list of acceptable certificates is empty.
4223 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4224 gave an unhelpful defer error. I have not managed to make this reject any
4225 supplied certificates, but the error message it gives is "no certificate
4226 supplied", which is not helpful.
4228 70. exigrep's output now also includes lines that are not associated with any
4229 message, but which match the given pattern. Implemented by a patch from
4230 Martin Sluka, which also tidied up the Perl a bit.
4232 71. Recipient callout verification, like sender verification, was using <> in
4233 the MAIL FROM command. This isn't really the right thing, since the actual
4234 sender may affect whether the remote host accepts the recipient or not. I
4235 have changed it to use the actual sender in the callout; this means that
4236 the cache record is now keyed on a recipient/sender pair, not just the
4237 recipient address. There doesn't seem to be a real danger of callout loops,
4238 since a callout by the remote host to check the sender would use <>.
4239 [SEE ABOVE: changed after hitting problems.]
4241 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4242 temporary errors. However, in the case of such a code being given after
4243 the end of a data transmission (i.e. after ".") Exim was failing to write
4244 a retry record for the message. (Yes, there was some broken host that was
4245 actually sending 8xx at this point.)
4247 73. An unknown lookup type in a host list could cause Exim to panic-die when
4248 the list was checked. (An example that provoked this was putting <; in the
4249 middle of a list instead of at the start.) If this happened during a DATA
4250 ACL check, a -D file could be left lying around. This kind of configuration
4251 error no longer causes Exim to die; instead it causes a defer errror. The
4252 incident is still logged to the main and panic logs.
4254 74. Buglet left over from Exim 3 conversion. The message "too many messages
4255 in one connection" was written to the rejectlog but not the mainlog, except
4256 when address rewriting (yes!) was being logged.
4258 75. Added write_rejectlog option.
4260 76. When a system filter was run not as root (that is, when system_filter_user
4261 was set), the values of the $n variables were not being returned to the
4262 main process; thus, they were not subsequently available in the $sn
4265 77. Added +return_path_on_delivery log selector.
4267 78. A connection timeout was being treated differently from recipients deferred
4268 when testing hosts_max_try with a message that was older than the host's
4269 retry timeout. (The host should not be counted, thus allowing all hosts to
4270 be tried at least once before bouncing.) This may have been the cause of an
4271 occasionally reported bug whereby a message would remain on the queue
4272 longer than the retry timeout, but would be bounced if a delivery was
4273 forced. I say "may" because I never totally pinned down the problem;
4274 setting up timeout/retry tests is difficult. See also the next item.
4276 79. The ultimate address timeout was not being applied to errors that involved
4277 a combination of host plus message (for example, a timeout on a MAIL
4278 command). When an address resolved to a number of possible hosts, and they
4279 were not all tried for each delivery (e.g. because of hosts_max_try), a
4280 message could remain on the queue longer than the retry timeout.
4282 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4285 81. Fixed an obscure SMTP outgoing bug which required at least the following
4286 conditions: (a) there was another message waiting for the same server;
4287 (b) the server returned 5xx to all RCPT commands in the first message so
4288 that the message was not completed; (c) the server dropped the connection
4289 or gave a negative response to the RSET that Exim sends to abort the
4290 transaction. The observed case was a dropped connection after DATA that had
4291 been sent in pipelining mode. That is, the server had advertised PIPELINING
4292 but was not implementing it correctly. The effect of the bug was incorrect
4293 behaviour, such as trying another host, and this could lead to a crash.
4299 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4300 and daemon.c were passed as pointers to ints; they should have been
4301 pointers to socklen_t variables (which are typically unsigned ints).
4303 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4306 3. Fixed a really odd bug that affected only the testing scheme; patching a
4307 certain fixed string in the binary changed the value of another string that
4308 happened to be identical to the end of the original first string.
4310 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4311 name", it returns that address as the IP address. On some operating
4312 systems (e.g. Solaris), it also passes back the IP address string as the
4313 "host name". However, on others (e.g. Linux), it passes back an empty
4314 string. Exim wasn't checking for this, and was changing the host name to an
4315 empty string, assuming it had been canonicized.
4317 5. Although rare, it is permitted to have more than one PTR record for a given
4318 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4319 all the names associated with an address, because they do in Solaris.
4320 However, it seems that they do not in Linux for data that comes from the
4321 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4322 I found this out when I moved to a new Linux workstation and tried to run
4323 the Exim test suite.
4325 To get round this problem I have changed the code so that it now does its
4326 own call to the DNS to look up PTR records when searching for a host name.
4327 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4328 addresses that are only in /etc/hosts are still found.
4330 This behaviour is, however, controlled by an option called host_lookup_
4331 order, which defaults to "bydns:byaddr". If people want to use the other
4332 order, or indeed, just use one or the other means of lookup, they can
4333 specify it in this variable.
4335 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4336 some operating systems, this comes back from gethostbyaddr() as an empty
4337 string, and this is what Exim used to test for. However, it seems that in
4338 other systems, "." is yielded. Exim now tests for this case too.
4340 7. The values of check_spool_space and check_log_space are now held internally
4341 as a number of kilobytes instead of an absolute number of bytes. If a
4342 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4343 kilobyte. This means that much larger values can be stored.
4345 8. Exim monitor: an attempt to get the action menu when not actually pointing
4346 at a message produces an empty menu entitled "No message selected". This
4347 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4348 no entries in it ("Shell widget menu has zero width and/or height"). So I
4349 have added a single, blank menu entry in this case.
4351 9. Added ${quote_local_part.
4353 10. MIME decoding is now applied to the contents of Subject: header lines when
4356 11. Now that a reference to $sender_host_address automatically causes a reverse
4357 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4358 host lookup before query-style lookups in lists that might use this
4359 variable. This has therefore been abolished, and the "net-" prefix is no
4360 longer necessary for query-style lookups.
4362 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4363 have been a typo for LFLAGS, so it has been changed.
4365 13. The install script calls Exim with "-C /dev/null" in order to find the
4366 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4367 to be output. Howeve, since Exim outputs its version number before the
4368 error, it didn't break the script. It just looked ugly. I fixed this by
4369 always allowing "-C /dev/null" if the caller is root.
4371 14. Ignore overlarge ACL variable number when reading spool file - insurance
4372 against a later release with more variables having written the file.
4374 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4375 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4376 this, as well as the form without the "IPv6" on the front.
4378 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4379 OS/Makefile-Darwin file.
4381 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4382 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4384 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4385 variable. However, some OS use ints for inodes. I've added cast to long int
4386 to get rid of the compiler warning.
4388 19. I had forgotten to lock out "/../" in configuration file names when
4389 ALT_CONFIG_PREFIX was set.
4391 20. Routers used for verification do not need to specify transports. However,
4392 if such a router generated a host list, and callout was configured, Exim
4393 crashed, because it could not find a port number from the (non-existent)
4394 transport. It now assumes port 25 in this circumstance.
4396 21. Added the -t option to exigrep.
4398 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4399 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4400 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4401 them is set, LOOKUP_LSEARCH is forced.
4403 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4405 24. Added sender and host information to the "rejected by local_scan()" log
4406 line; previously there was no indication of these.
4408 25. Added .include_if_exists.
4410 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4411 Linux. It turns out that not all file systems support this. Apparently some
4412 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4413 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4414 not supported on the file descriptor, is now ignored when Exim is trying to
4415 sync a directory. This applies only to Linux.
4417 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4419 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4420 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4423 29. Added support for maildirsize files from supplied patch (modified a bit).
4425 30. The use of :fail: followed by an empty string could lead Exim to respond to
4426 sender verification failures with (e.g.):
4428 550 Verification failed for <xxx>
4429 550 Sender verify failed
4431 where the first response line was missing the '-' that indicates it is not
4432 the final line of the response.
4434 31. The loop for finding the name of the user that called Exim had a hardwired
4435 limit of 10; it now uses the value of finduser_retries, which is used for
4436 all other user lookups.
4438 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4441 33. Exim was neglecting to zero errno before one call of strtol() when
4442 expanding a string and expecting an integer value. On some systems this
4443 resulted in spurious "integer overflow" errors. Also, it was casting the
4444 result into an int without checking.
4446 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4447 did not work. The code looks as if it has *never* worked, though it appears
4448 to have been documented since at least releast 1.62. I have made it work.
4450 35. The "timeout_DNS" error in retry rules, also documented since at least
4451 1.62, also never worked. As it isn't clear exactly what this means, and
4452 clearly it isn't a major issue, I have abolished the feature by treating it
4453 as "timeout", and writing a warning to the main and panic logs.
4455 36. The display of retry rules for -brt wasn't always showing the error code
4458 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4459 timeout_connect_A, timeout_connect_MX.
4461 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4462 to the empty sender.
4464 39. The daemon was not analysing the content of -oX till after it had closed
4465 stderr and disconnected from the controlling terminal. This meant that any
4466 syntax errors were only noted on the panic log, and the return code from
4467 the command was 0. By re-arranging the code a little, I've made the
4468 decoding happen first, so such errors now appear on stderr, and the return
4469 code is 1. However, the actual setting up of the sockets still happens in
4470 the disconnected process, so errors there are still only recorded on the
4473 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4474 connections (as happens on some IP stacks) was logged at start up time as
4475 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4476 it from "IPv6 and IPv4", which means that two separate sockets are being
4479 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4480 says whether AF_INET or AF_INET6 was passed as an argument.
4482 42. Exiwhat output was messed up when time zones were included in log
4485 43. Exiwhat now gives more information about the daemon's listening ports,
4486 and whether -tls-on-connect was used.
4488 44. The "port" option of the smtp transport is now expanded.
4490 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4491 silently ignored. Now an error message is written to the main and panic
4494 46. There's a new ACL modifier called "logwrite" which writes to a log file
4495 as soon as it is encountered.
4497 47. Added $local_user_uid and $local_user_gid at routing time.
4499 48. Exim crashed when trying to verify a sender address that was being
4502 49. Exim was recognizing only a space character after ".include". It now also
4503 recognizes a tab character.
4505 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4506 extracting the relevant information from the specification. The man page no
4507 longer contains scrambled data for the -d option, and I've added a section
4508 at the front about calling Exim under different names.
4510 51. Added "extra_headers" argument to the "mail" command in filter files.
4512 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4515 53. Installed eximstats 1.29.
4517 54. Added transport_filter_timeout as a generic transport option.
4519 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4520 Cc: header lines. This was required by RFC 822, but it not required by RFC
4523 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4524 incoming messages that did not have them. Now it does so only if the
4525 message originates locally, that is, if there is no associated remote host
4526 address. When Resent- header lines are present, this applies to the Resent-
4527 lines rather than the non-Resent- lines.
4529 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4530 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4532 58. Messages for configuration errors now include the name of the main
4533 configuration file - useful now that there may be more than one file in a
4534 list (.included file names were always shown).
4536 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4537 for those rare installations that do not start the daemon as root or run it
4538 setuid root. I've cut out the call to initgroups() if the daemon is not
4541 60. The Exim user and group can now be bound into the binary as text strings
4542 that are looked up at the start of Exim's processing.
4544 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4546 62. Added $mailstore_basename variable.
4548 63. Installed patch to sieve.c from Michael Haardt.
4550 64. When Exim failed to open the panic log after failing to open the main log,
4551 the original message it was trying to log was written to stderr and debug
4552 output, but if they were not available (the usual case in production), it
4553 was lost. Now it is written to syslog before the two lines that record the
4554 failures to open the logs.
4556 65. Users' Exim filters run in subprocesses under the user's uid. It is
4557 possible for a "deliver" command or an alias in a "personal" command to
4558 provoke an address rewrite. If logging of address rewriting is configured,
4559 this fails because the process is not running as root or exim. There may be
4560 a better way of dealing with this, but for the moment (because 4.30 needs
4561 to be released), I have disabled address rewrite logging when running a
4562 filter in a non-root, non-exim process.
4568 1. The buildconfig auxiliary program wasn't quoting the value set for
4569 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4570 not defined. This bug was masked in 4.22 by the effect that was fixed in
4573 2. Some messages that were rejected after a message id was allocated were
4574 shown as "incomplete" by exigrep. It no longer does this for messages that
4575 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4577 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4578 have allow_domain_literals set, the ID did not get logged in the <= line.
4579 Domain literals are now always recognized in Message-ID: header lines.
4581 4. The first argument for a ${extract expansion item is the key name or field
4582 number. Leading and trailing spaces in this item were not being ignored,
4583 causing some misleading effects.
4585 5. When deliver_drop_privilege was set, single queue runner processes started
4586 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4587 same command in the process it spins off) were not dropping privilege.
4589 6. When the daemon running as "exim" started a queue runner, it always
4590 re-executed Exim in the spun-off process. This is a waste of effort when
4591 deliver_drop_privilege is set. The new process now just calls the
4592 queue-runner function directly.
4598 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4601 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4602 ignored. Though the use of -oP was forcing the writing of a pid file, it
4603 was always written to the default place.
4605 3. If the message "no IP address found for host xxxx" is generated during
4606 incoming verification, it is now followed by identification of the incoming
4607 connection (so you can more easily find what provoked it).
4609 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4611 5. Added some features to "harden" Exim a bit more against certain attacks:
4613 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4614 be put in Local/Makefile. This is like the never_users runtime option,
4615 but it cannot be overridden. The default setting is "root".
4617 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4618 prefix string with which any file named in a -C command line option
4621 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4622 is retained for -C and -D only if the caller of Exim is root. Without
4623 it, the exim user may also use -C and -D and retain privilege.
4625 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4626 command line option is disabled.
4628 6. Macro names set by the -D option must start with an upper case letter, just
4629 like macro names defined in the configuration file.
4631 7. Added "dereference=" facility to LDAP.
4633 8. Two instances of the typo "uknown" in the source files are fixed.
4635 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4636 the Configure-Makefile script screwed up while processing it.
4638 10. Incorporated PCRE 4.4.
4640 11. The SMTP synchronization check was not operating right at the start of an
4641 SMTP session. For example, it could not catch a HELO sent before the client
4642 waited for the greeting. There is now a check for outstanding input at the
4643 point when the greeting is written. Because of the duplex, asynchronous
4644 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4645 way, but not yet received, when the check is performed.
4647 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4648 on TCP/IP sockets, because this apparently causes some broken clients to
4651 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4652 unchanged) from the Cygwin maintainer.
4654 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4655 was supported instead of testing for mbx. Effectively a typo.
4657 15. The spa authenticator server code was not checking that the input it
4658 received was valid base64.
4660 16. The debug output line for the "set" modifier in ACLs was not showing the
4661 name of the variable that was being set.
4663 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4665 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4666 lookup on its own. Something else was needed to trigger the lookup. For
4667 example, a match in host_lookup or the need for a host name in a host list.
4668 Now, if $sender_host_name is referenced and the host name has not yet been
4669 looked up, a lookup is performed. If the lookup fails, the variable remains
4670 empty, and $host_lookup_failed is set to "1".
4672 19. Added "eqi" as a case-independent comparison operator.
4674 20. The saslauthd authentication condition could segfault if neither service
4675 nor realm was specified.
4677 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4678 error message that was logged was misleading, and incoming SMTP
4679 connections were dropped. The message is now more accurate, and temporary
4680 errors are given to SMTP connections.
4682 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4683 (except RSET) with a 421 error, until QUIT is received. However, it was
4684 failing to send a response to QUIT.
4686 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4687 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4688 ACL is now run after the helo_try_verify_hosts code.
4690 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4691 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4692 case-independent, so other case variants are also recognized). Apparently
4693 some systems use these upper case variants.
4695 25. If more than two messages were waiting for the same host, and a transport
4696 filter was specified for the transport, Exim sent two messages over the
4697 same TCP/IP connection, and then failed with "socket operation on non-
4698 socket" when it tried to send the third.
4700 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4702 27. The extern definition of crypt16() in expand.c was not being excluded when
4703 the OS had its own crypt16() function.
4705 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4706 as a preferred synonym for return_size_limit, both as an option and as an
4709 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4711 30. Changed the default configuration ACL to relax the local part checking rule
4712 for addresses that are not in any local domains. For these addresses,
4713 slashes and pipe symbols are allowed within local parts, but the sequence
4714 /../ is explicitly forbidden.
4716 31. SPA server authentication was not clearing the challenge buffer before
4719 32. log_message in a "warn" ACL statement was writing to the reject log as
4720 well as to the main log, which contradicts the documentation and doesn't
4721 seem right (because no rejection is happening). So I have stopped it.
4723 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4724 However, I am unable to do any testing of this.
4726 34. Fixed an infelicity in the appendfile transport. When checking directories
4727 for a mailbox, to see if any needed to be created, it was accidentally
4728 using path names with one or more superfluous leading slashes; tracing
4729 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4731 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4732 contents are added to the log line that is written for every discarded
4733 recipient. (Previously a log_message setting was ignored.)
4735 36. The ${quote: operator now quotes the string if it is empty.
4737 37. The install script runs exim in order to find its version number. If for
4738 some reason other than non-existence or emptiness, which it checks, it
4739 could not run './exim', it was installing it with an empty version number,
4740 i.e. as "exim-". This error state is now caught, and the installation is
4743 38. An argument was missing from the function that creates an error message
4744 when Exim fails to connect to the socket for saslauthd authentication.
4745 This could cause Exim to crash, or give a corrupted message.
4747 39. Added isip, isip4, and isip6 to ${if conditions.
4749 40. The ACL variables $acl_xx are now saved with the message, and can be
4750 accessed later in routers, transports, and filters.
4752 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4753 strings in the file are not string-expanded.
4755 42. If a MAIL command specified a SIZE value that was too large to fit into an
4756 int variable, the check against message_size_limit failed. Such values are
4757 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4758 day this will have to be increased, but I don't think I want to be around
4759 when emails are that large.
4766 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4767 iconv() is not standard in FreeBSD.
4769 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4770 IPv6 enabled. The observed symptom was a segmentation fault on return from
4771 the function os_common_find_running_interfaces() in src/os.c.
4773 3. In the check_special_case() function in daemon.c I had used "errno" as an
4774 argument name, which causes warnings on some systems. This was basically a
4775 typo, since it was named "eno" in the comments!
4777 4. The code that waits for the clock to tick (at a resolution of some fraction
4778 of a second) so as to ensure message-id uniqueness was always waiting for
4779 at least one whole tick, when it could have waited for less. [This is
4780 almost certainly not relevant at current processor speeds, where it is
4781 unlikely to ever wait at all. But we try to future-proof.]
4783 5. The function that sleeps for a time interval that includes fractions of a
4784 second contained a race. It did not block SIGALRM between setting the
4785 timer, and suspending (a couple of lines later). If the interval was short
4786 and the sigsuspend() was delayed until after it had expired, the suspension
4787 never ended. On busy systems this could lead to processes getting stuck for
4790 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4791 process, before it forks any delivery processes. The open lookup caching
4792 mechanism meant that the open file or database connection was passed into
4793 the delivery process. The problem was that delivery processes always tidy
4794 up cached lookup data. This could cause a problem for the next delivery
4795 process started by the queue runner, because the external queue runner
4796 process does not know about the closure. So the next delivery process
4797 still has data in the lookup cache. In the case of a file lookup, there was
4798 no problem because closing a file descriptor in a subprocess doesn't affect
4799 the parent. However, if the lookup was caching a connection to a database,
4800 the connection was closed, and the second delivery process was likely to
4801 see errors such as "PGSQL: query failed: server closed the connection
4802 unexpectedly". The problem has been fixed by closing all cached lookups
4803 in a queue runner before running a delivery process.
4805 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4806 seem to have the "const" qualifier which it has on other OS. I've
4809 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4810 *of the same type* (client or server) with the same public name that an
4811 error should be diagnosed.
4813 9. When Exim looked up a host name for an IP address, but failed to find the
4814 original IP address when looking up the host name (a safety check), it
4815 output the message "<ip address> does not match any IP for NULL", which was
4816 confusing, to say the least. The bug was that the host name should have
4817 appeared instead of "NULL".
4819 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4820 user that is built into the binary, and the -C or -D options is used, root
4821 privilege is dropped before the configuration file is read. In addition,
4822 logging is switched to stderr instead of the normal log files. If the
4823 configuration then re-defines the Exim user, the unprivileged environment
4824 is probably not what is expected, so Exim logs a panic warning message (but
4827 However, if deliver_drop_privilege is set, the unprivileged state may well
4828 be exactly what is intended, so the warning has been cut out in that case,
4829 and Exim is allowed to try to write to its normal log files.
4835 1. smtp_return_error_details was not giving details for temporary sender
4836 or receiver verification errors.
4838 2. Diagnose a configuration error if two authenticators have the same public
4841 3. Exim used not to create the message log file for a message until the first
4842 delivery attempt. This could be confusing when incoming messages were held
4843 for policy or load reasons. The message log file is now created at the time
4844 the message is received, and an initial "Received" line is written to it.
4846 4. The automatically generated man page for command line options had a minor
4847 bug that caused no ill effects; however, a more serious problem was that
4848 the procedure for building the man page automatically didn't always
4849 operate. Consequently, release 4.20 contains an out-of-date version. This
4850 shouldn't happen again.
4852 5. When building Exim with embedded Perl support, the script that builds the
4853 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4854 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4856 6. The freeze_tell option was not being used for messages that were frozen on
4857 arrival, either by an ACL or by local_scan().
4859 7. Added the smtp_incomplete_transaction log selector.
4861 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4862 was accepting AUTH without a new EHLO.
4864 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4865 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4868 10. Logging of TCP/IP connections (when configured) now happens in the main
4869 daemon process instead of the child process, so that the TCP/IP connection
4870 count is more accurate (but it can never be perfect).
4872 11. The use of "drop" in a nested ACL was not being handled correctly in the
4873 outer ACL. Now, if condition failure induced by the nested "drop" causes
4874 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4875 or "require"), the connection is dropped.
4877 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4878 that yield "discard" can now be used with an "accept" or a "discard" verb,
4879 but an error is generated for any others (because I can't see a useful way
4880 to define what should happen).
4882 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4883 are now processed in the same way as lines in the Exim configuration file.
4884 In particular, continuation lines are supported.
4886 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4888 15. Added -ti meaning -t -i.
4890 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4891 domains, and warn by logging if others are found.
4893 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4894 returned from GIFCONF when Exim is trying to find the list of interfaces on
4895 a host. The code in os.c has been modified to copy each ifreq to an aligned
4896 structure in all cases.
4898 Also, in some cases, the returned ifreq's were being copied to a 'struct
4899 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4900 means the last couple of bytes of an IPv6 address could be chopped if the
4901 ifreq contained only a normal sockaddr (14 bytes storage).
4903 18. Named domain lists were not supported in the hosts_treat_as_local option.
4904 An entry such as +xxxx was not recognized, and was treated as a literal
4907 19. Ensure that header lines added by a DATA ACL are included in the reject log
4908 if the ACL subsequently rejects the message.
4910 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4911 MD5 (which is deprecated).
4913 21. When testing a filter file using -bf, Exim was writing a message when it
4914 took the sender from a "From " line in the message, but it was not doing so
4915 when it took $return_path from a Return-Path: header line. It now does.
4917 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4918 with a valid header line field name (a series of printing characters
4919 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4921 23. Changed "disc" in the source to "disk" to conform to the documentation and
4922 the book and for uniformity.
4924 24. Ignore Sendmail's -Ooption=value command line item.
4926 25. When execve() failed while trying to run a command in a pipe transport,
4927 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4928 could be confused with a return value of 69 from the command itself. This
4929 has been changed to 127, the value the shell returns if it is asked to run
4930 a non-existent command. The wording for the related log line suggests a
4931 non-existent command as the problem.
4933 26. If received_header_text expands to an empty string, do not add a Received:
4934 header line to the message. (Well, it adds a token one on the spool, but
4935 marks it "old" so that it doesn't get used or transmitted.)
4937 27. Installed eximstats 1.28 (addition of -nt option).
4939 28. There was no check for failure on the call to getsockname() in the daemon
4940 code. This can fail if there is a shortage of resources on the system, with
4941 ENOMEM, for example. A temporary error is now given on failure.
4943 29. Contrary to the C standard, it seems that in some environments, the
4944 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4945 program. Exim now does this explicitly; it affects the formatting of
4946 timestamps using strftime().
4948 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4949 complaints. I've now initialized all the variables, to avoid this.
4951 32. Header lines added by a system filter were not being "seen" during
4952 transport-time rewrites.
4954 33. The info_callback() function passed to OpenSSL is set up with type void
4955 (*)(SSL *, int, int), as described somewhere. However, when calling the
4956 function (actually a macro) that sets it up, the type void(*)() is
4957 expected. I've put in a cast to prevent warnings from picky compilers.
4959 34. If a DNS black list lookup found a CNAME record, but there were no A
4960 records associated with the domain it pointed at, Exim crashed.
4962 35. If a DNS black list lookup returned more than one A record, Exim ignored
4963 all but the first. It now scans all returned addresses if a particular IP
4964 value is being sought. In this situation, the contents of the
4965 $dnslist_value variable are a list of all the addresses, separated by a
4968 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4969 to accept a host name and all its aliases if the forward lookup for any of
4970 them yielded the IP address of the incoming connection. Now it accepts only
4971 those names whose forward lookup yields the correct IP address. Any other
4972 names are discarded. This closes a loophole whereby a rogue DNS
4973 administrator could create reverse DNS records to break through a
4974 wildcarded host restriction in an ACL.
4976 37. If a user filter or a system filter that ran in a subprocess used any of
4977 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4978 the wrong values were passed to the pipe command ($thisaddress had the
4979 value of $0, $0 had the value of $1, etc). This bug was introduced by
4980 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4982 38. Improved the line breaking for long SMTP error messages from ACLs.
4983 Previously, if there was no break point between 40 and 75 characters, Exim
4984 left the rest of the message alone. Two changes have been made: (a) I've
4985 reduced the minimum length to 35 characters; (b) if it can't find a break
4986 point between 35 and 75 characters, it looks ahead and uses the first one
4987 that it finds. This may give the occasional overlong line, but at least the
4988 remaining text gets split now.
4990 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4991 file descriptors might be low, and that setting 1000 would always raise it.
4992 It turns out that in some environments, the limit is already over 1000 and
4993 that lowering it causes trouble. So now Exim takes care not to decrease it.
4995 40. When delivering a message, the value of $return_path is set to $sender_
4996 address at the start of routing (routers may change the value). By an
4997 oversight, this default was not being set up when an address was tested by
4998 -bt or -bv, which affected the outcome if any router or filter referred to
5001 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5002 log only when "message" or "log_message" are set. However, if one of the
5003 conditions was an address verification, or a call to a nested ACL, the
5004 messages generated by the underlying test were being passed through. This
5005 no longer happens. The underlying message is available in $acl_verify_
5006 message for both "message" and "log_message" expansions, so it can be
5007 passed through if needed.
5009 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5010 new expansion $bh_ to give the encoded byte string without charset
5011 translation. Translation happens only if iconv() is available; HAVE_ICONV
5012 indicates this at build time. HEADERS_CHARSET gives the charset to
5013 translate to; headers_charset can change it in the configuration, and
5014 "headers charset" can change it in an individual filter file.
5016 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5017 that creates RFC 2047 encoded "words" labels them as that charset instead
5018 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5019 expansion operator; (ii) when Exim creates a From: line for a local
5020 message; (iii) when a header line is rewritten to include a "phrase" part.
5022 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5023 buggy, causing it to skip the first lines of messages whose message ID
5024 ended in 'D'. This would not have bitten before Exim release 4.14, because
5025 message IDs were unlikely to end in 'D' before then. The effect was to have
5026 incorrect size information for certain domains.
5028 45. #include "config.h" was missing at the start of the crypt16.c module. This
5029 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5032 46. If there was a timeout during a "random" callout check, Exim treated it as
5033 a failure of the random address, and carried on sending RSET and the real
5034 address. If the delay was just some slowness somewhere, the response to the
5035 original RCPT would be taken as a response to RSET and so on, causing
5036 mayhem of various kinds.
5038 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5039 when I implemented it. It didn't allow for the fact that some option values
5040 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5041 the right test for positive values.
5043 48. Domain names in DNS records are case-independent. Exim always looks them up
5044 in lower case. Some resolvers return domain names in exactly the case they
5045 appear in the zone file, that is, they may contain uppercase letters. Not
5046 all resolvers do this - some return always lower case. Exim was treating a
5047 change of case by a resolver as a change of domain, similar to a widening
5048 of a domain abbreviation. This triggered its re-routing code and so it was
5049 trying to route what was effectively the same domain again. This normally
5050 caused routing to fail (because the router wouldn't handle the domain
5051 twice). Now Exim checks for this case specially, and just changes the
5052 casing of the domain that it ultimately uses when it transmits the message
5055 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5058 50. If a filter generated a file delivery with a non-absolute name (possible if
5059 no home directory exists for the router), the forbid_file option was not
5062 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5063 the existing equality matching.
5065 52. Exim was using ints instead of ino_t variables in some places where it was
5066 dealing with inode numbers.
5068 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5069 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5070 called TMPDIR, and if it finds it is different, it changes its value.
5072 54. The smtp_printf() function is now made available to local_scan() so
5073 additional output lines can be written before returning. There is also an
5074 smtp_fflush() function to enable the detection of a dropped connection.
5075 The variables smtp_input and smtp_batched_input are exported to
5078 55. Changed the default runtime configuration: the message "Unknown user"
5079 has been removed from the ACL, and instead placed on the localuser router,
5080 using the cannot_route_message feature. This means that any verification
5081 failures that generate their own messages won't get overridden. Similarly,
5082 the "Unrouteable address" message that was in the ACL for unverifiable
5083 relay addresses has also been removed.
5085 56. Added hosts_avoid_esmtp to the smtp transport.
5087 57. The exicyclog script was not checking for the esoteric option
5088 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5089 will work only if exicyclog is run under the appropriate euid.
5091 58. Following a discussion on the list, the rules by which Exim recognises line
5092 endings on incoming messages have been changed. The -dropcr and drop_cr
5093 options are now no-ops, retained only for backwards compatibility. The
5094 following line terminators are recognized: LF CRLF CR. However, special
5095 processing applies to CR:
5097 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5098 nor a local message in the state where . is a terminator.
5100 (ii) If a bare CR is encountered in a header line, an extra space is added
5101 after the line terminator so as not to end the header. The reasoning
5102 behind this is that bare CRs in header lines are most likely either
5103 to be mistakes, or people trying to play silly games.
5105 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5106 was being incorrectly given as 18 bytes larger than it should have been.
5107 This is a VOB (very old bug).
5109 60. This may never have affected anything current, but just in case it has:
5110 When the local host is found other than at the start of a list of hosts,
5111 the local host, those with the same MX, and any that follow, are discarded.
5112 When the list in question was part of a longer list of hosts, the following
5113 hosts (not currently being processed) were also being discarded. This no
5114 longer happens. I'm not sure if this situation could ever has previously
5117 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5120 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5121 header. This is something that is recommended in a new Internet Draft, and
5122 is something that is documented as being done by Sendmail. There are two
5123 possible values. For messages generated by the autoreply transport, Exim
5126 Auto-Submitted: auto-replied
5128 whereas for all other generated messages (e.g. bounces) it adds
5130 Auto-Submitted: auto-generated
5132 63. The "personal" condition in filters now includes a test for the
5133 Auto-Submitted: header. If it contains the string "auto-" the message it
5134 not considered personal.
5136 64. Added rcpt_include_affixes as a generic transport option.
5138 65. Added queue_only_override (default true).
5140 66. Added the syslog_duplication option.
5142 67. If what should have been the first header line of a message consisted of
5143 a space followed by a colon, Exim was mis-interpreting it as a header line.
5144 It isn't of course - it is syntactically invalid and should therefore be
5145 treated as the start of the message body. The misbehaviour could have
5146 caused a number of strange effects, including loss of data in subsequent
5147 header lines, and spool format errors.
5149 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5150 client host had authenticated. This control can now be exercised by an ACL
5151 for more flexibility.
5153 69. By default, callouts do not happen when testing with -bh. There is now a
5154 variant, -bhc, which does actually run the callout code, including
5155 consulting and updating the callout cache.
5157 70. Added support for saslauthd authentication, courtesy of Alexander
5160 71. If statvfs() failed on the spool or log directories while checking their
5161 size for availability, Exim confusingly gave the error "space shortage".
5162 Furthermore, in debugging mode it crashed with a floating point exception.
5163 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5164 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5165 serious problem, Exim now writes to the main and panic logs when this
5166 happens, with details of the failure. It then refuses to accept the
5167 incoming message, giving the message "spool directory problem" or "log
5168 directory problem" with a 421 code for SMTP messages.
5170 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5171 0, 1, and 2 exist, because some OS complain for execs without them (see
5172 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5173 descriptors. However, the code omitted to check that the open succeeded,
5174 causing mysterious errors if for some reason the permissions on /dev/null
5175 got screwed. Now Exim writes a message to the main and panic logs, and
5176 bombs out if it can't open /dev/null.
5178 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5179 interact so that it is all more flexible. It is supposed to remain
5180 backwards compatible. Also added extra_local_interfaces.
5182 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5183 to bomb out with an assertion failure - to the client this appears as a
5184 connection drop. This problem occurs in the part of the code that was taken
5185 from the Samba project. Fortunately, the assertion is in a very simple
5186 function, so I have fixed this by reproducing the function inline in the
5187 one place where it is called, and arranging for authentication to fail
5188 instead of killing the process with assert().
5190 75. The SPA client code was not working when the server requested OEM rather
5191 than Unicode encoding.
5193 76. Added code to make require_files with a specific uid setting more usable in
5194 the case where statting the file as root fails - usually a non-root-mounted
5195 NFS file system. When this happens and the failure is EACCES, Exim now
5196 forks a subprocess and does the per-uid checking as the relevant uid.
5198 77. Added process_log_path.
5200 78. If log_file_path was not explicitly set, a setting of check_log_space or
5201 check_log_inodes was ignored.
5203 79. If a space check for the spool or log partitions fails, the incident is now
5204 logged. Of course, in the latter case the data may get lost...
5206 80. Added the %p formatting code to string_format() so that it can be used to
5207 print addresses in debug_print(). Adjusted all the address printing in the
5208 debugging in store.c to use %p rather than %d.
5210 81. There was a concern that a line of code in smtp_in.c could overflow a
5211 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5212 initially expressed, the concern was not well-founded, because trailing
5213 spaces are removed early. However, if the trailing spaces were followed by
5214 a NULL, they did not get removed, so the overflow was possible. Two fixes
5217 (a) I re-wrote the offending code in a cleaner fashion.
5218 (b) If an incoming SMTP command contains a NULL character, it is rejected
5221 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5222 runs initgroups(), so that if the Exim user is in any additional groups,
5223 they will be used during message reception.
5229 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob