1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
41 PP/01 Solaris build fix for Oracle's LDAP libraries.
42 Bugzilla 1109, patch from Stephen Usher.
44 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
46 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
49 TF/02 Fix a couple more cases where we did not log the error message
50 when unlink() failed. See also change 4.74-TF/03.
52 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
53 lock up or crash if it happened to be inside a call to libc when it
54 got a SIGUSR1 from exiwhat.
56 The SIGUSR1 handler appends the current process status to the process
57 log which is later printed by exiwhat. It used to use the general
58 purpose logging code to do this, but several functions it calls are
61 The new output code in the SIGUSR1 handler is specific to the process
62 log, and simple enough that it's easy to inspect for signal safety.
63 Removing some special cases also simplifies the general logging code.
64 Removing the spurious timestamps from the process log simplifies
67 TF/04 Improved ratelimit ACL condition.
69 The /noupdate option has been deprecated in favour of /readonly which
70 has clearer semantics. The /leaky, /strict, and /readonly update modes
71 are mutually exclusive. The update mode is no longer included in the
72 database key; it just determines when the database is updated. (This
73 means that when you upgrde Exim will forget old rate measurements.)
75 Exim now checks that the per_* options are used with an update mode that
76 makes sense for the current ACL. For example, when Exim is processing a
77 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
78 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
79 must specify per_mail/readonly. If you omit the update mode it defaults to
80 /leaky where that makes sense (as before) or /readonly where required.
82 The /noupdate option is now undocumented but still supported for
83 backwards compatibility. It is equivalent to /readonly except that in
84 ACLs where /readonly is required you may specify /leaky/noupdate or
85 /strict/noupdate which are treated the same as /readonly.
87 A useful new feature is the /count= option. This is a generalization
88 of the per_byte option, so that you can measure the throughput of other
89 aggregate values. For example, the per_byte option is now equivalent
90 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
92 The per_rcpt option has been generalized using the /count= mechanism
93 (though it's more complicated than the per_byte equivalence). When it is
94 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
95 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
96 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
97 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
98 Note that using per_rcpt with a non-readonly update mode in more than
99 one ACL will cause the recipients to be double-counted. (The per_mail
100 and per_byte options don't have this problem.)
102 The handling of very low rates has changed slightly. If the computed rate
103 is less than the event's count (usually one) then this event is the first
104 after a long gap. In this case the rate is set to the same as this event's
105 count, so that the first message of a spam run is counted properly.
107 The major new feature is a mechanism for counting the rate of unique
108 events. The new per_addr option counts the number of different
109 recipients that someone has sent messages to in the last time period. It
110 behaves like per_rcpt if all the recipient addresses are different, but
111 duplicate recipient addresses do not increase the measured rate. Like
112 the /count= option this is a general mechanism, so the per_addr option
113 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
114 example, measure the rate that a client uses different sender addresses
115 with the options per_mail/unique=$sender_address. There are further
116 details in the main documentation.
118 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
120 TF/06 Removed a few PCRE remnants.
122 TF/07 Automatically extract Exim's version number from tags in the git
123 repository when doing development or release builds.
125 PP/02 Raise smtp_cmd_buffer_size to 16kB.
126 Bugzilla 879. Patch from Paul Fisher.
128 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
129 Heavily based on revision 40f9a89a from Simon Arlott's tree.
132 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
134 PP/05 Variable $av_failed, true if the AV scanner deferred.
135 Bugzilla 1078. Patch from John Horne.
137 PP/06 Stop make process more reliably on build failure.
138 Bugzilla 1087. Patch from Heiko Schlittermann.
140 PP/07 Make maildir_use_size_file an _expandable_ boolean.
141 Bugzilla 1089. Patch from Heiko Schlittermann.
143 PP/08 Handle ${run} returning more data than OS pipe buffer size.
144 Bugzilla 1131. Patch from Holger Weiß.
146 PP/09 Handle IPv6 addresses with SPF.
147 Bugzilla 860. Patch from Wolfgang Breyha.
149 PP/10 GnuTLS: support TLS 1.2 & 1.1.
151 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
154 PP/11 match_* no longer expand right-hand-side by default.
155 New compile-time build option, EXPAND_LISTMATCH_RHS.
156 New expansion conditions, "inlist", "inlisti".
158 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
160 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
162 PP/14 fix log_write() format string regression from TF/03.
163 Bugzilla 1152. Patch from Dmitry Isaikin.
169 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
171 PP/02 Harmonised TLS library version reporting; only show if debugging.
172 Layout now matches that introduced for other libraries in 4.74 PP/03.
174 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
176 PP/04 New "dns_use_edns0" global option.
178 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
181 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
182 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
184 TK/01 Updated PolarSSL code to 0.14.2.
185 Bugzilla 1097. Patch from Andreas Metzler.
187 PP/07 Catch divide-by-zero in ${eval:...}.
190 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
193 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
194 format-string attack -- SECURITY: remote arbitrary code execution.
196 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
197 time unintentionally subject to list matching rules, letting the header
198 cause arbitrary Exim lookups (of items which can occur in lists, *not*
199 arbitrary string expansion). This allowed for information disclosure.
201 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
202 INT_MIN/-1 -- value coerced to INT_MAX.
208 NM/01 Workround for PCRE version dependency in version reporting
211 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
212 This fixes portability to compilers other than gcc, notably
213 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
215 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
216 makefiles for portability to HP-UX and POSIX correctness.
218 PP/01 Permit LOOKUP_foo enabling on the make command-line.
219 Also via indented variable definition in the Makefile.
220 (Debugging by Oliver Heesakkers).
222 PP/02 Restore caching of spamd results with expanded spamd_address.
223 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
225 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
226 Improves build reliability. Fix from: Frank Elsner
228 NM/02 Fix wide character breakage in the rfc2047 coding
229 Fixes bug 1064. Patch from Andrey N. Oktyabrski
231 NM/03 Allow underscore in dnslist lookups
232 Fixes bug 1026. Patch from Graeme Fowler
234 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
235 Code patches from Adam Ciarcinski of NetBSD.
237 NM/04 Fixed exiqgrep to cope with mailq missing size issue
240 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
241 is logged, to avoid truncation. Patch from John Horne.
243 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
244 Patch from Jakob Hirsch.
246 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
247 SQL string expansion failure details.
248 Patch from Andrey Oktyabrski.
250 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
251 Patch from Simon Arlott.
253 PP/09 New lookups functionality failed to compile on old gcc which rejects
254 extern declarations in function scope.
255 Patch from Oliver Fleischmann
257 PP/10 Use sig_atomic_t for flags set from signal handlers.
258 Check getgroups() return and improve debugging.
259 Fixed developed for diagnosis in bug 927 (which turned out to be
262 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
263 Patch from Mark Zealey.
265 PP/12 Bugzilla 1056: Improved spamd server selection.
266 Patch from Mark Zealey.
268 PP/13 Bugzilla 1086: Deal with maildir quota file races.
269 Based on patch from Heiko Schlittermann.
271 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
272 Patch from Uwe Doering, sign-off by Michael Haardt.
274 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
275 variable declaration deep within a block. Bug and patch from
278 PP/15 lookups-Makefile IRIX compatibilty coercion.
280 PP/16 Make DISABLE_DKIM build knob functional.
282 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
283 Patch by Simon Arlott
285 TF/03 Fix valgrind.h portability to C89 compilers that do not support
286 variable argument macros. Our copy now differs from upstream.
292 TF/01 Failure to get a lock on a hints database can have serious
293 consequences so log it to the panic log.
295 TF/02 Log LMTP confirmation messages in the same way as SMTP,
296 controlled using the smtp_confirmation log selector.
298 TF/03 Include the error message when we fail to unlink a spool file.
300 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
301 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
302 for maintaining out-of-tree patches for some time.
304 PP/01 Bugzilla 139: Documentation and portability issues.
305 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
306 Handle per-OS dynamic-module compilation flags.
308 PP/02 Let /dev/null have normal permissions.
309 The 4.73 fixes were a little too stringent and complained about the
310 permissions on /dev/null. Exempt it from some checks.
311 Reported by Andreas M. Kirchwitz.
313 PP/03 Report version information for many libraries, including
314 Exim version information for dynamically loaded libraries. Created
315 version.h, now support a version extension string for distributors
316 who patch heavily. Dynamic module ABI change.
318 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
319 privilege escalation vulnerability whereby the Exim run-time user
320 can cause root to append content of the attacker's choosing to
323 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
326 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
327 If dropping privileges for untrusted macros, we disabled normal logging
328 on the basis that it would fail; for the Exim run-time user, this is not
329 the case, and it resulted in successful deliveries going unlogged.
330 Fixed. Reported by Andreas Metzler.
336 PP/01 Date: & Message-Id: revert to normally being appended to a message,
337 only prepend for the Resent-* case. Fixes regression introduced in
338 Exim 4.70 by NM/22 for Bugzilla 607.
340 PP/02 Include check_rfc2047_length in configure.default because we're seeing
341 increasing numbers of administrators be bitten by this.
343 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
345 PP/03 Bugzilla 994: added openssl_options main configuration option.
347 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
349 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
351 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
353 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
354 without a peer certificate, leading to a segfault because of an
355 assumption that peers always have certificates. Be a little more
356 paranoid. Problem reported by Martin Tscholak.
358 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
359 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
360 NB: ClamAV planning to remove STREAM in "middle of 2010".
361 CL also introduces -bmalware, various -d+acl logging additions and
362 more caution in buffer sizes.
364 PP/09 Implemented reverse_ip expansion operator.
366 PP/10 Bugzilla 937: provide a "debug" ACL control.
368 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
370 PP/12 Bugzilla 973: Implement --version.
372 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
374 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
376 PP/15 Bugzilla 816: support multiple condition rules on Routers.
378 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
379 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
380 ignore trailing whitespace.
382 JJ/02 prevent non-panic DKIM error from being sent to paniclog
384 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
387 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
388 Notification from Dr Andrew Aitchison.
390 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
391 ExtendedDetectionInfo response format.
392 Notification from John Horne.
394 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
397 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
398 XSL and documented dependency on system catalogs, with examples of how
401 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
404 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
405 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
406 configuration file which is writeable by the Exim user or group.
408 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
409 of configuration files to cover files specified with the -C option if
410 they are going to be used with root privileges, not just the default
413 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
414 option (effectively making it always true).
416 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
417 files to be used while preserving root privileges.
419 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
420 that rogue child processes cannot use them.
422 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
423 run-time user, instead of root.
425 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
426 Exim run-time user without dropping privileges.
428 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
429 result string, instead of calling string_vformat() twice with the same
432 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
433 for other users. Others should always drop root privileges if they use
434 -C on the command line, even for a whitelisted configure file.
436 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
438 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
444 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
445 $data_path, and $header_path variables; fixed documentation bugs and
448 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
449 exipick to access non-standard spools, including the "frozen" queue
452 NM/01 Bugzilla 965: Support mysql stored procedures.
453 Patch from Alain Williams
455 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
457 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
458 Patch from Andreas Metzler
460 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
461 Patch from Kirill Miazine
463 NM/05 Bugzilla 671: Added umask to procmail example.
465 JJ/03 installed exipick 20100323.0, fixing doc bug
467 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
468 directory. Notification and patch from Dan Rosenberg.
470 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
472 TK/02 Improve log output when DKIM signing operation fails.
474 MH/01 Treat the transport option dkim_domain as a colon separated
475 list, not as a single string, and sign the message with each element,
476 omitting multiple occurences of the same signer.
478 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
479 Bugzilla 985, 986. Patch by Simon Arlott
481 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
482 Patch by Simon Arlott
484 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
485 MBX locking. Notification from Dan Rosenberg.
491 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
493 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
495 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
497 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
499 NM/04 Bugzilla 915: Flag broken perl installation during build.
505 TK/01 Added patch by Johannes Berg that expands the main option
506 "spamd_address" if it starts with a dollar sign.
508 TK/02 Write list of recipients to X-Envelope-Sender header when building
509 the mbox-format spool file for content scanning (suggested by Jakob
512 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
513 (http://www.dcc-servers.net/) support via dccifd. Activated by
514 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
516 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
517 by Mark Daniel Reidel <mr@df.eu>.
519 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
520 When building exim an external PCRE library is now needed -
521 PCRE is a system library on the majority of modern systems.
522 See entry on PCRE_LIBS in EDITME file.
524 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
525 conversation. Added nologin parameter to request.
526 Patch contributed by Kirill Miazine.
528 TF/01 Do not log submission mode rewrites if they do not change the address.
530 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
532 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
533 log files in place. Contributed by Roberto Lima.
535 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
537 TF/03 Bugzilla 615: When checking the local_parts router precondition
538 after a local_part_suffix or local_part_prefix option, Exim now
539 does not use the address's named list lookup cache, since this
540 contains cached lookups for the whole local part.
542 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
543 Robert Millan. Documentation is in experimental-spec.txt.
545 TF/04 Bugzilla 668: Fix parallel build (make -j).
547 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
549 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
550 Patch provided by Jan Srzednicki.
552 TF/05 Leading white space used to be stripped from $spam_report which
553 wrecked the formatting. Now it is preserved.
555 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
556 that they are available at delivery time.
558 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
560 TF/08 TLS error reporting now respects the incoming_interface and
561 incoming_port log selectors.
563 TF/09 Produce a more useful error message if an SMTP transport's hosts
564 setting expands to an empty string.
566 NM/06 Bugzilla 744: EXPN did not work under TLS.
567 Patch provided by Phil Pennock.
569 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
570 Patch provided by Richard Godbee.
572 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
573 acl_smtp_notquit, added index entry.
575 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
576 Patch provided by Eugene Bujak.
578 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
579 accept(). Patch provided by Maxim Dounin.
581 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
582 Patch provided by Phil Pennock.
584 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
586 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
587 Patch provided by Brad "anomie" Jorsch.
589 NM/14 Bugzilla 622: Added timeout setting to transport filter.
590 Patch provided by Dean Brooks.
592 TK/05 Add native DKIM support (does not depend on external libraries).
594 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
595 Patch provided by Graeme Fowler.
597 NM/16 Bugzilla 851: Documentation example syntax fix.
599 NM/17 Changed NOTICE file to remove references to embedded PCRE.
601 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
604 NM/19 Bugzilla 745: TLS version reporting.
605 Patch provided by Phil Pennock.
607 NM/20 Bugzilla 167: bool: condition support.
608 Patch provided by Phil Pennock.
610 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
611 clients. Patch provided by Phil Pennock.
613 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
614 Patch provided by Brad "anomie" Jorsch.
616 NM/23 Bugzilla 687: Fix misparses in eximstats.
617 Patch provided by Heiko Schlittermann.
619 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
620 Patch provided by Heiko Schlittermann.
622 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
623 plus update to original patch.
625 NM/26 Bugzilla 799: Documentation correction for ratelimit.
627 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
628 Patch provided by David Brownlee.
630 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
632 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
634 NM/30 Bugzilla 888: TLS documentation bugfixes.
636 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
638 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
639 Unlike the original bugzilla I have changed all shell scripts in src tree.
641 NM/33 Bugzilla 898: Transport filter timeout fix.
642 Patch by Todd Rinaldo.
644 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
645 Patch by Serge Demonchaux.
647 NM/35 Bugzilla 39: Base64 decode bug fixes.
648 Patch by Jakob Hirsch.
650 NM/36 Bugzilla 909: Correct connect() call in dcc code.
652 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
654 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
656 NM/39 Bugzilla 911: Fixed MakeLinks build script.
662 TK/01 Add preliminary DKIM support. Currently requires a forked version of
663 ALT-N's libdkim that I have put here:
664 http://duncanthrax.net/exim-experimental/
666 Note to Michael Haardt: I had to rename some vars in sieve.c. They
667 were called 'true' and it seems that C99 defines that as a reserved
668 keyword to be used with 'bool' variable types. That means you could
669 not include C99-style headers which use bools without triggering
670 build errors in sieve.c.
672 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
673 as mailq or other aliases. Changed the --help handling significantly
674 to do whats expected. exim_usage() emits usage/help information.
676 SC/01 Added the -bylocaldomain option to eximstats.
678 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
680 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
682 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
688 PH/01 Another patch from the Sieve maintainer.
690 PH/02 When an IPv6 address is converted to a string for single-key lookup
691 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
692 dots are used instead of colons so that keys in lsearch files need not
693 contain colons. This was done some time before quoting was made available
694 in lsearch files. However, iplsearch files do require colons in IPv6 keys
695 (notated using the quote facility) so as to distinguish them from IPv4
696 keys. This meant that lookups for IP addresses in host lists did not work
697 for iplsearch lookups.
699 This has been fixed by arranging for IPv6 addresses to be expressed with
700 colons if the lookup type is iplsearch. This is not incompatible, because
701 previously such lookups could never work.
703 The situation is now rather anomolous, since one *can* have colons in
704 ordinary lsearch keys. However, making the change in all cases is
705 incompatible and would probably break a number of configurations.
707 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
710 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
711 conversion specification without a maximum field width, thereby enabling
712 a rogue spamd server to cause a buffer overflow. While nobody in their
713 right mind would setup Exim to query an untrusted spamd server, an
714 attacker that gains access to a server running spamd could potentially
715 exploit this vulnerability to run arbitrary code as the Exim user.
717 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
718 $primary_hostname instead of what libspf2 thinks the hosts name is.
720 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
721 a directory entry by the name of the lookup key. Previously, if a
722 symlink pointed to a non-existing file or a file in a directory that
723 Exim lacked permissions to read, a lookup for a key matching that
724 symlink would fail. Now it is enough that a matching directory entry
725 exists, symlink or not. (Bugzilla 503.)
727 PH/03 The body_linecount and body_zerocount variables are now exported in the
730 PH/04 Added the $dnslist_matched variable.
732 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
733 This means they are set thereafter only if the connection becomes
736 PH/06 Added the client_condition to authenticators so that some can be skipped
737 by clients under certain conditions.
739 PH/07 The error message for a badly-placed control=no_multiline_responses left
740 "_responses" off the end of the name.
742 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
744 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
745 (without spaces) instead of just copying the configuration text.
747 PH/10 Added the /noupdate option to the ratelimit ACL condition.
749 PH/11 Added $max_received_linelength.
751 PH/12 Added +ignore_defer and +include_defer to host lists.
753 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
754 way in which PCRE > 7.0 is built.
756 PH/14 Implemented queue_only_load_latch.
758 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
759 MAIL command. The effect was to mangle the value on 64-bit systems.
761 PH/16 Another patch from the Sieve maintainer.
763 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
765 PH/18 If a system quota error occurred while trying to create the file for
766 a maildir delivery, the message "Mailbox is full" was not appended to the
767 bounce if the delivery eventually timed out. Change 4.67/27 below applied
768 only to a quota excession during the actual writing of the file.
770 PH/19 It seems that peer DN values may contain newlines (and other non-printing
771 characters?) which causes problems in log lines. The DN values are now
772 passed through string_printing() before being added to log lines.
774 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
775 and InterBase are left for another time.)
777 PH/21 Added message_body_newlines option.
779 PH/22 Guard against possible overflow in moan_check_errorcopy().
781 PH/23 POSIX allows open() to be a macro; guard against that.
783 PH/24 If the recipient of an error message contained an @ in the local part
784 (suitably quoted, of course), incorrect values were put in $domain and
785 $local_part during the evaluation of errors_copy.
791 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
792 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
795 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
796 issue a MAIL command.
798 PH/02 In an ACL statement such as
800 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
802 if a client was not listed at all, or was listed with a value other than
803 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
804 the condition was not true (as it should be), so access was not denied.
805 The bug was that the ! inversion was incorrectly passed on to the second
806 item. This has been fixed.
808 PH/03 Added additional dnslists conditions == and =& which are different from
809 = and & when the dns lookup returns more than one IP address.
811 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
812 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
814 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
815 FSYNC, which compiles an option called disable_fsync that allows for
816 bypassing fsync(). The documentation is heavily laced with warnings.
818 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
820 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
821 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
822 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
823 including adding "make clean"; (3) Added -fPIC when compiling the test
824 dynamically loaded module, to get rid of a warning.
826 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
827 message fails, move_frozen_messages = true and ignore_bounce_errors_after
828 = 0s. The bug is otherwise harmless.
830 PH/07 There was a bug in the dovecot authenticator such that the value of
831 $auth1 could be overwritten, and so not correctly preserved, after a
832 successful authentication. This usually meant that the value preserved by
833 the server_setid option was incorrect.
835 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
837 PH/09 Installed PCRE release 7.0.
839 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
840 run for batched SMTP input. It is now run at the start of every message
841 in the batch. While fixing this I discovered that the process information
842 (output by running exiwhat) was not always getting set for -bs and -bS
843 input. This is fixed, and it now also says "batched" for BSMTP.
845 PH/11 Added control=no_pipelining.
847 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
848 patch, slightly modified), and move the expansion of helo_data till after
849 the connection is made in the smtp transport (so it can use these
852 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
854 PH/14 Added log_selector = +pid.
856 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
858 PH/16 Add ${if forany and ${if forall.
860 PH/17 Added dsn_from option to vary the From: line in DSNs.
862 PH/18 Flush SMTP output before performing a callout, unless control =
863 no_callout_flush is set.
865 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
866 was true (the default) a successful delivery failed to delete the retry
867 item, thus causing premature timeout of the address. The bug is now
870 PH/20 Added hosts_avoid_pipelining to the smtp transport.
872 PH/21 Long custom messages for fakedefer and fakereject are now split up
873 into multiline reponses in the same way that messages for "deny" and
874 other ACL rejections are.
876 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
877 with slight modification.
879 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
880 draft, changing the syntax and factoring some duplicate code".
882 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
883 for deliveries of the second and subsequent messages over the same SMTP
886 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
887 ${reduce, with only minor "tidies".
889 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
891 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
892 expansion side effects.
894 PH/27 When a message times out after an over-quota error from an Exim-imposed
895 quota, the bounce message says "mailbox is full". This message was not
896 being given when it was a system quota that was exceeded. It now should
899 MH/03 Made $recipients available in local_scan(). local_scan() already has
900 better access to the recipient list through recipients_list[], but
901 $recipients can be useful in postmaster-provided expansion strings.
903 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
904 in the case of a MAIL command with additional options following the
905 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
906 were accidentally chopped off.
908 PH/29 SMTP synchronization checks are implemented when a command is read -
909 there is a check that no more input is waiting when there shouldn't be
910 any. However, for some commands, a delay in an ACL can mean that it is
911 some time before the response is written. In this time, more input might
912 arrive, invalidly. So now there are extra checks after an ACL has run for
913 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
914 pipelining has not been advertised.
916 PH/30 MH's patch to allow iscntrl() characters to be list separators.
918 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
919 returned in the SMTP response when smtp_return_error_details was false.
922 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
923 instead of the C I/O that was originally supplied, because problems were
926 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
927 Exim which did not show up earlier: it was assuming that a call to
928 SSL_CTX_set_info_callback() might give an error value. In fact, there is
929 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
930 was a macro that became an assignment, so it seemed to work. This has
931 changed to a proper function call with a void return, hence the compile
932 error. Exim's code has been fixed.
934 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
937 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
939 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
941 PH/37 If a message is not accepted after it has had an id assigned (e.g.
942 because it turns out to be too big or there is a timeout) there is no
943 "Completed" line in the log. When some messages of this type were
944 selected by exigrep, they were listed as "not completed". Others were
945 picked up by some special patterns. I have improved the selection
946 criteria to be more general.
948 PH/38 The host_find_failed option in the manualroute router can now be set
949 to "ignore", to completely ignore a host whose IP address cannot be
950 found. If all hosts are ignored, the behaviour is controlled by the new
951 host_all_ignored option.
953 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
954 homing or because of multiple MX records with /mx) generated more than
955 one IP address, and the following item turned out to be the local host,
956 all the secondary addresses of the first item were incorrectly removed
957 from the list, along with the local host and any following hosts (which
958 is what is supposed to happen).
960 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
961 whoever called Exim into the -H file. In the case of the daemon it was
962 behaving confusingly. When first started, it used values for whoever
963 started the daemon, but after a SIGHUP it used the Exim user (because it
964 calls itself on a restart). I have changed the code so that it now always
967 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
968 message are rejected with the same error (e.g. no authentication or bad
969 sender address), and a DATA command is nevertheless sent (as can happen
970 with PIPELINING or a stupid MUA), the error message that was given to the
971 RCPT commands is included in the rejection of the DATA command. This is
972 intended to be helpful for MUAs that show only the final error to their
975 PH/42 Another patch from the Sieve maintainer.
977 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
978 Eximstats - Fixed some broken HTML links and added missing column headers
980 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
981 columns for Rejects, Temp Rejects, Ham, and Spam rows.
983 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
985 PH/43 Yet another patch from the Sieve maintainer.
987 PH/44 I found a way to check for a TCP/IP connection going away before sending
988 the response to the final '.' that terminates a message, but only in the
989 case where the client has not sent further data following the '.'
990 (unfortunately, this is allowed). However, in many cases there won't be
991 any further data because there won't be any more messages to send. A call
992 to select() can be used: if it shows that the input is "ready", there is
993 either input waiting, or the socket has been closed. An attempt to read
994 the next input character can distinguish the two cases. Previously, Exim
995 would have sent an OK response which the client would never have see.
996 This could lead to message repetition. This fix should cure that, at
997 least in a lot of common cases.
999 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1000 advertised in response to EHLO.
1006 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1007 fixed by 4.65/MH/01 (is this a record?) are fixed:
1009 (i) An empty string was always treated as zero by the numeric comparison
1010 operators. This behaviour has been restored.
1012 (ii) It is documented that the numeric comparison operators always treat
1013 their arguments as decimal numbers. This was broken in that numbers
1014 starting with 0 were being interpreted as octal.
1016 While fixing these problems I realized that there was another issue that
1017 hadn't been noticed. Values of message_size_limit (both the global option
1018 and the transport option) were treated as octal if they started with 0.
1019 The documentation was vague. These values are now always treated as
1020 decimal, and I will make that clear in the documentation.
1026 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1027 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1030 MH/01 Don't check that the operands of numeric comparison operators are
1031 integers when their expansion is in "skipping" mode (fixes bug
1032 introduced by 4.64-PH/07).
1034 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1035 child addresses, Exim now panics and dies. Previously, because the count
1036 is held in a short int, deliveries were likely to be lost. As such a
1037 large number of recipients for a single message is ridiculous
1038 (performance will be very, very poor), I have chosen to impose a limit
1039 rather than extend the field.
1045 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1046 leftover -K file (the existence of which was triggered by #402).
1047 While we were at it, introduced process PID as part of the -K
1048 filename. This should rule out race conditions when creating
1051 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1052 processing considerably. Previous code took too long for large mails,
1053 triggering a timeout which in turn triggers #401.
1055 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1056 in the DK code in transports.c. sendfile() is not really portable,
1057 hence the _LINUX specificness.
1059 TF/01 In the add_headers option to the mail command in an Exim filter,
1060 there was a bug that Exim would claim a syntax error in any
1061 header after the first one which had an odd number of characters
1064 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1065 callout verification, Exim cached a "reject" for the entire domain. This
1066 is correct for most verifications, but it is not correct for a recipient
1067 verification with use_sender or use_postmaster set, because in that case
1068 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1069 case of MAIL FROM:<> rejection from other early rejections (e.g.
1070 rejection of HELO). When verifying a recipient using a non-null MAIL
1071 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1072 Whatever the result of the callout, the value of the domain cache is
1073 left unchanged (for any other kind of callout, getting as far as trying
1074 RCPT means that the domain itself is ok).
1076 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1079 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1080 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1081 session, the daemon ignores ECONNECT errors and logs others; it now
1082 ignores EPIPE as well.
1084 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1085 (quoted-printable decoding).
1087 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1088 later the small subsequent patch to fix an introduced bug.
1090 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1092 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1094 PH/08 An error is now given if message_size_limit is specified negative.
1096 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1097 to be given (somewhat) arbitrary names.
1099 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1102 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1103 miscellaneous code fixes
1105 PH/10 Added the log_reject_target ACL modifier to specify where to log
1108 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1109 hostname. This is wrong, because it relates to the incoming message (and
1110 probably the interface on which it is arriving) and not to the outgoing
1111 callout (which could be using a different interface). This has been
1112 changed to use the value of the helo_data option from the smtp transport
1113 instead - this is what is used when a message is actually being sent. If
1114 there is no remote transport (possible with a router that sets up host
1115 addresses), $smtp_active_hostname is used.
1117 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1118 tweaks were necessary in order to get it to work (see also 21 below):
1119 (a) The code assumed that strncpy() returns a negative number on buffer
1120 overflow, which isn't the case. Replaced with Exim's string_format()
1122 (b) There were several signed/unsigned issues. I just did the minimum
1123 hacking in of casts. There is scope for a larger refactoring.
1124 (c) The code used strcasecmp() which is not a standard C function.
1125 Replaced with Exim's strcmpic() function.
1126 (d) The code set only $1; it now sets $auth1 as well.
1127 (e) A simple test gave the error "authentication client didn't specify
1128 service in request". It would seem that Dovecot has changed its
1129 interface. Fortunately there's a specification; I followed it and
1130 changed what the client sends and it appears to be working now.
1132 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1135 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1136 address A is aliased to B and C, where B exists and C does not. Without
1137 -v the output is "A verified" because verification stops after a
1138 successful redirection if more than one address is generated. However,
1139 with -v the child addresses are also verified. Exim was outputting "A
1140 failed to verify" and then showing the successful verification for C,
1141 with its parentage. It now outputs "B failed to verify", showing B's
1142 parentage before showing the successful verification of C.
1144 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1145 look up a TXT record in a specific list after matching in a combined
1148 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1149 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1150 they consult the DNS. I had assumed they would set it the way they
1151 wanted; and indeed my experiments on Linux seem to show that in some
1152 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1153 To be on the safe side, however, I have now made the interface to
1154 host_find_byname() similar to host_find_bydns(), with an argument
1155 containing the DNS resolver options. The host_find_byname() function now
1156 sets these options at its start, just as host_find_bydns() does. The smtp
1157 transport options dns_qualify_single and dns_search_parents are passed to
1158 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1159 of host_find_byname() use the default settings of RES_DEFNAMES
1160 (qualify_single) but not RES_DNSRCH (search_parents).
1162 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1163 spool_read_header() do less string testing, by means of a preliminary
1164 switch on the second character of optional "-foo" lines. (This is
1165 overdue, caused by the large number of possibilities that now exist.
1166 Originally there were few.) While I was there, I also converted the
1167 str(n)cmp tests so they don't re-test the leading "-" and the first
1168 character, in the hope this might squeeze out yet more improvement.
1170 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1171 flag allowing group syntax was set by the header_syntax check but not
1172 turned off, possible causing trouble later; (2) The flag was not being
1173 set at all for the header_verify test, causing "group"-style headers to
1174 be rejected. I have now set it in this case, and also caused header_
1175 verify to ignore an empty address taken from a group. While doing this, I
1176 came across some other cases where the code for allowing group syntax
1177 while scanning a header line wasn't quite right (mostly, not resetting
1178 the flag correctly in the right place). These bugs could have caused
1179 trouble for malformed header lines. I hope it is now all correct.
1181 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1182 with the "reply" argument non-NULL. The code, however (which originally
1183 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1184 but it didn't always do it. This confused somebody who was copying the
1185 code for some other use. I have removed all the tests.
1187 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1188 feature that was used to support insecure browsers during the U.S. crypto
1189 embargo. It requires special client support, and Exim is probably the
1190 only MTA that supported it -- and would never use it because real RSA is
1191 always available. This code has been removed, because it had the bad
1192 effect of slowing Exim down by computing (never used) parameters for the
1193 RSA_EXPORT functionality.
1195 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1196 authenticator to fail if there's a tab character in the incoming data
1197 (there should never be unless someone is messing about, as it's supposed
1198 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1199 if the connection is using TLS or if the remote IP is the same as the
1200 local IP, and the "valid-client-cert option" if a client certificate has
1203 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1204 authenticators. This can be used for authorization after authentication
1205 succeeds. (In the case of plaintext, it servers for both authentication
1208 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1209 if any retry times were supplied.
1211 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1212 connection, where there is no client IP address to check. In this
1213 situation, the verify now always succeeds.
1215 PH/25 Applied John Jetmore's -Mset patch.
1217 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1219 PH/27 In a string expansion for a processed (not raw) header when multiple
1220 headers of the same name were present, leading whitespace was being
1221 removed from all of them, but trailing whitespace was being removed only
1222 from the last one. Now trailing whitespace is removed from each header
1223 before concatenation. Completely empty headers in a concatenation (as
1224 before) are ignored.
1226 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1227 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1229 PH/29 [Removed. This was a change that I later backed out, and forgot to
1230 correct the ChangeLog entry (that I had efficiently created) before
1231 committing the later change.]
1233 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1234 address errors (4xx response to RCPT) over the same connection as other
1235 messages routed to the same hosts. Such deliveries are always "forced",
1236 so retry times are not inspected. This resulted in far too many retries
1237 for the affected addresses. The effect occurred only when there were more
1238 hosts than the hosts_max_try setting in the smtp transport when it had
1239 the 4xx errors. Those hosts that it had tried were not added to the list
1240 of hosts for which the message was waiting, so if all were tried, there
1241 was no problem. Two fixes have been applied:
1243 (i) If there are any address or message errors in an SMTP delivery, none
1244 of the hosts (tried or untried) are now added to the list of hosts
1245 for which the message is waiting, so the message should not be a
1246 candidate for sending over the same connection that was used for a
1247 successful delivery of some other message. This seems entirely
1248 reasonable: after all the message is NOT "waiting for some host".
1249 This is so "obvious" that I'm not sure why it wasn't done
1250 previously. Hope I haven't missed anything, but it can't do any
1251 harm, as the worst effect is to miss an optimization.
1253 (ii) If, despite (i), such a delivery is accidentally attempted, the
1254 routing retry time is respected, so at least it doesn't keep
1255 hammering the server.
1257 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1258 in ${readsocket because some servers need this prod.
1260 PH/32 Added some extra debug output when updating a wait-xxx database.
1262 PH/33 The hint "could be header name not terminated by colon", which has been
1263 given for certain expansion errors for a long time, was not being given
1264 for the ${if def:h_colon_omitted{... case.
1266 PH/34 The spec says: "With one important exception, whenever a domain list is
1267 being scanned, $domain contains the subject domain." There was at least
1268 one case where this was not true.
1270 PH/35 The error "getsockname() failed: connection reset by peer" was being
1271 written to the panic log as well as the main log, but it isn't really
1272 panic-worthy as it just means the connection died rather early on. I have
1273 removed the panic log writing for the ECONNRESET error when getsockname()
1276 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1277 runs only) independently of the message's sender address. This meant
1278 that, if the 4xx error was in fact related to the sender, a different
1279 message to the same recipient with a different sender could confuse
1280 things. In particualar, this can happen when sending to a greylisting
1281 server, but other circumstances could also provoke similar problems.
1282 I have changed the default so that the retry time for these errors is now
1283 based a combination of the sender and recipient addresses. This change
1284 can be overridden by setting address_retry_include_sender=false in the
1287 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1288 remote server are returned as part of bounce messages. This was not
1289 happening for LMTP over a pipe (the lmtp transport), but now it is the
1290 same for both kinds of LMTP.
1292 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1293 in header lines that were in fact CNAMEs. This is no longer the case.
1295 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1296 and queue runs started by the daemon processed all messages. This has
1297 been fixed so that -R and -S can now usefully be given with -q<time>.
1299 PH/40 Import PCRE release 6.7 (fixes some bugs).
1301 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1303 PH/42 Give an error if -q is specified more than once.
1305 PH/43 Renamed the variables $interface_address and $interface_port as
1306 $received_ip_address and $received_port, to make it clear that these
1307 values apply to message reception, and not to the outgoing interface when
1308 a message is delivered. (The old names remain recognized, of course.)
1310 PH/44 There was no timeout on the connect() call when using a Unix domain
1311 socket in the ${readsocket expansion. There now is.
1313 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1314 be meaningful with "accept".
1316 SC/01 Eximstats V1.43
1317 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1319 SC/02 Eximstats V1.44
1320 Use a glob alias rather than an array ref in the generated
1321 parser. This improves both readability and performance.
1323 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1324 Collect SpamAssassin and rejection statistics.
1325 Don't display local sender or destination tables unless
1326 there is data to show.
1327 Added average volumes into the top table text output.
1329 SC/04 Eximstats V1.46
1330 Collect data on the number of addresses (recipients)
1331 as well as the number of messages.
1333 SC/05 Eximstats V1.47
1334 Added 'Message too big' to the list of mail rejection
1335 reasons (thanks to Marco Gaiarin).
1337 SC/06 Eximstats V1.48
1338 Mainlog lines which have GMT offsets and are too short to
1339 have a flag are now skipped.
1341 SC/07 Eximstats V1.49 (Alain Williams)
1342 Added the -emptyok flag.
1344 SC/08 Eximstats V1.50
1345 Fixes for obtaining the IP address from reject messages.
1347 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1348 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1349 whitesspace changes from 4.64-PH/27
1351 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1354 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1355 are found, allow negative numbers in numeric criteria)
1357 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1359 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1360 to match changes made in 4.64-PH/43
1362 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1364 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1365 "domain" is actually an IP address, and force a failure. This locks out
1366 those revolvers/nameservers that support "A-for-A" lookups, in
1367 contravention of the specifications.
1369 PH/48 When a host name was looked up from an IP address, and the subsequent
1370 forward lookup of the name timed out, the host name was left in
1371 $sender_host_name, contrary to the specification.
1373 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1374 restricted to single-key lookups, Exim was not diagnosing an error if
1375 * or *@ was used with a query-style lookup.
1377 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1379 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1380 long ago, but noone interested enough thought of it. Let's just say that
1381 the "1.1" means that there are some new functions that weren't there at
1382 some point in the past.
1384 PH/51 Error processing for expansion failure of helo_data from an smtp
1385 transport during callout processing was broken.
1387 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1388 tested/used via the -bh/-bhc/-bs options.
1390 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1391 bug, fixed in subsequent PCRE releases).
1393 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1394 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1396 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1402 SC/01 Use a glob alias rather than an array ref in eximstats generated
1403 parser. This improves both readability and performance.
1405 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1406 Don't display local sender or destination tables in eximstats unless
1407 there is data to show.
1408 Added average volumes into the eximstats top table text output.
1410 SC/03 Collect data on the number of addresses (recipients) as well
1411 as the number of messages in eximstats.
1413 TF/01 Correct an error in the documentation for the redirect router. Exim
1414 does (usually) call initgroups() when daemonizing.
1416 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1417 with consistent privilege compared to when running as a daemon.
1419 TF/03 Note in the spec that $authenticated_id is not set for local
1420 submissions from trusted users.
1422 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1423 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1425 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1426 by adding some example configuration directives to the default
1427 configuration file. A little bit of work is required to uncomment the
1428 directives and define how usernames and passwords are checked, but
1429 there is now a framework to start from.
1431 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1432 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1433 without this. I don't know how relevant this is to other LDAP libraries.
1435 PH/02 Add the verb name to the "unknown ACL verb" error.
1437 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1439 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1441 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1442 directory not expanded when it should be if an expanded home directory
1443 was set for the address (which is overridden by the transport).
1445 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1448 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1449 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1450 because it is too late at that time, and has no effect.
1452 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1453 security issue with \' (bugzilla #107). I could not use the
1454 PQescapeStringConn() function, because it needs a PGconn value as one of
1457 PH/08 When testing addresses using -bt, indicate those final addresses that
1458 are duplicates that would not cause an additional delivery. At least one
1459 person was confused, thinking that -bt output corresponded to deliveries.
1460 (Suppressing duplicates isn't a good idea as you lose the information
1461 about possibly different redirections that led to the duplicates.)
1463 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1464 systems where poll() doesn't work, in particular OS X.
1466 PH/10 Added more information to debugging output for retry time not reached.
1468 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1469 operations in malware.c.
1471 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1474 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1475 syslog_duplication set false, log lines that would normally be written
1476 both the the main log and to the reject log were not written to syslog at
1479 PH/14 In the default configuration, change the use of "message" in ACL warn
1480 statements to "add_header".
1482 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1483 not followed by a command (e.g. "seen endif").
1485 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1486 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1489 PH/17 Added extra conditions to the default value of delay_warning_condition
1493 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1494 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1495 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1498 The Auto-Submitted: and various List- headers are standardised, whereas I
1499 don't think Precedence: ever was.
1501 PH/18 Refactored debugging code in route_finduser() to show more information,
1502 in particular, the error code if getpwnam() issues one.
1504 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1505 This is apparently needed in addition to the PH/07 change above to avoid
1506 any possible encoding problems.
1508 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1509 but not after initializing Perl.
1511 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1512 output them only if debugging. By default they are written stderr,
1513 apparently, which is not desirable.
1515 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1518 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1521 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1523 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1524 authenticated or an ident call has been made. Suppress the default
1525 values for $authenticated_id and $authenticated_sender (but permit -oMai
1526 and -oMas) when testing with -bh.
1528 PH/24 Re-jigged the order of the tests in the default configuration so that the
1529 tests for valid domains and recipients precede the DNS black list and CSA
1530 tests, on the grounds that those ones are more expensive.
1532 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1533 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1534 This bug exists in every version of Exim that I still have, right back to
1537 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1538 However, an attempt to turn on case-sensitivity in a regex key by
1539 including (?-i) didn't work because the subject string was already
1540 lowercased, and the effects were non-intuitive. It turns out that a
1541 one-line patch can be used to allow (?-i) to work as expected.
1547 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1548 other effects) broke the use of negated acl sub-conditions.
1550 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1553 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1554 "Deny" causes Exim to reject the incoming connection with a 554 error.
1555 Unfortunately, if there is a major crisis, such as a disk failure,
1556 tcp-wrappers gives "deny", whereas what one would like would be some
1557 kind of temporary error. A kludge has been added to help with this.
1558 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1559 554 error is used if errno is still zero or contains ENOENT (which occurs
1560 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1563 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1565 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1566 errors. Otherwise a message that provokes a temporary error (when other
1567 messages do not) can cause a whole host to time out.
1569 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1570 addresses were routed directly to files or pipes from a redirect router.
1571 File deliveries just didn't batch; pipe deliveries might have suffered
1574 PH/06 A failure to get a lock for a hints database would erroneously always say
1575 "Failed to get write lock", even when it was really a read lock.
1577 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1578 of 0600. This has been changed to use the value of the lockfile_mode
1579 option (which defaults to 0600).
1581 PH/08 Applied small patch from the Sieve maintainer.
1583 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1584 folder from quota calculations, a direct delivery into this folder messed
1585 up the contents of the maildirsize file. This was because the regex was
1586 used only to exclude .Trash (or whatever) when the size of the mailbox
1587 was calculated. There was no check that a delivery was happening into an
1588 excluded directory. This bug has been fixed by ignoring all quota
1589 processing for deliveries into excluded directories.
1591 PH/10 Added the maildirfolder_create_regex option to appendfile.
1597 PH/01 The code for finding all the local interface addresses on a FreeBSD
1598 system running IPv6 was broken. This may well have applied to all BSD
1599 systems, as well as to others that have similar system calls. The broken
1600 code found IPv4 interfaces correctly, but gave incorrect values for the
1601 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1602 that it would not match correctly against @[] and not recognize the IPv6
1605 PH/02 The ipliteral router was not recognizing addresses of the form user@
1606 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1608 PH/03 Added disable_ipv6.
1610 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1611 decoded form, because it is most often used to construct To: headers
1612 lines in autoreplies, and the decoded form may well be syntactically
1613 invalid. However, $reply_address has leading white space removed, and all
1614 newlines turned into spaces so that the autoreply transport does not
1617 PH/05 If group was specified without a user on a router, and no group or user
1618 was specified on a transport, the group from the router was ignored.
1620 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1621 for visible compile-time settings that can be used to change these
1622 numbers, for those that want even more. Backwards compatibility with old
1623 spool files has been maintained. However, going back to a previous Exim
1624 release will lost any variables that are in spool files.
1626 PH/07 Two small changes when running in the test harness: increase delay when
1627 passing a TCP/IP connection to a new process, in case the original
1628 process has to generate a bounce, and remove special handling of
1629 127.0.0.2 (sic), which is no longer necessary.
1631 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1632 be the same on different OS.
1634 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1637 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1638 whether --show-vars was specified or not
1640 JJ/02 exipick: Added support for new ACL variable spool format introduced
1643 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1644 syntactically invalid From: or Reply-to: line, and a filter used this to
1645 generate an autoreply, and therefore failed to obtain an address for the
1646 autoreply, Exim could try to deliver to a non-existent relative file
1647 name, causing unrelated and misleading errors. What now happens is that
1648 it logs this as a hard delivery error, but does not attempt to create a
1651 PH/11 The exinext utility has a -C option for testing purposes, but although
1652 the given file was scanned by exinext itself; it wasn't being passed on
1653 when Exim was called.
1655 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1656 an end-of-file indication when reading a command response.
1658 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1659 compiled. In many other places in Exim, IPv6 addresses are always
1660 recognized, so I have changed this. It also means that IPv4 domain
1661 literals of the form [IPV4:n.n.n.n] are now always recognized.
1663 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1664 used if the router is not running as root, for example, when verifying at
1665 ACL time, or when using -bh. The debugging output from this situation was
1666 non-existent - all you got was a failure to exec. I have made two
1669 (a) Failures to set uid/gid, the current directory, or a process leader
1670 in a subprocess such as that created by queryprogram now generate
1671 suitable debugging ouput when -d is set.
1673 (b) The queryprogram router detects when it is not running as root,
1674 outputs suitable debugging information if -d is set, and then runs
1675 the subprocess without attempting to change uid/gid.
1677 PH/15 Minor change to Makefile for building test_host (undocumented testing
1680 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1681 additional section of a DNS packet that returns MX or SRV records.
1682 Instead, it always explicitly searches for A/AAAA records. This avoids
1683 major problems that occur when a DNS server includes only records of one
1684 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1685 fixed another bug: if SRV records were looked up and the corresponding
1686 address records were *not* found in the additional section, the port
1687 values from the SRV records were lost.
1689 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1690 using the correct key (the original address) when searching the retry
1691 rules in order to find which one to use for generating the retry hint.
1693 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1694 adding the default one. Similarly, if it contains a Reply-To: header, the
1695 errors_reply_to option, if set, is not used.
1697 PH/19 When calculating a retry time, Exim used to measure the "time since
1698 failure" by looking at the "first failed" field in the retry record. Now
1699 it does not use this if it is later than than the arrival time of the
1700 message. Instead it uses the arrival time. This makes for better
1701 behaviour in cases where some deliveries succeed, thus re-setting the
1702 "first failed" field. An example is a quota failure for a huge message
1703 when small messages continue to be delivered. Without this change, the
1704 "time since failure" will always be short, possible causing more frequent
1705 delivery attempts for the huge message than are intended.
1706 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1708 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1709 $1, $2, $3) because the numerical variables can be reset during some
1710 expansion items (e.g. "match"), thereby losing the authentication data.
1712 PH/21 Make -bV show the size of off_t variables so that the test suite can
1713 decide whether to run tests for quotas > 2G.
1715 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1716 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1717 filecount value is greater than 2G or if a quota value is greater than 2G
1718 on a system where the size of off_t is not greater than 4, a panic error
1721 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1722 never match. The debug and -bh output now contains an explicit error
1723 message indicating a malformed IPv4 address or mask.
1725 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1726 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1727 PH/23 above applies.
1729 PH/25 Do not write to syslog when running in the test harness. The only
1730 occasion when this arises is a failure to open the main or panic logs
1731 (for which there is an explicit test).
1733 PH/26 Added the /no_tell option to "control=freeze".
1735 PH/27 If a host name lookup failed very early in a connection, for example, if
1736 the IP address matched host_lookup and the reverse lookup yielded a name
1737 that did not have a forward lookup, an error message of the form "no IP
1738 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1739 could be logged. Now it outputs the IP address instead of "NULL".
1741 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1742 allows the sender and the authenticated sender to be set when
1743 submitting a message from within Exim. Since child_open_exim() is
1744 documented for local_scan(), the new function should be too.
1746 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1747 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1748 results in an empty string is now treated as unset.
1750 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1752 PH/31 Added sender_verify_fail logging option.
1754 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1755 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1756 tidied the source and removed it altogether.
1758 PH/33 When a queue run was abandoned because the load average was too high, a
1759 log line was always written; now it is written only if the queue_run log
1760 selector is set. In addition, the log line for abandonment now contains
1761 information about the queue run such as the pid. This is always present
1762 in "start" and "stop" lines but was omitted from the "abandon" line.
1764 PH/34 Omit spaces between a header name and the colon in the error message that
1765 is given when verify = headers_syntax fails (if there are lots of them,
1766 the message gets confusing).
1768 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1769 names, as there are now some PTR records that contain slashes. This check
1770 is only to protect against broken name servers that fall over on strange
1771 characters, so the fact that it applies to all lookups doesn't matter.
1773 PH/36 Now that the new test suite is complete, we can remove some of the
1774 special code in Exim that was needed for the old test suite. For example,
1775 sorting DNS records because real resolvers return them in an arbitrary
1776 order. The new test suite's fake resolver always returns records in the
1779 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1780 bounces) except when queue_only is set, to avoid logging races between
1781 the different processes.
1783 PH/38 Panic-die if .include specifies a non-absolute path.
1785 PH/39 A tweak to the "H" retry rule from its user.
1787 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1788 a label. They prevented compilation on older perls.
1790 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1791 a warning to be raised on newish perls.
1793 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1794 on queue. Changes to match documented behaviour of showing count of
1795 messages matching specified criteria.
1797 PH/40 Changed the default ident timeout from 30s to 5s.
1799 PH/41 Added support for the use of login_cap features, on those BSD systems
1800 that have them, for controlling the resources used by pipe deliveries.
1802 PH/42 The content-scanning code uses fopen() to create files in which to put
1803 message data. Previously it was not paying any attention to the mode of
1804 the files. Exim runs with umask(0) because the rest of the code creates
1805 files with open(), and sets the required mode explicitly. Thus, these
1806 files were ending up world-writeable. This was not a big issue, because,
1807 being within the spool directory, they were not world-accessible. I have
1808 created a function called modefopen, which takes an additional mode
1809 argument. It sets umask(777), creates the file, chmods it to the required
1810 mode, then resets the umask. All the relevant calls to fopen() in the
1811 content scanning code have been changed to use this function.
1813 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1814 to 24 hours. This avoids potential overflow problems when processing G
1815 and H retry rules. I suspect nobody ever tinkers with this value.
1817 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1819 PH/45 When the plaintext authenticator is running as a client, the server's
1820 challenges are checked to ensure they are valid base64 strings. By
1821 default, the authentication attempt is cancelled if an invalid string is
1822 received. Setting client_ignore_invalid_base64 true ignores these errors.
1823 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1824 they are received. Thus, the responses can be made to depend on the
1825 challenges. If an invalid string is ignored, an empty string is placed in
1828 PH/46 Messages that are created by the autoreply transport now contains a
1829 References: header, in accordance with RFCs 2822 and 3834.
1831 PH/47 Added authenticated_sender_force to the smtp transport.
1833 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1835 PH/49 Installed latest patch from the Sieve maintainer.
1837 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1838 was also set, the appendfile transport was unnecessarily scanning a
1839 directory of message files (e.g. for maildir delivery) to find the count
1840 of files (along with the size), even though it did not need this
1841 information. It now does the scan only if it needs to find either the
1842 size of the count of files.
1844 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1846 PH/52 Two bugs concerned with error handling when the smtp transport is
1849 (i) Exim was not creating retry information for temporary errors given
1850 for individual recipients after the DATA command when the smtp transport
1851 was used in LMTP mode. This meant that they could be retried too
1852 frequently, and not timed out correctly.
1854 (ii) Exim was setting the flag that allows error details to be returned
1855 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1856 recipients that were returned after the DATA command.
1858 PH/53 This is related to PH/52, but is more general: for any failing address,
1859 when detailed error information was permitted to be returned to the
1860 sender, but the error was temporary, then after the final timeout, only
1861 "retry timeout exceeded" was returned. Now it returns the full error as
1862 well as "retry timeout exceeded".
1864 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1865 do this, and (what is worse) MTAs that accept it.
1867 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1868 will now be deprecated.
1870 PH/56 New os.c-cygwin from the Cygwin maintainer.
1872 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1873 formats (previously only available in exim formats via -bpr, -bpru,
1874 and -bpra. Now also available in native and exiqgrep formats)
1876 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1877 with very large, slow to parse queues
1879 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1881 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1883 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1884 responses to authentication challenges, though it was showing the
1885 challenges; (ii) I've removed the CR characters from the debug output for
1888 PH/58 Allow for the insertion of a newline as well as a space when a string
1889 is turned into more than one encoded-word during RFC 2047 encoding. The
1890 Sieve code now uses this.
1892 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1893 data_4xx, lost_connection, tls_required.
1895 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1896 message was being sent as an SMTP response.
1898 PH/61 Add -l and -k options to exicyclog.
1900 PH/62 When verifying, if an address was redirected to one new address, so that
1901 verification continued, and the new address failed or deferred after
1902 having set something in $address_data, the value of $address_data was not
1903 passed back to the ACL. This was different to the case when no
1904 redirection occurred. The value is now passed back in both cases.
1906 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1907 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1908 use login_cap.h, so on its own it isn't the distinguishing feature. The
1909 new name refers directly to the setclassresources() function.
1911 PH/65 Added configuration files for NetBSD3.
1913 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1915 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1916 is preferred over IPv4.
1918 PH/68 The bounce_return_message and bounce_return_body options were not being
1919 honoured for bounces generated during the reception of non-SMTP messages.
1920 In particular, this applied to messages rejected by the ACL. This bug has
1921 been fixed. However, if bounce_return_message is true and bounce_return_
1922 body is false, the headers that are returned for a non-SMTP message
1923 include only those that have been read before the error was detected.
1924 (In the case of an ACL rejection, they have all been read.)
1926 PH/69 The HTML version of the specification is now built in a directory called
1927 spec_html instead of spec.html, because the latter looks like a path with
1928 a MIME-type, and this confuses some software.
1930 PH/70 Catch two compiler warnings in sieve.c.
1932 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1933 function verify_get_ident() calls ip_connect() to connect a socket, but
1934 if the "connect()" function timed out, ip_connect() used to close the
1935 socket. However, verify_get_ident() also closes the socket later, and in
1936 between Exim writes to the log, which may get opened at this point. When
1937 the socket was closed in ip_connect(), the log could get the same file
1938 descriptor number as the socket. This naturally causes chaos. The fix is
1939 not to close the socket in ip_connect(); the socket should be closed by
1940 the function that creates it. There was only one place in the code where
1941 this was missing, in the iplookup router, which I don't think anybody now
1942 uses, but I've fixed it anyway.
1944 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1945 well as to direct DNS lookups. Otherwise the handling of names in host
1946 lists is inconsistent and therefore confusing.
1952 PH/01 Two changes to the default runtime configuration:
1954 (1) Move the checks for relay_from_hosts and authenticated clients from
1955 after to before the (commented out) DNS black list checks.
1957 (2) Add control=submission to the relay_from_hosts and authenticated
1958 clients checks, on the grounds that messages accepted by these
1959 statements are most likely to be submissions.
1961 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1963 (1) Generate an error if the third argument for the ${prvs expansion is
1966 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1969 (3) Reset the variables that are obtained from the first argument of
1970 ${prvscheck and used in the second argument before leaving the code,
1971 because their memory is reclaimed, so using them afterwards may do
1974 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1975 one (it's much easier than Tom thought :-).
1977 (5) Because of (4), we can now allow for the use of $prvscheck_result
1978 inside the third argument.
1980 PH/03 For some reason, the default setting of PATH when running a command from
1981 a pipe transport was just "/usr/bin". I have changed it to
1984 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1985 anything to be listed in the output from -bV.
1987 PH/05 When a filter generated an autoreply, the entire To: header line was
1988 quoted in the delivery log line, like this:
1990 => >A.N.Other <ano@some.domain> <original@ddress> ...
1992 This has been changed so that it extracts the operative address. There
1993 may be more than one such address. If so, they are comma-separated, like
1996 => >ano@some.domain,ona@other.domain <original@ddress> ...
1998 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1999 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2000 not being looked up in the rDNS to get a host name, Exim was showing the
2001 IP address twice in Received: lines, even though the IP addresses were
2002 identical. For example:
2004 Received: from [1.2.3.4] (helo=[1.2.3.4])
2006 However, if the real host name was known, it was omitting the HELO data
2007 if it matched the actual IP address. This has been tidied up so that it
2008 doesn't show the same IP address twice.
2010 PH/07 When both +timestamp and +memory debugging was on, the value given by
2011 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2012 called by the debug printing, thereby overwriting the timestamp buffer.
2013 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2015 PH/08 When the original message was included in an autoreply transport, it
2016 always said "this is a copy of the message, including all the headers",
2017 even if body_only or headers_only was set. It now gives an appropriate
2020 PH/09 Applied a patch from the Sieve maintainer which:
2022 o fixes some comments
2023 o adds the (disabled) notify extension core
2024 o adds some debug output for the result of if/elsif tests
2025 o points to the current vacation draft in the documentation
2026 and documents the missing references header update
2030 o fixes a bug in processing the envelope test (when testing
2031 multiple envelope elements, the last element determinted the
2034 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2035 Electronic Mail") by including:
2037 Auto-submitted: auto-generated
2039 in the messages that it generates (bounce messages and others, such as
2040 warnings). In the case of bounce messages for non-SMTP mesages, there was
2041 also a typo: it was using "Auto_submitted" (underscore instead of
2042 hyphen). Since every message generated by Exim is necessarily in response
2043 to another message, thes have all been changed to:
2045 Auto-Submitted: auto-replied
2047 in accordance with these statements in the RFC:
2049 The auto-replied keyword:
2051 - SHOULD be used on messages sent in direct response to another
2052 message by an automatic process,
2054 - MUST NOT be used on manually-generated messages,
2056 - MAY be used on Delivery Status Notifications (DSNs) and Message
2057 Disposition Notifications (MDNs),
2059 - MUST NOT be used on messages generated by automatic or periodic
2060 processes, except for messages which are automatic responses to
2063 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2064 to the default Received: header definition.
2066 PH/12 Added log selector acl_warn_skipped (default on).
2068 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2069 variables because (a) they are in the wrong storage pool and (b) even if
2070 they were copied, it wouldn't work properly because of the caching.
2072 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2073 checking when decoding. Apparently there are clients that generate
2074 overlong encoded strings. Why am I not surprised?
2076 PH/15 If the first argument of "${if match_address" was not empty, but did not
2077 contain an "@" character, Exim crashed. Now it writes a panic log message
2078 and treats the condition as false.
2080 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2082 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2083 extension "envelope-auth". The code is finished and in agreement with
2084 other implementations, but there is no documentation so far and in fact,
2085 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2086 not changing the active code.
2088 Print executed "if" and "elsif" statements when debugging is used. This
2089 helps a great deal to understand what a filter does.
2091 Document more things not specified clearly in RFC3028. I had all this
2092 sorted out, when out of a sudden new issues came to my mind. Oops."
2094 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2097 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2098 canonical form (as documented). However, after a host name lookup from
2099 the IP address, check_host() was doing a simple string comparison with
2100 addresses acquired from the DNS when checking that the found name did
2101 have the original IP as one of its addresses. Since any found IPv6
2102 addresses are likely to be in abbreviated form, the comparison could
2103 fail. Luckily, there already exists a function for doing the comparison
2104 by converting both addresses to binary, so now that is used instead of
2105 the text comparison.
2107 PH/20 There was another similar case to PH/19, when a complete host name was
2108 given in a host list; looking up its IP address could give an abbreviated
2109 form, whereas the current host's name might or might not be abbreviated.
2110 The same fix has been applied.
2116 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2117 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2120 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2121 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2123 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2125 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2126 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2127 Also, for "Auto-Submitted", treat anything other than "no" as
2128 non-personal, in accordance with RFC 3834. (Previously it treated
2129 anything starting "auto-" as non-personal.)
2131 TF/01 The control=submission/name=... option had a problem with syntax
2132 errors if the name included a slash character. The /name= option
2133 now slurps the rest of the string, so it can include any characters
2134 but it must come last in the list of options (after /sender_retain
2137 PH/05 Some modifications to the interface to the fake nameserver for the new
2145 TK/01 Added the "success_on_redirect" address verification option. See
2146 NewStuff for rationale and an example.
2148 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2150 PH/02 Patch to exigrep to allow it to work on syslog lines.
2152 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2153 fread() to skip over the body file's header line, because in Cygwin the
2154 header line is locked and is inaccessible.
2156 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2157 co-exist for some time) to make it clear that it is the Exim ID that is
2158 referenced, not the Message-ID: header line.
2160 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2161 string_format() function, because snprintf() does not exist on all
2164 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2165 ${stat: expansion item.
2167 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2168 protocol synchronization error", to keep the pedants happy.
2170 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2171 well as for IRIX systems, when gcc is being used. See the host.c source
2174 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2176 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2179 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2180 transport and to the smtp transport in LMTP mode.
2182 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2184 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2185 run a filter in a subprocess. This could lead to confusion in subsequent
2186 lookups in the parent process. There should also be a search_tidyup() at
2187 the end of the subprocess.
2189 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2190 only if the host matched helo_try_verify_hosts, which caused the
2191 verification to occur when the EHLO/HELO command was issued. The ACL just
2192 tested the remembered result. Now, if a previous verification attempt has
2193 not happened, "verify = helo" does it there and then.
2195 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2197 TK/03 Fix log output including CR from clamd.
2199 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2200 exist provoked a memory error which could cause a segfault.
2202 PH/15 Installed PCRE 6.2
2204 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2206 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2207 of the problem. Specifically, suggested +O2 rather than +O1 for the
2210 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2212 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2213 record, the port was not correctly logged when the outgoing_port log
2214 selector was set (it logged the transort's default port).
2216 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2217 fallback_hosts, and "hosts" in the smtp transport.
2219 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2220 host errors such as "Connection refused".
2222 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2223 authentication with radiusclient 0.4.9:
2225 - Error returned from rc_read_config was caught wrongly
2226 - Username/password not passed on to radius server due to wrong length.
2228 The presumption is that some radiusclient API changes for 4.51/PH/17
2229 were not taken care of correctly. The code is still untested by me (my
2230 Linux distribution still has 0.3.2 of radiusclient), but it was
2231 contributed by a Radius user.
2233 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2234 expanding the "port" option of the smtp transport.
2236 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2237 while reading a MIME header. Thanks to Tom Hughes for a patch.
2239 PH/24 Include config.h inside local_scan.h so that configuration settings are
2242 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2243 that in an ACL for RCPT (for example), you can examine exactly what was
2246 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2247 commands, but it was not correctly comparing the address with the actual
2248 client host address. Thus, it would show the EHLO address in Received:
2249 header lines when this was not necessary.
2251 PH/27 Added the % operator to ${eval:}.
2253 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2254 it should be ignoring failures (because with -C, for example, it has lost
2255 privilege). It wasn't ignoring creation failures other than "already
2258 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2261 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2262 because an input line was too long, either on its own, or by virtue of
2263 too many continuations, the temporary file was not being removed, and the
2264 return code was incorrect.
2266 PH/31 Missing "BOOL" in function definition in filtertest.c.
2268 PH/32 Applied Sieve patches from the maintainer.
2270 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2272 PH/33 Added "verify = not_blind".
2274 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2275 Local/Makefile (with some defaults set). These are used in built scripts
2276 such as exicyclog, but they have never been used in the exim_install
2277 script (though there are many overriding facilities there). I have
2278 arranged that the exim_install script now takes note of these two
2281 PH/35 Installed configuration files for Dragonfly.
2283 PH/36 When a locally submitted message by a trusted user did not contain a
2284 From: header, and the sender address was obtained from -f or from an SMTP
2285 MAIL command, and the trusted user did not use -F to supply a sender
2286 name, $originator_name was incorrectly used when constructing a From:
2287 header. Furthermore, $originator_name was used for submission mode
2288 messages from external hosts without From: headers in a similar way,
2289 which is clearly wrong.
2291 PH/37 Added control=suppress_local_fixups.
2293 PH/38 When log_selector = +received_sender was set, and the addition of the
2294 sender made the log line's construction buffer exactly full, or one byte
2295 less than full, an overflow happened when the terminating "\n" was
2298 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2299 when the result of a list match is failure because a DNS lookup failed.
2301 PH/40 RM_COMMAND is now used in the building process.
2303 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2304 the "build-* directories that it finds.
2306 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2307 address in a domain literal was a prefix of an interface address.
2309 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2310 when verifying a sender address, unless rewrite_headers is false.
2312 PH/44 Wrote a long comment about why errors_to addresses are verified as
2313 recipients, not senders.
2315 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2316 the ratelimit ACL was added.
2318 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2320 PH/46 Added extra information about PostgreSQL errors to the error string.
2322 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2323 avoiding the need to install special zones in a real server. This is
2324 backwards compatible; if it can't find the fake resolver, it drops back.
2325 Thus, both old and new test suites can be run.
2327 TF/02 Added util/ratelimit.pl
2329 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2330 clock is set back in time.
2332 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2333 Candler <B.Candler@pobox.com>.
2335 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2336 OK for addresses that are the result of redirections.
2338 PH/48 A number of further additions for the benefit of the new test suite,
2339 including a fake gethostbyname() that interfaces to the fake DNS resolver
2342 TF/06 The fix for widen_domains has also been applied to qualify_single and
2343 search_parents which are the other dnslookup options that can cause
2346 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2349 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2351 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2352 service" if it encountered an MX record with an empty target hostname.
2353 The message is now "an MX or SRV record indicated no SMTP service".
2355 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2356 if the DNS of the sending site is misconfigured. This is quite a
2357 common situation. This change restores the behaviour of treating a
2358 helo verification defer as a failure.
2360 PH/51 If self=fail was set on a router, the bounce message did not include the
2361 actual error message.
2367 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2369 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2370 command itself ended in error, the underlying message about the transport
2371 filter timeout was being overwritten with the pipe command error. Now the
2372 underlying error message should be appended to the second error message.
2374 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2376 PH/02 Reduce the amount of output that "make" produces by default. Full output
2377 can still be requested.
2379 PH/03 The warning log line about a condition test deferring for a "warn" verb
2380 was being output only once per connection, rather than after each
2381 occurrence (because it was using the same function as for successful
2382 "warn" verbs). This seems wrong, so I have changed it.
2384 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2385 it should not have, which might have caused a crash in the right
2386 circumstances, but probably never did.
2388 PH/04 Installed a modified version of Tony Finch's patch to make submission
2389 mode fix the return path as well as the Sender: header line, and to
2390 add a /name= option so that you can make the user's friendly name appear
2393 TF/03 Added the control = fakedefer ACL modifier.
2395 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2396 Mark Lowes for thorough testing.
2398 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2400 TK/03 Merged latest SRS patch from Miles Wilton.
2402 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2403 with the definition in sysexits.h (which is #included earlier).
2404 Fortunately, Exim does not actually use EX_OK. The code used to try to
2405 preserve the sysexits.h value, by assumimg that macro definitions were
2406 scanned for macro replacements. I have been disabused of this notion,
2407 so now the code just undefines EX_OK before #including unistd.h.
2409 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2410 in the smtp transport. When a block could not be written in a single
2411 write() function, the timeout was being re-applied to each part-write.
2412 This seems wrong - if the receiver was accepting one byte at a time it
2413 would take for ever. The timeout is now adjusted when this happens. It
2414 doesn't have to be particularly precise.
2416 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2417 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2419 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2420 without a domain if the check to <postmaster@domain> fails.
2422 SC/01 Eximstats: added -xls and the ability to specify output files
2423 (patch written by Frank Heydlauf).
2425 SC/02 Eximstats: use FileHandles for outputing results.
2427 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2429 SC/04 Eximstats: fixed display of large numbers with -nvr option
2431 SC/05 Eximstats: fixed merging of reports with empty tables.
2433 SC/06 Eximstats: added the -include_original_destination flag
2435 SC/07 Eximstats: removed tabs and trailing whitespace.
2437 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2439 TK/06 MBOX spool code: Add real "From " MBOX separator line
2440 so the .eml file is really in mbox format (even though
2441 most programs do not really care). Patch from Alex Miller.
2443 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2444 The latter is generated from $received_to and is only set if the
2445 message has one envelope recipient. SA can use these headers,
2446 obviously out-of-the-box. Patch from Alex Miller.
2448 PH/08 The ${def test on a variable was returning false if the variable's
2449 value was "0", contrary to what the specification has always said!
2450 The result should be true unless the variable is empty.
2452 PH/09 The syntax error of a character other than { following "${if
2453 def:variable_name" (after optional whitespace) was not being diagnosed.
2454 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2455 accidental colon was present, for example, could give incorrect results.
2457 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2458 result is used (not including appendfile, where other changes are about
2461 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2462 This involved changing a lot of size variables from int to off_t. It
2463 should work with maildirs and everything.
2465 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2466 spamd dying while we are connected to it.
2468 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2471 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2472 function for BDB 4.3.
2474 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2476 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2477 the sources for PCRE have been re-arranged and more files are now
2480 PH/15 The code I had for printing potentially long long variables in PH/11
2481 above was not the best (it lost precision). The length of off_t variables
2482 is now inspected at build time, and an appropriate printing format (%ld
2483 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2484 to be "long long int" or "long int". This is needed for the internal
2485 formatting function string_vformat().
2487 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2488 the configuration file to be ":syslog", then the script "guesses" where
2489 the logs files are, rather than using the compiled in default. In our
2490 case the guess is not the same as the compiled default, so the script
2491 suddenly stopped working when I started to use syslog. The patch checks
2492 to see if log_file_path is "". If so, it attempts to read it from exim
2493 with no configuration file to get the compiled in version, before it
2494 falls back to the previous guessing code."
2496 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2497 implementing BATV in an Exim configuration. See NewStuff for the gory
2500 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2501 Makefile that are specific to HP-UX.
2503 PH/18 If the "use_postmaster" option was set for a recipient callout together
2504 with the "random" option, the postmaster address was used as the MAIL
2505 FROM address for the random test, but not for the subsequent recipient
2506 test. It is now used for both.
2508 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2509 patch removes a few documentation additions to RFC 3028, because the
2510 latest draft now contains them. It adds the new en;ascii-case comparator
2511 and a new error check for 8bit text in MIME parts. Comparator and
2512 require names are now matched exactly. I enabled the subaddress
2513 extension, but it is not well tested yet (read: it works for me)."
2515 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2516 rework some of the code of TK/09 above to avoid the hardwired use of
2517 "%lld" and "long long". Replaced the call to snprintf() with a call to
2520 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2521 records point to non-existent hosts", "retry timeout exceeded", and
2522 "retry time not reached for any host after a long failure period".
2524 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2525 experimental DomainKeys support:
2527 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2528 (2) On an error such as an illegally used "control", the wrong name for
2529 the control was given.
2531 These problems did NOT occur unless DomainKeys support was compiled.
2533 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2535 PH/24 Added ${if match_ip condition.
2537 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2538 databases so that it will be absolutely obvious if a crash occurs in the
2539 DB library. This is a regular occurrence (often caused by mis-matched
2542 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2543 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2544 picked up on a user's system that detects such things. There doesn't seem
2545 to be a gcc warning option for this - only an attribute that has to be
2546 put on the function's prototype. It seems that in Fedora Core 4 they have
2547 set this on a number of new functions. No doubt there will be more in due
2550 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2551 specify a transport. However, if an address that was verified by such a
2552 router was the subject of a callout, Exim crashed because it tried to
2553 read the rcpt_include_affixes from the non-existent transport. Now it
2554 just assumes that the setting of that option is false. This bug was
2555 introduced by 4.51/PH/31.
2557 PH/28 Changed -d+all to exclude +memory, because that information is very
2558 rarely of interest, but it makes the output a lot bigger. People tend to
2559 do -d+all out of habit.
2561 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2562 code in os-type was giving problems when libc.so lives in lib64, like on
2565 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2566 aren't the modern standard, and it seems that some systems' include files
2567 don't always have them. Exim was already checking for some of the newer
2568 ones like T_AAAA, and defining it itself. I've added checks for all the
2569 record types that Exim uses.
2571 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2572 not automatically generating a new one, as it is supposed to. This
2573 prevented TLS from working. If the file did exist, but contained invalid
2574 data, a new version was generated, as expected. It was only the case of a
2575 non-existent file that was broken.
2577 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2578 with a change in libdomainkeys > 0.64.
2580 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2581 from DNS. If the selector record carries the flag, it now has
2582 precedence over the domain-wide flag.
2584 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2586 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2587 the use of a port name that isn't defined in /etc/services) occurred, the
2588 message was deferred as in a normal delivery, and thus remained on the
2589 spool, instead of being failed because of the mua_wrapper setting. This
2590 is now fixed, and I tidied up some of the mua_wrapper messages at the
2593 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2594 the messages in a hash of arrays rather than using individual hashes.
2595 This is a bit cleaner and results in dramatic memory savings, albeit
2596 at a slight CPU cost.
2598 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2599 as requested by Marc Sherman.
2601 SC/10 Eximstats: added histograms for user specified patterns as requested
2604 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2606 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2607 fopen() in the content-scanning modules that did not already have it.
2613 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2614 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2616 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2618 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2620 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2621 to test Sieve filters that use "vacation".
2623 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2624 that changes the way the GnuTLS parameters are stored in the cache file.
2625 The new format can be generated externally. For backward compatibility,
2626 if the data in the cache doesn't make sense, Exim assumes it has read an
2627 old-format file, and it generates new data and writes a new file. This
2628 means that you can't go back to an older release without removing the
2631 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2632 work if there are any delivery delays because "one_time" forces the
2633 parent to be marked "delivered", so its unseen clone is never tried
2634 again. For this reason, Exim now forbids the simultaneous setting of
2637 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2638 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2639 entry if you want to know the details. The fix, however, neglected to
2640 consider the case when local delivery batching is involved. The test for
2641 "previously delivered" was not happening when checking to see if an
2642 address could be batched with a previous (undelivered) one; under
2643 certain circumstances this could lead to multiple deliveries to the same
2646 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2647 in its include files, and this causes problems building Exim.
2649 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2650 header_syntax) but Exim was just ignoring anything given after a slash.
2651 In particular, this caused confusion with an attempt to use "verify =
2652 reverse_host_lookup/defer_ok". An error is now given when options are
2653 supplied for verify items that do not have them. (Maybe reverse_host_
2654 lookup should have a defer_ok option, but that's a different point.)
2656 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2657 defined by RFC 821) to 2048, because there were problems with some AUTH
2658 commands, and RFC 1869 says the size should be increased for extended
2659 SMTP commands that take arguments.
2661 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2664 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2665 "unknown" error; now it says that the functionality isn't in the binary.
2667 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2668 an address' error message when a string expansion fails (syntax or
2669 whatever). Otherwise the password may appear in the log. Following change
2670 PH/42 below, there is no longer a chance of it appearing in a bounce
2673 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2675 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2676 of its addresses was ever tried. (Bugzilla bug #2.)
2678 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2679 the result incorrectly in the debug output. (It correctly added a newline
2680 to what was transported.)
2682 TF/01 Added $received_time.
2684 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2685 commented out examples of how to interface to a virus scanner and to
2686 SpamAssassin. Also added commented examples of av_scanner and
2687 spamd_address settings.
2689 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2690 and controls are allowed in which ACLs. There were a couple of minor
2691 errors. Some of the entries in the conditions table (which is a table of
2692 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2693 negation of where the condition IS allowed.
2695 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2697 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2698 header file does not have a version number, so I've had to invent a new
2699 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2700 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2701 radiusclient), but it was contributed by a Radius user.
2703 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2704 files or directories, for OpenSSL.
2706 PH/19 When an Exim process that is running as root has to create an Exim log
2707 file, it does so in a subprocess that runs as exim:exim so as to get the
2708 ownership right at creation (otherwise, other Exim processes might see
2709 the file with the wrong ownership). There was no test for failure of this
2710 fork() call, which would lead to the process getting stuck as it waited
2711 for a non-existent subprocess. Forks do occasionally fail when resources
2712 run out. I reviewed all the other calls to fork(); they all seem to check
2715 PH/20 When checking for unexpected SMTP input at connect time (before writing
2716 the banner), Exim was not dealing correctly with a non-positive return
2717 from the read() function. If the client had disconnected by this time,
2718 the result was a log entry for a synchronization error with an empty
2719 string after "input=" when read() returned zero. If read() returned -1
2720 (an event I could not check), uninitialized data bytes were printed.
2721 There were reports of junk text (parts of files, etc) appearing after
2724 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2726 PH/22 Added support for macro redefinition, and (re)definition in between
2727 driver and ACL definitions.
2729 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2730 forgetting to use the resulting value; it was using the unexpanded value.
2732 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2733 hadn't been configured. The fix is from Juergen Kreileder, who
2734 understands it better than I do:
2736 "Here's what I see happening with three configured cyrus_sasl
2737 authenticators configured (plain, login, cram-md5):
2739 On startup auth_cyrus_sasl_init() gets called for each of these.
2740 This means three calls to sasl_listmech() without a specified mech_list.
2741 => SASL tests which mechs of all available mechs actually work
2742 => three warnings about OTP not working
2743 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2745 With the patch, sasl_listmech() also gets called three times. But now
2746 SASL's mech_list option is set to the server_mech specified in the the
2747 authenticator. Or in other words, the answer from sasl_listmech()
2748 gets limited to just the mech you're testing for (which is different
2750 => the return list contains just 'plain' or 'login', 'cram-md5' or
2751 nothing depending on the value of ob->server_mech.
2753 I've just tested the patch: Authentication still works fine,
2754 unavailable mechs specified in the exim configuration are still
2755 caught, and the auth.log warnings about OTP are gone."
2757 PH/25 When debugging is enabled, the contents of the command line are added
2758 to the debugging output, even when log_selector=+arguments is not
2761 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2762 answer is "GNU", and only if the return is "GNU/something" is the answer
2765 PH/27 $acl_verify_message is now set immediately after the failure of a
2766 verification in an ACL, and so is available in subsequent modifiers. In
2767 particular, the message can be preserved by coding like this:
2769 warn !verify = sender
2770 set acl_m0 = $acl_verify_message
2772 Previously, $acl_verify_message was set only while expanding "message"
2773 and "log_message" when a very denied access.
2775 PH/28 Modified OS/os.c-Linux with
2777 -#ifndef OS_LOAD_AVERAGE
2778 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2780 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2781 nomenclature these days.)
2783 PH/29 Installed patch from the Sieve maintainer that adds the options
2784 sieve_useraddress and sieve_subaddress to the redirect router.
2786 PH/30 In these circumstances:
2787 . Two addresses routed to the same list of hosts;
2788 . First host does not offer TLS;
2789 . First host accepts first address;
2790 . First host gives temporary error to second address;
2791 . Second host offers TLS and a TLS session is established;
2792 . Second host accepts second address.
2793 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2794 and peerdn, if requested) that were in fact used only for the second
2797 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2798 attention to any local part prefix or suffix that was matched by the
2799 router that accepted the address. It now behaves in the same way as it
2800 does for delivery: the affixes are removed from the local part unless
2801 rcpt_include_affixes is set on the transport.
2803 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2804 timeout during the DATA phase of an incoming message.
2806 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2807 applied a patch sanctioned by the Sieve maintainer.
2809 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2810 the uid or gid is negative. A case of a negative gid caused this to be
2811 noticed. The fix allows for either to be negative.
2813 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2814 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2817 PH/36 The change PH/12 above was broken. Fixed it.
2819 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2820 the grounds that routing the same address twice would always produce the
2821 same answer. This might have been true once, but it is certainly no
2822 longer true now. Routing a child address may depend on the previous
2823 routing that produced that child. Some complicated redirection strategies
2824 went wrong when messages had multiple recipients, and made Exim's
2825 behaviour dependent on the order in which the addresses were given.
2827 I have moved the duplicate checking until after the routing is complete.
2828 Exim scans the addresses that are assigned to local and remote
2829 transports, and removes any duplicates. This means that more work will be
2830 done, as duplicates will always all be routed, but duplicates are
2831 presumably rare, so I don't expect this is of any significance.
2833 For deliveries to pipes, files, and autoreplies, the duplicate checking
2834 still happens during the routing process, since they are not going to be
2837 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2838 It corrects a timeout issue with spamd. This is Ian's comment: "The
2839 background is that sometimes spamd either never reads data from a
2840 connection it has accepted, or it never writes response data. The exiscan
2841 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2842 blindly assumes that writes won't block so it may never time out."
2844 PH/39 Allow G after quota size as well as K and M.
2846 PH/40 The value set for $authenticated_id in an authenticator may not contain
2847 binary zeroes or newlines because the value is written to log lines and
2848 to spool files. There was no check on this. Now the value is run through
2849 the string_printing() function so that such characters are converted to
2850 printable escape sequences.
2852 PH/41 $message_linecount is a new variable that contains the total number of
2853 lines in the message. Compare $body_linecount, which is the count for the
2856 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2857 bounce and delay warning messages, except in certain special cases, which
2860 (a) An SMTP error message from a remote host;
2861 (b) A message specified in a :fail: redirection;
2862 (c) A message specified in a "fail" command in a system filter;
2863 (d) A message specified in a FAIL return from the queryprogram router;
2864 (e) A message specified by the cannot_route_message router option.
2866 In these cases only, Exim does include the error details in bounce and
2867 warning messages. There are also a few cases where bland messages such
2868 as "unrouteable address" or "local delivery error" are given.
2870 PH/43 $value is now also set for the "else" part of a ${run expansion.
2872 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2873 being worked on, but at least Exim now implements the latest version to
2876 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2877 process to complete was treated as a delivery failure, a timeout while
2878 writing the message to the pipe was logged, but erroneously treated as a
2879 successful delivery. Such timeouts include transport filter timeouts. For
2880 consistency with the overall process timeout, these timeouts are now
2881 treated as errors, giving rise to delivery failures by default. However,
2882 there is now a new Boolean option for the pipe transport called
2883 timeout_defer, which, if set TRUE, converts the failures into defers for
2884 both kinds of timeout. A transport filter timeout is now identified in
2887 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2888 systems where "make" and "gmake" are different, calling "gmake" at top
2889 level broke things. I've arranged for the value of $(MAKE) to be passed
2890 from the Makefile to this script so that it can call the same version of
2894 A note about Exim versions 4.44 and 4.50
2895 ----------------------------------------
2897 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2898 changes of various kinds. As a consequence, a big documentation update was
2899 needed. This delayed the release for rather longer than seemed good, especially
2900 in the light of a couple of (minor) security issues. Therefore, the changes
2901 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2902 release. So 4.44 and 4.50 are in effect two different branches that both start
2905 I have left the 4.50 change log unchanged; it contains all the changes since
2906 4.43. The change log for 4.44 is below; many of its items are identical to
2907 those for 4.50. This seems to be the most sensible way to preserve the
2908 historical information.
2914 1. Minor wording change to the doc/README.SIEVE file.
2916 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2917 computation of the current number of files was incorrect.
2919 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2920 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2923 4. Give more explanation in the error message when the command for a transport
2924 filter fails to execute.
2926 5. There are several places where Exim runs a non-Exim command in a
2927 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2928 was being done only for the command run by the queryprogram router. It is
2929 now done for all such subprocesses. The other cases are: ${run, transport
2930 filters, and the commands run by the lmtp and pipe transports.
2932 6. Added CONFIGURE_GROUP build-time option.
2934 7. Some older OS have a limit of 256 on the maximum number of file
2935 descriptors. Exim was using setrlimit() to set 1000 as a large value
2936 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2937 systems. I've change it so that if it can't get 1000, it tries for 256.
2939 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2940 was an oversight, and furthermore, ever since the addition of extra
2941 controls (e.g. 4.43/32), the checks on when to allow different forms of
2942 "control" were broken. There should now be diagnostics for all cases when a
2943 control that does not make sense is encountered.
2945 9. Added the /retain_sender option to "control=submission".
2947 10. $recipients is now available in the predata ACL (oversight).
2949 11. Tidy the search cache before the fork to do a delivery from a message
2950 received from the command line. Otherwise the child will trigger a lookup
2951 failure and thereby defer the delivery if it tries to use (for example) a
2952 cached ldap connection that the parent has called unbind on.
2954 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2955 of $address_data from the recipient verification was clobbered by the
2956 sender verification.
2958 13. The value of address_data from a sender verification is now available in
2959 $sender_address_data in subsequent conditions in the ACL statement.
2961 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2963 15. Added a new option "connect=<time>" to callout options, to set a different
2966 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2967 was its contents. (It was OK if the option was not defined at all.)
2969 17. A "Completed" log line is now written for messages that are removed from
2970 the spool by the -Mrm option.
2972 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2973 information about exactly what failed.
2975 19. Added -dd to debug only the daemon process.
2977 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2978 handles timeouts, both on the server side and network timeouts. Renamed the
2979 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2981 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2982 It is now set to "smtps".
2984 22. $host_address is now set to the target address during the checking of
2985 ignore_target_hosts.
2987 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2988 being passed; this would have caused $sender_host_name to have been used if
2989 matching the list had actually called for a host name (not very likely,
2990 since this list is usually IP addresses). A host name is now passed as
2993 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2994 code that specifies a non-restarting handler (typically sigaction() in
2995 modern systems) in an attempt to fix a rare and obscure crash bug.
2997 25. Narrowed the window for a race in the daemon that could cause it to ignore
2998 SIGCHLD signals. This is not a major problem, because they are used only to
2999 wake it up if nothing else does.
3001 26. A malformed maildirsize file could cause Exim to calculate negative values
3002 for the mailbox size or file count. Odd effects could occur as a result.
3003 The maildirsize information is now recalculated if the size or filecount
3006 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3007 support for a long time. Removed HAVE_SYS_VFS_H.
3009 28. Installed the latest version of exipick from John Jetmore.
3011 29. In an address list, if the pattern was not a regular expression, an empty
3012 subject address (from a bounce message) matched only if the pattern was an
3013 empty string. Non-empty patterns were not even tested. This was the wrong
3014 because it is perfectly reasonable to use an empty address as part of a
3015 database query. An empty address is now tested by patterns that are
3016 lookups. However, all the other forms of pattern expect the subject to
3017 contain a local part and a domain, and therefore, for them, an empty
3018 address still always fails if the pattern is not itself empty.
3020 30. Exim went into a mad DNS loop when attempting to do a callout where the
3021 host was specified on an smtp transport, and looking it up yielded more
3022 than one IP address.
3024 31. Re-factored the code for checking spool and log partition space into a
3025 function that finds that data and another that does the check. The former
3026 is then used to implement four new variables: $spool_space, $log_space,
3027 $spool_inodes, and $log_inodes.
3029 32. The RFC2047 encoding function was originally intended for short strings
3030 such as real names; it was not keeping to the 75-character limit for
3031 encoded words that the RFC imposes. It now respects the limit, and
3032 generates multiple encoded words if necessary. To be on the safe side, I
3033 have increased the buffer size for the ${rfc2047: expansion operator from
3036 33. It is now permitted to omit both strings after an "if" condition; if the
3037 condition is true, the result is "true". As before, when the second string
3038 is omitted, a false condition yields an empty string. This makes it less
3039 cumbersome to write custom ACL and router conditions.
3041 34. Failure to deliver a bounce message always caused it to be frozen, even if
3042 there was an errors_to setting on the router. The errors_to setting is now
3045 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3046 canonical form (fully expanded) before being placed in
3047 $sender_host_address.
3049 36. The table in the code that translates DNS record types into text (T_A to
3050 "A" for instance) was missing entries for NS and CNAME. It is just possible
3051 that this could have caused confusion if both these types were looked up
3052 for the same domain, because the text type is used as part of Exim's
3053 per-process caching. But the chance of anyone hitting this buglet seems
3056 37. The dnsdb lookup has been extended in a number of ways.
3058 (1) There is a new type, "zns", which walks up the domain tree until it
3059 finds some nameserver records. It should be used with care.
3061 (2) There is a new type, "mxh", which is like "mx" except that it returns
3062 just the host names, not the priorities.
3064 (3) It is now possible to give a list of domains (or IP addresses) to be
3065 looked up. The behaviour when one of the lookups defers can be
3066 controlled by a keyword.
3068 (4) It is now possible to specify the separator character for use when
3069 multiple records are returned.
3071 38. The dnslists ACL condition has been extended: it is now possible to supply
3072 a list of IP addresses and/or domains to be looked up in a particular DNS
3075 39. Added log_selector=+queue_time_overall.
3077 40. When running the queue in the test harness, wait just a tad after forking a
3078 delivery process, to get repeatability of debugging output.
3080 41. Include certificate and key file names in error message when GnuTLS fails
3081 to set them up, because the GnuTLS error message doesn't include the name
3082 of the failing file when there is a problem reading it.
3084 42. Allow both -bf and -bF in the same test run.
3086 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3088 44. The "Exiscan patch" is now merged into the mainline Exim source.
3090 45. Sometimes the final signoff response after QUIT could fail to get
3091 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3092 < 0 before doing a fflush(). This bug looks as though it goes back to the
3093 introduction of TLS in release 3.20, but "sometimes" must have been rare
3094 because the tests only now provoked it.
3096 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3097 (this can affect the format of dates).
3099 47. exim_tidydb, when checking for the continued existence of a message for
3100 which it has found a message-specific retry record, was not finding
3101 messages that were in split spool directories. Consequently, it was
3102 deleting retry records that should have stayed in existence.
3104 48. Steve fixed some bugs in eximstats.
3106 49. The SPA authentication driver was not abandoning authentication and moving
3107 on to the next authenticator when an expansion was forced to fail,
3108 contradicting the general specification for all authenticators. Instead it
3109 was generating a temporary error. It now behaves as specified.
3111 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3112 (the order specifies the preference for clients). The order is now AES256,
3113 AES128, 3DES, ARCFOUR128.
3115 51. Small patch to Sieve code - explicitly set From: when generating an
3118 52. Exim crashed if a remote delivery caused a very long error message to be
3119 recorded - for instance if somebody sent an entire SpamAssassin report back
3120 as a large number of 550 error lines. This bug was coincidentally fixed by
3121 increasing the size of one of Exim's internal buffers (big_buffer) that
3122 happened as part of the Exiscan merge. However, to be on the safe side, I
3123 have made the code more robust (and fixed the comments that describe what
3126 53. Now that there can be additional text after "Completed" in log lines (if
3127 the queue_time_overall log selector is set), a one-byte patch to exigrep
3128 was needed to allow it to recognize "Completed" as not the last thing in
3131 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3132 patch that reportedly fixes this has been added. I am not expert enough to
3133 create a test for it. This is what the patch creator wrote:
3135 "I found a little strange behaviour of ldap code when working with
3136 Windows 2003 AD Domain, where users was placed in more than one
3137 Organization Units. When I tried to give exim partial DN, the exit code
3138 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3139 But simultaneously result of request was absolutely normal ldap result,
3140 so I produce this patch..."
3142 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3143 REFERENCE, so I have modified the code to exclude the patch when that macro
3146 55. Some experimental protocols are using DNS PTR records for new purposes. The
3147 keys for these records are domain names, not reversed IP addresses. The
3148 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3149 leaves it alone. Component reversal etc. now happens only for IP addresses.
3152 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3154 57. Double the size of the debug message buffer (to 2048) so that more of very
3155 long debug lines gets shown.
3157 58. The exicyclog utility now does better if the number of log files to keep
3158 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3160 59. Two changes related to the smtp_active_hostname option:
3162 (1) $smtp_active_hostname is now available as a variable.
3163 (2) The default for smtp_banner uses $smtp_active_hostname instead
3164 of $primary_hostname.
3166 60. The host_aton() function is supposed to be passed a string that is known
3167 to be a valid IP address. However, in the case of IPv6 addresses, it was
3168 not checking this. This is a hostage to fortune. Exim now panics and dies
3169 if the condition is not met. A case was found where this could be provoked
3170 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3171 components; fortuitously, this particular loophole had already been fixed
3172 by change 4.50/55 above.
3174 If there are any other similar loopholes, the new check in host_aton()
3175 itself should stop them being exploited. The report I received stated that
3176 data on the command line could provoke the exploit when Exim was running as
3177 exim, but did not say which command line option was involved. All I could
3178 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3179 running as the user.
3182 61. There was a buffer overflow vulnerability in the SPA authentication code
3183 (which came originally from the Samba project). I have added a test to the
3184 spa_base64_to_bits() function which I hope fixes it.
3187 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3188 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3190 63. The daemon start-up calls getloadavg() while still root for those OS that
3191 need the first call to be done as root, but it missed one case: when
3192 deliver_queue_load_max is set with deliver_drop_privilege. This is
3193 necessary for the benefit of the queue runner, because there is no re-exec
3194 when deliver_drop_privilege is set.
3196 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3197 This has been fixed.
3199 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3200 was in use, was not putting the data itself into the right store pool;
3201 consequently, it could be overwritten for a subsequent message in the same
3202 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3205 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3207 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3208 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3209 the function were treating the return as a boolean value, which happened to
3210 work because 0=false and not-0=true, but is not correct code.
3212 68. The host_aton() function was not handling scoped IPv6 addresses (those
3213 with, for example, "%eth0" on the end) correctly.
3215 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3216 negated items (that is, ~something) in unsigned ints. Some compilers
3217 apparently mutter when there is no cast.
3219 70. If an address verification called from an ACL failed, and did not produce a
3220 user-specific message (i.e. there was only a "system" message), nothing was
3221 put in $acl_verify_message. In this situation, it now puts the system
3224 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3225 session; change 4.31/43 added the unwanted input to the log line - except
3226 that it did not do this in the start of session case. It now does.
3228 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3229 This is wrong and can cause the other end to generate a synchronization
3230 error if it is another Exim or anything else that does the synchronization
3231 check. A QUIT command is no longer sent after a timeout.
3233 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3234 during host lookups.
3236 74. The defer_ok option of callout verification was not working if it was used
3237 when verifying addresses in header lines, that is, for this case:
3239 verify = header_sender/callout=defer_ok
3241 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3242 those file descriptors could be used for SMTP connections. If anything
3243 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3244 could be sent to the SMTP client, causing chaos. The daemon now opens
3245 stdin, stdout, and stderr to /dev/null when it puts itself into the
3248 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3249 log by default. The user can override this with suitable Perl magic.
3251 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3252 the log message when discard triggers, was not working for the DATA ACL or
3253 for the non-SMTP ACL.
3255 78. Error message wording change in sieve.c.
3257 79. If smtp_accept_max_per_host was set, the number of connections could be
3258 restricted to fewer than expected, because the daemon was trying to set up
3259 a new connection before checking whether the processes handling previous
3260 connections had finished. The check for completed processes is now done
3261 earlier. On busy systems, this bug wouldn't be noticed because something
3262 else would have woken the daemon, and it would have reaped the completed
3265 80. If a message was submitted locally by a user whose login name contained one
3266 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3267 It caused a spool format error. I have fixed the spool reading code. A
3268 related problem was that the "from" clause in the Received: line became
3269 illegal because of the space(s). It is now covered by ${quote_local_part.
3271 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3274 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3276 83. Patch from Sieve maintainer to fix unterminated string problem in
3277 "vacation" handling.
3279 84. Some minor changes to the Linux configuration files to help with other
3280 OS variants using glibc.
3282 85. One more patch for Sieve to update vacation handling to latest spec.
3285 ----------------------------------------------------
3286 See the note above about the 4.44 and 4.50 releases.
3287 ----------------------------------------------------
3293 1. Change 4.43/35 introduced a bug that caused file counts to be
3294 incorrectly computed when quota_filecount was set in an appendfile
3297 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3298 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3301 3. Give more explanation in the error message when the command for a transport
3302 filter fails to execute.
3304 4. There are several places where Exim runs a non-Exim command in a
3305 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3306 was being done only for the command run by the queryprogram router. It is
3307 now done for all such subprocesses. The other cases are: ${run, transport
3308 filters, and the commands run by the lmtp and pipe transports.
3310 5. Some older OS have a limit of 256 on the maximum number of file
3311 descriptors. Exim was using setrlimit() to set 1000 as a large value
3312 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3313 systems. I've change it so that if it can't get 1000, it tries for 256.
3315 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3316 was an oversight, and furthermore, ever since the addition of extra
3317 controls (e.g. 4.43/32), the checks on when to allow different forms of
3318 "control" were broken. There should now be diagnostics for all cases when a
3319 control that does not make sense is encountered.
3321 7. $recipients is now available in the predata ACL (oversight).
3323 8. Tidy the search cache before the fork to do a delivery from a message
3324 received from the command line. Otherwise the child will trigger a lookup
3325 failure and thereby defer the delivery if it tries to use (for example) a
3326 cached ldap connection that the parent has called unbind on.
3328 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3329 of $address_data from the recipient verification was clobbered by the
3330 sender verification.
3332 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3333 was its contents. (It was OK if the option was not defined at all.)
3335 11. A "Completed" log line is now written for messages that are removed from
3336 the spool by the -Mrm option.
3338 12. $host_address is now set to the target address during the checking of
3339 ignore_target_hosts.
3341 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3342 being passed; this would have caused $sender_host_name to have been used if
3343 matching the list had actually called for a host name (not very likely,
3344 since this list is usually IP addresses). A host name is now passed as
3347 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3348 code that specifies a non-restarting handler (typically sigaction() in
3349 modern systems) in an attempt to fix a rare and obscure crash bug.
3351 15. Narrowed the window for a race in the daemon that could cause it to ignore
3352 SIGCHLD signals. This is not a major problem, because they are used only to
3353 wake it up if nothing else does.
3355 16. A malformed maildirsize file could cause Exim to calculate negative values
3356 for the mailbox size or file count. Odd effects could occur as a result.
3357 The maildirsize information is now recalculated if the size or filecount
3360 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3361 support for a long time. Removed HAVE_SYS_VFS_H.
3363 18. Updated exipick to current release from John Jetmore.
3365 19. Allow an empty sender to be matched against a lookup in an address list.
3366 Previously the only cases considered were a regular expression, or an
3369 20. Exim went into a mad DNS lookup loop when doing a callout where the
3370 host was specified on the transport, if the DNS lookup yielded more than
3373 21. The RFC2047 encoding function was originally intended for short strings
3374 such as real names; it was not keeping to the 75-character limit for
3375 encoded words that the RFC imposes. It now respects the limit, and
3376 generates multiple encoded words if necessary. To be on the safe side, I
3377 have increased the buffer size for the ${rfc2047: expansion operator from
3380 22. Failure to deliver a bounce message always caused it to be frozen, even if
3381 there was an errors_to setting on the router. The errors_to setting is now
3384 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3385 canonical form (fully expanded) before being placed in
3386 $sender_host_address.
3388 24. Updated eximstats to version 1.33
3390 25. Include certificate and key file names in error message when GnuTLS fails
3391 to set them up, because the GnuTLS error message doesn't include the name
3392 of the failing file when there is a problem reading it.
3394 26. Expand error message when OpenSSL has problems setting up cert/key files.
3397 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3398 (this can affect the format of dates).
3400 28. exim_tidydb, when checking for the continued existence of a message for
3401 which it has found a message-specific retry record, was not finding
3402 messages that were in split spool directories. Consequently, it was
3403 deleting retry records that should have stayed in existence.
3405 29. eximstats updated to version 1.35
3406 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3407 1.35 - bugfix such that pie charts by volume are generated correctly
3409 30. The SPA authentication driver was not abandoning authentication and moving
3410 on to the next authenticator when an expansion was forced to fail,
3411 contradicting the general specification for all authenticators. Instead it
3412 was generating a temporary error. It now behaves as specified.
3414 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3415 (the order specifies the preference for clients). The order is now AES256,
3416 AES128, 3DES, ARCFOUR128.
3418 31. Small patch to Sieve code - explicitly set From: when generating an
3421 32. Exim crashed if a remote delivery caused a very long error message to be
3422 recorded - for instance if somebody sent an entire SpamAssassin report back
3423 as a large number of 550 error lines. This bug was coincidentally fixed by
3424 increasing the size of one of Exim's internal buffers (big_buffer) that
3425 happened as part of the Exiscan merge. However, to be on the safe side, I
3426 have made the code more robust (and fixed the comments that describe what
3429 33. Some experimental protocols are using DNS PTR records for new purposes. The
3430 keys for these records are domain names, not reversed IP addresses. The
3431 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3432 leaves it alone. Component reversal etc. now happens only for IP addresses.
3435 34. The host_aton() function is supposed to be passed a string that is known
3436 to be a valid IP address. However, in the case of IPv6 addresses, it was
3437 not checking this. This is a hostage to fortune. Exim now panics and dies
3438 if the condition is not met. A case was found where this could be provoked
3439 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3440 components; fortuitously, this particular loophole had already been fixed
3441 by change 4.50/55 or 4.44/33 above.
3443 If there are any other similar loopholes, the new check in host_aton()
3444 itself should stop them being exploited. The report I received stated that
3445 data on the command line could provoke the exploit when Exim was running as
3446 exim, but did not say which command line option was involved. All I could
3447 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3448 running as the user.
3451 35. There was a buffer overflow vulnerability in the SPA authentication code
3452 (which came originally from the Samba project). I have added a test to the
3453 spa_base64_to_bits() function which I hope fixes it.
3456 36. The daemon start-up calls getloadavg() while still root for those OS that
3457 need the first call to be done as root, but it missed one case: when
3458 deliver_queue_load_max is set with deliver_drop_privilege. This is
3459 necessary for the benefit of the queue runner, because there is no re-exec
3460 when deliver_drop_privilege is set.
3462 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3463 was in use, was not putting the data itself into the right store pool;
3464 consequently, it could be overwritten for a subsequent message in the same
3465 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3468 38. Sometimes the final signoff response after QUIT could fail to get
3469 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3470 < 0 before doing a fflush(). This bug looks as though it goes back to the
3471 introduction of TLS in release 3.20, but "sometimes" must have been rare
3472 because the tests only now provoked it.
3478 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3479 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3480 Now it returns an integer. A number of calls were still expecting a T/F
3481 return. Fortuitously, in all cases, the tests worked in OK situations,
3482 which is the norm. However, things would have gone wrong on any write
3483 failures on the smtp file descriptor. This function is used when sending
3484 messages over SMTP and also when doing verify callouts.
3486 2. When Exim is called to do synchronous delivery of a locally submitted
3487 message (the -odf or -odi options), it no longer closes stderr before doing
3490 3. Implemented the mua_wrapper option.
3492 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3494 5. Implemented the functions header_remove(), header_testname(),
3495 header_add_at_position(), and receive_remove_recipient(), and exported them
3498 6. If an ACL "warn" statement specified the addition of headers, Exim already
3499 inserted X-ACL-Warn: at the start if there was no header name. However, it
3500 was not making this test for the second and subsequent header lines if
3501 there were newlines in the string. This meant that an invalid header could
3502 be inserted if Exim was badly configured.
3504 7. Allow an ACL "warn" statement to add header lines at the start or after all
3505 the Received: headers, as well as at the end.
3507 8. Added the rcpt_4xx retry error code.
3509 9. Added postmaster_mailfrom=xxx to callout verification option.
3511 10. Added mailfrom=xxxx to the callout verification option, for verify=
3514 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3515 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3517 12. Inserted some casts to stop certain compilers warning when using pointer
3518 differences as field lengths or precisions in printf-type calls (mostly
3519 affecting debugging statements).
3521 13. Added optional readline() support for -be (dynamically loaded).
3523 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3524 same clock tick as a message's arrival, so that its received time was the
3525 same as the "first fail" time on the retry record, and that message
3526 remained on the queue past the ultimate address timeout, every queue runner
3527 would try a delivery (because it was past the ultimate address timeout) but
3528 after another failure, the ultimate address timeout, which should have then
3529 bounced the address, did not kick in. This was a "< instead of <=" error;
3530 in most cases the first failure would have been in the next clock tick
3531 after the received time, and all would be well.
3533 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3534 being recognized when the domain list was tested by the match_domain
3535 condition in an expansion string.
3537 16. Added the ${str2b64: operator.
3539 17. Exim was always calling setrlimit() to set a large limit for the number of
3540 processes, without checking whether the existing limit was already
3541 adequate. (It did check for the limit on file descriptors.) Furthermore,
3542 errors from getrlimit() and setrlimit() were being ignored. Now they are
3543 logged to the main and panic logs, but Exim does carry on, to try to do its
3544 job under whatever limits there are.
3546 18. Imported PCRE 5.0.
3548 19. Trivial typo in log message " temporarily refused connection" (the leading
3551 20. If the log selector return_path_on_delivery was set and an address was
3552 redirected to /dev/null, the delivery process crashed because it assumed
3553 that a return path would always be set for a "successful" delivery. In this
3554 case, the whole delivery is bypassed as an optimization, and therefore no
3557 21. Internal re-arrangement: the function for sending a challenge and reading
3558 a response while authentication was assuming a zero-terminated challenge
3559 string. It's now changed to take a pointer and a length, to allow for
3560 binary data in such strings.
3562 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3564 23. Exim was not respecting finduser_retries when seeking the login of the
3565 uid under which it was called; it was always trying 10 times. (The default
3566 setting of finduser_retries is zero.) Also, it was sleeping after the final
3567 failure, which is pointless.
3569 24. Implemented tls_on_connect_ports.
3571 25. Implemented acl_smtp_predata.
3573 26. If the domain in control=submission is set empty, Exim assumes that the
3574 authenticated id is a complete email address when it generates From: or
3575 Sender: header lines.
3577 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3578 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3579 chown and chgrp in /bin and hostname in /usr/bin.
3581 28. Exim was keeping the "process log" file open after each use, just as it
3582 does for the main log. This opens the possibility of it remaining open for
3583 long periods when the USR1 signal hits a daemon. Occasional processlog
3584 errors were reported, that could have been caused by this. Anyway, it seems
3585 much more sensible not to leave this file open at all, so that is what now
3588 29. The long-running daemon process does not normally write to the log once it
3589 has entered its main loop, and it closes the log before doing so. This is
3590 so that log files can straightforwardly be renamed and moved. However,
3591 there are a couple of unusual error situations where the daemon does write
3592 log entries, and I had neglected to close the log afterwards.
3594 30. The text of an SMTP error response that was received during a remote
3595 delivery was being truncated at 512 bytes. This is too short for some of
3596 the long messages that one sometimes sees. I've increased the limit to
3599 31. It is now possible to make retry rules that apply only when a message has a
3600 specific sender, in particular, an empty sender.
3602 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3603 it possible to be selective about when SMTP synchronization is enforced.
3605 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3607 32. Implemented hosts_connection_nolog.
3609 33. Added an ACL for QUIT.
3611 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3614 35. Added mailbox_size and mailbox_filecount to appendfile.
3616 36. Added control = no_multiline_responses to ACLs.
3618 37. There was a bug in the logic of the code that waits for the clock to tick
3619 in the case where the clock went backwards by a substantial amount such
3620 that the microsecond fraction of "now" was more than the microsecond
3621 fraction of "then" (but the whole seconds number was less).
3623 38. Added support for the libradius Radius client library this is found on
3624 FreeBSD (previously only the radiusclient library was supported).
3630 1. When certain lookups returned multiple values in the form name=value, the
3631 quoting of the values was not always being done properly. Specifically:
3632 (a) If the value started with a double quote, but contained no whitespace,
3634 (b) If the value contained whitespace other than a space character (i.e.
3635 tabs or newlines or carriage returns) it was not quoted.
3636 This fix has been applied to the mysql and pgsql lookups by writing a
3637 separate quoting function and calling it from the lookup code. The fix
3638 should probably also be applied to nisplus, ibase and oracle lookups, but
3639 since I cannot test any of those, I have not disturbed their existing code.
3641 2. A hit in the callout cache for a specific address caused a log line with no
3642 reason for rejecting RCPT. Now it says "Previous (cached) callout
3643 verification failure".
3645 3. There was an off-by-one bug in the queryprogram router. An over-long
3646 return line was truncated at 256 instead of 255 characters, thereby
3647 overflowing its buffer with the terminating zero. As well as fixing this, I
3648 have increased the buffer size to 1024 (and made a note to document this).
3650 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3651 when Exim is waiting for an SMTP response from a remote server, Exim
3652 restarts its select() call on the socket, thereby resetting its timeout.
3653 This is not a problem when such interrupts are rare. Somebody set up a cron
3654 job to run exiwhat every 2 minutes, which is less than the normal select()
3655 timeout (5 or 10 minutes). This meant that the select() timeout never
3656 kicked in because it was always reset. I have fixed this by comparing the
3657 time when an interrupt arrives with the time at the start of the first call
3658 to select(). If more time than the timeout has elapsed, the interrupt is
3659 treated as a timeout.
3661 5. Some internal re-factoring in preparation for the addition of Sieve
3662 extensions (by MH). In particular, the "personal" test is moved to a
3663 separate function, and given an option for scanning Cc: and Bcc: (which is
3664 not set for Exim filters).
3666 6. When Exim created an email address using the login of the caller as the
3667 local part (e.g. when creating a From: or Sender: header line), it was not
3668 quoting the local part when it contained special characters such as @.
3670 7. Installed new OpenBSD configuration files.
3672 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3673 try to make them clearer.
3675 9. Callout options, other than the timeout value, were being ignored when
3676 verifying sender addresses in header lines. For example, when using
3678 verify = header_sender/callout=no_cache
3680 the cache was (incorrectly) being used.
3682 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3683 only the Cygwin environment.
3685 11. When return_path_on_delivery was set as a log selector, if different remote
3686 addresses in the same message used different return paths and parallel
3687 remote delivery occurred, the wrong values would sometimes be logged.
3688 (Whenever a remote delivery process finished, the return path value from
3689 the most recently started remote delivery process was logged.)
3691 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3692 lines when AUTH and/or TLS are in use. This is the "received protocol"
3693 field. Exim used to use "asmtp" for authenticated SMTP, without any
3694 indication (in the protocol name) for TLS use. Now it follows the RFC and
3695 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3696 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3697 names appear in log lines as well as in Received: header lines.
3699 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3700 extensions, and comparison tests, and to fix some bugs.
3702 14. Changes to the "personal" filter test:
3704 (1) The test was buggy in that it was just doing the equivalent of
3705 "contains" tests on header lines. For example, if a user's address was
3706 anne@some.where, the "personal" test would incorrectly be true for
3708 To: susanne@some.where
3710 This test is now done by extracting each address from the header in turn,
3711 and checking the entire address. Other tests that are part of "personal"
3712 are now done using regular expressions (for example, to check local parts
3713 of addresses in From: header lines).
3715 (2) The list of non-personal local parts in From: addresses has been
3716 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3717 taken from the Sieve specification recommendations.
3719 (3) If the message contains any header line starting with "List-" it is
3720 treated as non-personal.
3722 (4) The test for "circular" in the Subject: header line has been removed
3723 because it now seems ill-conceived.
3725 15. Minor typos in src/EDITME comments corrected.
3727 16. Installed latest exipick from John Jetmore.
3729 17. If headers_add on a router specified a text string that was too long for
3730 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3731 of string_sprintf() is now avoided.
3733 18. $message_body_size was not set (it was always zero) when running the DATA
3734 ACL and the local_scan() function.
3736 19. For the "mail" command in an Exim filter, no default was being set for
3737 the once_repeat time, causing a random time value to be used if "once" was
3738 specified. (If the value happened to be <= 0, no repeat happened.) The
3739 default is now 0s, meaning "never repeat". The "vacation" command was OK
3740 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3741 (I found it when inspecting the code).
3743 20. There is now an overall timeout for performing a callout verification. It
3744 defaults to 4 times the callout timeout, which applies to individual SMTP
3745 commands during the callout. The overall timeout applies when there is more
3746 than one host that can be tried. The timeout is checked before trying the
3747 next host. This prevents very long delays if there are a large number of
3748 hosts and all are timing out (e.g. when the network connections are timing
3749 out). The value of the overall timeout can be changed by specifying an
3750 additional sub-option for "callout", called "maxwait". For example:
3752 verify = sender/callout=5s,maxwait=20s
3754 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3755 to the end before writing, but this should make it even safer).
3757 22. Exim was forgetting that it had advertised PIPELINING for the second and
3758 subsequent messages on an SMTP connection. It was also not resetting its
3759 memory on STARTTLS and an internal HELO.
3761 23. When Exim logs an SMTP synchronization error within a session, it now
3762 records whether PIPELINING has been advertised or not.
3764 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3765 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3766 rather than long int.
3768 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3770 26. Added the never_mail option to autoreply.
3776 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3777 crash if the getsockname() call failed; this can happen if a connection is
3778 closed very soon after it is established. The problem was simply in the
3779 order in which certain operations were done, causing Exim to try to write
3780 to the SMTP stream before it had set up the file descriptor. The bug has
3781 been fixed by making things happen in the correct order.
3787 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3788 before the connection was closed, thus losing the rejection response.
3790 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3791 some early Solaris releases, but causes trouble in current releases where
3792 socklen_t is defined.
3794 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3797 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3800 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3801 that releasing the top of it at the end releases what was used for sub-
3802 expansions (unless the block got too big). However, discard this block if
3803 the first thing is a variable or header, so that we can use its block when
3804 it is dynamic (useful for very large $message_headers, for example).
3806 6. Lookups now cache *every* query, not just the most recent. A new, separate
3807 store pool is used for this. It can be recovered when all lookup caches are
3808 flushed. Lookups now release memory at the end of their result strings.
3809 This has involved some general refactoring of the lookup sources.
3811 7. Some code has been added to the store_xxx() functions to reduce the amount
3812 of flapping under certain conditions.
3814 8. log_incoming_interface used to affect only the <= reception log lines. Now
3815 it causes the local interface and port to be added to several more SMTP log
3816 lines, for example "SMTP connection from", and rejection lines.
3818 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3820 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3822 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3823 could be overwritten at the end of the current message (or the start of a
3824 new message if it was set in a HELO ACL). The value is now preserved for
3825 the duration of the SMTP connection.
3827 12. If a transport had a headers_rewrite setting, and a matching header line
3828 contained an unqualified address, that address was qualified, even if it
3829 did not match any rewriting rules. The underlying bug was that the values
3830 of the flags that permit the existence of unqualified sender and recipient
3831 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3832 non-local messages, and by -bnq for local messages) were not being
3833 preserved with the message after it was received.
3835 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3836 "next input=" as part of the text comprising the host identity instead of
3837 the correct text. The code was using the same buffer for two different
3838 strings. However, depending on which order the printing function evaluated
3839 its arguments, the bug did not always show up. Under Linux, for example, my
3840 test suite worked just fine.
3842 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3843 doesn't work with some older versions of Perl. It has been changed to "my",
3844 which in any case is probably the better facility to use.
3846 15. A really picky compiler found some instances of statements for creating
3847 error messages that either had too many or two few arguments for the format
3850 16. The size of the buffer for calls to the DNS resolver has been increased
3851 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3852 for addresses that have a lot of PTR records. This alleviates a problem; it
3853 does not fully solve it.
3855 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3856 buffer now truncates the list and logs the incident, which is the same
3857 action as happens when Exim is looking up a host name and its aliases.
3858 Previously in this situation something unpredictable would happen;
3859 sometimes it was "internal error: store_reset failed".
3861 18. If a server dropped the connection unexpectedly when an Exim client was
3862 using GnuTLS and trying to read a response, the client delivery process
3863 crashed while trying to generate an error log message.
3865 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3866 string, for example:
3868 warn message = H1: something\nH2: something
3870 the text was added as a single header line from Exim's point of view
3871 though it ended up OK in the delivered message. However, searching for the
3872 second and subsequent header lines using $h_h2: did not work. This has been
3873 fixed. Similarly, if a system filter added multiple headers in this way,
3874 the routers could not see them.
3876 20. Expanded the error message when iplsearch is called with an invalid key to
3877 suggest using net-iplsearch in a host list.
3879 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3880 ACLs are no longer actually imposed (and a message to that effect is
3883 22. If a "gecos" field in a passwd entry contained escaped characters, in
3884 particular, if it contained a \" sequence, Exim got it wrong when building
3885 a From: or a Sender: header from that name. A second bug also caused
3886 incorrect handling when an unquoted " was present following a character
3887 that needed quoting.
3889 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3890 was not being matched caselessly.
3892 24. Arranged for all hyphens in the exim.8 source to be escaped with
3895 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3896 properly. Recipient callout cache records were still being keyed to include
3897 the sender, even when use_sender was set false. This led to far more
3898 callouts that were necessary. The sender is no longer included in the key
3899 when use_sender is false.
3901 26. Added "control = submission" modifier to ACLs.
3903 27. Added the ${base62d: operator to decode base 62 numbers.
3905 28. dnsdb lookups can now access SRV records.
3907 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3908 the configuration file.
3910 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3911 (-v) mode. This makes the output for a verbose queue run more intelligible.
3913 31. Added a use_postmaster feature to recipient callouts.
3915 32. Added the $body_zerocount variable, containing the number of binary zero
3916 bytes in the message body.
3918 33. The time of last modification of the "new" subdirectory is now used as the
3919 "mailbox time last read" when there is a quota error for a maildir
3922 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3924 35. Added +ignore_unknown as a special item in host lists.
3926 36. Code for decoding IPv6 addresses in host lists is now included, even if
3927 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3928 address was recognized as an IP address, but was then not correctly decoded
3929 into binary, causing unexpected and incorrect effects when compared with
3936 1. Very minor rewording of debugging text in manualroute to say "list of
3937 hosts" instead of "hostlist".
3939 2. If verify=header_syntax was set, and a header line with an unqualified
3940 address (no domain) and a large number of spaces between the end of the
3941 name and the colon was received, the reception process suffered a buffer
3942 overflow, and (when I tested it) crashed. This was caused by some obsolete
3943 code that should have been removed. The fix is to remove it!
3945 3. When running in the test harness, delay a bit after writing a bounce
3946 message to get a bit more predictability in the log output.
3948 4. Added a call to search_tidyup() just before forking a reception process. In
3949 theory, someone could use a lookup in the expansion of smtp_accept_max_
3950 per_host which, without the tidyup, could leave open a database connection.
3952 5. Added the variables $recipient_data and $sender_data which get set from a
3953 lookup success in an ACL "recipients" or "senders" condition, or a router
3954 "senders" option, similar to $domain_data and $local_part_data.
3956 6. Moved the writing of debug_print from before to after the "senders" test
3959 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3960 problems for message scanning, either using a data ACL, or using
3961 local_scan() because the Received: header was not generated till after they
3962 were called (in order to set the time as the time of reception completion).
3963 I have revised the way this works. The header is now generated after the
3964 body is received, but before the ACL or local_scan() are called. After they
3965 are run, the timestamp in the header is updated.
3971 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3972 before starting a queue runner without re-exec. This happened only when
3973 deliver_drop_privilege was set or when the Exim user was set to root. The
3974 effect of the bug was that timeouts during subsequent deliveries caused
3975 crashes instead of being properly handled. The handler is now left at its
3976 default (and expected) setting.
3978 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3979 message, again when deliver_drop_privilege is set or Exim is run as root.
3980 The bug described in (1) was not present in this case, but the tidying up
3981 of the other signals was missing. I have made the two cases consistent.
3983 3. The ignore_target_hosts setting on a manualroute router was being ignored
3984 for hosts that were looked up using the /MX notation.
3986 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3989 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3990 operated on the sender address. After changing the $sender_address to <>
3991 for the sender address verify, Exim was re-instated it as the original
3992 (before rewriting) address, but remembering that it had rewritten it, so it
3993 wasn't rewriting it again. This bug also had the effect of breaking the
3994 sender address verification caching when the sender address was rewritten.
3996 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3997 This has been changed so that if the ip literal address matches
3998 ignore_target_hosts, the router declines.
4000 7. Added expansion conditions match_domain, match_address, and match_local_
4001 part (NOT match_host).
4003 8. The placeholder for the Received: header didn't have a length field set.
4005 9. Added code to Exim itself and to exim_lock to test for a specific race
4006 condition that could lead to file corruption when using MBX delivery. The
4007 issue is with the lockfile that is created in /tmp. If this file is removed
4008 after a process has opened it but before that process has acquired a lock,
4009 there is the potential for a second process to recreate the file and also
4010 acquire a lock. This could lead to two Exim processes writing to the file
4011 at the same time. The added code performs the same test as UW imapd; it
4012 checks after acquiring the lock that its file descriptor still refers to
4013 the same named file.
4015 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4016 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4019 11. Added the smtp_active_hostname option. If used, this will typically be made
4020 to depend on the incoming interface address. Because $interface_address is
4021 not set up until the daemon has forked a reception process, error responses
4022 that can happen earlier (such as "too many connections") no longer contain
4025 12. If an expansion in a condition on a "warn" statement fails because a lookup
4026 defers, the "warn" statement is abandoned, and the next ACL statement is
4027 processed. Previously this caused the whole ACL to be aborted.
4029 13. Added the iplsearch lookup type.
4031 14. Added ident_timeout as a log selector.
4033 15. Added tls_certificate_verified as a log selector.
4035 16. Added a global option tls_require_ciphers (compare the smtp transport
4036 option of the same name). This controls incoming TLS connections.
4038 17. I finally figured out how to make tls_require_ciphers do a similar thing
4039 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4040 before starting the TLS session.
4042 18. Tabs are now shown as \t in -bP output.
4044 19. If the log selector return_path_on_delivery was set, Exim crashed when
4045 bouncing a message because it had too many Received: header lines.
4047 20. If two routers both had headers_remove settings, and the first one included
4048 a superfluous trailing colon, the final name in the first list and the
4049 first name in the second list were incorrectly joined into one item (with a
4050 colon in the middle).
4056 1. Added -C and -D options to the exinext utility, mainly to make it easier
4057 to include in the automated testing, but these could be helpful when
4058 multiple configurations are in use.
4060 2. The exinext utility was not formatting the output nicely when there was
4061 an alternate port involved in the retry record key, nor when there was a
4062 message id as well (for retries that were specific to a specific message
4063 and a specific host). It was also confused by IPv6 addresses, because of
4064 the additional colons they contain. I have fixed the IPv4 problem, and
4065 patched it up to do a reasonable job for IPv6.
4067 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4068 delivery, the log line now contains "pipelined" if PIPELINING was used.
4070 4. An SMTP transport process used to panic and die if the bind() call to set
4071 an explicit outgoing interface failed. This has been changed; it is now
4072 treated in the same way as a connect() failure.
4074 5. A reference to $sender_host_name in the part of a conditional expansion
4075 that was being skipped was still causing a DNS lookup. This no longer
4078 6. The def: expansion condition was not recognizing references to header lines
4079 that used bh_ and bheader_.
4081 7. Added the _cache feature to named lists.
4083 8. The code for checking quota_filecount in the appendfile transport was
4084 allowing one more file than it should have been.
4086 9. For compatibility with Sendmail, the command line option
4094 and sets the incoming protocol and host name (for trusted callers). The
4095 host name and its colon can be omitted when only the protocol is to be set.
4096 Note the Exim already has two private options, -pd and -ps, that refer to
4097 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4098 "s", but I don't think that's a major issue.
4100 10. A number of refactoring changes to the code, none of which should affect
4103 (a) The number of logging options was getting close to filling up the
4104 32-bit word that was used as a bit map. I have split them into two classes:
4105 those that are passed in the argument to log_write(), and those that are
4106 only ever tested independently outside of that function. These are now in
4107 separate 32-bit words, so there is plenty of room for expansion again.
4108 There is no change in the user interface or the logging behaviour.
4110 (b) When building, for example, log lines, the code previously used a
4111 macro that called string_cat() twice, in order to add two strings. This is
4112 not really sufficiently general. Furthermore, there was one instance where
4113 it was actually wrong because one of the argument was used twice, and in
4114 one call a function was used. (As it happened, calling the function twice
4115 did not affect the overall behaviour.) The macro has been replaced by a
4116 function that can join an arbitrary number of extra strings onto a growing
4119 (c) The code for expansion conditions now uses a table and a binary chop
4120 instead of a serial search (which was left over from when there were very
4121 few conditions). Also, it now recognizes conditions like "pam" even when
4122 the relevant support is not compiled in: a suitably worded error message is
4123 given if an attempt is made to use such a condition.
4125 11. Added ${time_interval:xxxxx}.
4127 12. A bug was causing one of the ddress fields not to be passed back correctly
4128 from remote delivery subprocesses. The field in question was not being
4129 subsequently used, so this caused to problems in practice.
4131 13. Added new log selectors queue_time and deliver_time.
4133 14. Might have fixed a bug in maildirsizefile handling that threw up
4134 "unexpected character" debug warnings, and recalculated the data
4135 unnecessarily. In any case, I expanded the warning message to give more
4138 15. Added the message "Restricted characters in address" to the statements in
4139 the default ACL that block characters like @ and % in local parts.
4141 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4142 Three changes have been made:
4144 (a) There was a serious bug; a negative response to MAIL caused the whole
4145 recipient domain to be cached as invalid, thereby blocking all messages
4146 to all local parts at the same domain, from all senders. This bug has
4147 been fixed. The domain is no longer cached after a negative response to
4148 MAIL if the sender used is not empty.
4150 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4153 (c) A new callout option, "use_sender" has been added for people who want
4154 the modified behaviour.
4160 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4163 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4164 indeed breaks things for older releases.
4166 3. Added additional logging to the case where there is a problem reading data
4167 from a filter that is running in a subprocess using a pipe, in order to
4168 try to track down a specific problem.
4170 4. Testing facility fudge: when running in the test harness and attempting
4171 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4172 getting "No route to host". Convert this to a timeout.
4174 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4177 6. Some OS don't have socklen_t but use size_t instead. This affects the
4178 fifth argument of getsockopt() amongst other things. This is now
4179 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4180 can be set for individual OS. I have set it for SunOS5, OSF1, and
4181 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4182 some earlier ones do not.
4184 7. Change 4.30/15 was not doing the test caselessly.
4186 8. The standard form for an IPv6 address literal was being rejected by address
4187 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4188 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4189 this, as well as the form without the "IPv6" on the front (but only when
4190 address literals are enabled, of course).
4192 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4194 10. Exim crashed if a message with an empty sender address specified by -f
4195 encountered a router with an errors_to setting. This could be provoked only
4196 by a command such as
4200 where an empty string was supplied; "<>" did not hit this bug.
4202 11. Installed PCRE release 4.5.
4204 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4205 remained set. It is now erased.
4207 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4208 times from message ids (which are base 36 rather than the normal 62).
4210 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4211 were being counted as actual protocol errors, and logged if the log
4212 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4213 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4214 and DATA following a set of rejected RCPTs do not count as protocol errors.
4215 In other words, Exim assumes they were pipelined, though this may not
4216 actually be the case. Of course, in all cases the client gets an
4217 appropriate error code.
4219 15. If a lookup fails in an ACL condition, a message about the failure may
4220 be available; it is used if testing the ACL cannot continue, because most
4221 such messages specify what the cause of the deferral is. However, some
4222 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4223 that caused an old message to be retained and used if a later statement
4224 caused a defer, replacing the real cause of the deferral.
4226 16. If an IP address had so many PTR records that the DNS lookup buffer
4227 was not large enough to hold them, Exim could crash while trying to process
4228 the truncated data. It now detects and logs this case.
4230 17. Further to 4.21/58, another change has been made: if (and only if) the
4231 first line of a message (the first header line) ends with CRLF, a bare LF
4232 in a subsequent header line has a space inserted after it, so as not to
4233 terminate the header.
4235 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4236 unnecessarily, used atoi() instead of strtol(), and didn't check the
4237 termination when getting file sizes from file names by regex.
4239 19. Completely re-implemented the support for maildirsize files, in the light
4240 of a number of problems with the previous contributed implementation
4241 (4.30/29). In particular:
4243 . If the quota is zero, the maildirsize file is maintained, but no quota is
4246 . If the maildir directory does not exist, it is created before any attempt
4247 to write a maildirsize file.
4249 . The quota value in the file is just a cache; if the quota is changed in
4250 the transport, the new value overrides.
4252 . A regular expression is available for excluding directories from the
4255 20. The autoreply transport checks the characters in options that define the
4256 message's headers; it allows continued headers, but it was checking with
4257 isspace() after an embedded newline instead of explicitly looking for a
4260 21. If all the "regular" hosts to which an address was routed had passed their
4261 expiry times, and had not reached their retry times, the address was
4262 bounced, even if fallback hosts were defined. Now Exim should go on to try
4265 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4266 equivalent code in the SMTP transport. Some hosts send humungous responses
4267 to HELO/EHLO, more than 1024 it seems.
4269 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4270 gives compiler warnings in some environments. I've now done it "properly",
4273 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4274 (because of problems with the built-in one) was declared to return uschar *
4275 instead of char *, causing compiler failure.
4277 25. Fixed a file descriptor leak when processing alias/forward files.
4279 26. Fixed a minor format string issue in dbfn.c.
4281 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4283 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4284 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4287 29. When a sender address is verified, it is cached, to save repeating the test
4288 when there is more than one recipient in a message. However, when the
4289 verification involves a callout, it is possible for different callout
4290 options to be set for different recipients. It is too complicated to keep
4291 track of this in the cache, so now Exim always runs a verification when a
4292 callout is required, relying on the callout cache for the optimization.
4293 The overhead is duplication of the address routing, but this should not be
4296 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4297 to be verified with callout=postmaster, and the main callout worked but the
4298 postmaster check failed, the verification correctly failed. However, if a
4299 subsequent RCPT command asked for sender verification *without* the
4300 postmaster check, incorrect caching caused this verification also to fail,
4303 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4304 it was not caching the DNS options (qualify_single, search_parents) that
4305 were used when the lookup failed. A subsequent lookup with different
4306 options therefore always gave the same answer, though there were cases
4307 where it should not have. (Example: a "domains = !$mx_any" option on a
4308 dnslookup router: the "domains" option is always processed without any
4309 widening, but the router might have qualify_single set.) Now Exim uses the
4310 cached value only when the same options are set.
4312 32. Added John Jetmore's "exipick" utility to the distribution.
4314 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4315 than a timeout (e.g. a certificate is required, and is not provided), an
4316 Exim server now closes the connection immediately. Previously it waited for
4317 the client to close - but if the client is SSL, it seems that they each
4318 wait for each other, leading to a delay before one of them times out.
4320 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4321 maintained 0.8.x compatibility because I don't think many are using it, and
4322 it is clearly obsolete.
4324 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4327 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4328 client certificate was expired. A simple patch fixes this, though I don't
4329 understand the full logic of why the verify callback is called multiple
4332 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4333 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4334 which causes problems with some clients (such as the Certicom SSL Plus
4335 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4336 disables the coutermeasure allowing Eudora to connect."
4338 38. Exim was not checking that a write() to a log file succeeded. This could
4339 lead to Bad Things if a log got too big, in particular if it hit a file
4340 size limit. Exim now panics and dies if it cannot write to a log file, just
4341 as it does if it cannot open a log file.
4343 39. Modified OS/Makefile-Linux so that it now contains
4345 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4347 The two -D definitions ensure that Exim is compiled with large file
4348 support, which makes it possible to handle log files that are bigger than
4351 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4352 instance) a domain was checked against a named list that involved a lookup,
4353 causing $domain_data to be set, then another domain was checked against the
4354 same list, then the first domain was re-checked, the value of $domain_data
4355 after the final check could be wrong. In particular, if the second check
4356 failed, it could be set empty. This bug probably also applied to
4359 41. The strip_trailing_dot option was not being applied to the address given
4360 with the -f command-line option.
4362 42. The code for reading a message's header from the spool was incrementing
4363 $received_count, but never initializing it. This meant that the value was
4364 incorrect (doubled) while delivering a message in the same process in which
4365 it was received. In the most common configuration of Exim, this never
4366 happens - a fresh exec is done - but it can happen when
4367 deliver_drop_privilege is set.
4369 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4370 it now includes up to 150 characters of the unexpected data in the log
4373 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4374 and building data strings. The size of both of these buffers was 10 000
4375 bytes - far larger than anybody would *ever* want, thought I. Needless to
4376 say, somebody hit the limit. I have increased the maximum line length to
4377 20 000 and the maximum data length of concatenated lines to 100 000. I have
4378 also fixed two bugs, because there was no checking on these buffers. Tsk,
4379 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4380 buffer is too small.
4382 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4383 lsearch lookups. Now it does.
4385 46. When parsing a route_list item in a manualroute router, a fixed-length
4386 buffer was used for the list of hosts. I made this 1024 bytes long,
4387 thinking that nobody would ever have a list of hosts that long. Wrong.
4388 Somebody had a whole pile of complicated expansion conditions, and the
4389 string was silently truncated, leading to an expansion error. It turns out
4390 that it is easier to change to an unlimited length (owing to other changes
4391 that have happened since this code was originally written) than to build
4392 structure for giving a limitation error. The length of the item that
4393 expands into the list of hosts is now unlimited.
4395 47. The lsearch lookup could not handle data where the length of text line was
4396 more than 4095 characters. Such lines were truncated, leading to shortened
4397 data being returned. It should now handle lines of any length.
4399 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4400 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4403 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4404 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4405 no longer happens in comment lines. A list of replacements is now placed
4406 at the head of all of the source files, except those whose only change is
4407 to replace PERL_COMMAND in the very first #! line.
4409 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4410 messages on the queue, with a bottom-up merge sort, using code contributed
4411 by Michael Haardt. This should make operations like -bp somewhat faster on
4412 large queues. It won't affect queue runners, except when queue_run_in_order
4415 51. Installed eximstats 1.31 in the distribution.
4417 52. Added support for SRV lookups to the dnslookup router.
4419 53. If an ACL referred to $message_body or $message_body_end, the value was not
4420 reset for any messages that followed in the same SMTP session.
4422 54. The store-handling optimization for building very long strings was not
4423 differentiating between the different store pools. I don't think this
4424 actually made any difference in practice, but I've tidied it.
4426 55. While running the routers to verify a sender address, $sender_address
4427 was still set to the sender address. This is wrong, because when routing to
4428 send a bounce to the sender, it would be empty. Therefore, I have changed
4429 it so that, while verifying a sender address, $sender_address is set to <>.
4430 (There is no change to what happens when verifying a recipient address.)
4432 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4433 target A or AAAA records (if not already returned) without resetting the
4434 qualify_single or search_parents options of the DNS resolver. These are
4435 inappropriate in this case because the targets of MX and SRV records must
4436 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4437 target that, when qualified, matched something in the local domain. These
4438 two options are now turned off when doing these lookups.
4440 57. It seems that at least some releases of Reiserfs (which does not have the
4441 concept of a fixed number of inodes) returns zero and not -1 for the
4442 number of available inodes. This interacted badly with check_spool_inodes,
4443 which assumed that -1 was the "no such thing" setting. What I have done is
4444 to check that the total number of inodes is greater than zero before doing
4445 the test of how many are available.
4447 58. When a "warn" ACL statement has a log_message modifier, the message is
4448 remembered, and not repeated. This is to avoid a lot of repetition when a
4449 message has many recipients that cause the same warning to be written.
4450 Howewer, Exim was preserving the list of already written lines for an
4451 entire SMTP session, which doesn't seem right. The memory is now reset if a
4452 new message is started.
4454 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4455 debugging output unless log_rewrite was also set.
4457 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4458 of (int)(handle) when we know that handle contains (void *)(-1).
4460 61. The Exim daemon panic-logs an error return when it closes the incoming
4461 connection. However "connection reset by peer" seems to be common, and
4462 isn't really an error worthy of noting specially, so that particular error
4465 62. When Exim is trying to find all the local interfaces, it used to panic and
4466 die if the ioctl to get the interface flags failed. However, it seems that
4467 on at least one OS (Solaris 9) it is possible to have an interface that is
4468 included in the list of interfaces, but for which you get a failure error
4469 for this call. This happens when the interface is not "plumbed" into a
4470 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4471 failure of the "get flags" call assumes that the interface is down.
4473 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4474 makes life easier for people who are doing arithmetic on fields extracted
4475 from dates, where you often get leading zeros that should not be
4476 interpreted as octal.
4478 64. Added qualify_domain to the redirect router, to override the global
4481 65. If a pathologically long header line contained very many addresses (the
4482 report of this problem mentioned 10 000) and each of them was rewritten,
4483 Exim could use up a very large amount of memory. (It kept on making new
4484 copies of the header line as it rewrote, and never released the old ones.)
4485 At the expense of a bit more processing, the header rewriting function has
4486 been changed so that it no longer eats memory in this way.
4488 66. The generation of the Received: header has been moved from the time that a
4489 message starts to be received, to the time that it finishes. The timestamp
4490 in the Received: header should now be very close to that of the <= log
4491 line. There are two side-effects of this change:
4493 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4494 logged header lines no longer include the local Received: line, because
4495 it has not yet been created. The same applies to a copy of the message
4496 that is returned to a non-SMTP sender when a message is rejected.
4498 (b) When a filter file is tested using -bf, no additional Received: header
4499 is added to the test message. After some thought, I decided that this
4502 This change does not affect the value of $received_for. It is still set
4503 after address rewriting, but before local_scan() is called.
4505 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4507 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4508 gave an unhelpful panic error message, and a defer error. I have managed to
4509 change this behaviour so that it now rejects any supplied certificate,
4510 which seems right, as the list of acceptable certificates is empty.
4512 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4513 gave an unhelpful defer error. I have not managed to make this reject any
4514 supplied certificates, but the error message it gives is "no certificate
4515 supplied", which is not helpful.
4517 70. exigrep's output now also includes lines that are not associated with any
4518 message, but which match the given pattern. Implemented by a patch from
4519 Martin Sluka, which also tidied up the Perl a bit.
4521 71. Recipient callout verification, like sender verification, was using <> in
4522 the MAIL FROM command. This isn't really the right thing, since the actual
4523 sender may affect whether the remote host accepts the recipient or not. I
4524 have changed it to use the actual sender in the callout; this means that
4525 the cache record is now keyed on a recipient/sender pair, not just the
4526 recipient address. There doesn't seem to be a real danger of callout loops,
4527 since a callout by the remote host to check the sender would use <>.
4528 [SEE ABOVE: changed after hitting problems.]
4530 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4531 temporary errors. However, in the case of such a code being given after
4532 the end of a data transmission (i.e. after ".") Exim was failing to write
4533 a retry record for the message. (Yes, there was some broken host that was
4534 actually sending 8xx at this point.)
4536 73. An unknown lookup type in a host list could cause Exim to panic-die when
4537 the list was checked. (An example that provoked this was putting <; in the
4538 middle of a list instead of at the start.) If this happened during a DATA
4539 ACL check, a -D file could be left lying around. This kind of configuration
4540 error no longer causes Exim to die; instead it causes a defer errror. The
4541 incident is still logged to the main and panic logs.
4543 74. Buglet left over from Exim 3 conversion. The message "too many messages
4544 in one connection" was written to the rejectlog but not the mainlog, except
4545 when address rewriting (yes!) was being logged.
4547 75. Added write_rejectlog option.
4549 76. When a system filter was run not as root (that is, when system_filter_user
4550 was set), the values of the $n variables were not being returned to the
4551 main process; thus, they were not subsequently available in the $sn
4554 77. Added +return_path_on_delivery log selector.
4556 78. A connection timeout was being treated differently from recipients deferred
4557 when testing hosts_max_try with a message that was older than the host's
4558 retry timeout. (The host should not be counted, thus allowing all hosts to
4559 be tried at least once before bouncing.) This may have been the cause of an
4560 occasionally reported bug whereby a message would remain on the queue
4561 longer than the retry timeout, but would be bounced if a delivery was
4562 forced. I say "may" because I never totally pinned down the problem;
4563 setting up timeout/retry tests is difficult. See also the next item.
4565 79. The ultimate address timeout was not being applied to errors that involved
4566 a combination of host plus message (for example, a timeout on a MAIL
4567 command). When an address resolved to a number of possible hosts, and they
4568 were not all tried for each delivery (e.g. because of hosts_max_try), a
4569 message could remain on the queue longer than the retry timeout.
4571 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4574 81. Fixed an obscure SMTP outgoing bug which required at least the following
4575 conditions: (a) there was another message waiting for the same server;
4576 (b) the server returned 5xx to all RCPT commands in the first message so
4577 that the message was not completed; (c) the server dropped the connection
4578 or gave a negative response to the RSET that Exim sends to abort the
4579 transaction. The observed case was a dropped connection after DATA that had
4580 been sent in pipelining mode. That is, the server had advertised PIPELINING
4581 but was not implementing it correctly. The effect of the bug was incorrect
4582 behaviour, such as trying another host, and this could lead to a crash.
4588 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4589 and daemon.c were passed as pointers to ints; they should have been
4590 pointers to socklen_t variables (which are typically unsigned ints).
4592 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4595 3. Fixed a really odd bug that affected only the testing scheme; patching a
4596 certain fixed string in the binary changed the value of another string that
4597 happened to be identical to the end of the original first string.
4599 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4600 name", it returns that address as the IP address. On some operating
4601 systems (e.g. Solaris), it also passes back the IP address string as the
4602 "host name". However, on others (e.g. Linux), it passes back an empty
4603 string. Exim wasn't checking for this, and was changing the host name to an
4604 empty string, assuming it had been canonicized.
4606 5. Although rare, it is permitted to have more than one PTR record for a given
4607 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4608 all the names associated with an address, because they do in Solaris.
4609 However, it seems that they do not in Linux for data that comes from the
4610 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4611 I found this out when I moved to a new Linux workstation and tried to run
4612 the Exim test suite.
4614 To get round this problem I have changed the code so that it now does its
4615 own call to the DNS to look up PTR records when searching for a host name.
4616 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4617 addresses that are only in /etc/hosts are still found.
4619 This behaviour is, however, controlled by an option called host_lookup_
4620 order, which defaults to "bydns:byaddr". If people want to use the other
4621 order, or indeed, just use one or the other means of lookup, they can
4622 specify it in this variable.
4624 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4625 some operating systems, this comes back from gethostbyaddr() as an empty
4626 string, and this is what Exim used to test for. However, it seems that in
4627 other systems, "." is yielded. Exim now tests for this case too.
4629 7. The values of check_spool_space and check_log_space are now held internally
4630 as a number of kilobytes instead of an absolute number of bytes. If a
4631 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4632 kilobyte. This means that much larger values can be stored.
4634 8. Exim monitor: an attempt to get the action menu when not actually pointing
4635 at a message produces an empty menu entitled "No message selected". This
4636 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4637 no entries in it ("Shell widget menu has zero width and/or height"). So I
4638 have added a single, blank menu entry in this case.
4640 9. Added ${quote_local_part.
4642 10. MIME decoding is now applied to the contents of Subject: header lines when
4645 11. Now that a reference to $sender_host_address automatically causes a reverse
4646 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4647 host lookup before query-style lookups in lists that might use this
4648 variable. This has therefore been abolished, and the "net-" prefix is no
4649 longer necessary for query-style lookups.
4651 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4652 have been a typo for LFLAGS, so it has been changed.
4654 13. The install script calls Exim with "-C /dev/null" in order to find the
4655 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4656 to be output. Howeve, since Exim outputs its version number before the
4657 error, it didn't break the script. It just looked ugly. I fixed this by
4658 always allowing "-C /dev/null" if the caller is root.
4660 14. Ignore overlarge ACL variable number when reading spool file - insurance
4661 against a later release with more variables having written the file.
4663 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4664 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4665 this, as well as the form without the "IPv6" on the front.
4667 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4668 OS/Makefile-Darwin file.
4670 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4671 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4673 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4674 variable. However, some OS use ints for inodes. I've added cast to long int
4675 to get rid of the compiler warning.
4677 19. I had forgotten to lock out "/../" in configuration file names when
4678 ALT_CONFIG_PREFIX was set.
4680 20. Routers used for verification do not need to specify transports. However,
4681 if such a router generated a host list, and callout was configured, Exim
4682 crashed, because it could not find a port number from the (non-existent)
4683 transport. It now assumes port 25 in this circumstance.
4685 21. Added the -t option to exigrep.
4687 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4688 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4689 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4690 them is set, LOOKUP_LSEARCH is forced.
4692 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4694 24. Added sender and host information to the "rejected by local_scan()" log
4695 line; previously there was no indication of these.
4697 25. Added .include_if_exists.
4699 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4700 Linux. It turns out that not all file systems support this. Apparently some
4701 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4702 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4703 not supported on the file descriptor, is now ignored when Exim is trying to
4704 sync a directory. This applies only to Linux.
4706 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4708 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4709 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4712 29. Added support for maildirsize files from supplied patch (modified a bit).
4714 30. The use of :fail: followed by an empty string could lead Exim to respond to
4715 sender verification failures with (e.g.):
4717 550 Verification failed for <xxx>
4718 550 Sender verify failed
4720 where the first response line was missing the '-' that indicates it is not
4721 the final line of the response.
4723 31. The loop for finding the name of the user that called Exim had a hardwired
4724 limit of 10; it now uses the value of finduser_retries, which is used for
4725 all other user lookups.
4727 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4730 33. Exim was neglecting to zero errno before one call of strtol() when
4731 expanding a string and expecting an integer value. On some systems this
4732 resulted in spurious "integer overflow" errors. Also, it was casting the
4733 result into an int without checking.
4735 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4736 did not work. The code looks as if it has *never* worked, though it appears
4737 to have been documented since at least releast 1.62. I have made it work.
4739 35. The "timeout_DNS" error in retry rules, also documented since at least
4740 1.62, also never worked. As it isn't clear exactly what this means, and
4741 clearly it isn't a major issue, I have abolished the feature by treating it
4742 as "timeout", and writing a warning to the main and panic logs.
4744 36. The display of retry rules for -brt wasn't always showing the error code
4747 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4748 timeout_connect_A, timeout_connect_MX.
4750 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4751 to the empty sender.
4753 39. The daemon was not analysing the content of -oX till after it had closed
4754 stderr and disconnected from the controlling terminal. This meant that any
4755 syntax errors were only noted on the panic log, and the return code from
4756 the command was 0. By re-arranging the code a little, I've made the
4757 decoding happen first, so such errors now appear on stderr, and the return
4758 code is 1. However, the actual setting up of the sockets still happens in
4759 the disconnected process, so errors there are still only recorded on the
4762 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4763 connections (as happens on some IP stacks) was logged at start up time as
4764 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4765 it from "IPv6 and IPv4", which means that two separate sockets are being
4768 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4769 says whether AF_INET or AF_INET6 was passed as an argument.
4771 42. Exiwhat output was messed up when time zones were included in log
4774 43. Exiwhat now gives more information about the daemon's listening ports,
4775 and whether -tls-on-connect was used.
4777 44. The "port" option of the smtp transport is now expanded.
4779 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4780 silently ignored. Now an error message is written to the main and panic
4783 46. There's a new ACL modifier called "logwrite" which writes to a log file
4784 as soon as it is encountered.
4786 47. Added $local_user_uid and $local_user_gid at routing time.
4788 48. Exim crashed when trying to verify a sender address that was being
4791 49. Exim was recognizing only a space character after ".include". It now also
4792 recognizes a tab character.
4794 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4795 extracting the relevant information from the specification. The man page no
4796 longer contains scrambled data for the -d option, and I've added a section
4797 at the front about calling Exim under different names.
4799 51. Added "extra_headers" argument to the "mail" command in filter files.
4801 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4804 53. Installed eximstats 1.29.
4806 54. Added transport_filter_timeout as a generic transport option.
4808 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4809 Cc: header lines. This was required by RFC 822, but it not required by RFC
4812 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4813 incoming messages that did not have them. Now it does so only if the
4814 message originates locally, that is, if there is no associated remote host
4815 address. When Resent- header lines are present, this applies to the Resent-
4816 lines rather than the non-Resent- lines.
4818 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4819 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4821 58. Messages for configuration errors now include the name of the main
4822 configuration file - useful now that there may be more than one file in a
4823 list (.included file names were always shown).
4825 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4826 for those rare installations that do not start the daemon as root or run it
4827 setuid root. I've cut out the call to initgroups() if the daemon is not
4830 60. The Exim user and group can now be bound into the binary as text strings
4831 that are looked up at the start of Exim's processing.
4833 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4835 62. Added $mailstore_basename variable.
4837 63. Installed patch to sieve.c from Michael Haardt.
4839 64. When Exim failed to open the panic log after failing to open the main log,
4840 the original message it was trying to log was written to stderr and debug
4841 output, but if they were not available (the usual case in production), it
4842 was lost. Now it is written to syslog before the two lines that record the
4843 failures to open the logs.
4845 65. Users' Exim filters run in subprocesses under the user's uid. It is
4846 possible for a "deliver" command or an alias in a "personal" command to
4847 provoke an address rewrite. If logging of address rewriting is configured,
4848 this fails because the process is not running as root or exim. There may be
4849 a better way of dealing with this, but for the moment (because 4.30 needs
4850 to be released), I have disabled address rewrite logging when running a
4851 filter in a non-root, non-exim process.
4857 1. The buildconfig auxiliary program wasn't quoting the value set for
4858 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4859 not defined. This bug was masked in 4.22 by the effect that was fixed in
4862 2. Some messages that were rejected after a message id was allocated were
4863 shown as "incomplete" by exigrep. It no longer does this for messages that
4864 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4866 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4867 have allow_domain_literals set, the ID did not get logged in the <= line.
4868 Domain literals are now always recognized in Message-ID: header lines.
4870 4. The first argument for a ${extract expansion item is the key name or field
4871 number. Leading and trailing spaces in this item were not being ignored,
4872 causing some misleading effects.
4874 5. When deliver_drop_privilege was set, single queue runner processes started
4875 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4876 same command in the process it spins off) were not dropping privilege.
4878 6. When the daemon running as "exim" started a queue runner, it always
4879 re-executed Exim in the spun-off process. This is a waste of effort when
4880 deliver_drop_privilege is set. The new process now just calls the
4881 queue-runner function directly.
4887 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4890 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4891 ignored. Though the use of -oP was forcing the writing of a pid file, it
4892 was always written to the default place.
4894 3. If the message "no IP address found for host xxxx" is generated during
4895 incoming verification, it is now followed by identification of the incoming
4896 connection (so you can more easily find what provoked it).
4898 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4900 5. Added some features to "harden" Exim a bit more against certain attacks:
4902 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4903 be put in Local/Makefile. This is like the never_users runtime option,
4904 but it cannot be overridden. The default setting is "root".
4906 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4907 prefix string with which any file named in a -C command line option
4910 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4911 is retained for -C and -D only if the caller of Exim is root. Without
4912 it, the exim user may also use -C and -D and retain privilege.
4914 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4915 command line option is disabled.
4917 6. Macro names set by the -D option must start with an upper case letter, just
4918 like macro names defined in the configuration file.
4920 7. Added "dereference=" facility to LDAP.
4922 8. Two instances of the typo "uknown" in the source files are fixed.
4924 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4925 the Configure-Makefile script screwed up while processing it.
4927 10. Incorporated PCRE 4.4.
4929 11. The SMTP synchronization check was not operating right at the start of an
4930 SMTP session. For example, it could not catch a HELO sent before the client
4931 waited for the greeting. There is now a check for outstanding input at the
4932 point when the greeting is written. Because of the duplex, asynchronous
4933 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4934 way, but not yet received, when the check is performed.
4936 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4937 on TCP/IP sockets, because this apparently causes some broken clients to
4940 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4941 unchanged) from the Cygwin maintainer.
4943 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4944 was supported instead of testing for mbx. Effectively a typo.
4946 15. The spa authenticator server code was not checking that the input it
4947 received was valid base64.
4949 16. The debug output line for the "set" modifier in ACLs was not showing the
4950 name of the variable that was being set.
4952 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4954 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4955 lookup on its own. Something else was needed to trigger the lookup. For
4956 example, a match in host_lookup or the need for a host name in a host list.
4957 Now, if $sender_host_name is referenced and the host name has not yet been
4958 looked up, a lookup is performed. If the lookup fails, the variable remains
4959 empty, and $host_lookup_failed is set to "1".
4961 19. Added "eqi" as a case-independent comparison operator.
4963 20. The saslauthd authentication condition could segfault if neither service
4964 nor realm was specified.
4966 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4967 error message that was logged was misleading, and incoming SMTP
4968 connections were dropped. The message is now more accurate, and temporary
4969 errors are given to SMTP connections.
4971 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4972 (except RSET) with a 421 error, until QUIT is received. However, it was
4973 failing to send a response to QUIT.
4975 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4976 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4977 ACL is now run after the helo_try_verify_hosts code.
4979 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4980 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4981 case-independent, so other case variants are also recognized). Apparently
4982 some systems use these upper case variants.
4984 25. If more than two messages were waiting for the same host, and a transport
4985 filter was specified for the transport, Exim sent two messages over the
4986 same TCP/IP connection, and then failed with "socket operation on non-
4987 socket" when it tried to send the third.
4989 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4991 27. The extern definition of crypt16() in expand.c was not being excluded when
4992 the OS had its own crypt16() function.
4994 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4995 as a preferred synonym for return_size_limit, both as an option and as an
4998 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5000 30. Changed the default configuration ACL to relax the local part checking rule
5001 for addresses that are not in any local domains. For these addresses,
5002 slashes and pipe symbols are allowed within local parts, but the sequence
5003 /../ is explicitly forbidden.
5005 31. SPA server authentication was not clearing the challenge buffer before
5008 32. log_message in a "warn" ACL statement was writing to the reject log as
5009 well as to the main log, which contradicts the documentation and doesn't
5010 seem right (because no rejection is happening). So I have stopped it.
5012 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5013 However, I am unable to do any testing of this.
5015 34. Fixed an infelicity in the appendfile transport. When checking directories
5016 for a mailbox, to see if any needed to be created, it was accidentally
5017 using path names with one or more superfluous leading slashes; tracing
5018 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5020 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5021 contents are added to the log line that is written for every discarded
5022 recipient. (Previously a log_message setting was ignored.)
5024 36. The ${quote: operator now quotes the string if it is empty.
5026 37. The install script runs exim in order to find its version number. If for
5027 some reason other than non-existence or emptiness, which it checks, it
5028 could not run './exim', it was installing it with an empty version number,
5029 i.e. as "exim-". This error state is now caught, and the installation is
5032 38. An argument was missing from the function that creates an error message
5033 when Exim fails to connect to the socket for saslauthd authentication.
5034 This could cause Exim to crash, or give a corrupted message.
5036 39. Added isip, isip4, and isip6 to ${if conditions.
5038 40. The ACL variables $acl_xx are now saved with the message, and can be
5039 accessed later in routers, transports, and filters.
5041 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5042 strings in the file are not string-expanded.
5044 42. If a MAIL command specified a SIZE value that was too large to fit into an
5045 int variable, the check against message_size_limit failed. Such values are
5046 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5047 day this will have to be increased, but I don't think I want to be around
5048 when emails are that large.
5055 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5056 iconv() is not standard in FreeBSD.
5058 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5059 IPv6 enabled. The observed symptom was a segmentation fault on return from
5060 the function os_common_find_running_interfaces() in src/os.c.
5062 3. In the check_special_case() function in daemon.c I had used "errno" as an
5063 argument name, which causes warnings on some systems. This was basically a
5064 typo, since it was named "eno" in the comments!
5066 4. The code that waits for the clock to tick (at a resolution of some fraction
5067 of a second) so as to ensure message-id uniqueness was always waiting for
5068 at least one whole tick, when it could have waited for less. [This is
5069 almost certainly not relevant at current processor speeds, where it is
5070 unlikely to ever wait at all. But we try to future-proof.]
5072 5. The function that sleeps for a time interval that includes fractions of a
5073 second contained a race. It did not block SIGALRM between setting the
5074 timer, and suspending (a couple of lines later). If the interval was short
5075 and the sigsuspend() was delayed until after it had expired, the suspension
5076 never ended. On busy systems this could lead to processes getting stuck for
5079 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5080 process, before it forks any delivery processes. The open lookup caching
5081 mechanism meant that the open file or database connection was passed into
5082 the delivery process. The problem was that delivery processes always tidy
5083 up cached lookup data. This could cause a problem for the next delivery
5084 process started by the queue runner, because the external queue runner
5085 process does not know about the closure. So the next delivery process
5086 still has data in the lookup cache. In the case of a file lookup, there was
5087 no problem because closing a file descriptor in a subprocess doesn't affect
5088 the parent. However, if the lookup was caching a connection to a database,
5089 the connection was closed, and the second delivery process was likely to
5090 see errors such as "PGSQL: query failed: server closed the connection
5091 unexpectedly". The problem has been fixed by closing all cached lookups
5092 in a queue runner before running a delivery process.
5094 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5095 seem to have the "const" qualifier which it has on other OS. I've
5098 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5099 *of the same type* (client or server) with the same public name that an
5100 error should be diagnosed.
5102 9. When Exim looked up a host name for an IP address, but failed to find the
5103 original IP address when looking up the host name (a safety check), it
5104 output the message "<ip address> does not match any IP for NULL", which was
5105 confusing, to say the least. The bug was that the host name should have
5106 appeared instead of "NULL".
5108 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5109 user that is built into the binary, and the -C or -D options is used, root
5110 privilege is dropped before the configuration file is read. In addition,
5111 logging is switched to stderr instead of the normal log files. If the
5112 configuration then re-defines the Exim user, the unprivileged environment
5113 is probably not what is expected, so Exim logs a panic warning message (but
5116 However, if deliver_drop_privilege is set, the unprivileged state may well
5117 be exactly what is intended, so the warning has been cut out in that case,
5118 and Exim is allowed to try to write to its normal log files.
5124 1. smtp_return_error_details was not giving details for temporary sender
5125 or receiver verification errors.
5127 2. Diagnose a configuration error if two authenticators have the same public
5130 3. Exim used not to create the message log file for a message until the first
5131 delivery attempt. This could be confusing when incoming messages were held
5132 for policy or load reasons. The message log file is now created at the time
5133 the message is received, and an initial "Received" line is written to it.
5135 4. The automatically generated man page for command line options had a minor
5136 bug that caused no ill effects; however, a more serious problem was that
5137 the procedure for building the man page automatically didn't always
5138 operate. Consequently, release 4.20 contains an out-of-date version. This
5139 shouldn't happen again.
5141 5. When building Exim with embedded Perl support, the script that builds the
5142 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5143 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5145 6. The freeze_tell option was not being used for messages that were frozen on
5146 arrival, either by an ACL or by local_scan().
5148 7. Added the smtp_incomplete_transaction log selector.
5150 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5151 was accepting AUTH without a new EHLO.
5153 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5154 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5157 10. Logging of TCP/IP connections (when configured) now happens in the main
5158 daemon process instead of the child process, so that the TCP/IP connection
5159 count is more accurate (but it can never be perfect).
5161 11. The use of "drop" in a nested ACL was not being handled correctly in the
5162 outer ACL. Now, if condition failure induced by the nested "drop" causes
5163 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5164 or "require"), the connection is dropped.
5166 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5167 that yield "discard" can now be used with an "accept" or a "discard" verb,
5168 but an error is generated for any others (because I can't see a useful way
5169 to define what should happen).
5171 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5172 are now processed in the same way as lines in the Exim configuration file.
5173 In particular, continuation lines are supported.
5175 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5177 15. Added -ti meaning -t -i.
5179 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5180 domains, and warn by logging if others are found.
5182 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5183 returned from GIFCONF when Exim is trying to find the list of interfaces on
5184 a host. The code in os.c has been modified to copy each ifreq to an aligned
5185 structure in all cases.
5187 Also, in some cases, the returned ifreq's were being copied to a 'struct
5188 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5189 means the last couple of bytes of an IPv6 address could be chopped if the
5190 ifreq contained only a normal sockaddr (14 bytes storage).
5192 18. Named domain lists were not supported in the hosts_treat_as_local option.
5193 An entry such as +xxxx was not recognized, and was treated as a literal
5196 19. Ensure that header lines added by a DATA ACL are included in the reject log
5197 if the ACL subsequently rejects the message.
5199 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5200 MD5 (which is deprecated).
5202 21. When testing a filter file using -bf, Exim was writing a message when it
5203 took the sender from a "From " line in the message, but it was not doing so
5204 when it took $return_path from a Return-Path: header line. It now does.
5206 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5207 with a valid header line field name (a series of printing characters
5208 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5210 23. Changed "disc" in the source to "disk" to conform to the documentation and
5211 the book and for uniformity.
5213 24. Ignore Sendmail's -Ooption=value command line item.
5215 25. When execve() failed while trying to run a command in a pipe transport,
5216 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5217 could be confused with a return value of 69 from the command itself. This
5218 has been changed to 127, the value the shell returns if it is asked to run
5219 a non-existent command. The wording for the related log line suggests a
5220 non-existent command as the problem.
5222 26. If received_header_text expands to an empty string, do not add a Received:
5223 header line to the message. (Well, it adds a token one on the spool, but
5224 marks it "old" so that it doesn't get used or transmitted.)
5226 27. Installed eximstats 1.28 (addition of -nt option).
5228 28. There was no check for failure on the call to getsockname() in the daemon
5229 code. This can fail if there is a shortage of resources on the system, with
5230 ENOMEM, for example. A temporary error is now given on failure.
5232 29. Contrary to the C standard, it seems that in some environments, the
5233 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5234 program. Exim now does this explicitly; it affects the formatting of
5235 timestamps using strftime().
5237 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5238 complaints. I've now initialized all the variables, to avoid this.
5240 32. Header lines added by a system filter were not being "seen" during
5241 transport-time rewrites.
5243 33. The info_callback() function passed to OpenSSL is set up with type void
5244 (*)(SSL *, int, int), as described somewhere. However, when calling the
5245 function (actually a macro) that sets it up, the type void(*)() is
5246 expected. I've put in a cast to prevent warnings from picky compilers.
5248 34. If a DNS black list lookup found a CNAME record, but there were no A
5249 records associated with the domain it pointed at, Exim crashed.
5251 35. If a DNS black list lookup returned more than one A record, Exim ignored
5252 all but the first. It now scans all returned addresses if a particular IP
5253 value is being sought. In this situation, the contents of the
5254 $dnslist_value variable are a list of all the addresses, separated by a
5257 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5258 to accept a host name and all its aliases if the forward lookup for any of
5259 them yielded the IP address of the incoming connection. Now it accepts only
5260 those names whose forward lookup yields the correct IP address. Any other
5261 names are discarded. This closes a loophole whereby a rogue DNS
5262 administrator could create reverse DNS records to break through a
5263 wildcarded host restriction in an ACL.
5265 37. If a user filter or a system filter that ran in a subprocess used any of
5266 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5267 the wrong values were passed to the pipe command ($thisaddress had the
5268 value of $0, $0 had the value of $1, etc). This bug was introduced by
5269 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5271 38. Improved the line breaking for long SMTP error messages from ACLs.
5272 Previously, if there was no break point between 40 and 75 characters, Exim
5273 left the rest of the message alone. Two changes have been made: (a) I've
5274 reduced the minimum length to 35 characters; (b) if it can't find a break
5275 point between 35 and 75 characters, it looks ahead and uses the first one
5276 that it finds. This may give the occasional overlong line, but at least the
5277 remaining text gets split now.
5279 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5280 file descriptors might be low, and that setting 1000 would always raise it.
5281 It turns out that in some environments, the limit is already over 1000 and
5282 that lowering it causes trouble. So now Exim takes care not to decrease it.
5284 40. When delivering a message, the value of $return_path is set to $sender_
5285 address at the start of routing (routers may change the value). By an
5286 oversight, this default was not being set up when an address was tested by
5287 -bt or -bv, which affected the outcome if any router or filter referred to
5290 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5291 log only when "message" or "log_message" are set. However, if one of the
5292 conditions was an address verification, or a call to a nested ACL, the
5293 messages generated by the underlying test were being passed through. This
5294 no longer happens. The underlying message is available in $acl_verify_
5295 message for both "message" and "log_message" expansions, so it can be
5296 passed through if needed.
5298 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5299 new expansion $bh_ to give the encoded byte string without charset
5300 translation. Translation happens only if iconv() is available; HAVE_ICONV
5301 indicates this at build time. HEADERS_CHARSET gives the charset to
5302 translate to; headers_charset can change it in the configuration, and
5303 "headers charset" can change it in an individual filter file.
5305 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5306 that creates RFC 2047 encoded "words" labels them as that charset instead
5307 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5308 expansion operator; (ii) when Exim creates a From: line for a local
5309 message; (iii) when a header line is rewritten to include a "phrase" part.
5311 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5312 buggy, causing it to skip the first lines of messages whose message ID
5313 ended in 'D'. This would not have bitten before Exim release 4.14, because
5314 message IDs were unlikely to end in 'D' before then. The effect was to have
5315 incorrect size information for certain domains.
5317 45. #include "config.h" was missing at the start of the crypt16.c module. This
5318 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5321 46. If there was a timeout during a "random" callout check, Exim treated it as
5322 a failure of the random address, and carried on sending RSET and the real
5323 address. If the delay was just some slowness somewhere, the response to the
5324 original RCPT would be taken as a response to RSET and so on, causing
5325 mayhem of various kinds.
5327 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5328 when I implemented it. It didn't allow for the fact that some option values
5329 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5330 the right test for positive values.
5332 48. Domain names in DNS records are case-independent. Exim always looks them up
5333 in lower case. Some resolvers return domain names in exactly the case they
5334 appear in the zone file, that is, they may contain uppercase letters. Not
5335 all resolvers do this - some return always lower case. Exim was treating a
5336 change of case by a resolver as a change of domain, similar to a widening
5337 of a domain abbreviation. This triggered its re-routing code and so it was
5338 trying to route what was effectively the same domain again. This normally
5339 caused routing to fail (because the router wouldn't handle the domain
5340 twice). Now Exim checks for this case specially, and just changes the
5341 casing of the domain that it ultimately uses when it transmits the message
5344 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5347 50. If a filter generated a file delivery with a non-absolute name (possible if
5348 no home directory exists for the router), the forbid_file option was not
5351 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5352 the existing equality matching.
5354 52. Exim was using ints instead of ino_t variables in some places where it was
5355 dealing with inode numbers.
5357 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5358 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5359 called TMPDIR, and if it finds it is different, it changes its value.
5361 54. The smtp_printf() function is now made available to local_scan() so
5362 additional output lines can be written before returning. There is also an
5363 smtp_fflush() function to enable the detection of a dropped connection.
5364 The variables smtp_input and smtp_batched_input are exported to
5367 55. Changed the default runtime configuration: the message "Unknown user"
5368 has been removed from the ACL, and instead placed on the localuser router,
5369 using the cannot_route_message feature. This means that any verification
5370 failures that generate their own messages won't get overridden. Similarly,
5371 the "Unrouteable address" message that was in the ACL for unverifiable
5372 relay addresses has also been removed.
5374 56. Added hosts_avoid_esmtp to the smtp transport.
5376 57. The exicyclog script was not checking for the esoteric option
5377 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5378 will work only if exicyclog is run under the appropriate euid.
5380 58. Following a discussion on the list, the rules by which Exim recognises line
5381 endings on incoming messages have been changed. The -dropcr and drop_cr
5382 options are now no-ops, retained only for backwards compatibility. The
5383 following line terminators are recognized: LF CRLF CR. However, special
5384 processing applies to CR:
5386 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5387 nor a local message in the state where . is a terminator.
5389 (ii) If a bare CR is encountered in a header line, an extra space is added
5390 after the line terminator so as not to end the header. The reasoning
5391 behind this is that bare CRs in header lines are most likely either
5392 to be mistakes, or people trying to play silly games.
5394 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5395 was being incorrectly given as 18 bytes larger than it should have been.
5396 This is a VOB (very old bug).
5398 60. This may never have affected anything current, but just in case it has:
5399 When the local host is found other than at the start of a list of hosts,
5400 the local host, those with the same MX, and any that follow, are discarded.
5401 When the list in question was part of a longer list of hosts, the following
5402 hosts (not currently being processed) were also being discarded. This no
5403 longer happens. I'm not sure if this situation could ever has previously
5406 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5409 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5410 header. This is something that is recommended in a new Internet Draft, and
5411 is something that is documented as being done by Sendmail. There are two
5412 possible values. For messages generated by the autoreply transport, Exim
5415 Auto-Submitted: auto-replied
5417 whereas for all other generated messages (e.g. bounces) it adds
5419 Auto-Submitted: auto-generated
5421 63. The "personal" condition in filters now includes a test for the
5422 Auto-Submitted: header. If it contains the string "auto-" the message it
5423 not considered personal.
5425 64. Added rcpt_include_affixes as a generic transport option.
5427 65. Added queue_only_override (default true).
5429 66. Added the syslog_duplication option.
5431 67. If what should have been the first header line of a message consisted of
5432 a space followed by a colon, Exim was mis-interpreting it as a header line.
5433 It isn't of course - it is syntactically invalid and should therefore be
5434 treated as the start of the message body. The misbehaviour could have
5435 caused a number of strange effects, including loss of data in subsequent
5436 header lines, and spool format errors.
5438 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5439 client host had authenticated. This control can now be exercised by an ACL
5440 for more flexibility.
5442 69. By default, callouts do not happen when testing with -bh. There is now a
5443 variant, -bhc, which does actually run the callout code, including
5444 consulting and updating the callout cache.
5446 70. Added support for saslauthd authentication, courtesy of Alexander
5449 71. If statvfs() failed on the spool or log directories while checking their
5450 size for availability, Exim confusingly gave the error "space shortage".
5451 Furthermore, in debugging mode it crashed with a floating point exception.
5452 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5453 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5454 serious problem, Exim now writes to the main and panic logs when this
5455 happens, with details of the failure. It then refuses to accept the
5456 incoming message, giving the message "spool directory problem" or "log
5457 directory problem" with a 421 code for SMTP messages.
5459 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5460 0, 1, and 2 exist, because some OS complain for execs without them (see
5461 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5462 descriptors. However, the code omitted to check that the open succeeded,
5463 causing mysterious errors if for some reason the permissions on /dev/null
5464 got screwed. Now Exim writes a message to the main and panic logs, and
5465 bombs out if it can't open /dev/null.
5467 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5468 interact so that it is all more flexible. It is supposed to remain
5469 backwards compatible. Also added extra_local_interfaces.
5471 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5472 to bomb out with an assertion failure - to the client this appears as a
5473 connection drop. This problem occurs in the part of the code that was taken
5474 from the Samba project. Fortunately, the assertion is in a very simple
5475 function, so I have fixed this by reproducing the function inline in the
5476 one place where it is called, and arranging for authentication to fail
5477 instead of killing the process with assert().
5479 75. The SPA client code was not working when the server requested OEM rather
5480 than Unicode encoding.
5482 76. Added code to make require_files with a specific uid setting more usable in
5483 the case where statting the file as root fails - usually a non-root-mounted
5484 NFS file system. When this happens and the failure is EACCES, Exim now
5485 forks a subprocess and does the per-uid checking as the relevant uid.
5487 77. Added process_log_path.
5489 78. If log_file_path was not explicitly set, a setting of check_log_space or
5490 check_log_inodes was ignored.
5492 79. If a space check for the spool or log partitions fails, the incident is now
5493 logged. Of course, in the latter case the data may get lost...
5495 80. Added the %p formatting code to string_format() so that it can be used to
5496 print addresses in debug_print(). Adjusted all the address printing in the
5497 debugging in store.c to use %p rather than %d.
5499 81. There was a concern that a line of code in smtp_in.c could overflow a
5500 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5501 initially expressed, the concern was not well-founded, because trailing
5502 spaces are removed early. However, if the trailing spaces were followed by
5503 a NULL, they did not get removed, so the overflow was possible. Two fixes
5506 (a) I re-wrote the offending code in a cleaner fashion.
5507 (b) If an incoming SMTP command contains a NULL character, it is rejected
5510 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5511 runs initgroups(), so that if the Exim user is in any additional groups,
5512 they will be used during message reception.
5518 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob