1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Solaris build fix for Oracle's LDAP libraries.
8 Bugzilla 1109, patch from Stephen Usher.
10 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
12 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
15 TF/02 Fix a couple more cases where we did not log the error message
16 when unlink() failed. See also change 4.74-TF/03.
18 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
19 lock up or crash if it happened to be inside a call to libc when it
20 got a SIGUSR1 from exiwhat.
22 The SIGUSR1 handler appends the current process status to the process
23 log which is later printed by exiwhat. It used to use the general
24 purpose logging code to do this, but several functions it calls are
27 The new output code in the SIGUSR1 handler is specific to the process
28 log, and simple enough that it's easy to inspect for signal safety.
29 Removing some special cases also simplifies the general logging code.
30 Removing the spurious timestamps from the process log simplifies
33 TF/04 Improved ratelimit ACL condition.
35 The /noupdate option has been deprecated in favour of /readonly which
36 has clearer semantics. The /leaky, /strict, and /readonly update modes
37 are mutually exclusive. The update mode is no longer included in the
38 database key; it just determines when the database is updated. (This
39 means that when you upgrde Exim will forget old rate measurements.)
41 Exim now checks that the per_* options are used with an update mode that
42 makes sense for the current ACL. For example, when Exim is processing a
43 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
44 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
45 must specify per_mail/readonly. If you omit the update mode it defaults to
46 /leaky where that makes sense (as before) or /readonly where required.
48 The /noupdate option is now undocumented but still supported for
49 backwards compatibility. It is equivalent to /readonly except that in
50 ACLs where /readonly is required you may specify /leaky/noupdate or
51 /strict/noupdate which are treated the same as /readonly.
53 A useful new feature is the /count= option. This is a generalization
54 of the per_byte option, so that you can measure the throughput of other
55 aggregate values. For example, the per_byte option is now equivalent
56 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
58 The per_rcpt option has been generalized using the /count= mechanism
59 (though it's more complicated than the per_byte equivalence). When it is
60 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
61 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
62 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
63 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
64 Note that using per_rcpt with a non-readonly update mode in more than
65 one ACL will cause the recipients to be double-counted. (The per_mail
66 and per_byte options don't have this problem.)
68 The handling of very low rates has changed slightly. If the computed rate
69 is less than the event's count (usually one) then this event is the first
70 after a long gap. In this case the rate is set to the same as this event's
71 count, so that the first message of a spam run is counted properly.
73 The major new feature is a mechanism for counting the rate of unique
74 events. The new per_addr option counts the number of different
75 recipients that someone has sent messages to in the last time period. It
76 behaves like per_rcpt if all the recipient addresses are different, but
77 duplicate recipient addresses do not increase the measured rate. Like
78 the /count= option this is a general mechanism, so the per_addr option
79 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
80 example, measure the rate that a client uses different sender addresses
81 with the options per_mail/unique=$sender_address. There are further
82 details in the main documentation.
84 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
86 TF/06 Removed a few PCRE remnants.
88 TF/07 Automatically extract Exim's version number from tags in the git
89 repository when doing development or release builds.
91 PP/02 Raise smtp_cmd_buffer_size to 16kB.
92 Bugzilla 879. Patch from Paul Fisher.
94 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
95 Heavily based on revision 40f9a89a from Simon Arlott's tree.
98 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
100 PP/05 Variable $av_failed, true if the AV scanner deferred.
101 Bugzilla 1078. Patch from John Horne.
103 PP/06 Stop make process more reliably on build failure.
104 Bugzilla 1087. Patch from Heiko Schlittermann.
106 PP/07 Make maildir_use_size_file an _expandable_ boolean.
107 Bugzilla 1089. Patch from Heiko Schlittermann.
109 PP/08 Handle ${run} returning more data than OS pipe buffer size.
110 Bugzilla 1131. Patch from Holger Weiß.
112 PP/09 Handle IPv6 addresses with SPF.
113 Bugzilla 860. Patch from Wolfgang Breyha.
115 PP/10 GnuTLS: support TLS 1.2 & 1.1.
122 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
124 PP/02 Harmonised TLS library version reporting; only show if debugging.
125 Layout now matches that introduced for other libraries in 4.74 PP/03.
127 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
129 PP/04 New "dns_use_edns0" global option.
131 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
134 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
135 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
137 TK/01 Updated PolarSSL code to 0.14.2.
138 Bugzilla 1097. Patch from Andreas Metzler.
140 PP/07 Catch divide-by-zero in ${eval:...}.
143 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
146 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
147 format-string attack -- SECURITY: remote arbitrary code execution.
149 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
150 time unintentionally subject to list matching rules, letting the header
151 cause arbitrary Exim lookups (of items which can occur in lists, *not*
152 arbitrary string expansion). This allowed for information disclosure.
154 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
155 INT_MIN/-1 -- value coerced to INT_MAX.
161 NM/01 Workround for PCRE version dependency in version reporting
164 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
165 This fixes portability to compilers other than gcc, notably
166 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
168 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
169 makefiles for portability to HP-UX and POSIX correctness.
171 PP/01 Permit LOOKUP_foo enabling on the make command-line.
172 Also via indented variable definition in the Makefile.
173 (Debugging by Oliver Heesakkers).
175 PP/02 Restore caching of spamd results with expanded spamd_address.
176 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
178 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
179 Improves build reliability. Fix from: Frank Elsner
181 NM/02 Fix wide character breakage in the rfc2047 coding
182 Fixes bug 1064. Patch from Andrey N. Oktyabrski
184 NM/03 Allow underscore in dnslist lookups
185 Fixes bug 1026. Patch from Graeme Fowler
187 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
188 Code patches from Adam Ciarcinski of NetBSD.
190 NM/04 Fixed exiqgrep to cope with mailq missing size issue
193 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
194 is logged, to avoid truncation. Patch from John Horne.
196 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
197 Patch from Jakob Hirsch.
199 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
200 SQL string expansion failure details.
201 Patch from Andrey Oktyabrski.
203 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
204 Patch from Simon Arlott.
206 PP/09 New lookups functionality failed to compile on old gcc which rejects
207 extern declarations in function scope.
208 Patch from Oliver Fleischmann
210 PP/10 Use sig_atomic_t for flags set from signal handlers.
211 Check getgroups() return and improve debugging.
212 Fixed developed for diagnosis in bug 927 (which turned out to be
215 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
216 Patch from Mark Zealey.
218 PP/12 Bugzilla 1056: Improved spamd server selection.
219 Patch from Mark Zealey.
221 PP/13 Bugzilla 1086: Deal with maildir quota file races.
222 Based on patch from Heiko Schlittermann.
224 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
225 Patch from Uwe Doering, sign-off by Michael Haardt.
227 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
228 variable declaration deep within a block. Bug and patch from
231 PP/15 lookups-Makefile IRIX compatibilty coercion.
233 PP/16 Make DISABLE_DKIM build knob functional.
235 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
236 Patch by Simon Arlott
238 TF/03 Fix valgrind.h portability to C89 compilers that do not support
239 variable argument macros. Our copy now differs from upstream.
245 TF/01 Failure to get a lock on a hints database can have serious
246 consequences so log it to the panic log.
248 TF/02 Log LMTP confirmation messages in the same way as SMTP,
249 controlled using the smtp_confirmation log selector.
251 TF/03 Include the error message when we fail to unlink a spool file.
253 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
254 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
255 for maintaining out-of-tree patches for some time.
257 PP/01 Bugzilla 139: Documentation and portability issues.
258 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
259 Handle per-OS dynamic-module compilation flags.
261 PP/02 Let /dev/null have normal permissions.
262 The 4.73 fixes were a little too stringent and complained about the
263 permissions on /dev/null. Exempt it from some checks.
264 Reported by Andreas M. Kirchwitz.
266 PP/03 Report version information for many libraries, including
267 Exim version information for dynamically loaded libraries. Created
268 version.h, now support a version extension string for distributors
269 who patch heavily. Dynamic module ABI change.
271 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
272 privilege escalation vulnerability whereby the Exim run-time user
273 can cause root to append content of the attacker's choosing to
276 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
279 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
280 If dropping privileges for untrusted macros, we disabled normal logging
281 on the basis that it would fail; for the Exim run-time user, this is not
282 the case, and it resulted in successful deliveries going unlogged.
283 Fixed. Reported by Andreas Metzler.
289 PP/01 Date: & Message-Id: revert to normally being appended to a message,
290 only prepend for the Resent-* case. Fixes regression introduced in
291 Exim 4.70 by NM/22 for Bugzilla 607.
293 PP/02 Include check_rfc2047_length in configure.default because we're seeing
294 increasing numbers of administrators be bitten by this.
296 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
298 PP/03 Bugzilla 994: added openssl_options main configuration option.
300 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
302 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
304 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
306 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
307 without a peer certificate, leading to a segfault because of an
308 assumption that peers always have certificates. Be a little more
309 paranoid. Problem reported by Martin Tscholak.
311 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
312 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
313 NB: ClamAV planning to remove STREAM in "middle of 2010".
314 CL also introduces -bmalware, various -d+acl logging additions and
315 more caution in buffer sizes.
317 PP/09 Implemented reverse_ip expansion operator.
319 PP/10 Bugzilla 937: provide a "debug" ACL control.
321 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
323 PP/12 Bugzilla 973: Implement --version.
325 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
327 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
329 PP/15 Bugzilla 816: support multiple condition rules on Routers.
331 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
332 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
333 ignore trailing whitespace.
335 JJ/02 prevent non-panic DKIM error from being sent to paniclog
337 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
340 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
341 Notification from Dr Andrew Aitchison.
343 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
344 ExtendedDetectionInfo response format.
345 Notification from John Horne.
347 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
350 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
351 XSL and documented dependency on system catalogs, with examples of how
354 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
357 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
358 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
359 configuration file which is writeable by the Exim user or group.
361 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
362 of configuration files to cover files specified with the -C option if
363 they are going to be used with root privileges, not just the default
366 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
367 option (effectively making it always true).
369 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
370 files to be used while preserving root privileges.
372 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
373 that rogue child processes cannot use them.
375 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
376 run-time user, instead of root.
378 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
379 Exim run-time user without dropping privileges.
381 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
382 result string, instead of calling string_vformat() twice with the same
385 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
386 for other users. Others should always drop root privileges if they use
387 -C on the command line, even for a whitelisted configure file.
389 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
391 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
397 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
398 $data_path, and $header_path variables; fixed documentation bugs and
401 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
402 exipick to access non-standard spools, including the "frozen" queue
405 NM/01 Bugzilla 965: Support mysql stored procedures.
406 Patch from Alain Williams
408 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
410 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
411 Patch from Andreas Metzler
413 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
414 Patch from Kirill Miazine
416 NM/05 Bugzilla 671: Added umask to procmail example.
418 JJ/03 installed exipick 20100323.0, fixing doc bug
420 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
421 directory. Notification and patch from Dan Rosenberg.
423 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
425 TK/02 Improve log output when DKIM signing operation fails.
427 MH/01 Treat the transport option dkim_domain as a colon separated
428 list, not as a single string, and sign the message with each element,
429 omitting multiple occurences of the same signer.
431 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
432 Bugzilla 985, 986. Patch by Simon Arlott
434 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
435 Patch by Simon Arlott
437 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
438 MBX locking. Notification from Dan Rosenberg.
444 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
446 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
448 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
450 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
452 NM/04 Bugzilla 915: Flag broken perl installation during build.
458 TK/01 Added patch by Johannes Berg that expands the main option
459 "spamd_address" if it starts with a dollar sign.
461 TK/02 Write list of recipients to X-Envelope-Sender header when building
462 the mbox-format spool file for content scanning (suggested by Jakob
465 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
466 (http://www.dcc-servers.net/) support via dccifd. Activated by
467 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
469 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
470 by Mark Daniel Reidel <mr@df.eu>.
472 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
473 When building exim an external PCRE library is now needed -
474 PCRE is a system library on the majority of modern systems.
475 See entry on PCRE_LIBS in EDITME file.
477 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
478 conversation. Added nologin parameter to request.
479 Patch contributed by Kirill Miazine.
481 TF/01 Do not log submission mode rewrites if they do not change the address.
483 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
485 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
486 log files in place. Contributed by Roberto Lima.
488 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
490 TF/03 Bugzilla 615: When checking the local_parts router precondition
491 after a local_part_suffix or local_part_prefix option, Exim now
492 does not use the address's named list lookup cache, since this
493 contains cached lookups for the whole local part.
495 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
496 Robert Millan. Documentation is in experimental-spec.txt.
498 TF/04 Bugzilla 668: Fix parallel build (make -j).
500 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
502 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
503 Patch provided by Jan Srzednicki.
505 TF/05 Leading white space used to be stripped from $spam_report which
506 wrecked the formatting. Now it is preserved.
508 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
509 that they are available at delivery time.
511 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
513 TF/08 TLS error reporting now respects the incoming_interface and
514 incoming_port log selectors.
516 TF/09 Produce a more useful error message if an SMTP transport's hosts
517 setting expands to an empty string.
519 NM/06 Bugzilla 744: EXPN did not work under TLS.
520 Patch provided by Phil Pennock.
522 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
523 Patch provided by Richard Godbee.
525 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
526 acl_smtp_notquit, added index entry.
528 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
529 Patch provided by Eugene Bujak.
531 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
532 accept(). Patch provided by Maxim Dounin.
534 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
535 Patch provided by Phil Pennock.
537 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
539 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
540 Patch provided by Brad "anomie" Jorsch.
542 NM/14 Bugzilla 622: Added timeout setting to transport filter.
543 Patch provided by Dean Brooks.
545 TK/05 Add native DKIM support (does not depend on external libraries).
547 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
548 Patch provided by Graeme Fowler.
550 NM/16 Bugzilla 851: Documentation example syntax fix.
552 NM/17 Changed NOTICE file to remove references to embedded PCRE.
554 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
557 NM/19 Bugzilla 745: TLS version reporting.
558 Patch provided by Phil Pennock.
560 NM/20 Bugzilla 167: bool: condition support.
561 Patch provided by Phil Pennock.
563 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
564 clients. Patch provided by Phil Pennock.
566 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
567 Patch provided by Brad "anomie" Jorsch.
569 NM/23 Bugzilla 687: Fix misparses in eximstats.
570 Patch provided by Heiko Schlittermann.
572 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
573 Patch provided by Heiko Schlittermann.
575 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
576 plus update to original patch.
578 NM/26 Bugzilla 799: Documentation correction for ratelimit.
580 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
581 Patch provided by David Brownlee.
583 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
585 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
587 NM/30 Bugzilla 888: TLS documentation bugfixes.
589 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
591 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
592 Unlike the original bugzilla I have changed all shell scripts in src tree.
594 NM/33 Bugzilla 898: Transport filter timeout fix.
595 Patch by Todd Rinaldo.
597 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
598 Patch by Serge Demonchaux.
600 NM/35 Bugzilla 39: Base64 decode bug fixes.
601 Patch by Jakob Hirsch.
603 NM/36 Bugzilla 909: Correct connect() call in dcc code.
605 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
607 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
609 NM/39 Bugzilla 911: Fixed MakeLinks build script.
615 TK/01 Add preliminary DKIM support. Currently requires a forked version of
616 ALT-N's libdkim that I have put here:
617 http://duncanthrax.net/exim-experimental/
619 Note to Michael Haardt: I had to rename some vars in sieve.c. They
620 were called 'true' and it seems that C99 defines that as a reserved
621 keyword to be used with 'bool' variable types. That means you could
622 not include C99-style headers which use bools without triggering
623 build errors in sieve.c.
625 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
626 as mailq or other aliases. Changed the --help handling significantly
627 to do whats expected. exim_usage() emits usage/help information.
629 SC/01 Added the -bylocaldomain option to eximstats.
631 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
633 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
635 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
641 PH/01 Another patch from the Sieve maintainer.
643 PH/02 When an IPv6 address is converted to a string for single-key lookup
644 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
645 dots are used instead of colons so that keys in lsearch files need not
646 contain colons. This was done some time before quoting was made available
647 in lsearch files. However, iplsearch files do require colons in IPv6 keys
648 (notated using the quote facility) so as to distinguish them from IPv4
649 keys. This meant that lookups for IP addresses in host lists did not work
650 for iplsearch lookups.
652 This has been fixed by arranging for IPv6 addresses to be expressed with
653 colons if the lookup type is iplsearch. This is not incompatible, because
654 previously such lookups could never work.
656 The situation is now rather anomolous, since one *can* have colons in
657 ordinary lsearch keys. However, making the change in all cases is
658 incompatible and would probably break a number of configurations.
660 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
663 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
664 conversion specification without a maximum field width, thereby enabling
665 a rogue spamd server to cause a buffer overflow. While nobody in their
666 right mind would setup Exim to query an untrusted spamd server, an
667 attacker that gains access to a server running spamd could potentially
668 exploit this vulnerability to run arbitrary code as the Exim user.
670 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
671 $primary_hostname instead of what libspf2 thinks the hosts name is.
673 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
674 a directory entry by the name of the lookup key. Previously, if a
675 symlink pointed to a non-existing file or a file in a directory that
676 Exim lacked permissions to read, a lookup for a key matching that
677 symlink would fail. Now it is enough that a matching directory entry
678 exists, symlink or not. (Bugzilla 503.)
680 PH/03 The body_linecount and body_zerocount variables are now exported in the
683 PH/04 Added the $dnslist_matched variable.
685 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
686 This means they are set thereafter only if the connection becomes
689 PH/06 Added the client_condition to authenticators so that some can be skipped
690 by clients under certain conditions.
692 PH/07 The error message for a badly-placed control=no_multiline_responses left
693 "_responses" off the end of the name.
695 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
697 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
698 (without spaces) instead of just copying the configuration text.
700 PH/10 Added the /noupdate option to the ratelimit ACL condition.
702 PH/11 Added $max_received_linelength.
704 PH/12 Added +ignore_defer and +include_defer to host lists.
706 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
707 way in which PCRE > 7.0 is built.
709 PH/14 Implemented queue_only_load_latch.
711 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
712 MAIL command. The effect was to mangle the value on 64-bit systems.
714 PH/16 Another patch from the Sieve maintainer.
716 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
718 PH/18 If a system quota error occurred while trying to create the file for
719 a maildir delivery, the message "Mailbox is full" was not appended to the
720 bounce if the delivery eventually timed out. Change 4.67/27 below applied
721 only to a quota excession during the actual writing of the file.
723 PH/19 It seems that peer DN values may contain newlines (and other non-printing
724 characters?) which causes problems in log lines. The DN values are now
725 passed through string_printing() before being added to log lines.
727 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
728 and InterBase are left for another time.)
730 PH/21 Added message_body_newlines option.
732 PH/22 Guard against possible overflow in moan_check_errorcopy().
734 PH/23 POSIX allows open() to be a macro; guard against that.
736 PH/24 If the recipient of an error message contained an @ in the local part
737 (suitably quoted, of course), incorrect values were put in $domain and
738 $local_part during the evaluation of errors_copy.
744 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
745 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
748 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
749 issue a MAIL command.
751 PH/02 In an ACL statement such as
753 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
755 if a client was not listed at all, or was listed with a value other than
756 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
757 the condition was not true (as it should be), so access was not denied.
758 The bug was that the ! inversion was incorrectly passed on to the second
759 item. This has been fixed.
761 PH/03 Added additional dnslists conditions == and =& which are different from
762 = and & when the dns lookup returns more than one IP address.
764 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
765 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
767 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
768 FSYNC, which compiles an option called disable_fsync that allows for
769 bypassing fsync(). The documentation is heavily laced with warnings.
771 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
773 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
774 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
775 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
776 including adding "make clean"; (3) Added -fPIC when compiling the test
777 dynamically loaded module, to get rid of a warning.
779 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
780 message fails, move_frozen_messages = true and ignore_bounce_errors_after
781 = 0s. The bug is otherwise harmless.
783 PH/07 There was a bug in the dovecot authenticator such that the value of
784 $auth1 could be overwritten, and so not correctly preserved, after a
785 successful authentication. This usually meant that the value preserved by
786 the server_setid option was incorrect.
788 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
790 PH/09 Installed PCRE release 7.0.
792 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
793 run for batched SMTP input. It is now run at the start of every message
794 in the batch. While fixing this I discovered that the process information
795 (output by running exiwhat) was not always getting set for -bs and -bS
796 input. This is fixed, and it now also says "batched" for BSMTP.
798 PH/11 Added control=no_pipelining.
800 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
801 patch, slightly modified), and move the expansion of helo_data till after
802 the connection is made in the smtp transport (so it can use these
805 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
807 PH/14 Added log_selector = +pid.
809 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
811 PH/16 Add ${if forany and ${if forall.
813 PH/17 Added dsn_from option to vary the From: line in DSNs.
815 PH/18 Flush SMTP output before performing a callout, unless control =
816 no_callout_flush is set.
818 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
819 was true (the default) a successful delivery failed to delete the retry
820 item, thus causing premature timeout of the address. The bug is now
823 PH/20 Added hosts_avoid_pipelining to the smtp transport.
825 PH/21 Long custom messages for fakedefer and fakereject are now split up
826 into multiline reponses in the same way that messages for "deny" and
827 other ACL rejections are.
829 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
830 with slight modification.
832 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
833 draft, changing the syntax and factoring some duplicate code".
835 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
836 for deliveries of the second and subsequent messages over the same SMTP
839 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
840 ${reduce, with only minor "tidies".
842 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
844 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
845 expansion side effects.
847 PH/27 When a message times out after an over-quota error from an Exim-imposed
848 quota, the bounce message says "mailbox is full". This message was not
849 being given when it was a system quota that was exceeded. It now should
852 MH/03 Made $recipients available in local_scan(). local_scan() already has
853 better access to the recipient list through recipients_list[], but
854 $recipients can be useful in postmaster-provided expansion strings.
856 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
857 in the case of a MAIL command with additional options following the
858 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
859 were accidentally chopped off.
861 PH/29 SMTP synchronization checks are implemented when a command is read -
862 there is a check that no more input is waiting when there shouldn't be
863 any. However, for some commands, a delay in an ACL can mean that it is
864 some time before the response is written. In this time, more input might
865 arrive, invalidly. So now there are extra checks after an ACL has run for
866 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
867 pipelining has not been advertised.
869 PH/30 MH's patch to allow iscntrl() characters to be list separators.
871 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
872 returned in the SMTP response when smtp_return_error_details was false.
875 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
876 instead of the C I/O that was originally supplied, because problems were
879 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
880 Exim which did not show up earlier: it was assuming that a call to
881 SSL_CTX_set_info_callback() might give an error value. In fact, there is
882 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
883 was a macro that became an assignment, so it seemed to work. This has
884 changed to a proper function call with a void return, hence the compile
885 error. Exim's code has been fixed.
887 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
890 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
892 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
894 PH/37 If a message is not accepted after it has had an id assigned (e.g.
895 because it turns out to be too big or there is a timeout) there is no
896 "Completed" line in the log. When some messages of this type were
897 selected by exigrep, they were listed as "not completed". Others were
898 picked up by some special patterns. I have improved the selection
899 criteria to be more general.
901 PH/38 The host_find_failed option in the manualroute router can now be set
902 to "ignore", to completely ignore a host whose IP address cannot be
903 found. If all hosts are ignored, the behaviour is controlled by the new
904 host_all_ignored option.
906 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
907 homing or because of multiple MX records with /mx) generated more than
908 one IP address, and the following item turned out to be the local host,
909 all the secondary addresses of the first item were incorrectly removed
910 from the list, along with the local host and any following hosts (which
911 is what is supposed to happen).
913 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
914 whoever called Exim into the -H file. In the case of the daemon it was
915 behaving confusingly. When first started, it used values for whoever
916 started the daemon, but after a SIGHUP it used the Exim user (because it
917 calls itself on a restart). I have changed the code so that it now always
920 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
921 message are rejected with the same error (e.g. no authentication or bad
922 sender address), and a DATA command is nevertheless sent (as can happen
923 with PIPELINING or a stupid MUA), the error message that was given to the
924 RCPT commands is included in the rejection of the DATA command. This is
925 intended to be helpful for MUAs that show only the final error to their
928 PH/42 Another patch from the Sieve maintainer.
930 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
931 Eximstats - Fixed some broken HTML links and added missing column headers
933 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
934 columns for Rejects, Temp Rejects, Ham, and Spam rows.
936 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
938 PH/43 Yet another patch from the Sieve maintainer.
940 PH/44 I found a way to check for a TCP/IP connection going away before sending
941 the response to the final '.' that terminates a message, but only in the
942 case where the client has not sent further data following the '.'
943 (unfortunately, this is allowed). However, in many cases there won't be
944 any further data because there won't be any more messages to send. A call
945 to select() can be used: if it shows that the input is "ready", there is
946 either input waiting, or the socket has been closed. An attempt to read
947 the next input character can distinguish the two cases. Previously, Exim
948 would have sent an OK response which the client would never have see.
949 This could lead to message repetition. This fix should cure that, at
950 least in a lot of common cases.
952 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
953 advertised in response to EHLO.
959 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
960 fixed by 4.65/MH/01 (is this a record?) are fixed:
962 (i) An empty string was always treated as zero by the numeric comparison
963 operators. This behaviour has been restored.
965 (ii) It is documented that the numeric comparison operators always treat
966 their arguments as decimal numbers. This was broken in that numbers
967 starting with 0 were being interpreted as octal.
969 While fixing these problems I realized that there was another issue that
970 hadn't been noticed. Values of message_size_limit (both the global option
971 and the transport option) were treated as octal if they started with 0.
972 The documentation was vague. These values are now always treated as
973 decimal, and I will make that clear in the documentation.
979 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
980 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
983 MH/01 Don't check that the operands of numeric comparison operators are
984 integers when their expansion is in "skipping" mode (fixes bug
985 introduced by 4.64-PH/07).
987 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
988 child addresses, Exim now panics and dies. Previously, because the count
989 is held in a short int, deliveries were likely to be lost. As such a
990 large number of recipients for a single message is ridiculous
991 (performance will be very, very poor), I have chosen to impose a limit
992 rather than extend the field.
998 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
999 leftover -K file (the existence of which was triggered by #402).
1000 While we were at it, introduced process PID as part of the -K
1001 filename. This should rule out race conditions when creating
1004 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1005 processing considerably. Previous code took too long for large mails,
1006 triggering a timeout which in turn triggers #401.
1008 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1009 in the DK code in transports.c. sendfile() is not really portable,
1010 hence the _LINUX specificness.
1012 TF/01 In the add_headers option to the mail command in an Exim filter,
1013 there was a bug that Exim would claim a syntax error in any
1014 header after the first one which had an odd number of characters
1017 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1018 callout verification, Exim cached a "reject" for the entire domain. This
1019 is correct for most verifications, but it is not correct for a recipient
1020 verification with use_sender or use_postmaster set, because in that case
1021 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1022 case of MAIL FROM:<> rejection from other early rejections (e.g.
1023 rejection of HELO). When verifying a recipient using a non-null MAIL
1024 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1025 Whatever the result of the callout, the value of the domain cache is
1026 left unchanged (for any other kind of callout, getting as far as trying
1027 RCPT means that the domain itself is ok).
1029 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1032 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1033 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1034 session, the daemon ignores ECONNECT errors and logs others; it now
1035 ignores EPIPE as well.
1037 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1038 (quoted-printable decoding).
1040 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1041 later the small subsequent patch to fix an introduced bug.
1043 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1045 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1047 PH/08 An error is now given if message_size_limit is specified negative.
1049 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1050 to be given (somewhat) arbitrary names.
1052 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1055 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1056 miscellaneous code fixes
1058 PH/10 Added the log_reject_target ACL modifier to specify where to log
1061 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1062 hostname. This is wrong, because it relates to the incoming message (and
1063 probably the interface on which it is arriving) and not to the outgoing
1064 callout (which could be using a different interface). This has been
1065 changed to use the value of the helo_data option from the smtp transport
1066 instead - this is what is used when a message is actually being sent. If
1067 there is no remote transport (possible with a router that sets up host
1068 addresses), $smtp_active_hostname is used.
1070 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1071 tweaks were necessary in order to get it to work (see also 21 below):
1072 (a) The code assumed that strncpy() returns a negative number on buffer
1073 overflow, which isn't the case. Replaced with Exim's string_format()
1075 (b) There were several signed/unsigned issues. I just did the minimum
1076 hacking in of casts. There is scope for a larger refactoring.
1077 (c) The code used strcasecmp() which is not a standard C function.
1078 Replaced with Exim's strcmpic() function.
1079 (d) The code set only $1; it now sets $auth1 as well.
1080 (e) A simple test gave the error "authentication client didn't specify
1081 service in request". It would seem that Dovecot has changed its
1082 interface. Fortunately there's a specification; I followed it and
1083 changed what the client sends and it appears to be working now.
1085 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1088 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1089 address A is aliased to B and C, where B exists and C does not. Without
1090 -v the output is "A verified" because verification stops after a
1091 successful redirection if more than one address is generated. However,
1092 with -v the child addresses are also verified. Exim was outputting "A
1093 failed to verify" and then showing the successful verification for C,
1094 with its parentage. It now outputs "B failed to verify", showing B's
1095 parentage before showing the successful verification of C.
1097 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1098 look up a TXT record in a specific list after matching in a combined
1101 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1102 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1103 they consult the DNS. I had assumed they would set it the way they
1104 wanted; and indeed my experiments on Linux seem to show that in some
1105 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1106 To be on the safe side, however, I have now made the interface to
1107 host_find_byname() similar to host_find_bydns(), with an argument
1108 containing the DNS resolver options. The host_find_byname() function now
1109 sets these options at its start, just as host_find_bydns() does. The smtp
1110 transport options dns_qualify_single and dns_search_parents are passed to
1111 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1112 of host_find_byname() use the default settings of RES_DEFNAMES
1113 (qualify_single) but not RES_DNSRCH (search_parents).
1115 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1116 spool_read_header() do less string testing, by means of a preliminary
1117 switch on the second character of optional "-foo" lines. (This is
1118 overdue, caused by the large number of possibilities that now exist.
1119 Originally there were few.) While I was there, I also converted the
1120 str(n)cmp tests so they don't re-test the leading "-" and the first
1121 character, in the hope this might squeeze out yet more improvement.
1123 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1124 flag allowing group syntax was set by the header_syntax check but not
1125 turned off, possible causing trouble later; (2) The flag was not being
1126 set at all for the header_verify test, causing "group"-style headers to
1127 be rejected. I have now set it in this case, and also caused header_
1128 verify to ignore an empty address taken from a group. While doing this, I
1129 came across some other cases where the code for allowing group syntax
1130 while scanning a header line wasn't quite right (mostly, not resetting
1131 the flag correctly in the right place). These bugs could have caused
1132 trouble for malformed header lines. I hope it is now all correct.
1134 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1135 with the "reply" argument non-NULL. The code, however (which originally
1136 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1137 but it didn't always do it. This confused somebody who was copying the
1138 code for some other use. I have removed all the tests.
1140 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1141 feature that was used to support insecure browsers during the U.S. crypto
1142 embargo. It requires special client support, and Exim is probably the
1143 only MTA that supported it -- and would never use it because real RSA is
1144 always available. This code has been removed, because it had the bad
1145 effect of slowing Exim down by computing (never used) parameters for the
1146 RSA_EXPORT functionality.
1148 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1149 authenticator to fail if there's a tab character in the incoming data
1150 (there should never be unless someone is messing about, as it's supposed
1151 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1152 if the connection is using TLS or if the remote IP is the same as the
1153 local IP, and the "valid-client-cert option" if a client certificate has
1156 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1157 authenticators. This can be used for authorization after authentication
1158 succeeds. (In the case of plaintext, it servers for both authentication
1161 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1162 if any retry times were supplied.
1164 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1165 connection, where there is no client IP address to check. In this
1166 situation, the verify now always succeeds.
1168 PH/25 Applied John Jetmore's -Mset patch.
1170 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1172 PH/27 In a string expansion for a processed (not raw) header when multiple
1173 headers of the same name were present, leading whitespace was being
1174 removed from all of them, but trailing whitespace was being removed only
1175 from the last one. Now trailing whitespace is removed from each header
1176 before concatenation. Completely empty headers in a concatenation (as
1177 before) are ignored.
1179 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1180 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1182 PH/29 [Removed. This was a change that I later backed out, and forgot to
1183 correct the ChangeLog entry (that I had efficiently created) before
1184 committing the later change.]
1186 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1187 address errors (4xx response to RCPT) over the same connection as other
1188 messages routed to the same hosts. Such deliveries are always "forced",
1189 so retry times are not inspected. This resulted in far too many retries
1190 for the affected addresses. The effect occurred only when there were more
1191 hosts than the hosts_max_try setting in the smtp transport when it had
1192 the 4xx errors. Those hosts that it had tried were not added to the list
1193 of hosts for which the message was waiting, so if all were tried, there
1194 was no problem. Two fixes have been applied:
1196 (i) If there are any address or message errors in an SMTP delivery, none
1197 of the hosts (tried or untried) are now added to the list of hosts
1198 for which the message is waiting, so the message should not be a
1199 candidate for sending over the same connection that was used for a
1200 successful delivery of some other message. This seems entirely
1201 reasonable: after all the message is NOT "waiting for some host".
1202 This is so "obvious" that I'm not sure why it wasn't done
1203 previously. Hope I haven't missed anything, but it can't do any
1204 harm, as the worst effect is to miss an optimization.
1206 (ii) If, despite (i), such a delivery is accidentally attempted, the
1207 routing retry time is respected, so at least it doesn't keep
1208 hammering the server.
1210 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1211 in ${readsocket because some servers need this prod.
1213 PH/32 Added some extra debug output when updating a wait-xxx database.
1215 PH/33 The hint "could be header name not terminated by colon", which has been
1216 given for certain expansion errors for a long time, was not being given
1217 for the ${if def:h_colon_omitted{... case.
1219 PH/34 The spec says: "With one important exception, whenever a domain list is
1220 being scanned, $domain contains the subject domain." There was at least
1221 one case where this was not true.
1223 PH/35 The error "getsockname() failed: connection reset by peer" was being
1224 written to the panic log as well as the main log, but it isn't really
1225 panic-worthy as it just means the connection died rather early on. I have
1226 removed the panic log writing for the ECONNRESET error when getsockname()
1229 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1230 runs only) independently of the message's sender address. This meant
1231 that, if the 4xx error was in fact related to the sender, a different
1232 message to the same recipient with a different sender could confuse
1233 things. In particualar, this can happen when sending to a greylisting
1234 server, but other circumstances could also provoke similar problems.
1235 I have changed the default so that the retry time for these errors is now
1236 based a combination of the sender and recipient addresses. This change
1237 can be overridden by setting address_retry_include_sender=false in the
1240 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1241 remote server are returned as part of bounce messages. This was not
1242 happening for LMTP over a pipe (the lmtp transport), but now it is the
1243 same for both kinds of LMTP.
1245 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1246 in header lines that were in fact CNAMEs. This is no longer the case.
1248 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1249 and queue runs started by the daemon processed all messages. This has
1250 been fixed so that -R and -S can now usefully be given with -q<time>.
1252 PH/40 Import PCRE release 6.7 (fixes some bugs).
1254 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1256 PH/42 Give an error if -q is specified more than once.
1258 PH/43 Renamed the variables $interface_address and $interface_port as
1259 $received_ip_address and $received_port, to make it clear that these
1260 values apply to message reception, and not to the outgoing interface when
1261 a message is delivered. (The old names remain recognized, of course.)
1263 PH/44 There was no timeout on the connect() call when using a Unix domain
1264 socket in the ${readsocket expansion. There now is.
1266 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1267 be meaningful with "accept".
1269 SC/01 Eximstats V1.43
1270 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1272 SC/02 Eximstats V1.44
1273 Use a glob alias rather than an array ref in the generated
1274 parser. This improves both readability and performance.
1276 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1277 Collect SpamAssassin and rejection statistics.
1278 Don't display local sender or destination tables unless
1279 there is data to show.
1280 Added average volumes into the top table text output.
1282 SC/04 Eximstats V1.46
1283 Collect data on the number of addresses (recipients)
1284 as well as the number of messages.
1286 SC/05 Eximstats V1.47
1287 Added 'Message too big' to the list of mail rejection
1288 reasons (thanks to Marco Gaiarin).
1290 SC/06 Eximstats V1.48
1291 Mainlog lines which have GMT offsets and are too short to
1292 have a flag are now skipped.
1294 SC/07 Eximstats V1.49 (Alain Williams)
1295 Added the -emptyok flag.
1297 SC/08 Eximstats V1.50
1298 Fixes for obtaining the IP address from reject messages.
1300 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1301 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1302 whitesspace changes from 4.64-PH/27
1304 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1307 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1308 are found, allow negative numbers in numeric criteria)
1310 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1312 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1313 to match changes made in 4.64-PH/43
1315 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1317 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1318 "domain" is actually an IP address, and force a failure. This locks out
1319 those revolvers/nameservers that support "A-for-A" lookups, in
1320 contravention of the specifications.
1322 PH/48 When a host name was looked up from an IP address, and the subsequent
1323 forward lookup of the name timed out, the host name was left in
1324 $sender_host_name, contrary to the specification.
1326 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1327 restricted to single-key lookups, Exim was not diagnosing an error if
1328 * or *@ was used with a query-style lookup.
1330 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1332 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1333 long ago, but noone interested enough thought of it. Let's just say that
1334 the "1.1" means that there are some new functions that weren't there at
1335 some point in the past.
1337 PH/51 Error processing for expansion failure of helo_data from an smtp
1338 transport during callout processing was broken.
1340 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1341 tested/used via the -bh/-bhc/-bs options.
1343 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1344 bug, fixed in subsequent PCRE releases).
1346 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1347 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1349 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1355 SC/01 Use a glob alias rather than an array ref in eximstats generated
1356 parser. This improves both readability and performance.
1358 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1359 Don't display local sender or destination tables in eximstats unless
1360 there is data to show.
1361 Added average volumes into the eximstats top table text output.
1363 SC/03 Collect data on the number of addresses (recipients) as well
1364 as the number of messages in eximstats.
1366 TF/01 Correct an error in the documentation for the redirect router. Exim
1367 does (usually) call initgroups() when daemonizing.
1369 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1370 with consistent privilege compared to when running as a daemon.
1372 TF/03 Note in the spec that $authenticated_id is not set for local
1373 submissions from trusted users.
1375 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1376 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1378 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1379 by adding some example configuration directives to the default
1380 configuration file. A little bit of work is required to uncomment the
1381 directives and define how usernames and passwords are checked, but
1382 there is now a framework to start from.
1384 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1385 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1386 without this. I don't know how relevant this is to other LDAP libraries.
1388 PH/02 Add the verb name to the "unknown ACL verb" error.
1390 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1392 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1394 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1395 directory not expanded when it should be if an expanded home directory
1396 was set for the address (which is overridden by the transport).
1398 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1401 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1402 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1403 because it is too late at that time, and has no effect.
1405 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1406 security issue with \' (bugzilla #107). I could not use the
1407 PQescapeStringConn() function, because it needs a PGconn value as one of
1410 PH/08 When testing addresses using -bt, indicate those final addresses that
1411 are duplicates that would not cause an additional delivery. At least one
1412 person was confused, thinking that -bt output corresponded to deliveries.
1413 (Suppressing duplicates isn't a good idea as you lose the information
1414 about possibly different redirections that led to the duplicates.)
1416 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1417 systems where poll() doesn't work, in particular OS X.
1419 PH/10 Added more information to debugging output for retry time not reached.
1421 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1422 operations in malware.c.
1424 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1427 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1428 syslog_duplication set false, log lines that would normally be written
1429 both the the main log and to the reject log were not written to syslog at
1432 PH/14 In the default configuration, change the use of "message" in ACL warn
1433 statements to "add_header".
1435 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1436 not followed by a command (e.g. "seen endif").
1438 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1439 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1442 PH/17 Added extra conditions to the default value of delay_warning_condition
1446 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1447 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1448 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1451 The Auto-Submitted: and various List- headers are standardised, whereas I
1452 don't think Precedence: ever was.
1454 PH/18 Refactored debugging code in route_finduser() to show more information,
1455 in particular, the error code if getpwnam() issues one.
1457 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1458 This is apparently needed in addition to the PH/07 change above to avoid
1459 any possible encoding problems.
1461 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1462 but not after initializing Perl.
1464 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1465 output them only if debugging. By default they are written stderr,
1466 apparently, which is not desirable.
1468 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1471 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1474 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1476 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1477 authenticated or an ident call has been made. Suppress the default
1478 values for $authenticated_id and $authenticated_sender (but permit -oMai
1479 and -oMas) when testing with -bh.
1481 PH/24 Re-jigged the order of the tests in the default configuration so that the
1482 tests for valid domains and recipients precede the DNS black list and CSA
1483 tests, on the grounds that those ones are more expensive.
1485 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1486 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1487 This bug exists in every version of Exim that I still have, right back to
1490 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1491 However, an attempt to turn on case-sensitivity in a regex key by
1492 including (?-i) didn't work because the subject string was already
1493 lowercased, and the effects were non-intuitive. It turns out that a
1494 one-line patch can be used to allow (?-i) to work as expected.
1500 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1501 other effects) broke the use of negated acl sub-conditions.
1503 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1506 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1507 "Deny" causes Exim to reject the incoming connection with a 554 error.
1508 Unfortunately, if there is a major crisis, such as a disk failure,
1509 tcp-wrappers gives "deny", whereas what one would like would be some
1510 kind of temporary error. A kludge has been added to help with this.
1511 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1512 554 error is used if errno is still zero or contains ENOENT (which occurs
1513 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1516 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1518 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1519 errors. Otherwise a message that provokes a temporary error (when other
1520 messages do not) can cause a whole host to time out.
1522 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1523 addresses were routed directly to files or pipes from a redirect router.
1524 File deliveries just didn't batch; pipe deliveries might have suffered
1527 PH/06 A failure to get a lock for a hints database would erroneously always say
1528 "Failed to get write lock", even when it was really a read lock.
1530 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1531 of 0600. This has been changed to use the value of the lockfile_mode
1532 option (which defaults to 0600).
1534 PH/08 Applied small patch from the Sieve maintainer.
1536 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1537 folder from quota calculations, a direct delivery into this folder messed
1538 up the contents of the maildirsize file. This was because the regex was
1539 used only to exclude .Trash (or whatever) when the size of the mailbox
1540 was calculated. There was no check that a delivery was happening into an
1541 excluded directory. This bug has been fixed by ignoring all quota
1542 processing for deliveries into excluded directories.
1544 PH/10 Added the maildirfolder_create_regex option to appendfile.
1550 PH/01 The code for finding all the local interface addresses on a FreeBSD
1551 system running IPv6 was broken. This may well have applied to all BSD
1552 systems, as well as to others that have similar system calls. The broken
1553 code found IPv4 interfaces correctly, but gave incorrect values for the
1554 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1555 that it would not match correctly against @[] and not recognize the IPv6
1558 PH/02 The ipliteral router was not recognizing addresses of the form user@
1559 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1561 PH/03 Added disable_ipv6.
1563 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1564 decoded form, because it is most often used to construct To: headers
1565 lines in autoreplies, and the decoded form may well be syntactically
1566 invalid. However, $reply_address has leading white space removed, and all
1567 newlines turned into spaces so that the autoreply transport does not
1570 PH/05 If group was specified without a user on a router, and no group or user
1571 was specified on a transport, the group from the router was ignored.
1573 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1574 for visible compile-time settings that can be used to change these
1575 numbers, for those that want even more. Backwards compatibility with old
1576 spool files has been maintained. However, going back to a previous Exim
1577 release will lost any variables that are in spool files.
1579 PH/07 Two small changes when running in the test harness: increase delay when
1580 passing a TCP/IP connection to a new process, in case the original
1581 process has to generate a bounce, and remove special handling of
1582 127.0.0.2 (sic), which is no longer necessary.
1584 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1585 be the same on different OS.
1587 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1590 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1591 whether --show-vars was specified or not
1593 JJ/02 exipick: Added support for new ACL variable spool format introduced
1596 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1597 syntactically invalid From: or Reply-to: line, and a filter used this to
1598 generate an autoreply, and therefore failed to obtain an address for the
1599 autoreply, Exim could try to deliver to a non-existent relative file
1600 name, causing unrelated and misleading errors. What now happens is that
1601 it logs this as a hard delivery error, but does not attempt to create a
1604 PH/11 The exinext utility has a -C option for testing purposes, but although
1605 the given file was scanned by exinext itself; it wasn't being passed on
1606 when Exim was called.
1608 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1609 an end-of-file indication when reading a command response.
1611 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1612 compiled. In many other places in Exim, IPv6 addresses are always
1613 recognized, so I have changed this. It also means that IPv4 domain
1614 literals of the form [IPV4:n.n.n.n] are now always recognized.
1616 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1617 used if the router is not running as root, for example, when verifying at
1618 ACL time, or when using -bh. The debugging output from this situation was
1619 non-existent - all you got was a failure to exec. I have made two
1622 (a) Failures to set uid/gid, the current directory, or a process leader
1623 in a subprocess such as that created by queryprogram now generate
1624 suitable debugging ouput when -d is set.
1626 (b) The queryprogram router detects when it is not running as root,
1627 outputs suitable debugging information if -d is set, and then runs
1628 the subprocess without attempting to change uid/gid.
1630 PH/15 Minor change to Makefile for building test_host (undocumented testing
1633 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1634 additional section of a DNS packet that returns MX or SRV records.
1635 Instead, it always explicitly searches for A/AAAA records. This avoids
1636 major problems that occur when a DNS server includes only records of one
1637 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1638 fixed another bug: if SRV records were looked up and the corresponding
1639 address records were *not* found in the additional section, the port
1640 values from the SRV records were lost.
1642 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1643 using the correct key (the original address) when searching the retry
1644 rules in order to find which one to use for generating the retry hint.
1646 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1647 adding the default one. Similarly, if it contains a Reply-To: header, the
1648 errors_reply_to option, if set, is not used.
1650 PH/19 When calculating a retry time, Exim used to measure the "time since
1651 failure" by looking at the "first failed" field in the retry record. Now
1652 it does not use this if it is later than than the arrival time of the
1653 message. Instead it uses the arrival time. This makes for better
1654 behaviour in cases where some deliveries succeed, thus re-setting the
1655 "first failed" field. An example is a quota failure for a huge message
1656 when small messages continue to be delivered. Without this change, the
1657 "time since failure" will always be short, possible causing more frequent
1658 delivery attempts for the huge message than are intended.
1659 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1661 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1662 $1, $2, $3) because the numerical variables can be reset during some
1663 expansion items (e.g. "match"), thereby losing the authentication data.
1665 PH/21 Make -bV show the size of off_t variables so that the test suite can
1666 decide whether to run tests for quotas > 2G.
1668 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1669 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1670 filecount value is greater than 2G or if a quota value is greater than 2G
1671 on a system where the size of off_t is not greater than 4, a panic error
1674 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1675 never match. The debug and -bh output now contains an explicit error
1676 message indicating a malformed IPv4 address or mask.
1678 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1679 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1680 PH/23 above applies.
1682 PH/25 Do not write to syslog when running in the test harness. The only
1683 occasion when this arises is a failure to open the main or panic logs
1684 (for which there is an explicit test).
1686 PH/26 Added the /no_tell option to "control=freeze".
1688 PH/27 If a host name lookup failed very early in a connection, for example, if
1689 the IP address matched host_lookup and the reverse lookup yielded a name
1690 that did not have a forward lookup, an error message of the form "no IP
1691 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1692 could be logged. Now it outputs the IP address instead of "NULL".
1694 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1695 allows the sender and the authenticated sender to be set when
1696 submitting a message from within Exim. Since child_open_exim() is
1697 documented for local_scan(), the new function should be too.
1699 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1700 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1701 results in an empty string is now treated as unset.
1703 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1705 PH/31 Added sender_verify_fail logging option.
1707 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1708 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1709 tidied the source and removed it altogether.
1711 PH/33 When a queue run was abandoned because the load average was too high, a
1712 log line was always written; now it is written only if the queue_run log
1713 selector is set. In addition, the log line for abandonment now contains
1714 information about the queue run such as the pid. This is always present
1715 in "start" and "stop" lines but was omitted from the "abandon" line.
1717 PH/34 Omit spaces between a header name and the colon in the error message that
1718 is given when verify = headers_syntax fails (if there are lots of them,
1719 the message gets confusing).
1721 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1722 names, as there are now some PTR records that contain slashes. This check
1723 is only to protect against broken name servers that fall over on strange
1724 characters, so the fact that it applies to all lookups doesn't matter.
1726 PH/36 Now that the new test suite is complete, we can remove some of the
1727 special code in Exim that was needed for the old test suite. For example,
1728 sorting DNS records because real resolvers return them in an arbitrary
1729 order. The new test suite's fake resolver always returns records in the
1732 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1733 bounces) except when queue_only is set, to avoid logging races between
1734 the different processes.
1736 PH/38 Panic-die if .include specifies a non-absolute path.
1738 PH/39 A tweak to the "H" retry rule from its user.
1740 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1741 a label. They prevented compilation on older perls.
1743 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1744 a warning to be raised on newish perls.
1746 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1747 on queue. Changes to match documented behaviour of showing count of
1748 messages matching specified criteria.
1750 PH/40 Changed the default ident timeout from 30s to 5s.
1752 PH/41 Added support for the use of login_cap features, on those BSD systems
1753 that have them, for controlling the resources used by pipe deliveries.
1755 PH/42 The content-scanning code uses fopen() to create files in which to put
1756 message data. Previously it was not paying any attention to the mode of
1757 the files. Exim runs with umask(0) because the rest of the code creates
1758 files with open(), and sets the required mode explicitly. Thus, these
1759 files were ending up world-writeable. This was not a big issue, because,
1760 being within the spool directory, they were not world-accessible. I have
1761 created a function called modefopen, which takes an additional mode
1762 argument. It sets umask(777), creates the file, chmods it to the required
1763 mode, then resets the umask. All the relevant calls to fopen() in the
1764 content scanning code have been changed to use this function.
1766 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1767 to 24 hours. This avoids potential overflow problems when processing G
1768 and H retry rules. I suspect nobody ever tinkers with this value.
1770 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1772 PH/45 When the plaintext authenticator is running as a client, the server's
1773 challenges are checked to ensure they are valid base64 strings. By
1774 default, the authentication attempt is cancelled if an invalid string is
1775 received. Setting client_ignore_invalid_base64 true ignores these errors.
1776 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1777 they are received. Thus, the responses can be made to depend on the
1778 challenges. If an invalid string is ignored, an empty string is placed in
1781 PH/46 Messages that are created by the autoreply transport now contains a
1782 References: header, in accordance with RFCs 2822 and 3834.
1784 PH/47 Added authenticated_sender_force to the smtp transport.
1786 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1788 PH/49 Installed latest patch from the Sieve maintainer.
1790 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1791 was also set, the appendfile transport was unnecessarily scanning a
1792 directory of message files (e.g. for maildir delivery) to find the count
1793 of files (along with the size), even though it did not need this
1794 information. It now does the scan only if it needs to find either the
1795 size of the count of files.
1797 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1799 PH/52 Two bugs concerned with error handling when the smtp transport is
1802 (i) Exim was not creating retry information for temporary errors given
1803 for individual recipients after the DATA command when the smtp transport
1804 was used in LMTP mode. This meant that they could be retried too
1805 frequently, and not timed out correctly.
1807 (ii) Exim was setting the flag that allows error details to be returned
1808 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1809 recipients that were returned after the DATA command.
1811 PH/53 This is related to PH/52, but is more general: for any failing address,
1812 when detailed error information was permitted to be returned to the
1813 sender, but the error was temporary, then after the final timeout, only
1814 "retry timeout exceeded" was returned. Now it returns the full error as
1815 well as "retry timeout exceeded".
1817 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1818 do this, and (what is worse) MTAs that accept it.
1820 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1821 will now be deprecated.
1823 PH/56 New os.c-cygwin from the Cygwin maintainer.
1825 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1826 formats (previously only available in exim formats via -bpr, -bpru,
1827 and -bpra. Now also available in native and exiqgrep formats)
1829 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1830 with very large, slow to parse queues
1832 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1834 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1836 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1837 responses to authentication challenges, though it was showing the
1838 challenges; (ii) I've removed the CR characters from the debug output for
1841 PH/58 Allow for the insertion of a newline as well as a space when a string
1842 is turned into more than one encoded-word during RFC 2047 encoding. The
1843 Sieve code now uses this.
1845 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1846 data_4xx, lost_connection, tls_required.
1848 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1849 message was being sent as an SMTP response.
1851 PH/61 Add -l and -k options to exicyclog.
1853 PH/62 When verifying, if an address was redirected to one new address, so that
1854 verification continued, and the new address failed or deferred after
1855 having set something in $address_data, the value of $address_data was not
1856 passed back to the ACL. This was different to the case when no
1857 redirection occurred. The value is now passed back in both cases.
1859 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1860 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1861 use login_cap.h, so on its own it isn't the distinguishing feature. The
1862 new name refers directly to the setclassresources() function.
1864 PH/65 Added configuration files for NetBSD3.
1866 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1868 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1869 is preferred over IPv4.
1871 PH/68 The bounce_return_message and bounce_return_body options were not being
1872 honoured for bounces generated during the reception of non-SMTP messages.
1873 In particular, this applied to messages rejected by the ACL. This bug has
1874 been fixed. However, if bounce_return_message is true and bounce_return_
1875 body is false, the headers that are returned for a non-SMTP message
1876 include only those that have been read before the error was detected.
1877 (In the case of an ACL rejection, they have all been read.)
1879 PH/69 The HTML version of the specification is now built in a directory called
1880 spec_html instead of spec.html, because the latter looks like a path with
1881 a MIME-type, and this confuses some software.
1883 PH/70 Catch two compiler warnings in sieve.c.
1885 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1886 function verify_get_ident() calls ip_connect() to connect a socket, but
1887 if the "connect()" function timed out, ip_connect() used to close the
1888 socket. However, verify_get_ident() also closes the socket later, and in
1889 between Exim writes to the log, which may get opened at this point. When
1890 the socket was closed in ip_connect(), the log could get the same file
1891 descriptor number as the socket. This naturally causes chaos. The fix is
1892 not to close the socket in ip_connect(); the socket should be closed by
1893 the function that creates it. There was only one place in the code where
1894 this was missing, in the iplookup router, which I don't think anybody now
1895 uses, but I've fixed it anyway.
1897 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1898 well as to direct DNS lookups. Otherwise the handling of names in host
1899 lists is inconsistent and therefore confusing.
1905 PH/01 Two changes to the default runtime configuration:
1907 (1) Move the checks for relay_from_hosts and authenticated clients from
1908 after to before the (commented out) DNS black list checks.
1910 (2) Add control=submission to the relay_from_hosts and authenticated
1911 clients checks, on the grounds that messages accepted by these
1912 statements are most likely to be submissions.
1914 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1916 (1) Generate an error if the third argument for the ${prvs expansion is
1919 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1922 (3) Reset the variables that are obtained from the first argument of
1923 ${prvscheck and used in the second argument before leaving the code,
1924 because their memory is reclaimed, so using them afterwards may do
1927 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1928 one (it's much easier than Tom thought :-).
1930 (5) Because of (4), we can now allow for the use of $prvscheck_result
1931 inside the third argument.
1933 PH/03 For some reason, the default setting of PATH when running a command from
1934 a pipe transport was just "/usr/bin". I have changed it to
1937 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1938 anything to be listed in the output from -bV.
1940 PH/05 When a filter generated an autoreply, the entire To: header line was
1941 quoted in the delivery log line, like this:
1943 => >A.N.Other <ano@some.domain> <original@ddress> ...
1945 This has been changed so that it extracts the operative address. There
1946 may be more than one such address. If so, they are comma-separated, like
1949 => >ano@some.domain,ona@other.domain <original@ddress> ...
1951 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1952 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1953 not being looked up in the rDNS to get a host name, Exim was showing the
1954 IP address twice in Received: lines, even though the IP addresses were
1955 identical. For example:
1957 Received: from [1.2.3.4] (helo=[1.2.3.4])
1959 However, if the real host name was known, it was omitting the HELO data
1960 if it matched the actual IP address. This has been tidied up so that it
1961 doesn't show the same IP address twice.
1963 PH/07 When both +timestamp and +memory debugging was on, the value given by
1964 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1965 called by the debug printing, thereby overwriting the timestamp buffer.
1966 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1968 PH/08 When the original message was included in an autoreply transport, it
1969 always said "this is a copy of the message, including all the headers",
1970 even if body_only or headers_only was set. It now gives an appropriate
1973 PH/09 Applied a patch from the Sieve maintainer which:
1975 o fixes some comments
1976 o adds the (disabled) notify extension core
1977 o adds some debug output for the result of if/elsif tests
1978 o points to the current vacation draft in the documentation
1979 and documents the missing references header update
1983 o fixes a bug in processing the envelope test (when testing
1984 multiple envelope elements, the last element determinted the
1987 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1988 Electronic Mail") by including:
1990 Auto-submitted: auto-generated
1992 in the messages that it generates (bounce messages and others, such as
1993 warnings). In the case of bounce messages for non-SMTP mesages, there was
1994 also a typo: it was using "Auto_submitted" (underscore instead of
1995 hyphen). Since every message generated by Exim is necessarily in response
1996 to another message, thes have all been changed to:
1998 Auto-Submitted: auto-replied
2000 in accordance with these statements in the RFC:
2002 The auto-replied keyword:
2004 - SHOULD be used on messages sent in direct response to another
2005 message by an automatic process,
2007 - MUST NOT be used on manually-generated messages,
2009 - MAY be used on Delivery Status Notifications (DSNs) and Message
2010 Disposition Notifications (MDNs),
2012 - MUST NOT be used on messages generated by automatic or periodic
2013 processes, except for messages which are automatic responses to
2016 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2017 to the default Received: header definition.
2019 PH/12 Added log selector acl_warn_skipped (default on).
2021 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2022 variables because (a) they are in the wrong storage pool and (b) even if
2023 they were copied, it wouldn't work properly because of the caching.
2025 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2026 checking when decoding. Apparently there are clients that generate
2027 overlong encoded strings. Why am I not surprised?
2029 PH/15 If the first argument of "${if match_address" was not empty, but did not
2030 contain an "@" character, Exim crashed. Now it writes a panic log message
2031 and treats the condition as false.
2033 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2035 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2036 extension "envelope-auth". The code is finished and in agreement with
2037 other implementations, but there is no documentation so far and in fact,
2038 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2039 not changing the active code.
2041 Print executed "if" and "elsif" statements when debugging is used. This
2042 helps a great deal to understand what a filter does.
2044 Document more things not specified clearly in RFC3028. I had all this
2045 sorted out, when out of a sudden new issues came to my mind. Oops."
2047 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2050 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2051 canonical form (as documented). However, after a host name lookup from
2052 the IP address, check_host() was doing a simple string comparison with
2053 addresses acquired from the DNS when checking that the found name did
2054 have the original IP as one of its addresses. Since any found IPv6
2055 addresses are likely to be in abbreviated form, the comparison could
2056 fail. Luckily, there already exists a function for doing the comparison
2057 by converting both addresses to binary, so now that is used instead of
2058 the text comparison.
2060 PH/20 There was another similar case to PH/19, when a complete host name was
2061 given in a host list; looking up its IP address could give an abbreviated
2062 form, whereas the current host's name might or might not be abbreviated.
2063 The same fix has been applied.
2069 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2070 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2073 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2074 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2076 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2078 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2079 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2080 Also, for "Auto-Submitted", treat anything other than "no" as
2081 non-personal, in accordance with RFC 3834. (Previously it treated
2082 anything starting "auto-" as non-personal.)
2084 TF/01 The control=submission/name=... option had a problem with syntax
2085 errors if the name included a slash character. The /name= option
2086 now slurps the rest of the string, so it can include any characters
2087 but it must come last in the list of options (after /sender_retain
2090 PH/05 Some modifications to the interface to the fake nameserver for the new
2098 TK/01 Added the "success_on_redirect" address verification option. See
2099 NewStuff for rationale and an example.
2101 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2103 PH/02 Patch to exigrep to allow it to work on syslog lines.
2105 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2106 fread() to skip over the body file's header line, because in Cygwin the
2107 header line is locked and is inaccessible.
2109 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2110 co-exist for some time) to make it clear that it is the Exim ID that is
2111 referenced, not the Message-ID: header line.
2113 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2114 string_format() function, because snprintf() does not exist on all
2117 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2118 ${stat: expansion item.
2120 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2121 protocol synchronization error", to keep the pedants happy.
2123 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2124 well as for IRIX systems, when gcc is being used. See the host.c source
2127 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2129 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2132 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2133 transport and to the smtp transport in LMTP mode.
2135 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2137 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2138 run a filter in a subprocess. This could lead to confusion in subsequent
2139 lookups in the parent process. There should also be a search_tidyup() at
2140 the end of the subprocess.
2142 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2143 only if the host matched helo_try_verify_hosts, which caused the
2144 verification to occur when the EHLO/HELO command was issued. The ACL just
2145 tested the remembered result. Now, if a previous verification attempt has
2146 not happened, "verify = helo" does it there and then.
2148 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2150 TK/03 Fix log output including CR from clamd.
2152 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2153 exist provoked a memory error which could cause a segfault.
2155 PH/15 Installed PCRE 6.2
2157 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2159 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2160 of the problem. Specifically, suggested +O2 rather than +O1 for the
2163 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2165 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2166 record, the port was not correctly logged when the outgoing_port log
2167 selector was set (it logged the transort's default port).
2169 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2170 fallback_hosts, and "hosts" in the smtp transport.
2172 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2173 host errors such as "Connection refused".
2175 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2176 authentication with radiusclient 0.4.9:
2178 - Error returned from rc_read_config was caught wrongly
2179 - Username/password not passed on to radius server due to wrong length.
2181 The presumption is that some radiusclient API changes for 4.51/PH/17
2182 were not taken care of correctly. The code is still untested by me (my
2183 Linux distribution still has 0.3.2 of radiusclient), but it was
2184 contributed by a Radius user.
2186 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2187 expanding the "port" option of the smtp transport.
2189 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2190 while reading a MIME header. Thanks to Tom Hughes for a patch.
2192 PH/24 Include config.h inside local_scan.h so that configuration settings are
2195 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2196 that in an ACL for RCPT (for example), you can examine exactly what was
2199 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2200 commands, but it was not correctly comparing the address with the actual
2201 client host address. Thus, it would show the EHLO address in Received:
2202 header lines when this was not necessary.
2204 PH/27 Added the % operator to ${eval:}.
2206 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2207 it should be ignoring failures (because with -C, for example, it has lost
2208 privilege). It wasn't ignoring creation failures other than "already
2211 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2214 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2215 because an input line was too long, either on its own, or by virtue of
2216 too many continuations, the temporary file was not being removed, and the
2217 return code was incorrect.
2219 PH/31 Missing "BOOL" in function definition in filtertest.c.
2221 PH/32 Applied Sieve patches from the maintainer.
2223 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2225 PH/33 Added "verify = not_blind".
2227 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2228 Local/Makefile (with some defaults set). These are used in built scripts
2229 such as exicyclog, but they have never been used in the exim_install
2230 script (though there are many overriding facilities there). I have
2231 arranged that the exim_install script now takes note of these two
2234 PH/35 Installed configuration files for Dragonfly.
2236 PH/36 When a locally submitted message by a trusted user did not contain a
2237 From: header, and the sender address was obtained from -f or from an SMTP
2238 MAIL command, and the trusted user did not use -F to supply a sender
2239 name, $originator_name was incorrectly used when constructing a From:
2240 header. Furthermore, $originator_name was used for submission mode
2241 messages from external hosts without From: headers in a similar way,
2242 which is clearly wrong.
2244 PH/37 Added control=suppress_local_fixups.
2246 PH/38 When log_selector = +received_sender was set, and the addition of the
2247 sender made the log line's construction buffer exactly full, or one byte
2248 less than full, an overflow happened when the terminating "\n" was
2251 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2252 when the result of a list match is failure because a DNS lookup failed.
2254 PH/40 RM_COMMAND is now used in the building process.
2256 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2257 the "build-* directories that it finds.
2259 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2260 address in a domain literal was a prefix of an interface address.
2262 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2263 when verifying a sender address, unless rewrite_headers is false.
2265 PH/44 Wrote a long comment about why errors_to addresses are verified as
2266 recipients, not senders.
2268 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2269 the ratelimit ACL was added.
2271 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2273 PH/46 Added extra information about PostgreSQL errors to the error string.
2275 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2276 avoiding the need to install special zones in a real server. This is
2277 backwards compatible; if it can't find the fake resolver, it drops back.
2278 Thus, both old and new test suites can be run.
2280 TF/02 Added util/ratelimit.pl
2282 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2283 clock is set back in time.
2285 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2286 Candler <B.Candler@pobox.com>.
2288 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2289 OK for addresses that are the result of redirections.
2291 PH/48 A number of further additions for the benefit of the new test suite,
2292 including a fake gethostbyname() that interfaces to the fake DNS resolver
2295 TF/06 The fix for widen_domains has also been applied to qualify_single and
2296 search_parents which are the other dnslookup options that can cause
2299 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2302 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2304 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2305 service" if it encountered an MX record with an empty target hostname.
2306 The message is now "an MX or SRV record indicated no SMTP service".
2308 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2309 if the DNS of the sending site is misconfigured. This is quite a
2310 common situation. This change restores the behaviour of treating a
2311 helo verification defer as a failure.
2313 PH/51 If self=fail was set on a router, the bounce message did not include the
2314 actual error message.
2320 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2322 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2323 command itself ended in error, the underlying message about the transport
2324 filter timeout was being overwritten with the pipe command error. Now the
2325 underlying error message should be appended to the second error message.
2327 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2329 PH/02 Reduce the amount of output that "make" produces by default. Full output
2330 can still be requested.
2332 PH/03 The warning log line about a condition test deferring for a "warn" verb
2333 was being output only once per connection, rather than after each
2334 occurrence (because it was using the same function as for successful
2335 "warn" verbs). This seems wrong, so I have changed it.
2337 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2338 it should not have, which might have caused a crash in the right
2339 circumstances, but probably never did.
2341 PH/04 Installed a modified version of Tony Finch's patch to make submission
2342 mode fix the return path as well as the Sender: header line, and to
2343 add a /name= option so that you can make the user's friendly name appear
2346 TF/03 Added the control = fakedefer ACL modifier.
2348 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2349 Mark Lowes for thorough testing.
2351 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2353 TK/03 Merged latest SRS patch from Miles Wilton.
2355 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2356 with the definition in sysexits.h (which is #included earlier).
2357 Fortunately, Exim does not actually use EX_OK. The code used to try to
2358 preserve the sysexits.h value, by assumimg that macro definitions were
2359 scanned for macro replacements. I have been disabused of this notion,
2360 so now the code just undefines EX_OK before #including unistd.h.
2362 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2363 in the smtp transport. When a block could not be written in a single
2364 write() function, the timeout was being re-applied to each part-write.
2365 This seems wrong - if the receiver was accepting one byte at a time it
2366 would take for ever. The timeout is now adjusted when this happens. It
2367 doesn't have to be particularly precise.
2369 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2370 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2372 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2373 without a domain if the check to <postmaster@domain> fails.
2375 SC/01 Eximstats: added -xls and the ability to specify output files
2376 (patch written by Frank Heydlauf).
2378 SC/02 Eximstats: use FileHandles for outputing results.
2380 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2382 SC/04 Eximstats: fixed display of large numbers with -nvr option
2384 SC/05 Eximstats: fixed merging of reports with empty tables.
2386 SC/06 Eximstats: added the -include_original_destination flag
2388 SC/07 Eximstats: removed tabs and trailing whitespace.
2390 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2392 TK/06 MBOX spool code: Add real "From " MBOX separator line
2393 so the .eml file is really in mbox format (even though
2394 most programs do not really care). Patch from Alex Miller.
2396 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2397 The latter is generated from $received_to and is only set if the
2398 message has one envelope recipient. SA can use these headers,
2399 obviously out-of-the-box. Patch from Alex Miller.
2401 PH/08 The ${def test on a variable was returning false if the variable's
2402 value was "0", contrary to what the specification has always said!
2403 The result should be true unless the variable is empty.
2405 PH/09 The syntax error of a character other than { following "${if
2406 def:variable_name" (after optional whitespace) was not being diagnosed.
2407 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2408 accidental colon was present, for example, could give incorrect results.
2410 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2411 result is used (not including appendfile, where other changes are about
2414 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2415 This involved changing a lot of size variables from int to off_t. It
2416 should work with maildirs and everything.
2418 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2419 spamd dying while we are connected to it.
2421 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2424 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2425 function for BDB 4.3.
2427 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2429 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2430 the sources for PCRE have been re-arranged and more files are now
2433 PH/15 The code I had for printing potentially long long variables in PH/11
2434 above was not the best (it lost precision). The length of off_t variables
2435 is now inspected at build time, and an appropriate printing format (%ld
2436 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2437 to be "long long int" or "long int". This is needed for the internal
2438 formatting function string_vformat().
2440 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2441 the configuration file to be ":syslog", then the script "guesses" where
2442 the logs files are, rather than using the compiled in default. In our
2443 case the guess is not the same as the compiled default, so the script
2444 suddenly stopped working when I started to use syslog. The patch checks
2445 to see if log_file_path is "". If so, it attempts to read it from exim
2446 with no configuration file to get the compiled in version, before it
2447 falls back to the previous guessing code."
2449 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2450 implementing BATV in an Exim configuration. See NewStuff for the gory
2453 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2454 Makefile that are specific to HP-UX.
2456 PH/18 If the "use_postmaster" option was set for a recipient callout together
2457 with the "random" option, the postmaster address was used as the MAIL
2458 FROM address for the random test, but not for the subsequent recipient
2459 test. It is now used for both.
2461 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2462 patch removes a few documentation additions to RFC 3028, because the
2463 latest draft now contains them. It adds the new en;ascii-case comparator
2464 and a new error check for 8bit text in MIME parts. Comparator and
2465 require names are now matched exactly. I enabled the subaddress
2466 extension, but it is not well tested yet (read: it works for me)."
2468 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2469 rework some of the code of TK/09 above to avoid the hardwired use of
2470 "%lld" and "long long". Replaced the call to snprintf() with a call to
2473 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2474 records point to non-existent hosts", "retry timeout exceeded", and
2475 "retry time not reached for any host after a long failure period".
2477 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2478 experimental DomainKeys support:
2480 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2481 (2) On an error such as an illegally used "control", the wrong name for
2482 the control was given.
2484 These problems did NOT occur unless DomainKeys support was compiled.
2486 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2488 PH/24 Added ${if match_ip condition.
2490 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2491 databases so that it will be absolutely obvious if a crash occurs in the
2492 DB library. This is a regular occurrence (often caused by mis-matched
2495 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2496 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2497 picked up on a user's system that detects such things. There doesn't seem
2498 to be a gcc warning option for this - only an attribute that has to be
2499 put on the function's prototype. It seems that in Fedora Core 4 they have
2500 set this on a number of new functions. No doubt there will be more in due
2503 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2504 specify a transport. However, if an address that was verified by such a
2505 router was the subject of a callout, Exim crashed because it tried to
2506 read the rcpt_include_affixes from the non-existent transport. Now it
2507 just assumes that the setting of that option is false. This bug was
2508 introduced by 4.51/PH/31.
2510 PH/28 Changed -d+all to exclude +memory, because that information is very
2511 rarely of interest, but it makes the output a lot bigger. People tend to
2512 do -d+all out of habit.
2514 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2515 code in os-type was giving problems when libc.so lives in lib64, like on
2518 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2519 aren't the modern standard, and it seems that some systems' include files
2520 don't always have them. Exim was already checking for some of the newer
2521 ones like T_AAAA, and defining it itself. I've added checks for all the
2522 record types that Exim uses.
2524 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2525 not automatically generating a new one, as it is supposed to. This
2526 prevented TLS from working. If the file did exist, but contained invalid
2527 data, a new version was generated, as expected. It was only the case of a
2528 non-existent file that was broken.
2530 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2531 with a change in libdomainkeys > 0.64.
2533 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2534 from DNS. If the selector record carries the flag, it now has
2535 precedence over the domain-wide flag.
2537 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2539 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2540 the use of a port name that isn't defined in /etc/services) occurred, the
2541 message was deferred as in a normal delivery, and thus remained on the
2542 spool, instead of being failed because of the mua_wrapper setting. This
2543 is now fixed, and I tidied up some of the mua_wrapper messages at the
2546 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2547 the messages in a hash of arrays rather than using individual hashes.
2548 This is a bit cleaner and results in dramatic memory savings, albeit
2549 at a slight CPU cost.
2551 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2552 as requested by Marc Sherman.
2554 SC/10 Eximstats: added histograms for user specified patterns as requested
2557 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2559 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2560 fopen() in the content-scanning modules that did not already have it.
2566 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2567 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2569 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2571 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2573 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2574 to test Sieve filters that use "vacation".
2576 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2577 that changes the way the GnuTLS parameters are stored in the cache file.
2578 The new format can be generated externally. For backward compatibility,
2579 if the data in the cache doesn't make sense, Exim assumes it has read an
2580 old-format file, and it generates new data and writes a new file. This
2581 means that you can't go back to an older release without removing the
2584 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2585 work if there are any delivery delays because "one_time" forces the
2586 parent to be marked "delivered", so its unseen clone is never tried
2587 again. For this reason, Exim now forbids the simultaneous setting of
2590 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2591 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2592 entry if you want to know the details. The fix, however, neglected to
2593 consider the case when local delivery batching is involved. The test for
2594 "previously delivered" was not happening when checking to see if an
2595 address could be batched with a previous (undelivered) one; under
2596 certain circumstances this could lead to multiple deliveries to the same
2599 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2600 in its include files, and this causes problems building Exim.
2602 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2603 header_syntax) but Exim was just ignoring anything given after a slash.
2604 In particular, this caused confusion with an attempt to use "verify =
2605 reverse_host_lookup/defer_ok". An error is now given when options are
2606 supplied for verify items that do not have them. (Maybe reverse_host_
2607 lookup should have a defer_ok option, but that's a different point.)
2609 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2610 defined by RFC 821) to 2048, because there were problems with some AUTH
2611 commands, and RFC 1869 says the size should be increased for extended
2612 SMTP commands that take arguments.
2614 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2617 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2618 "unknown" error; now it says that the functionality isn't in the binary.
2620 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2621 an address' error message when a string expansion fails (syntax or
2622 whatever). Otherwise the password may appear in the log. Following change
2623 PH/42 below, there is no longer a chance of it appearing in a bounce
2626 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2628 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2629 of its addresses was ever tried. (Bugzilla bug #2.)
2631 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2632 the result incorrectly in the debug output. (It correctly added a newline
2633 to what was transported.)
2635 TF/01 Added $received_time.
2637 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2638 commented out examples of how to interface to a virus scanner and to
2639 SpamAssassin. Also added commented examples of av_scanner and
2640 spamd_address settings.
2642 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2643 and controls are allowed in which ACLs. There were a couple of minor
2644 errors. Some of the entries in the conditions table (which is a table of
2645 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2646 negation of where the condition IS allowed.
2648 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2650 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2651 header file does not have a version number, so I've had to invent a new
2652 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2653 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2654 radiusclient), but it was contributed by a Radius user.
2656 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2657 files or directories, for OpenSSL.
2659 PH/19 When an Exim process that is running as root has to create an Exim log
2660 file, it does so in a subprocess that runs as exim:exim so as to get the
2661 ownership right at creation (otherwise, other Exim processes might see
2662 the file with the wrong ownership). There was no test for failure of this
2663 fork() call, which would lead to the process getting stuck as it waited
2664 for a non-existent subprocess. Forks do occasionally fail when resources
2665 run out. I reviewed all the other calls to fork(); they all seem to check
2668 PH/20 When checking for unexpected SMTP input at connect time (before writing
2669 the banner), Exim was not dealing correctly with a non-positive return
2670 from the read() function. If the client had disconnected by this time,
2671 the result was a log entry for a synchronization error with an empty
2672 string after "input=" when read() returned zero. If read() returned -1
2673 (an event I could not check), uninitialized data bytes were printed.
2674 There were reports of junk text (parts of files, etc) appearing after
2677 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2679 PH/22 Added support for macro redefinition, and (re)definition in between
2680 driver and ACL definitions.
2682 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2683 forgetting to use the resulting value; it was using the unexpanded value.
2685 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2686 hadn't been configured. The fix is from Juergen Kreileder, who
2687 understands it better than I do:
2689 "Here's what I see happening with three configured cyrus_sasl
2690 authenticators configured (plain, login, cram-md5):
2692 On startup auth_cyrus_sasl_init() gets called for each of these.
2693 This means three calls to sasl_listmech() without a specified mech_list.
2694 => SASL tests which mechs of all available mechs actually work
2695 => three warnings about OTP not working
2696 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2698 With the patch, sasl_listmech() also gets called three times. But now
2699 SASL's mech_list option is set to the server_mech specified in the the
2700 authenticator. Or in other words, the answer from sasl_listmech()
2701 gets limited to just the mech you're testing for (which is different
2703 => the return list contains just 'plain' or 'login', 'cram-md5' or
2704 nothing depending on the value of ob->server_mech.
2706 I've just tested the patch: Authentication still works fine,
2707 unavailable mechs specified in the exim configuration are still
2708 caught, and the auth.log warnings about OTP are gone."
2710 PH/25 When debugging is enabled, the contents of the command line are added
2711 to the debugging output, even when log_selector=+arguments is not
2714 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2715 answer is "GNU", and only if the return is "GNU/something" is the answer
2718 PH/27 $acl_verify_message is now set immediately after the failure of a
2719 verification in an ACL, and so is available in subsequent modifiers. In
2720 particular, the message can be preserved by coding like this:
2722 warn !verify = sender
2723 set acl_m0 = $acl_verify_message
2725 Previously, $acl_verify_message was set only while expanding "message"
2726 and "log_message" when a very denied access.
2728 PH/28 Modified OS/os.c-Linux with
2730 -#ifndef OS_LOAD_AVERAGE
2731 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2733 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2734 nomenclature these days.)
2736 PH/29 Installed patch from the Sieve maintainer that adds the options
2737 sieve_useraddress and sieve_subaddress to the redirect router.
2739 PH/30 In these circumstances:
2740 . Two addresses routed to the same list of hosts;
2741 . First host does not offer TLS;
2742 . First host accepts first address;
2743 . First host gives temporary error to second address;
2744 . Second host offers TLS and a TLS session is established;
2745 . Second host accepts second address.
2746 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2747 and peerdn, if requested) that were in fact used only for the second
2750 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2751 attention to any local part prefix or suffix that was matched by the
2752 router that accepted the address. It now behaves in the same way as it
2753 does for delivery: the affixes are removed from the local part unless
2754 rcpt_include_affixes is set on the transport.
2756 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2757 timeout during the DATA phase of an incoming message.
2759 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2760 applied a patch sanctioned by the Sieve maintainer.
2762 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2763 the uid or gid is negative. A case of a negative gid caused this to be
2764 noticed. The fix allows for either to be negative.
2766 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2767 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2770 PH/36 The change PH/12 above was broken. Fixed it.
2772 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2773 the grounds that routing the same address twice would always produce the
2774 same answer. This might have been true once, but it is certainly no
2775 longer true now. Routing a child address may depend on the previous
2776 routing that produced that child. Some complicated redirection strategies
2777 went wrong when messages had multiple recipients, and made Exim's
2778 behaviour dependent on the order in which the addresses were given.
2780 I have moved the duplicate checking until after the routing is complete.
2781 Exim scans the addresses that are assigned to local and remote
2782 transports, and removes any duplicates. This means that more work will be
2783 done, as duplicates will always all be routed, but duplicates are
2784 presumably rare, so I don't expect this is of any significance.
2786 For deliveries to pipes, files, and autoreplies, the duplicate checking
2787 still happens during the routing process, since they are not going to be
2790 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2791 It corrects a timeout issue with spamd. This is Ian's comment: "The
2792 background is that sometimes spamd either never reads data from a
2793 connection it has accepted, or it never writes response data. The exiscan
2794 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2795 blindly assumes that writes won't block so it may never time out."
2797 PH/39 Allow G after quota size as well as K and M.
2799 PH/40 The value set for $authenticated_id in an authenticator may not contain
2800 binary zeroes or newlines because the value is written to log lines and
2801 to spool files. There was no check on this. Now the value is run through
2802 the string_printing() function so that such characters are converted to
2803 printable escape sequences.
2805 PH/41 $message_linecount is a new variable that contains the total number of
2806 lines in the message. Compare $body_linecount, which is the count for the
2809 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2810 bounce and delay warning messages, except in certain special cases, which
2813 (a) An SMTP error message from a remote host;
2814 (b) A message specified in a :fail: redirection;
2815 (c) A message specified in a "fail" command in a system filter;
2816 (d) A message specified in a FAIL return from the queryprogram router;
2817 (e) A message specified by the cannot_route_message router option.
2819 In these cases only, Exim does include the error details in bounce and
2820 warning messages. There are also a few cases where bland messages such
2821 as "unrouteable address" or "local delivery error" are given.
2823 PH/43 $value is now also set for the "else" part of a ${run expansion.
2825 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2826 being worked on, but at least Exim now implements the latest version to
2829 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2830 process to complete was treated as a delivery failure, a timeout while
2831 writing the message to the pipe was logged, but erroneously treated as a
2832 successful delivery. Such timeouts include transport filter timeouts. For
2833 consistency with the overall process timeout, these timeouts are now
2834 treated as errors, giving rise to delivery failures by default. However,
2835 there is now a new Boolean option for the pipe transport called
2836 timeout_defer, which, if set TRUE, converts the failures into defers for
2837 both kinds of timeout. A transport filter timeout is now identified in
2840 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2841 systems where "make" and "gmake" are different, calling "gmake" at top
2842 level broke things. I've arranged for the value of $(MAKE) to be passed
2843 from the Makefile to this script so that it can call the same version of
2847 A note about Exim versions 4.44 and 4.50
2848 ----------------------------------------
2850 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2851 changes of various kinds. As a consequence, a big documentation update was
2852 needed. This delayed the release for rather longer than seemed good, especially
2853 in the light of a couple of (minor) security issues. Therefore, the changes
2854 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2855 release. So 4.44 and 4.50 are in effect two different branches that both start
2858 I have left the 4.50 change log unchanged; it contains all the changes since
2859 4.43. The change log for 4.44 is below; many of its items are identical to
2860 those for 4.50. This seems to be the most sensible way to preserve the
2861 historical information.
2867 1. Minor wording change to the doc/README.SIEVE file.
2869 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2870 computation of the current number of files was incorrect.
2872 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2873 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2876 4. Give more explanation in the error message when the command for a transport
2877 filter fails to execute.
2879 5. There are several places where Exim runs a non-Exim command in a
2880 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2881 was being done only for the command run by the queryprogram router. It is
2882 now done for all such subprocesses. The other cases are: ${run, transport
2883 filters, and the commands run by the lmtp and pipe transports.
2885 6. Added CONFIGURE_GROUP build-time option.
2887 7. Some older OS have a limit of 256 on the maximum number of file
2888 descriptors. Exim was using setrlimit() to set 1000 as a large value
2889 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2890 systems. I've change it so that if it can't get 1000, it tries for 256.
2892 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2893 was an oversight, and furthermore, ever since the addition of extra
2894 controls (e.g. 4.43/32), the checks on when to allow different forms of
2895 "control" were broken. There should now be diagnostics for all cases when a
2896 control that does not make sense is encountered.
2898 9. Added the /retain_sender option to "control=submission".
2900 10. $recipients is now available in the predata ACL (oversight).
2902 11. Tidy the search cache before the fork to do a delivery from a message
2903 received from the command line. Otherwise the child will trigger a lookup
2904 failure and thereby defer the delivery if it tries to use (for example) a
2905 cached ldap connection that the parent has called unbind on.
2907 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2908 of $address_data from the recipient verification was clobbered by the
2909 sender verification.
2911 13. The value of address_data from a sender verification is now available in
2912 $sender_address_data in subsequent conditions in the ACL statement.
2914 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2916 15. Added a new option "connect=<time>" to callout options, to set a different
2919 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2920 was its contents. (It was OK if the option was not defined at all.)
2922 17. A "Completed" log line is now written for messages that are removed from
2923 the spool by the -Mrm option.
2925 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2926 information about exactly what failed.
2928 19. Added -dd to debug only the daemon process.
2930 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2931 handles timeouts, both on the server side and network timeouts. Renamed the
2932 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2934 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2935 It is now set to "smtps".
2937 22. $host_address is now set to the target address during the checking of
2938 ignore_target_hosts.
2940 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2941 being passed; this would have caused $sender_host_name to have been used if
2942 matching the list had actually called for a host name (not very likely,
2943 since this list is usually IP addresses). A host name is now passed as
2946 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2947 code that specifies a non-restarting handler (typically sigaction() in
2948 modern systems) in an attempt to fix a rare and obscure crash bug.
2950 25. Narrowed the window for a race in the daemon that could cause it to ignore
2951 SIGCHLD signals. This is not a major problem, because they are used only to
2952 wake it up if nothing else does.
2954 26. A malformed maildirsize file could cause Exim to calculate negative values
2955 for the mailbox size or file count. Odd effects could occur as a result.
2956 The maildirsize information is now recalculated if the size or filecount
2959 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2960 support for a long time. Removed HAVE_SYS_VFS_H.
2962 28. Installed the latest version of exipick from John Jetmore.
2964 29. In an address list, if the pattern was not a regular expression, an empty
2965 subject address (from a bounce message) matched only if the pattern was an
2966 empty string. Non-empty patterns were not even tested. This was the wrong
2967 because it is perfectly reasonable to use an empty address as part of a
2968 database query. An empty address is now tested by patterns that are
2969 lookups. However, all the other forms of pattern expect the subject to
2970 contain a local part and a domain, and therefore, for them, an empty
2971 address still always fails if the pattern is not itself empty.
2973 30. Exim went into a mad DNS loop when attempting to do a callout where the
2974 host was specified on an smtp transport, and looking it up yielded more
2975 than one IP address.
2977 31. Re-factored the code for checking spool and log partition space into a
2978 function that finds that data and another that does the check. The former
2979 is then used to implement four new variables: $spool_space, $log_space,
2980 $spool_inodes, and $log_inodes.
2982 32. The RFC2047 encoding function was originally intended for short strings
2983 such as real names; it was not keeping to the 75-character limit for
2984 encoded words that the RFC imposes. It now respects the limit, and
2985 generates multiple encoded words if necessary. To be on the safe side, I
2986 have increased the buffer size for the ${rfc2047: expansion operator from
2989 33. It is now permitted to omit both strings after an "if" condition; if the
2990 condition is true, the result is "true". As before, when the second string
2991 is omitted, a false condition yields an empty string. This makes it less
2992 cumbersome to write custom ACL and router conditions.
2994 34. Failure to deliver a bounce message always caused it to be frozen, even if
2995 there was an errors_to setting on the router. The errors_to setting is now
2998 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2999 canonical form (fully expanded) before being placed in
3000 $sender_host_address.
3002 36. The table in the code that translates DNS record types into text (T_A to
3003 "A" for instance) was missing entries for NS and CNAME. It is just possible
3004 that this could have caused confusion if both these types were looked up
3005 for the same domain, because the text type is used as part of Exim's
3006 per-process caching. But the chance of anyone hitting this buglet seems
3009 37. The dnsdb lookup has been extended in a number of ways.
3011 (1) There is a new type, "zns", which walks up the domain tree until it
3012 finds some nameserver records. It should be used with care.
3014 (2) There is a new type, "mxh", which is like "mx" except that it returns
3015 just the host names, not the priorities.
3017 (3) It is now possible to give a list of domains (or IP addresses) to be
3018 looked up. The behaviour when one of the lookups defers can be
3019 controlled by a keyword.
3021 (4) It is now possible to specify the separator character for use when
3022 multiple records are returned.
3024 38. The dnslists ACL condition has been extended: it is now possible to supply
3025 a list of IP addresses and/or domains to be looked up in a particular DNS
3028 39. Added log_selector=+queue_time_overall.
3030 40. When running the queue in the test harness, wait just a tad after forking a
3031 delivery process, to get repeatability of debugging output.
3033 41. Include certificate and key file names in error message when GnuTLS fails
3034 to set them up, because the GnuTLS error message doesn't include the name
3035 of the failing file when there is a problem reading it.
3037 42. Allow both -bf and -bF in the same test run.
3039 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3041 44. The "Exiscan patch" is now merged into the mainline Exim source.
3043 45. Sometimes the final signoff response after QUIT could fail to get
3044 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3045 < 0 before doing a fflush(). This bug looks as though it goes back to the
3046 introduction of TLS in release 3.20, but "sometimes" must have been rare
3047 because the tests only now provoked it.
3049 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3050 (this can affect the format of dates).
3052 47. exim_tidydb, when checking for the continued existence of a message for
3053 which it has found a message-specific retry record, was not finding
3054 messages that were in split spool directories. Consequently, it was
3055 deleting retry records that should have stayed in existence.
3057 48. Steve fixed some bugs in eximstats.
3059 49. The SPA authentication driver was not abandoning authentication and moving
3060 on to the next authenticator when an expansion was forced to fail,
3061 contradicting the general specification for all authenticators. Instead it
3062 was generating a temporary error. It now behaves as specified.
3064 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3065 (the order specifies the preference for clients). The order is now AES256,
3066 AES128, 3DES, ARCFOUR128.
3068 51. Small patch to Sieve code - explicitly set From: when generating an
3071 52. Exim crashed if a remote delivery caused a very long error message to be
3072 recorded - for instance if somebody sent an entire SpamAssassin report back
3073 as a large number of 550 error lines. This bug was coincidentally fixed by
3074 increasing the size of one of Exim's internal buffers (big_buffer) that
3075 happened as part of the Exiscan merge. However, to be on the safe side, I
3076 have made the code more robust (and fixed the comments that describe what
3079 53. Now that there can be additional text after "Completed" in log lines (if
3080 the queue_time_overall log selector is set), a one-byte patch to exigrep
3081 was needed to allow it to recognize "Completed" as not the last thing in
3084 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3085 patch that reportedly fixes this has been added. I am not expert enough to
3086 create a test for it. This is what the patch creator wrote:
3088 "I found a little strange behaviour of ldap code when working with
3089 Windows 2003 AD Domain, where users was placed in more than one
3090 Organization Units. When I tried to give exim partial DN, the exit code
3091 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3092 But simultaneously result of request was absolutely normal ldap result,
3093 so I produce this patch..."
3095 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3096 REFERENCE, so I have modified the code to exclude the patch when that macro
3099 55. Some experimental protocols are using DNS PTR records for new purposes. The
3100 keys for these records are domain names, not reversed IP addresses. The
3101 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3102 leaves it alone. Component reversal etc. now happens only for IP addresses.
3105 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3107 57. Double the size of the debug message buffer (to 2048) so that more of very
3108 long debug lines gets shown.
3110 58. The exicyclog utility now does better if the number of log files to keep
3111 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3113 59. Two changes related to the smtp_active_hostname option:
3115 (1) $smtp_active_hostname is now available as a variable.
3116 (2) The default for smtp_banner uses $smtp_active_hostname instead
3117 of $primary_hostname.
3119 60. The host_aton() function is supposed to be passed a string that is known
3120 to be a valid IP address. However, in the case of IPv6 addresses, it was
3121 not checking this. This is a hostage to fortune. Exim now panics and dies
3122 if the condition is not met. A case was found where this could be provoked
3123 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3124 components; fortuitously, this particular loophole had already been fixed
3125 by change 4.50/55 above.
3127 If there are any other similar loopholes, the new check in host_aton()
3128 itself should stop them being exploited. The report I received stated that
3129 data on the command line could provoke the exploit when Exim was running as
3130 exim, but did not say which command line option was involved. All I could
3131 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3132 running as the user.
3135 61. There was a buffer overflow vulnerability in the SPA authentication code
3136 (which came originally from the Samba project). I have added a test to the
3137 spa_base64_to_bits() function which I hope fixes it.
3140 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3141 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3143 63. The daemon start-up calls getloadavg() while still root for those OS that
3144 need the first call to be done as root, but it missed one case: when
3145 deliver_queue_load_max is set with deliver_drop_privilege. This is
3146 necessary for the benefit of the queue runner, because there is no re-exec
3147 when deliver_drop_privilege is set.
3149 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3150 This has been fixed.
3152 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3153 was in use, was not putting the data itself into the right store pool;
3154 consequently, it could be overwritten for a subsequent message in the same
3155 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3158 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3160 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3161 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3162 the function were treating the return as a boolean value, which happened to
3163 work because 0=false and not-0=true, but is not correct code.
3165 68. The host_aton() function was not handling scoped IPv6 addresses (those
3166 with, for example, "%eth0" on the end) correctly.
3168 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3169 negated items (that is, ~something) in unsigned ints. Some compilers
3170 apparently mutter when there is no cast.
3172 70. If an address verification called from an ACL failed, and did not produce a
3173 user-specific message (i.e. there was only a "system" message), nothing was
3174 put in $acl_verify_message. In this situation, it now puts the system
3177 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3178 session; change 4.31/43 added the unwanted input to the log line - except
3179 that it did not do this in the start of session case. It now does.
3181 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3182 This is wrong and can cause the other end to generate a synchronization
3183 error if it is another Exim or anything else that does the synchronization
3184 check. A QUIT command is no longer sent after a timeout.
3186 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3187 during host lookups.
3189 74. The defer_ok option of callout verification was not working if it was used
3190 when verifying addresses in header lines, that is, for this case:
3192 verify = header_sender/callout=defer_ok
3194 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3195 those file descriptors could be used for SMTP connections. If anything
3196 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3197 could be sent to the SMTP client, causing chaos. The daemon now opens
3198 stdin, stdout, and stderr to /dev/null when it puts itself into the
3201 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3202 log by default. The user can override this with suitable Perl magic.
3204 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3205 the log message when discard triggers, was not working for the DATA ACL or
3206 for the non-SMTP ACL.
3208 78. Error message wording change in sieve.c.
3210 79. If smtp_accept_max_per_host was set, the number of connections could be
3211 restricted to fewer than expected, because the daemon was trying to set up
3212 a new connection before checking whether the processes handling previous
3213 connections had finished. The check for completed processes is now done
3214 earlier. On busy systems, this bug wouldn't be noticed because something
3215 else would have woken the daemon, and it would have reaped the completed
3218 80. If a message was submitted locally by a user whose login name contained one
3219 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3220 It caused a spool format error. I have fixed the spool reading code. A
3221 related problem was that the "from" clause in the Received: line became
3222 illegal because of the space(s). It is now covered by ${quote_local_part.
3224 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3227 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3229 83. Patch from Sieve maintainer to fix unterminated string problem in
3230 "vacation" handling.
3232 84. Some minor changes to the Linux configuration files to help with other
3233 OS variants using glibc.
3235 85. One more patch for Sieve to update vacation handling to latest spec.
3238 ----------------------------------------------------
3239 See the note above about the 4.44 and 4.50 releases.
3240 ----------------------------------------------------
3246 1. Change 4.43/35 introduced a bug that caused file counts to be
3247 incorrectly computed when quota_filecount was set in an appendfile
3250 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3251 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3254 3. Give more explanation in the error message when the command for a transport
3255 filter fails to execute.
3257 4. There are several places where Exim runs a non-Exim command in a
3258 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3259 was being done only for the command run by the queryprogram router. It is
3260 now done for all such subprocesses. The other cases are: ${run, transport
3261 filters, and the commands run by the lmtp and pipe transports.
3263 5. Some older OS have a limit of 256 on the maximum number of file
3264 descriptors. Exim was using setrlimit() to set 1000 as a large value
3265 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3266 systems. I've change it so that if it can't get 1000, it tries for 256.
3268 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3269 was an oversight, and furthermore, ever since the addition of extra
3270 controls (e.g. 4.43/32), the checks on when to allow different forms of
3271 "control" were broken. There should now be diagnostics for all cases when a
3272 control that does not make sense is encountered.
3274 7. $recipients is now available in the predata ACL (oversight).
3276 8. Tidy the search cache before the fork to do a delivery from a message
3277 received from the command line. Otherwise the child will trigger a lookup
3278 failure and thereby defer the delivery if it tries to use (for example) a
3279 cached ldap connection that the parent has called unbind on.
3281 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3282 of $address_data from the recipient verification was clobbered by the
3283 sender verification.
3285 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3286 was its contents. (It was OK if the option was not defined at all.)
3288 11. A "Completed" log line is now written for messages that are removed from
3289 the spool by the -Mrm option.
3291 12. $host_address is now set to the target address during the checking of
3292 ignore_target_hosts.
3294 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3295 being passed; this would have caused $sender_host_name to have been used if
3296 matching the list had actually called for a host name (not very likely,
3297 since this list is usually IP addresses). A host name is now passed as
3300 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3301 code that specifies a non-restarting handler (typically sigaction() in
3302 modern systems) in an attempt to fix a rare and obscure crash bug.
3304 15. Narrowed the window for a race in the daemon that could cause it to ignore
3305 SIGCHLD signals. This is not a major problem, because they are used only to
3306 wake it up if nothing else does.
3308 16. A malformed maildirsize file could cause Exim to calculate negative values
3309 for the mailbox size or file count. Odd effects could occur as a result.
3310 The maildirsize information is now recalculated if the size or filecount
3313 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3314 support for a long time. Removed HAVE_SYS_VFS_H.
3316 18. Updated exipick to current release from John Jetmore.
3318 19. Allow an empty sender to be matched against a lookup in an address list.
3319 Previously the only cases considered were a regular expression, or an
3322 20. Exim went into a mad DNS lookup loop when doing a callout where the
3323 host was specified on the transport, if the DNS lookup yielded more than
3326 21. The RFC2047 encoding function was originally intended for short strings
3327 such as real names; it was not keeping to the 75-character limit for
3328 encoded words that the RFC imposes. It now respects the limit, and
3329 generates multiple encoded words if necessary. To be on the safe side, I
3330 have increased the buffer size for the ${rfc2047: expansion operator from
3333 22. Failure to deliver a bounce message always caused it to be frozen, even if
3334 there was an errors_to setting on the router. The errors_to setting is now
3337 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3338 canonical form (fully expanded) before being placed in
3339 $sender_host_address.
3341 24. Updated eximstats to version 1.33
3343 25. Include certificate and key file names in error message when GnuTLS fails
3344 to set them up, because the GnuTLS error message doesn't include the name
3345 of the failing file when there is a problem reading it.
3347 26. Expand error message when OpenSSL has problems setting up cert/key files.
3350 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3351 (this can affect the format of dates).
3353 28. exim_tidydb, when checking for the continued existence of a message for
3354 which it has found a message-specific retry record, was not finding
3355 messages that were in split spool directories. Consequently, it was
3356 deleting retry records that should have stayed in existence.
3358 29. eximstats updated to version 1.35
3359 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3360 1.35 - bugfix such that pie charts by volume are generated correctly
3362 30. The SPA authentication driver was not abandoning authentication and moving
3363 on to the next authenticator when an expansion was forced to fail,
3364 contradicting the general specification for all authenticators. Instead it
3365 was generating a temporary error. It now behaves as specified.
3367 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3368 (the order specifies the preference for clients). The order is now AES256,
3369 AES128, 3DES, ARCFOUR128.
3371 31. Small patch to Sieve code - explicitly set From: when generating an
3374 32. Exim crashed if a remote delivery caused a very long error message to be
3375 recorded - for instance if somebody sent an entire SpamAssassin report back
3376 as a large number of 550 error lines. This bug was coincidentally fixed by
3377 increasing the size of one of Exim's internal buffers (big_buffer) that
3378 happened as part of the Exiscan merge. However, to be on the safe side, I
3379 have made the code more robust (and fixed the comments that describe what
3382 33. Some experimental protocols are using DNS PTR records for new purposes. The
3383 keys for these records are domain names, not reversed IP addresses. The
3384 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3385 leaves it alone. Component reversal etc. now happens only for IP addresses.
3388 34. The host_aton() function is supposed to be passed a string that is known
3389 to be a valid IP address. However, in the case of IPv6 addresses, it was
3390 not checking this. This is a hostage to fortune. Exim now panics and dies
3391 if the condition is not met. A case was found where this could be provoked
3392 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3393 components; fortuitously, this particular loophole had already been fixed
3394 by change 4.50/55 or 4.44/33 above.
3396 If there are any other similar loopholes, the new check in host_aton()
3397 itself should stop them being exploited. The report I received stated that
3398 data on the command line could provoke the exploit when Exim was running as
3399 exim, but did not say which command line option was involved. All I could
3400 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3401 running as the user.
3404 35. There was a buffer overflow vulnerability in the SPA authentication code
3405 (which came originally from the Samba project). I have added a test to the
3406 spa_base64_to_bits() function which I hope fixes it.
3409 36. The daemon start-up calls getloadavg() while still root for those OS that
3410 need the first call to be done as root, but it missed one case: when
3411 deliver_queue_load_max is set with deliver_drop_privilege. This is
3412 necessary for the benefit of the queue runner, because there is no re-exec
3413 when deliver_drop_privilege is set.
3415 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3416 was in use, was not putting the data itself into the right store pool;
3417 consequently, it could be overwritten for a subsequent message in the same
3418 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3421 38. Sometimes the final signoff response after QUIT could fail to get
3422 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3423 < 0 before doing a fflush(). This bug looks as though it goes back to the
3424 introduction of TLS in release 3.20, but "sometimes" must have been rare
3425 because the tests only now provoked it.
3431 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3432 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3433 Now it returns an integer. A number of calls were still expecting a T/F
3434 return. Fortuitously, in all cases, the tests worked in OK situations,
3435 which is the norm. However, things would have gone wrong on any write
3436 failures on the smtp file descriptor. This function is used when sending
3437 messages over SMTP and also when doing verify callouts.
3439 2. When Exim is called to do synchronous delivery of a locally submitted
3440 message (the -odf or -odi options), it no longer closes stderr before doing
3443 3. Implemented the mua_wrapper option.
3445 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3447 5. Implemented the functions header_remove(), header_testname(),
3448 header_add_at_position(), and receive_remove_recipient(), and exported them
3451 6. If an ACL "warn" statement specified the addition of headers, Exim already
3452 inserted X-ACL-Warn: at the start if there was no header name. However, it
3453 was not making this test for the second and subsequent header lines if
3454 there were newlines in the string. This meant that an invalid header could
3455 be inserted if Exim was badly configured.
3457 7. Allow an ACL "warn" statement to add header lines at the start or after all
3458 the Received: headers, as well as at the end.
3460 8. Added the rcpt_4xx retry error code.
3462 9. Added postmaster_mailfrom=xxx to callout verification option.
3464 10. Added mailfrom=xxxx to the callout verification option, for verify=
3467 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3468 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3470 12. Inserted some casts to stop certain compilers warning when using pointer
3471 differences as field lengths or precisions in printf-type calls (mostly
3472 affecting debugging statements).
3474 13. Added optional readline() support for -be (dynamically loaded).
3476 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3477 same clock tick as a message's arrival, so that its received time was the
3478 same as the "first fail" time on the retry record, and that message
3479 remained on the queue past the ultimate address timeout, every queue runner
3480 would try a delivery (because it was past the ultimate address timeout) but
3481 after another failure, the ultimate address timeout, which should have then
3482 bounced the address, did not kick in. This was a "< instead of <=" error;
3483 in most cases the first failure would have been in the next clock tick
3484 after the received time, and all would be well.
3486 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3487 being recognized when the domain list was tested by the match_domain
3488 condition in an expansion string.
3490 16. Added the ${str2b64: operator.
3492 17. Exim was always calling setrlimit() to set a large limit for the number of
3493 processes, without checking whether the existing limit was already
3494 adequate. (It did check for the limit on file descriptors.) Furthermore,
3495 errors from getrlimit() and setrlimit() were being ignored. Now they are
3496 logged to the main and panic logs, but Exim does carry on, to try to do its
3497 job under whatever limits there are.
3499 18. Imported PCRE 5.0.
3501 19. Trivial typo in log message " temporarily refused connection" (the leading
3504 20. If the log selector return_path_on_delivery was set and an address was
3505 redirected to /dev/null, the delivery process crashed because it assumed
3506 that a return path would always be set for a "successful" delivery. In this
3507 case, the whole delivery is bypassed as an optimization, and therefore no
3510 21. Internal re-arrangement: the function for sending a challenge and reading
3511 a response while authentication was assuming a zero-terminated challenge
3512 string. It's now changed to take a pointer and a length, to allow for
3513 binary data in such strings.
3515 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3517 23. Exim was not respecting finduser_retries when seeking the login of the
3518 uid under which it was called; it was always trying 10 times. (The default
3519 setting of finduser_retries is zero.) Also, it was sleeping after the final
3520 failure, which is pointless.
3522 24. Implemented tls_on_connect_ports.
3524 25. Implemented acl_smtp_predata.
3526 26. If the domain in control=submission is set empty, Exim assumes that the
3527 authenticated id is a complete email address when it generates From: or
3528 Sender: header lines.
3530 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3531 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3532 chown and chgrp in /bin and hostname in /usr/bin.
3534 28. Exim was keeping the "process log" file open after each use, just as it
3535 does for the main log. This opens the possibility of it remaining open for
3536 long periods when the USR1 signal hits a daemon. Occasional processlog
3537 errors were reported, that could have been caused by this. Anyway, it seems
3538 much more sensible not to leave this file open at all, so that is what now
3541 29. The long-running daemon process does not normally write to the log once it
3542 has entered its main loop, and it closes the log before doing so. This is
3543 so that log files can straightforwardly be renamed and moved. However,
3544 there are a couple of unusual error situations where the daemon does write
3545 log entries, and I had neglected to close the log afterwards.
3547 30. The text of an SMTP error response that was received during a remote
3548 delivery was being truncated at 512 bytes. This is too short for some of
3549 the long messages that one sometimes sees. I've increased the limit to
3552 31. It is now possible to make retry rules that apply only when a message has a
3553 specific sender, in particular, an empty sender.
3555 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3556 it possible to be selective about when SMTP synchronization is enforced.
3558 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3560 32. Implemented hosts_connection_nolog.
3562 33. Added an ACL for QUIT.
3564 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3567 35. Added mailbox_size and mailbox_filecount to appendfile.
3569 36. Added control = no_multiline_responses to ACLs.
3571 37. There was a bug in the logic of the code that waits for the clock to tick
3572 in the case where the clock went backwards by a substantial amount such
3573 that the microsecond fraction of "now" was more than the microsecond
3574 fraction of "then" (but the whole seconds number was less).
3576 38. Added support for the libradius Radius client library this is found on
3577 FreeBSD (previously only the radiusclient library was supported).
3583 1. When certain lookups returned multiple values in the form name=value, the
3584 quoting of the values was not always being done properly. Specifically:
3585 (a) If the value started with a double quote, but contained no whitespace,
3587 (b) If the value contained whitespace other than a space character (i.e.
3588 tabs or newlines or carriage returns) it was not quoted.
3589 This fix has been applied to the mysql and pgsql lookups by writing a
3590 separate quoting function and calling it from the lookup code. The fix
3591 should probably also be applied to nisplus, ibase and oracle lookups, but
3592 since I cannot test any of those, I have not disturbed their existing code.
3594 2. A hit in the callout cache for a specific address caused a log line with no
3595 reason for rejecting RCPT. Now it says "Previous (cached) callout
3596 verification failure".
3598 3. There was an off-by-one bug in the queryprogram router. An over-long
3599 return line was truncated at 256 instead of 255 characters, thereby
3600 overflowing its buffer with the terminating zero. As well as fixing this, I
3601 have increased the buffer size to 1024 (and made a note to document this).
3603 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3604 when Exim is waiting for an SMTP response from a remote server, Exim
3605 restarts its select() call on the socket, thereby resetting its timeout.
3606 This is not a problem when such interrupts are rare. Somebody set up a cron
3607 job to run exiwhat every 2 minutes, which is less than the normal select()
3608 timeout (5 or 10 minutes). This meant that the select() timeout never
3609 kicked in because it was always reset. I have fixed this by comparing the
3610 time when an interrupt arrives with the time at the start of the first call
3611 to select(). If more time than the timeout has elapsed, the interrupt is
3612 treated as a timeout.
3614 5. Some internal re-factoring in preparation for the addition of Sieve
3615 extensions (by MH). In particular, the "personal" test is moved to a
3616 separate function, and given an option for scanning Cc: and Bcc: (which is
3617 not set for Exim filters).
3619 6. When Exim created an email address using the login of the caller as the
3620 local part (e.g. when creating a From: or Sender: header line), it was not
3621 quoting the local part when it contained special characters such as @.
3623 7. Installed new OpenBSD configuration files.
3625 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3626 try to make them clearer.
3628 9. Callout options, other than the timeout value, were being ignored when
3629 verifying sender addresses in header lines. For example, when using
3631 verify = header_sender/callout=no_cache
3633 the cache was (incorrectly) being used.
3635 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3636 only the Cygwin environment.
3638 11. When return_path_on_delivery was set as a log selector, if different remote
3639 addresses in the same message used different return paths and parallel
3640 remote delivery occurred, the wrong values would sometimes be logged.
3641 (Whenever a remote delivery process finished, the return path value from
3642 the most recently started remote delivery process was logged.)
3644 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3645 lines when AUTH and/or TLS are in use. This is the "received protocol"
3646 field. Exim used to use "asmtp" for authenticated SMTP, without any
3647 indication (in the protocol name) for TLS use. Now it follows the RFC and
3648 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3649 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3650 names appear in log lines as well as in Received: header lines.
3652 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3653 extensions, and comparison tests, and to fix some bugs.
3655 14. Changes to the "personal" filter test:
3657 (1) The test was buggy in that it was just doing the equivalent of
3658 "contains" tests on header lines. For example, if a user's address was
3659 anne@some.where, the "personal" test would incorrectly be true for
3661 To: susanne@some.where
3663 This test is now done by extracting each address from the header in turn,
3664 and checking the entire address. Other tests that are part of "personal"
3665 are now done using regular expressions (for example, to check local parts
3666 of addresses in From: header lines).
3668 (2) The list of non-personal local parts in From: addresses has been
3669 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3670 taken from the Sieve specification recommendations.
3672 (3) If the message contains any header line starting with "List-" it is
3673 treated as non-personal.
3675 (4) The test for "circular" in the Subject: header line has been removed
3676 because it now seems ill-conceived.
3678 15. Minor typos in src/EDITME comments corrected.
3680 16. Installed latest exipick from John Jetmore.
3682 17. If headers_add on a router specified a text string that was too long for
3683 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3684 of string_sprintf() is now avoided.
3686 18. $message_body_size was not set (it was always zero) when running the DATA
3687 ACL and the local_scan() function.
3689 19. For the "mail" command in an Exim filter, no default was being set for
3690 the once_repeat time, causing a random time value to be used if "once" was
3691 specified. (If the value happened to be <= 0, no repeat happened.) The
3692 default is now 0s, meaning "never repeat". The "vacation" command was OK
3693 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3694 (I found it when inspecting the code).
3696 20. There is now an overall timeout for performing a callout verification. It
3697 defaults to 4 times the callout timeout, which applies to individual SMTP
3698 commands during the callout. The overall timeout applies when there is more
3699 than one host that can be tried. The timeout is checked before trying the
3700 next host. This prevents very long delays if there are a large number of
3701 hosts and all are timing out (e.g. when the network connections are timing
3702 out). The value of the overall timeout can be changed by specifying an
3703 additional sub-option for "callout", called "maxwait". For example:
3705 verify = sender/callout=5s,maxwait=20s
3707 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3708 to the end before writing, but this should make it even safer).
3710 22. Exim was forgetting that it had advertised PIPELINING for the second and
3711 subsequent messages on an SMTP connection. It was also not resetting its
3712 memory on STARTTLS and an internal HELO.
3714 23. When Exim logs an SMTP synchronization error within a session, it now
3715 records whether PIPELINING has been advertised or not.
3717 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3718 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3719 rather than long int.
3721 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3723 26. Added the never_mail option to autoreply.
3729 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3730 crash if the getsockname() call failed; this can happen if a connection is
3731 closed very soon after it is established. The problem was simply in the
3732 order in which certain operations were done, causing Exim to try to write
3733 to the SMTP stream before it had set up the file descriptor. The bug has
3734 been fixed by making things happen in the correct order.
3740 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3741 before the connection was closed, thus losing the rejection response.
3743 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3744 some early Solaris releases, but causes trouble in current releases where
3745 socklen_t is defined.
3747 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3750 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3753 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3754 that releasing the top of it at the end releases what was used for sub-
3755 expansions (unless the block got too big). However, discard this block if
3756 the first thing is a variable or header, so that we can use its block when
3757 it is dynamic (useful for very large $message_headers, for example).
3759 6. Lookups now cache *every* query, not just the most recent. A new, separate
3760 store pool is used for this. It can be recovered when all lookup caches are
3761 flushed. Lookups now release memory at the end of their result strings.
3762 This has involved some general refactoring of the lookup sources.
3764 7. Some code has been added to the store_xxx() functions to reduce the amount
3765 of flapping under certain conditions.
3767 8. log_incoming_interface used to affect only the <= reception log lines. Now
3768 it causes the local interface and port to be added to several more SMTP log
3769 lines, for example "SMTP connection from", and rejection lines.
3771 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3773 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3775 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3776 could be overwritten at the end of the current message (or the start of a
3777 new message if it was set in a HELO ACL). The value is now preserved for
3778 the duration of the SMTP connection.
3780 12. If a transport had a headers_rewrite setting, and a matching header line
3781 contained an unqualified address, that address was qualified, even if it
3782 did not match any rewriting rules. The underlying bug was that the values
3783 of the flags that permit the existence of unqualified sender and recipient
3784 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3785 non-local messages, and by -bnq for local messages) were not being
3786 preserved with the message after it was received.
3788 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3789 "next input=" as part of the text comprising the host identity instead of
3790 the correct text. The code was using the same buffer for two different
3791 strings. However, depending on which order the printing function evaluated
3792 its arguments, the bug did not always show up. Under Linux, for example, my
3793 test suite worked just fine.
3795 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3796 doesn't work with some older versions of Perl. It has been changed to "my",
3797 which in any case is probably the better facility to use.
3799 15. A really picky compiler found some instances of statements for creating
3800 error messages that either had too many or two few arguments for the format
3803 16. The size of the buffer for calls to the DNS resolver has been increased
3804 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3805 for addresses that have a lot of PTR records. This alleviates a problem; it
3806 does not fully solve it.
3808 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3809 buffer now truncates the list and logs the incident, which is the same
3810 action as happens when Exim is looking up a host name and its aliases.
3811 Previously in this situation something unpredictable would happen;
3812 sometimes it was "internal error: store_reset failed".
3814 18. If a server dropped the connection unexpectedly when an Exim client was
3815 using GnuTLS and trying to read a response, the client delivery process
3816 crashed while trying to generate an error log message.
3818 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3819 string, for example:
3821 warn message = H1: something\nH2: something
3823 the text was added as a single header line from Exim's point of view
3824 though it ended up OK in the delivered message. However, searching for the
3825 second and subsequent header lines using $h_h2: did not work. This has been
3826 fixed. Similarly, if a system filter added multiple headers in this way,
3827 the routers could not see them.
3829 20. Expanded the error message when iplsearch is called with an invalid key to
3830 suggest using net-iplsearch in a host list.
3832 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3833 ACLs are no longer actually imposed (and a message to that effect is
3836 22. If a "gecos" field in a passwd entry contained escaped characters, in
3837 particular, if it contained a \" sequence, Exim got it wrong when building
3838 a From: or a Sender: header from that name. A second bug also caused
3839 incorrect handling when an unquoted " was present following a character
3840 that needed quoting.
3842 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3843 was not being matched caselessly.
3845 24. Arranged for all hyphens in the exim.8 source to be escaped with
3848 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3849 properly. Recipient callout cache records were still being keyed to include
3850 the sender, even when use_sender was set false. This led to far more
3851 callouts that were necessary. The sender is no longer included in the key
3852 when use_sender is false.
3854 26. Added "control = submission" modifier to ACLs.
3856 27. Added the ${base62d: operator to decode base 62 numbers.
3858 28. dnsdb lookups can now access SRV records.
3860 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3861 the configuration file.
3863 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3864 (-v) mode. This makes the output for a verbose queue run more intelligible.
3866 31. Added a use_postmaster feature to recipient callouts.
3868 32. Added the $body_zerocount variable, containing the number of binary zero
3869 bytes in the message body.
3871 33. The time of last modification of the "new" subdirectory is now used as the
3872 "mailbox time last read" when there is a quota error for a maildir
3875 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3877 35. Added +ignore_unknown as a special item in host lists.
3879 36. Code for decoding IPv6 addresses in host lists is now included, even if
3880 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3881 address was recognized as an IP address, but was then not correctly decoded
3882 into binary, causing unexpected and incorrect effects when compared with
3889 1. Very minor rewording of debugging text in manualroute to say "list of
3890 hosts" instead of "hostlist".
3892 2. If verify=header_syntax was set, and a header line with an unqualified
3893 address (no domain) and a large number of spaces between the end of the
3894 name and the colon was received, the reception process suffered a buffer
3895 overflow, and (when I tested it) crashed. This was caused by some obsolete
3896 code that should have been removed. The fix is to remove it!
3898 3. When running in the test harness, delay a bit after writing a bounce
3899 message to get a bit more predictability in the log output.
3901 4. Added a call to search_tidyup() just before forking a reception process. In
3902 theory, someone could use a lookup in the expansion of smtp_accept_max_
3903 per_host which, without the tidyup, could leave open a database connection.
3905 5. Added the variables $recipient_data and $sender_data which get set from a
3906 lookup success in an ACL "recipients" or "senders" condition, or a router
3907 "senders" option, similar to $domain_data and $local_part_data.
3909 6. Moved the writing of debug_print from before to after the "senders" test
3912 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3913 problems for message scanning, either using a data ACL, or using
3914 local_scan() because the Received: header was not generated till after they
3915 were called (in order to set the time as the time of reception completion).
3916 I have revised the way this works. The header is now generated after the
3917 body is received, but before the ACL or local_scan() are called. After they
3918 are run, the timestamp in the header is updated.
3924 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3925 before starting a queue runner without re-exec. This happened only when
3926 deliver_drop_privilege was set or when the Exim user was set to root. The
3927 effect of the bug was that timeouts during subsequent deliveries caused
3928 crashes instead of being properly handled. The handler is now left at its
3929 default (and expected) setting.
3931 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3932 message, again when deliver_drop_privilege is set or Exim is run as root.
3933 The bug described in (1) was not present in this case, but the tidying up
3934 of the other signals was missing. I have made the two cases consistent.
3936 3. The ignore_target_hosts setting on a manualroute router was being ignored
3937 for hosts that were looked up using the /MX notation.
3939 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3942 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3943 operated on the sender address. After changing the $sender_address to <>
3944 for the sender address verify, Exim was re-instated it as the original
3945 (before rewriting) address, but remembering that it had rewritten it, so it
3946 wasn't rewriting it again. This bug also had the effect of breaking the
3947 sender address verification caching when the sender address was rewritten.
3949 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3950 This has been changed so that if the ip literal address matches
3951 ignore_target_hosts, the router declines.
3953 7. Added expansion conditions match_domain, match_address, and match_local_
3954 part (NOT match_host).
3956 8. The placeholder for the Received: header didn't have a length field set.
3958 9. Added code to Exim itself and to exim_lock to test for a specific race
3959 condition that could lead to file corruption when using MBX delivery. The
3960 issue is with the lockfile that is created in /tmp. If this file is removed
3961 after a process has opened it but before that process has acquired a lock,
3962 there is the potential for a second process to recreate the file and also
3963 acquire a lock. This could lead to two Exim processes writing to the file
3964 at the same time. The added code performs the same test as UW imapd; it
3965 checks after acquiring the lock that its file descriptor still refers to
3966 the same named file.
3968 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3969 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3972 11. Added the smtp_active_hostname option. If used, this will typically be made
3973 to depend on the incoming interface address. Because $interface_address is
3974 not set up until the daemon has forked a reception process, error responses
3975 that can happen earlier (such as "too many connections") no longer contain
3978 12. If an expansion in a condition on a "warn" statement fails because a lookup
3979 defers, the "warn" statement is abandoned, and the next ACL statement is
3980 processed. Previously this caused the whole ACL to be aborted.
3982 13. Added the iplsearch lookup type.
3984 14. Added ident_timeout as a log selector.
3986 15. Added tls_certificate_verified as a log selector.
3988 16. Added a global option tls_require_ciphers (compare the smtp transport
3989 option of the same name). This controls incoming TLS connections.
3991 17. I finally figured out how to make tls_require_ciphers do a similar thing
3992 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3993 before starting the TLS session.
3995 18. Tabs are now shown as \t in -bP output.
3997 19. If the log selector return_path_on_delivery was set, Exim crashed when
3998 bouncing a message because it had too many Received: header lines.
4000 20. If two routers both had headers_remove settings, and the first one included
4001 a superfluous trailing colon, the final name in the first list and the
4002 first name in the second list were incorrectly joined into one item (with a
4003 colon in the middle).
4009 1. Added -C and -D options to the exinext utility, mainly to make it easier
4010 to include in the automated testing, but these could be helpful when
4011 multiple configurations are in use.
4013 2. The exinext utility was not formatting the output nicely when there was
4014 an alternate port involved in the retry record key, nor when there was a
4015 message id as well (for retries that were specific to a specific message
4016 and a specific host). It was also confused by IPv6 addresses, because of
4017 the additional colons they contain. I have fixed the IPv4 problem, and
4018 patched it up to do a reasonable job for IPv6.
4020 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4021 delivery, the log line now contains "pipelined" if PIPELINING was used.
4023 4. An SMTP transport process used to panic and die if the bind() call to set
4024 an explicit outgoing interface failed. This has been changed; it is now
4025 treated in the same way as a connect() failure.
4027 5. A reference to $sender_host_name in the part of a conditional expansion
4028 that was being skipped was still causing a DNS lookup. This no longer
4031 6. The def: expansion condition was not recognizing references to header lines
4032 that used bh_ and bheader_.
4034 7. Added the _cache feature to named lists.
4036 8. The code for checking quota_filecount in the appendfile transport was
4037 allowing one more file than it should have been.
4039 9. For compatibility with Sendmail, the command line option
4047 and sets the incoming protocol and host name (for trusted callers). The
4048 host name and its colon can be omitted when only the protocol is to be set.
4049 Note the Exim already has two private options, -pd and -ps, that refer to
4050 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4051 "s", but I don't think that's a major issue.
4053 10. A number of refactoring changes to the code, none of which should affect
4056 (a) The number of logging options was getting close to filling up the
4057 32-bit word that was used as a bit map. I have split them into two classes:
4058 those that are passed in the argument to log_write(), and those that are
4059 only ever tested independently outside of that function. These are now in
4060 separate 32-bit words, so there is plenty of room for expansion again.
4061 There is no change in the user interface or the logging behaviour.
4063 (b) When building, for example, log lines, the code previously used a
4064 macro that called string_cat() twice, in order to add two strings. This is
4065 not really sufficiently general. Furthermore, there was one instance where
4066 it was actually wrong because one of the argument was used twice, and in
4067 one call a function was used. (As it happened, calling the function twice
4068 did not affect the overall behaviour.) The macro has been replaced by a
4069 function that can join an arbitrary number of extra strings onto a growing
4072 (c) The code for expansion conditions now uses a table and a binary chop
4073 instead of a serial search (which was left over from when there were very
4074 few conditions). Also, it now recognizes conditions like "pam" even when
4075 the relevant support is not compiled in: a suitably worded error message is
4076 given if an attempt is made to use such a condition.
4078 11. Added ${time_interval:xxxxx}.
4080 12. A bug was causing one of the ddress fields not to be passed back correctly
4081 from remote delivery subprocesses. The field in question was not being
4082 subsequently used, so this caused to problems in practice.
4084 13. Added new log selectors queue_time and deliver_time.
4086 14. Might have fixed a bug in maildirsizefile handling that threw up
4087 "unexpected character" debug warnings, and recalculated the data
4088 unnecessarily. In any case, I expanded the warning message to give more
4091 15. Added the message "Restricted characters in address" to the statements in
4092 the default ACL that block characters like @ and % in local parts.
4094 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4095 Three changes have been made:
4097 (a) There was a serious bug; a negative response to MAIL caused the whole
4098 recipient domain to be cached as invalid, thereby blocking all messages
4099 to all local parts at the same domain, from all senders. This bug has
4100 been fixed. The domain is no longer cached after a negative response to
4101 MAIL if the sender used is not empty.
4103 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4106 (c) A new callout option, "use_sender" has been added for people who want
4107 the modified behaviour.
4113 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4116 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4117 indeed breaks things for older releases.
4119 3. Added additional logging to the case where there is a problem reading data
4120 from a filter that is running in a subprocess using a pipe, in order to
4121 try to track down a specific problem.
4123 4. Testing facility fudge: when running in the test harness and attempting
4124 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4125 getting "No route to host". Convert this to a timeout.
4127 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4130 6. Some OS don't have socklen_t but use size_t instead. This affects the
4131 fifth argument of getsockopt() amongst other things. This is now
4132 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4133 can be set for individual OS. I have set it for SunOS5, OSF1, and
4134 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4135 some earlier ones do not.
4137 7. Change 4.30/15 was not doing the test caselessly.
4139 8. The standard form for an IPv6 address literal was being rejected by address
4140 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4141 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4142 this, as well as the form without the "IPv6" on the front (but only when
4143 address literals are enabled, of course).
4145 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4147 10. Exim crashed if a message with an empty sender address specified by -f
4148 encountered a router with an errors_to setting. This could be provoked only
4149 by a command such as
4153 where an empty string was supplied; "<>" did not hit this bug.
4155 11. Installed PCRE release 4.5.
4157 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4158 remained set. It is now erased.
4160 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4161 times from message ids (which are base 36 rather than the normal 62).
4163 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4164 were being counted as actual protocol errors, and logged if the log
4165 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4166 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4167 and DATA following a set of rejected RCPTs do not count as protocol errors.
4168 In other words, Exim assumes they were pipelined, though this may not
4169 actually be the case. Of course, in all cases the client gets an
4170 appropriate error code.
4172 15. If a lookup fails in an ACL condition, a message about the failure may
4173 be available; it is used if testing the ACL cannot continue, because most
4174 such messages specify what the cause of the deferral is. However, some
4175 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4176 that caused an old message to be retained and used if a later statement
4177 caused a defer, replacing the real cause of the deferral.
4179 16. If an IP address had so many PTR records that the DNS lookup buffer
4180 was not large enough to hold them, Exim could crash while trying to process
4181 the truncated data. It now detects and logs this case.
4183 17. Further to 4.21/58, another change has been made: if (and only if) the
4184 first line of a message (the first header line) ends with CRLF, a bare LF
4185 in a subsequent header line has a space inserted after it, so as not to
4186 terminate the header.
4188 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4189 unnecessarily, used atoi() instead of strtol(), and didn't check the
4190 termination when getting file sizes from file names by regex.
4192 19. Completely re-implemented the support for maildirsize files, in the light
4193 of a number of problems with the previous contributed implementation
4194 (4.30/29). In particular:
4196 . If the quota is zero, the maildirsize file is maintained, but no quota is
4199 . If the maildir directory does not exist, it is created before any attempt
4200 to write a maildirsize file.
4202 . The quota value in the file is just a cache; if the quota is changed in
4203 the transport, the new value overrides.
4205 . A regular expression is available for excluding directories from the
4208 20. The autoreply transport checks the characters in options that define the
4209 message's headers; it allows continued headers, but it was checking with
4210 isspace() after an embedded newline instead of explicitly looking for a
4213 21. If all the "regular" hosts to which an address was routed had passed their
4214 expiry times, and had not reached their retry times, the address was
4215 bounced, even if fallback hosts were defined. Now Exim should go on to try
4218 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4219 equivalent code in the SMTP transport. Some hosts send humungous responses
4220 to HELO/EHLO, more than 1024 it seems.
4222 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4223 gives compiler warnings in some environments. I've now done it "properly",
4226 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4227 (because of problems with the built-in one) was declared to return uschar *
4228 instead of char *, causing compiler failure.
4230 25. Fixed a file descriptor leak when processing alias/forward files.
4232 26. Fixed a minor format string issue in dbfn.c.
4234 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4236 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4237 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4240 29. When a sender address is verified, it is cached, to save repeating the test
4241 when there is more than one recipient in a message. However, when the
4242 verification involves a callout, it is possible for different callout
4243 options to be set for different recipients. It is too complicated to keep
4244 track of this in the cache, so now Exim always runs a verification when a
4245 callout is required, relying on the callout cache for the optimization.
4246 The overhead is duplication of the address routing, but this should not be
4249 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4250 to be verified with callout=postmaster, and the main callout worked but the
4251 postmaster check failed, the verification correctly failed. However, if a
4252 subsequent RCPT command asked for sender verification *without* the
4253 postmaster check, incorrect caching caused this verification also to fail,
4256 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4257 it was not caching the DNS options (qualify_single, search_parents) that
4258 were used when the lookup failed. A subsequent lookup with different
4259 options therefore always gave the same answer, though there were cases
4260 where it should not have. (Example: a "domains = !$mx_any" option on a
4261 dnslookup router: the "domains" option is always processed without any
4262 widening, but the router might have qualify_single set.) Now Exim uses the
4263 cached value only when the same options are set.
4265 32. Added John Jetmore's "exipick" utility to the distribution.
4267 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4268 than a timeout (e.g. a certificate is required, and is not provided), an
4269 Exim server now closes the connection immediately. Previously it waited for
4270 the client to close - but if the client is SSL, it seems that they each
4271 wait for each other, leading to a delay before one of them times out.
4273 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4274 maintained 0.8.x compatibility because I don't think many are using it, and
4275 it is clearly obsolete.
4277 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4280 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4281 client certificate was expired. A simple patch fixes this, though I don't
4282 understand the full logic of why the verify callback is called multiple
4285 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4286 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4287 which causes problems with some clients (such as the Certicom SSL Plus
4288 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4289 disables the coutermeasure allowing Eudora to connect."
4291 38. Exim was not checking that a write() to a log file succeeded. This could
4292 lead to Bad Things if a log got too big, in particular if it hit a file
4293 size limit. Exim now panics and dies if it cannot write to a log file, just
4294 as it does if it cannot open a log file.
4296 39. Modified OS/Makefile-Linux so that it now contains
4298 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4300 The two -D definitions ensure that Exim is compiled with large file
4301 support, which makes it possible to handle log files that are bigger than
4304 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4305 instance) a domain was checked against a named list that involved a lookup,
4306 causing $domain_data to be set, then another domain was checked against the
4307 same list, then the first domain was re-checked, the value of $domain_data
4308 after the final check could be wrong. In particular, if the second check
4309 failed, it could be set empty. This bug probably also applied to
4312 41. The strip_trailing_dot option was not being applied to the address given
4313 with the -f command-line option.
4315 42. The code for reading a message's header from the spool was incrementing
4316 $received_count, but never initializing it. This meant that the value was
4317 incorrect (doubled) while delivering a message in the same process in which
4318 it was received. In the most common configuration of Exim, this never
4319 happens - a fresh exec is done - but it can happen when
4320 deliver_drop_privilege is set.
4322 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4323 it now includes up to 150 characters of the unexpected data in the log
4326 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4327 and building data strings. The size of both of these buffers was 10 000
4328 bytes - far larger than anybody would *ever* want, thought I. Needless to
4329 say, somebody hit the limit. I have increased the maximum line length to
4330 20 000 and the maximum data length of concatenated lines to 100 000. I have
4331 also fixed two bugs, because there was no checking on these buffers. Tsk,
4332 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4333 buffer is too small.
4335 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4336 lsearch lookups. Now it does.
4338 46. When parsing a route_list item in a manualroute router, a fixed-length
4339 buffer was used for the list of hosts. I made this 1024 bytes long,
4340 thinking that nobody would ever have a list of hosts that long. Wrong.
4341 Somebody had a whole pile of complicated expansion conditions, and the
4342 string was silently truncated, leading to an expansion error. It turns out
4343 that it is easier to change to an unlimited length (owing to other changes
4344 that have happened since this code was originally written) than to build
4345 structure for giving a limitation error. The length of the item that
4346 expands into the list of hosts is now unlimited.
4348 47. The lsearch lookup could not handle data where the length of text line was
4349 more than 4095 characters. Such lines were truncated, leading to shortened
4350 data being returned. It should now handle lines of any length.
4352 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4353 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4356 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4357 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4358 no longer happens in comment lines. A list of replacements is now placed
4359 at the head of all of the source files, except those whose only change is
4360 to replace PERL_COMMAND in the very first #! line.
4362 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4363 messages on the queue, with a bottom-up merge sort, using code contributed
4364 by Michael Haardt. This should make operations like -bp somewhat faster on
4365 large queues. It won't affect queue runners, except when queue_run_in_order
4368 51. Installed eximstats 1.31 in the distribution.
4370 52. Added support for SRV lookups to the dnslookup router.
4372 53. If an ACL referred to $message_body or $message_body_end, the value was not
4373 reset for any messages that followed in the same SMTP session.
4375 54. The store-handling optimization for building very long strings was not
4376 differentiating between the different store pools. I don't think this
4377 actually made any difference in practice, but I've tidied it.
4379 55. While running the routers to verify a sender address, $sender_address
4380 was still set to the sender address. This is wrong, because when routing to
4381 send a bounce to the sender, it would be empty. Therefore, I have changed
4382 it so that, while verifying a sender address, $sender_address is set to <>.
4383 (There is no change to what happens when verifying a recipient address.)
4385 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4386 target A or AAAA records (if not already returned) without resetting the
4387 qualify_single or search_parents options of the DNS resolver. These are
4388 inappropriate in this case because the targets of MX and SRV records must
4389 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4390 target that, when qualified, matched something in the local domain. These
4391 two options are now turned off when doing these lookups.
4393 57. It seems that at least some releases of Reiserfs (which does not have the
4394 concept of a fixed number of inodes) returns zero and not -1 for the
4395 number of available inodes. This interacted badly with check_spool_inodes,
4396 which assumed that -1 was the "no such thing" setting. What I have done is
4397 to check that the total number of inodes is greater than zero before doing
4398 the test of how many are available.
4400 58. When a "warn" ACL statement has a log_message modifier, the message is
4401 remembered, and not repeated. This is to avoid a lot of repetition when a
4402 message has many recipients that cause the same warning to be written.
4403 Howewer, Exim was preserving the list of already written lines for an
4404 entire SMTP session, which doesn't seem right. The memory is now reset if a
4405 new message is started.
4407 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4408 debugging output unless log_rewrite was also set.
4410 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4411 of (int)(handle) when we know that handle contains (void *)(-1).
4413 61. The Exim daemon panic-logs an error return when it closes the incoming
4414 connection. However "connection reset by peer" seems to be common, and
4415 isn't really an error worthy of noting specially, so that particular error
4418 62. When Exim is trying to find all the local interfaces, it used to panic and
4419 die if the ioctl to get the interface flags failed. However, it seems that
4420 on at least one OS (Solaris 9) it is possible to have an interface that is
4421 included in the list of interfaces, but for which you get a failure error
4422 for this call. This happens when the interface is not "plumbed" into a
4423 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4424 failure of the "get flags" call assumes that the interface is down.
4426 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4427 makes life easier for people who are doing arithmetic on fields extracted
4428 from dates, where you often get leading zeros that should not be
4429 interpreted as octal.
4431 64. Added qualify_domain to the redirect router, to override the global
4434 65. If a pathologically long header line contained very many addresses (the
4435 report of this problem mentioned 10 000) and each of them was rewritten,
4436 Exim could use up a very large amount of memory. (It kept on making new
4437 copies of the header line as it rewrote, and never released the old ones.)
4438 At the expense of a bit more processing, the header rewriting function has
4439 been changed so that it no longer eats memory in this way.
4441 66. The generation of the Received: header has been moved from the time that a
4442 message starts to be received, to the time that it finishes. The timestamp
4443 in the Received: header should now be very close to that of the <= log
4444 line. There are two side-effects of this change:
4446 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4447 logged header lines no longer include the local Received: line, because
4448 it has not yet been created. The same applies to a copy of the message
4449 that is returned to a non-SMTP sender when a message is rejected.
4451 (b) When a filter file is tested using -bf, no additional Received: header
4452 is added to the test message. After some thought, I decided that this
4455 This change does not affect the value of $received_for. It is still set
4456 after address rewriting, but before local_scan() is called.
4458 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4460 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4461 gave an unhelpful panic error message, and a defer error. I have managed to
4462 change this behaviour so that it now rejects any supplied certificate,
4463 which seems right, as the list of acceptable certificates is empty.
4465 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4466 gave an unhelpful defer error. I have not managed to make this reject any
4467 supplied certificates, but the error message it gives is "no certificate
4468 supplied", which is not helpful.
4470 70. exigrep's output now also includes lines that are not associated with any
4471 message, but which match the given pattern. Implemented by a patch from
4472 Martin Sluka, which also tidied up the Perl a bit.
4474 71. Recipient callout verification, like sender verification, was using <> in
4475 the MAIL FROM command. This isn't really the right thing, since the actual
4476 sender may affect whether the remote host accepts the recipient or not. I
4477 have changed it to use the actual sender in the callout; this means that
4478 the cache record is now keyed on a recipient/sender pair, not just the
4479 recipient address. There doesn't seem to be a real danger of callout loops,
4480 since a callout by the remote host to check the sender would use <>.
4481 [SEE ABOVE: changed after hitting problems.]
4483 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4484 temporary errors. However, in the case of such a code being given after
4485 the end of a data transmission (i.e. after ".") Exim was failing to write
4486 a retry record for the message. (Yes, there was some broken host that was
4487 actually sending 8xx at this point.)
4489 73. An unknown lookup type in a host list could cause Exim to panic-die when
4490 the list was checked. (An example that provoked this was putting <; in the
4491 middle of a list instead of at the start.) If this happened during a DATA
4492 ACL check, a -D file could be left lying around. This kind of configuration
4493 error no longer causes Exim to die; instead it causes a defer errror. The
4494 incident is still logged to the main and panic logs.
4496 74. Buglet left over from Exim 3 conversion. The message "too many messages
4497 in one connection" was written to the rejectlog but not the mainlog, except
4498 when address rewriting (yes!) was being logged.
4500 75. Added write_rejectlog option.
4502 76. When a system filter was run not as root (that is, when system_filter_user
4503 was set), the values of the $n variables were not being returned to the
4504 main process; thus, they were not subsequently available in the $sn
4507 77. Added +return_path_on_delivery log selector.
4509 78. A connection timeout was being treated differently from recipients deferred
4510 when testing hosts_max_try with a message that was older than the host's
4511 retry timeout. (The host should not be counted, thus allowing all hosts to
4512 be tried at least once before bouncing.) This may have been the cause of an
4513 occasionally reported bug whereby a message would remain on the queue
4514 longer than the retry timeout, but would be bounced if a delivery was
4515 forced. I say "may" because I never totally pinned down the problem;
4516 setting up timeout/retry tests is difficult. See also the next item.
4518 79. The ultimate address timeout was not being applied to errors that involved
4519 a combination of host plus message (for example, a timeout on a MAIL
4520 command). When an address resolved to a number of possible hosts, and they
4521 were not all tried for each delivery (e.g. because of hosts_max_try), a
4522 message could remain on the queue longer than the retry timeout.
4524 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4527 81. Fixed an obscure SMTP outgoing bug which required at least the following
4528 conditions: (a) there was another message waiting for the same server;
4529 (b) the server returned 5xx to all RCPT commands in the first message so
4530 that the message was not completed; (c) the server dropped the connection
4531 or gave a negative response to the RSET that Exim sends to abort the
4532 transaction. The observed case was a dropped connection after DATA that had
4533 been sent in pipelining mode. That is, the server had advertised PIPELINING
4534 but was not implementing it correctly. The effect of the bug was incorrect
4535 behaviour, such as trying another host, and this could lead to a crash.
4541 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4542 and daemon.c were passed as pointers to ints; they should have been
4543 pointers to socklen_t variables (which are typically unsigned ints).
4545 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4548 3. Fixed a really odd bug that affected only the testing scheme; patching a
4549 certain fixed string in the binary changed the value of another string that
4550 happened to be identical to the end of the original first string.
4552 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4553 name", it returns that address as the IP address. On some operating
4554 systems (e.g. Solaris), it also passes back the IP address string as the
4555 "host name". However, on others (e.g. Linux), it passes back an empty
4556 string. Exim wasn't checking for this, and was changing the host name to an
4557 empty string, assuming it had been canonicized.
4559 5. Although rare, it is permitted to have more than one PTR record for a given
4560 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4561 all the names associated with an address, because they do in Solaris.
4562 However, it seems that they do not in Linux for data that comes from the
4563 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4564 I found this out when I moved to a new Linux workstation and tried to run
4565 the Exim test suite.
4567 To get round this problem I have changed the code so that it now does its
4568 own call to the DNS to look up PTR records when searching for a host name.
4569 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4570 addresses that are only in /etc/hosts are still found.
4572 This behaviour is, however, controlled by an option called host_lookup_
4573 order, which defaults to "bydns:byaddr". If people want to use the other
4574 order, or indeed, just use one or the other means of lookup, they can
4575 specify it in this variable.
4577 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4578 some operating systems, this comes back from gethostbyaddr() as an empty
4579 string, and this is what Exim used to test for. However, it seems that in
4580 other systems, "." is yielded. Exim now tests for this case too.
4582 7. The values of check_spool_space and check_log_space are now held internally
4583 as a number of kilobytes instead of an absolute number of bytes. If a
4584 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4585 kilobyte. This means that much larger values can be stored.
4587 8. Exim monitor: an attempt to get the action menu when not actually pointing
4588 at a message produces an empty menu entitled "No message selected". This
4589 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4590 no entries in it ("Shell widget menu has zero width and/or height"). So I
4591 have added a single, blank menu entry in this case.
4593 9. Added ${quote_local_part.
4595 10. MIME decoding is now applied to the contents of Subject: header lines when
4598 11. Now that a reference to $sender_host_address automatically causes a reverse
4599 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4600 host lookup before query-style lookups in lists that might use this
4601 variable. This has therefore been abolished, and the "net-" prefix is no
4602 longer necessary for query-style lookups.
4604 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4605 have been a typo for LFLAGS, so it has been changed.
4607 13. The install script calls Exim with "-C /dev/null" in order to find the
4608 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4609 to be output. Howeve, since Exim outputs its version number before the
4610 error, it didn't break the script. It just looked ugly. I fixed this by
4611 always allowing "-C /dev/null" if the caller is root.
4613 14. Ignore overlarge ACL variable number when reading spool file - insurance
4614 against a later release with more variables having written the file.
4616 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4617 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4618 this, as well as the form without the "IPv6" on the front.
4620 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4621 OS/Makefile-Darwin file.
4623 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4624 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4626 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4627 variable. However, some OS use ints for inodes. I've added cast to long int
4628 to get rid of the compiler warning.
4630 19. I had forgotten to lock out "/../" in configuration file names when
4631 ALT_CONFIG_PREFIX was set.
4633 20. Routers used for verification do not need to specify transports. However,
4634 if such a router generated a host list, and callout was configured, Exim
4635 crashed, because it could not find a port number from the (non-existent)
4636 transport. It now assumes port 25 in this circumstance.
4638 21. Added the -t option to exigrep.
4640 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4641 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4642 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4643 them is set, LOOKUP_LSEARCH is forced.
4645 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4647 24. Added sender and host information to the "rejected by local_scan()" log
4648 line; previously there was no indication of these.
4650 25. Added .include_if_exists.
4652 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4653 Linux. It turns out that not all file systems support this. Apparently some
4654 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4655 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4656 not supported on the file descriptor, is now ignored when Exim is trying to
4657 sync a directory. This applies only to Linux.
4659 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4661 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4662 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4665 29. Added support for maildirsize files from supplied patch (modified a bit).
4667 30. The use of :fail: followed by an empty string could lead Exim to respond to
4668 sender verification failures with (e.g.):
4670 550 Verification failed for <xxx>
4671 550 Sender verify failed
4673 where the first response line was missing the '-' that indicates it is not
4674 the final line of the response.
4676 31. The loop for finding the name of the user that called Exim had a hardwired
4677 limit of 10; it now uses the value of finduser_retries, which is used for
4678 all other user lookups.
4680 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4683 33. Exim was neglecting to zero errno before one call of strtol() when
4684 expanding a string and expecting an integer value. On some systems this
4685 resulted in spurious "integer overflow" errors. Also, it was casting the
4686 result into an int without checking.
4688 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4689 did not work. The code looks as if it has *never* worked, though it appears
4690 to have been documented since at least releast 1.62. I have made it work.
4692 35. The "timeout_DNS" error in retry rules, also documented since at least
4693 1.62, also never worked. As it isn't clear exactly what this means, and
4694 clearly it isn't a major issue, I have abolished the feature by treating it
4695 as "timeout", and writing a warning to the main and panic logs.
4697 36. The display of retry rules for -brt wasn't always showing the error code
4700 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4701 timeout_connect_A, timeout_connect_MX.
4703 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4704 to the empty sender.
4706 39. The daemon was not analysing the content of -oX till after it had closed
4707 stderr and disconnected from the controlling terminal. This meant that any
4708 syntax errors were only noted on the panic log, and the return code from
4709 the command was 0. By re-arranging the code a little, I've made the
4710 decoding happen first, so such errors now appear on stderr, and the return
4711 code is 1. However, the actual setting up of the sockets still happens in
4712 the disconnected process, so errors there are still only recorded on the
4715 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4716 connections (as happens on some IP stacks) was logged at start up time as
4717 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4718 it from "IPv6 and IPv4", which means that two separate sockets are being
4721 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4722 says whether AF_INET or AF_INET6 was passed as an argument.
4724 42. Exiwhat output was messed up when time zones were included in log
4727 43. Exiwhat now gives more information about the daemon's listening ports,
4728 and whether -tls-on-connect was used.
4730 44. The "port" option of the smtp transport is now expanded.
4732 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4733 silently ignored. Now an error message is written to the main and panic
4736 46. There's a new ACL modifier called "logwrite" which writes to a log file
4737 as soon as it is encountered.
4739 47. Added $local_user_uid and $local_user_gid at routing time.
4741 48. Exim crashed when trying to verify a sender address that was being
4744 49. Exim was recognizing only a space character after ".include". It now also
4745 recognizes a tab character.
4747 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4748 extracting the relevant information from the specification. The man page no
4749 longer contains scrambled data for the -d option, and I've added a section
4750 at the front about calling Exim under different names.
4752 51. Added "extra_headers" argument to the "mail" command in filter files.
4754 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4757 53. Installed eximstats 1.29.
4759 54. Added transport_filter_timeout as a generic transport option.
4761 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4762 Cc: header lines. This was required by RFC 822, but it not required by RFC
4765 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4766 incoming messages that did not have them. Now it does so only if the
4767 message originates locally, that is, if there is no associated remote host
4768 address. When Resent- header lines are present, this applies to the Resent-
4769 lines rather than the non-Resent- lines.
4771 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4772 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4774 58. Messages for configuration errors now include the name of the main
4775 configuration file - useful now that there may be more than one file in a
4776 list (.included file names were always shown).
4778 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4779 for those rare installations that do not start the daemon as root or run it
4780 setuid root. I've cut out the call to initgroups() if the daemon is not
4783 60. The Exim user and group can now be bound into the binary as text strings
4784 that are looked up at the start of Exim's processing.
4786 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4788 62. Added $mailstore_basename variable.
4790 63. Installed patch to sieve.c from Michael Haardt.
4792 64. When Exim failed to open the panic log after failing to open the main log,
4793 the original message it was trying to log was written to stderr and debug
4794 output, but if they were not available (the usual case in production), it
4795 was lost. Now it is written to syslog before the two lines that record the
4796 failures to open the logs.
4798 65. Users' Exim filters run in subprocesses under the user's uid. It is
4799 possible for a "deliver" command or an alias in a "personal" command to
4800 provoke an address rewrite. If logging of address rewriting is configured,
4801 this fails because the process is not running as root or exim. There may be
4802 a better way of dealing with this, but for the moment (because 4.30 needs
4803 to be released), I have disabled address rewrite logging when running a
4804 filter in a non-root, non-exim process.
4810 1. The buildconfig auxiliary program wasn't quoting the value set for
4811 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4812 not defined. This bug was masked in 4.22 by the effect that was fixed in
4815 2. Some messages that were rejected after a message id was allocated were
4816 shown as "incomplete" by exigrep. It no longer does this for messages that
4817 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4819 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4820 have allow_domain_literals set, the ID did not get logged in the <= line.
4821 Domain literals are now always recognized in Message-ID: header lines.
4823 4. The first argument for a ${extract expansion item is the key name or field
4824 number. Leading and trailing spaces in this item were not being ignored,
4825 causing some misleading effects.
4827 5. When deliver_drop_privilege was set, single queue runner processes started
4828 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4829 same command in the process it spins off) were not dropping privilege.
4831 6. When the daemon running as "exim" started a queue runner, it always
4832 re-executed Exim in the spun-off process. This is a waste of effort when
4833 deliver_drop_privilege is set. The new process now just calls the
4834 queue-runner function directly.
4840 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4843 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4844 ignored. Though the use of -oP was forcing the writing of a pid file, it
4845 was always written to the default place.
4847 3. If the message "no IP address found for host xxxx" is generated during
4848 incoming verification, it is now followed by identification of the incoming
4849 connection (so you can more easily find what provoked it).
4851 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4853 5. Added some features to "harden" Exim a bit more against certain attacks:
4855 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4856 be put in Local/Makefile. This is like the never_users runtime option,
4857 but it cannot be overridden. The default setting is "root".
4859 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4860 prefix string with which any file named in a -C command line option
4863 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4864 is retained for -C and -D only if the caller of Exim is root. Without
4865 it, the exim user may also use -C and -D and retain privilege.
4867 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4868 command line option is disabled.
4870 6. Macro names set by the -D option must start with an upper case letter, just
4871 like macro names defined in the configuration file.
4873 7. Added "dereference=" facility to LDAP.
4875 8. Two instances of the typo "uknown" in the source files are fixed.
4877 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4878 the Configure-Makefile script screwed up while processing it.
4880 10. Incorporated PCRE 4.4.
4882 11. The SMTP synchronization check was not operating right at the start of an
4883 SMTP session. For example, it could not catch a HELO sent before the client
4884 waited for the greeting. There is now a check for outstanding input at the
4885 point when the greeting is written. Because of the duplex, asynchronous
4886 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4887 way, but not yet received, when the check is performed.
4889 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4890 on TCP/IP sockets, because this apparently causes some broken clients to
4893 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4894 unchanged) from the Cygwin maintainer.
4896 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4897 was supported instead of testing for mbx. Effectively a typo.
4899 15. The spa authenticator server code was not checking that the input it
4900 received was valid base64.
4902 16. The debug output line for the "set" modifier in ACLs was not showing the
4903 name of the variable that was being set.
4905 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4907 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4908 lookup on its own. Something else was needed to trigger the lookup. For
4909 example, a match in host_lookup or the need for a host name in a host list.
4910 Now, if $sender_host_name is referenced and the host name has not yet been
4911 looked up, a lookup is performed. If the lookup fails, the variable remains
4912 empty, and $host_lookup_failed is set to "1".
4914 19. Added "eqi" as a case-independent comparison operator.
4916 20. The saslauthd authentication condition could segfault if neither service
4917 nor realm was specified.
4919 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4920 error message that was logged was misleading, and incoming SMTP
4921 connections were dropped. The message is now more accurate, and temporary
4922 errors are given to SMTP connections.
4924 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4925 (except RSET) with a 421 error, until QUIT is received. However, it was
4926 failing to send a response to QUIT.
4928 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4929 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4930 ACL is now run after the helo_try_verify_hosts code.
4932 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4933 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4934 case-independent, so other case variants are also recognized). Apparently
4935 some systems use these upper case variants.
4937 25. If more than two messages were waiting for the same host, and a transport
4938 filter was specified for the transport, Exim sent two messages over the
4939 same TCP/IP connection, and then failed with "socket operation on non-
4940 socket" when it tried to send the third.
4942 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4944 27. The extern definition of crypt16() in expand.c was not being excluded when
4945 the OS had its own crypt16() function.
4947 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4948 as a preferred synonym for return_size_limit, both as an option and as an
4951 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4953 30. Changed the default configuration ACL to relax the local part checking rule
4954 for addresses that are not in any local domains. For these addresses,
4955 slashes and pipe symbols are allowed within local parts, but the sequence
4956 /../ is explicitly forbidden.
4958 31. SPA server authentication was not clearing the challenge buffer before
4961 32. log_message in a "warn" ACL statement was writing to the reject log as
4962 well as to the main log, which contradicts the documentation and doesn't
4963 seem right (because no rejection is happening). So I have stopped it.
4965 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4966 However, I am unable to do any testing of this.
4968 34. Fixed an infelicity in the appendfile transport. When checking directories
4969 for a mailbox, to see if any needed to be created, it was accidentally
4970 using path names with one or more superfluous leading slashes; tracing
4971 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4973 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4974 contents are added to the log line that is written for every discarded
4975 recipient. (Previously a log_message setting was ignored.)
4977 36. The ${quote: operator now quotes the string if it is empty.
4979 37. The install script runs exim in order to find its version number. If for
4980 some reason other than non-existence or emptiness, which it checks, it
4981 could not run './exim', it was installing it with an empty version number,
4982 i.e. as "exim-". This error state is now caught, and the installation is
4985 38. An argument was missing from the function that creates an error message
4986 when Exim fails to connect to the socket for saslauthd authentication.
4987 This could cause Exim to crash, or give a corrupted message.
4989 39. Added isip, isip4, and isip6 to ${if conditions.
4991 40. The ACL variables $acl_xx are now saved with the message, and can be
4992 accessed later in routers, transports, and filters.
4994 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4995 strings in the file are not string-expanded.
4997 42. If a MAIL command specified a SIZE value that was too large to fit into an
4998 int variable, the check against message_size_limit failed. Such values are
4999 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5000 day this will have to be increased, but I don't think I want to be around
5001 when emails are that large.
5008 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5009 iconv() is not standard in FreeBSD.
5011 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5012 IPv6 enabled. The observed symptom was a segmentation fault on return from
5013 the function os_common_find_running_interfaces() in src/os.c.
5015 3. In the check_special_case() function in daemon.c I had used "errno" as an
5016 argument name, which causes warnings on some systems. This was basically a
5017 typo, since it was named "eno" in the comments!
5019 4. The code that waits for the clock to tick (at a resolution of some fraction
5020 of a second) so as to ensure message-id uniqueness was always waiting for
5021 at least one whole tick, when it could have waited for less. [This is
5022 almost certainly not relevant at current processor speeds, where it is
5023 unlikely to ever wait at all. But we try to future-proof.]
5025 5. The function that sleeps for a time interval that includes fractions of a
5026 second contained a race. It did not block SIGALRM between setting the
5027 timer, and suspending (a couple of lines later). If the interval was short
5028 and the sigsuspend() was delayed until after it had expired, the suspension
5029 never ended. On busy systems this could lead to processes getting stuck for
5032 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5033 process, before it forks any delivery processes. The open lookup caching
5034 mechanism meant that the open file or database connection was passed into
5035 the delivery process. The problem was that delivery processes always tidy
5036 up cached lookup data. This could cause a problem for the next delivery
5037 process started by the queue runner, because the external queue runner
5038 process does not know about the closure. So the next delivery process
5039 still has data in the lookup cache. In the case of a file lookup, there was
5040 no problem because closing a file descriptor in a subprocess doesn't affect
5041 the parent. However, if the lookup was caching a connection to a database,
5042 the connection was closed, and the second delivery process was likely to
5043 see errors such as "PGSQL: query failed: server closed the connection
5044 unexpectedly". The problem has been fixed by closing all cached lookups
5045 in a queue runner before running a delivery process.
5047 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5048 seem to have the "const" qualifier which it has on other OS. I've
5051 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5052 *of the same type* (client or server) with the same public name that an
5053 error should be diagnosed.
5055 9. When Exim looked up a host name for an IP address, but failed to find the
5056 original IP address when looking up the host name (a safety check), it
5057 output the message "<ip address> does not match any IP for NULL", which was
5058 confusing, to say the least. The bug was that the host name should have
5059 appeared instead of "NULL".
5061 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5062 user that is built into the binary, and the -C or -D options is used, root
5063 privilege is dropped before the configuration file is read. In addition,
5064 logging is switched to stderr instead of the normal log files. If the
5065 configuration then re-defines the Exim user, the unprivileged environment
5066 is probably not what is expected, so Exim logs a panic warning message (but
5069 However, if deliver_drop_privilege is set, the unprivileged state may well
5070 be exactly what is intended, so the warning has been cut out in that case,
5071 and Exim is allowed to try to write to its normal log files.
5077 1. smtp_return_error_details was not giving details for temporary sender
5078 or receiver verification errors.
5080 2. Diagnose a configuration error if two authenticators have the same public
5083 3. Exim used not to create the message log file for a message until the first
5084 delivery attempt. This could be confusing when incoming messages were held
5085 for policy or load reasons. The message log file is now created at the time
5086 the message is received, and an initial "Received" line is written to it.
5088 4. The automatically generated man page for command line options had a minor
5089 bug that caused no ill effects; however, a more serious problem was that
5090 the procedure for building the man page automatically didn't always
5091 operate. Consequently, release 4.20 contains an out-of-date version. This
5092 shouldn't happen again.
5094 5. When building Exim with embedded Perl support, the script that builds the
5095 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5096 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5098 6. The freeze_tell option was not being used for messages that were frozen on
5099 arrival, either by an ACL or by local_scan().
5101 7. Added the smtp_incomplete_transaction log selector.
5103 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5104 was accepting AUTH without a new EHLO.
5106 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5107 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5110 10. Logging of TCP/IP connections (when configured) now happens in the main
5111 daemon process instead of the child process, so that the TCP/IP connection
5112 count is more accurate (but it can never be perfect).
5114 11. The use of "drop" in a nested ACL was not being handled correctly in the
5115 outer ACL. Now, if condition failure induced by the nested "drop" causes
5116 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5117 or "require"), the connection is dropped.
5119 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5120 that yield "discard" can now be used with an "accept" or a "discard" verb,
5121 but an error is generated for any others (because I can't see a useful way
5122 to define what should happen).
5124 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5125 are now processed in the same way as lines in the Exim configuration file.
5126 In particular, continuation lines are supported.
5128 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5130 15. Added -ti meaning -t -i.
5132 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5133 domains, and warn by logging if others are found.
5135 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5136 returned from GIFCONF when Exim is trying to find the list of interfaces on
5137 a host. The code in os.c has been modified to copy each ifreq to an aligned
5138 structure in all cases.
5140 Also, in some cases, the returned ifreq's were being copied to a 'struct
5141 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5142 means the last couple of bytes of an IPv6 address could be chopped if the
5143 ifreq contained only a normal sockaddr (14 bytes storage).
5145 18. Named domain lists were not supported in the hosts_treat_as_local option.
5146 An entry such as +xxxx was not recognized, and was treated as a literal
5149 19. Ensure that header lines added by a DATA ACL are included in the reject log
5150 if the ACL subsequently rejects the message.
5152 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5153 MD5 (which is deprecated).
5155 21. When testing a filter file using -bf, Exim was writing a message when it
5156 took the sender from a "From " line in the message, but it was not doing so
5157 when it took $return_path from a Return-Path: header line. It now does.
5159 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5160 with a valid header line field name (a series of printing characters
5161 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5163 23. Changed "disc" in the source to "disk" to conform to the documentation and
5164 the book and for uniformity.
5166 24. Ignore Sendmail's -Ooption=value command line item.
5168 25. When execve() failed while trying to run a command in a pipe transport,
5169 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5170 could be confused with a return value of 69 from the command itself. This
5171 has been changed to 127, the value the shell returns if it is asked to run
5172 a non-existent command. The wording for the related log line suggests a
5173 non-existent command as the problem.
5175 26. If received_header_text expands to an empty string, do not add a Received:
5176 header line to the message. (Well, it adds a token one on the spool, but
5177 marks it "old" so that it doesn't get used or transmitted.)
5179 27. Installed eximstats 1.28 (addition of -nt option).
5181 28. There was no check for failure on the call to getsockname() in the daemon
5182 code. This can fail if there is a shortage of resources on the system, with
5183 ENOMEM, for example. A temporary error is now given on failure.
5185 29. Contrary to the C standard, it seems that in some environments, the
5186 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5187 program. Exim now does this explicitly; it affects the formatting of
5188 timestamps using strftime().
5190 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5191 complaints. I've now initialized all the variables, to avoid this.
5193 32. Header lines added by a system filter were not being "seen" during
5194 transport-time rewrites.
5196 33. The info_callback() function passed to OpenSSL is set up with type void
5197 (*)(SSL *, int, int), as described somewhere. However, when calling the
5198 function (actually a macro) that sets it up, the type void(*)() is
5199 expected. I've put in a cast to prevent warnings from picky compilers.
5201 34. If a DNS black list lookup found a CNAME record, but there were no A
5202 records associated with the domain it pointed at, Exim crashed.
5204 35. If a DNS black list lookup returned more than one A record, Exim ignored
5205 all but the first. It now scans all returned addresses if a particular IP
5206 value is being sought. In this situation, the contents of the
5207 $dnslist_value variable are a list of all the addresses, separated by a
5210 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5211 to accept a host name and all its aliases if the forward lookup for any of
5212 them yielded the IP address of the incoming connection. Now it accepts only
5213 those names whose forward lookup yields the correct IP address. Any other
5214 names are discarded. This closes a loophole whereby a rogue DNS
5215 administrator could create reverse DNS records to break through a
5216 wildcarded host restriction in an ACL.
5218 37. If a user filter or a system filter that ran in a subprocess used any of
5219 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5220 the wrong values were passed to the pipe command ($thisaddress had the
5221 value of $0, $0 had the value of $1, etc). This bug was introduced by
5222 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5224 38. Improved the line breaking for long SMTP error messages from ACLs.
5225 Previously, if there was no break point between 40 and 75 characters, Exim
5226 left the rest of the message alone. Two changes have been made: (a) I've
5227 reduced the minimum length to 35 characters; (b) if it can't find a break
5228 point between 35 and 75 characters, it looks ahead and uses the first one
5229 that it finds. This may give the occasional overlong line, but at least the
5230 remaining text gets split now.
5232 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5233 file descriptors might be low, and that setting 1000 would always raise it.
5234 It turns out that in some environments, the limit is already over 1000 and
5235 that lowering it causes trouble. So now Exim takes care not to decrease it.
5237 40. When delivering a message, the value of $return_path is set to $sender_
5238 address at the start of routing (routers may change the value). By an
5239 oversight, this default was not being set up when an address was tested by
5240 -bt or -bv, which affected the outcome if any router or filter referred to
5243 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5244 log only when "message" or "log_message" are set. However, if one of the
5245 conditions was an address verification, or a call to a nested ACL, the
5246 messages generated by the underlying test were being passed through. This
5247 no longer happens. The underlying message is available in $acl_verify_
5248 message for both "message" and "log_message" expansions, so it can be
5249 passed through if needed.
5251 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5252 new expansion $bh_ to give the encoded byte string without charset
5253 translation. Translation happens only if iconv() is available; HAVE_ICONV
5254 indicates this at build time. HEADERS_CHARSET gives the charset to
5255 translate to; headers_charset can change it in the configuration, and
5256 "headers charset" can change it in an individual filter file.
5258 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5259 that creates RFC 2047 encoded "words" labels them as that charset instead
5260 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5261 expansion operator; (ii) when Exim creates a From: line for a local
5262 message; (iii) when a header line is rewritten to include a "phrase" part.
5264 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5265 buggy, causing it to skip the first lines of messages whose message ID
5266 ended in 'D'. This would not have bitten before Exim release 4.14, because
5267 message IDs were unlikely to end in 'D' before then. The effect was to have
5268 incorrect size information for certain domains.
5270 45. #include "config.h" was missing at the start of the crypt16.c module. This
5271 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5274 46. If there was a timeout during a "random" callout check, Exim treated it as
5275 a failure of the random address, and carried on sending RSET and the real
5276 address. If the delay was just some slowness somewhere, the response to the
5277 original RCPT would be taken as a response to RSET and so on, causing
5278 mayhem of various kinds.
5280 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5281 when I implemented it. It didn't allow for the fact that some option values
5282 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5283 the right test for positive values.
5285 48. Domain names in DNS records are case-independent. Exim always looks them up
5286 in lower case. Some resolvers return domain names in exactly the case they
5287 appear in the zone file, that is, they may contain uppercase letters. Not
5288 all resolvers do this - some return always lower case. Exim was treating a
5289 change of case by a resolver as a change of domain, similar to a widening
5290 of a domain abbreviation. This triggered its re-routing code and so it was
5291 trying to route what was effectively the same domain again. This normally
5292 caused routing to fail (because the router wouldn't handle the domain
5293 twice). Now Exim checks for this case specially, and just changes the
5294 casing of the domain that it ultimately uses when it transmits the message
5297 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5300 50. If a filter generated a file delivery with a non-absolute name (possible if
5301 no home directory exists for the router), the forbid_file option was not
5304 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5305 the existing equality matching.
5307 52. Exim was using ints instead of ino_t variables in some places where it was
5308 dealing with inode numbers.
5310 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5311 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5312 called TMPDIR, and if it finds it is different, it changes its value.
5314 54. The smtp_printf() function is now made available to local_scan() so
5315 additional output lines can be written before returning. There is also an
5316 smtp_fflush() function to enable the detection of a dropped connection.
5317 The variables smtp_input and smtp_batched_input are exported to
5320 55. Changed the default runtime configuration: the message "Unknown user"
5321 has been removed from the ACL, and instead placed on the localuser router,
5322 using the cannot_route_message feature. This means that any verification
5323 failures that generate their own messages won't get overridden. Similarly,
5324 the "Unrouteable address" message that was in the ACL for unverifiable
5325 relay addresses has also been removed.
5327 56. Added hosts_avoid_esmtp to the smtp transport.
5329 57. The exicyclog script was not checking for the esoteric option
5330 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5331 will work only if exicyclog is run under the appropriate euid.
5333 58. Following a discussion on the list, the rules by which Exim recognises line
5334 endings on incoming messages have been changed. The -dropcr and drop_cr
5335 options are now no-ops, retained only for backwards compatibility. The
5336 following line terminators are recognized: LF CRLF CR. However, special
5337 processing applies to CR:
5339 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5340 nor a local message in the state where . is a terminator.
5342 (ii) If a bare CR is encountered in a header line, an extra space is added
5343 after the line terminator so as not to end the header. The reasoning
5344 behind this is that bare CRs in header lines are most likely either
5345 to be mistakes, or people trying to play silly games.
5347 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5348 was being incorrectly given as 18 bytes larger than it should have been.
5349 This is a VOB (very old bug).
5351 60. This may never have affected anything current, but just in case it has:
5352 When the local host is found other than at the start of a list of hosts,
5353 the local host, those with the same MX, and any that follow, are discarded.
5354 When the list in question was part of a longer list of hosts, the following
5355 hosts (not currently being processed) were also being discarded. This no
5356 longer happens. I'm not sure if this situation could ever has previously
5359 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5362 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5363 header. This is something that is recommended in a new Internet Draft, and
5364 is something that is documented as being done by Sendmail. There are two
5365 possible values. For messages generated by the autoreply transport, Exim
5368 Auto-Submitted: auto-replied
5370 whereas for all other generated messages (e.g. bounces) it adds
5372 Auto-Submitted: auto-generated
5374 63. The "personal" condition in filters now includes a test for the
5375 Auto-Submitted: header. If it contains the string "auto-" the message it
5376 not considered personal.
5378 64. Added rcpt_include_affixes as a generic transport option.
5380 65. Added queue_only_override (default true).
5382 66. Added the syslog_duplication option.
5384 67. If what should have been the first header line of a message consisted of
5385 a space followed by a colon, Exim was mis-interpreting it as a header line.
5386 It isn't of course - it is syntactically invalid and should therefore be
5387 treated as the start of the message body. The misbehaviour could have
5388 caused a number of strange effects, including loss of data in subsequent
5389 header lines, and spool format errors.
5391 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5392 client host had authenticated. This control can now be exercised by an ACL
5393 for more flexibility.
5395 69. By default, callouts do not happen when testing with -bh. There is now a
5396 variant, -bhc, which does actually run the callout code, including
5397 consulting and updating the callout cache.
5399 70. Added support for saslauthd authentication, courtesy of Alexander
5402 71. If statvfs() failed on the spool or log directories while checking their
5403 size for availability, Exim confusingly gave the error "space shortage".
5404 Furthermore, in debugging mode it crashed with a floating point exception.
5405 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5406 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5407 serious problem, Exim now writes to the main and panic logs when this
5408 happens, with details of the failure. It then refuses to accept the
5409 incoming message, giving the message "spool directory problem" or "log
5410 directory problem" with a 421 code for SMTP messages.
5412 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5413 0, 1, and 2 exist, because some OS complain for execs without them (see
5414 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5415 descriptors. However, the code omitted to check that the open succeeded,
5416 causing mysterious errors if for some reason the permissions on /dev/null
5417 got screwed. Now Exim writes a message to the main and panic logs, and
5418 bombs out if it can't open /dev/null.
5420 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5421 interact so that it is all more flexible. It is supposed to remain
5422 backwards compatible. Also added extra_local_interfaces.
5424 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5425 to bomb out with an assertion failure - to the client this appears as a
5426 connection drop. This problem occurs in the part of the code that was taken
5427 from the Samba project. Fortunately, the assertion is in a very simple
5428 function, so I have fixed this by reproducing the function inline in the
5429 one place where it is called, and arranging for authentication to fail
5430 instead of killing the process with assert().
5432 75. The SPA client code was not working when the server requested OEM rather
5433 than Unicode encoding.
5435 76. Added code to make require_files with a specific uid setting more usable in
5436 the case where statting the file as root fails - usually a non-root-mounted
5437 NFS file system. When this happens and the failure is EACCES, Exim now
5438 forks a subprocess and does the per-uid checking as the relevant uid.
5440 77. Added process_log_path.
5442 78. If log_file_path was not explicitly set, a setting of check_log_space or
5443 check_log_inodes was ignored.
5445 79. If a space check for the spool or log partitions fails, the incident is now
5446 logged. Of course, in the latter case the data may get lost...
5448 80. Added the %p formatting code to string_format() so that it can be used to
5449 print addresses in debug_print(). Adjusted all the address printing in the
5450 debugging in store.c to use %p rather than %d.
5452 81. There was a concern that a line of code in smtp_in.c could overflow a
5453 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5454 initially expressed, the concern was not well-founded, because trailing
5455 spaces are removed early. However, if the trailing spaces were followed by
5456 a NULL, they did not get removed, so the overflow was possible. Two fixes
5459 (a) I re-wrote the offending code in a cleaner fashion.
5460 (b) If an incoming SMTP command contains a NULL character, it is rejected
5463 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5464 runs initgroups(), so that if the Exim user is in any additional groups,
5465 they will be used during message reception.
5471 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob