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 New expansion variable $tls_bits; Cyrus SASL server connection
19 properties get this fed in as external SSF. A number of robustness
20 and debugging improvements to the cyrus_sasl authenticator.
26 PP/01 Solaris build fix for Oracle's LDAP libraries.
27 Bugzilla 1109, patch from Stephen Usher.
29 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
31 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
34 TF/02 Fix a couple more cases where we did not log the error message
35 when unlink() failed. See also change 4.74-TF/03.
37 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
38 lock up or crash if it happened to be inside a call to libc when it
39 got a SIGUSR1 from exiwhat.
41 The SIGUSR1 handler appends the current process status to the process
42 log which is later printed by exiwhat. It used to use the general
43 purpose logging code to do this, but several functions it calls are
46 The new output code in the SIGUSR1 handler is specific to the process
47 log, and simple enough that it's easy to inspect for signal safety.
48 Removing some special cases also simplifies the general logging code.
49 Removing the spurious timestamps from the process log simplifies
52 TF/04 Improved ratelimit ACL condition.
54 The /noupdate option has been deprecated in favour of /readonly which
55 has clearer semantics. The /leaky, /strict, and /readonly update modes
56 are mutually exclusive. The update mode is no longer included in the
57 database key; it just determines when the database is updated. (This
58 means that when you upgrde Exim will forget old rate measurements.)
60 Exim now checks that the per_* options are used with an update mode that
61 makes sense for the current ACL. For example, when Exim is processing a
62 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
63 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
64 must specify per_mail/readonly. If you omit the update mode it defaults to
65 /leaky where that makes sense (as before) or /readonly where required.
67 The /noupdate option is now undocumented but still supported for
68 backwards compatibility. It is equivalent to /readonly except that in
69 ACLs where /readonly is required you may specify /leaky/noupdate or
70 /strict/noupdate which are treated the same as /readonly.
72 A useful new feature is the /count= option. This is a generalization
73 of the per_byte option, so that you can measure the throughput of other
74 aggregate values. For example, the per_byte option is now equivalent
75 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
77 The per_rcpt option has been generalized using the /count= mechanism
78 (though it's more complicated than the per_byte equivalence). When it is
79 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
80 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
81 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
82 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
83 Note that using per_rcpt with a non-readonly update mode in more than
84 one ACL will cause the recipients to be double-counted. (The per_mail
85 and per_byte options don't have this problem.)
87 The handling of very low rates has changed slightly. If the computed rate
88 is less than the event's count (usually one) then this event is the first
89 after a long gap. In this case the rate is set to the same as this event's
90 count, so that the first message of a spam run is counted properly.
92 The major new feature is a mechanism for counting the rate of unique
93 events. The new per_addr option counts the number of different
94 recipients that someone has sent messages to in the last time period. It
95 behaves like per_rcpt if all the recipient addresses are different, but
96 duplicate recipient addresses do not increase the measured rate. Like
97 the /count= option this is a general mechanism, so the per_addr option
98 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
99 example, measure the rate that a client uses different sender addresses
100 with the options per_mail/unique=$sender_address. There are further
101 details in the main documentation.
103 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
105 TF/06 Removed a few PCRE remnants.
107 TF/07 Automatically extract Exim's version number from tags in the git
108 repository when doing development or release builds.
110 PP/02 Raise smtp_cmd_buffer_size to 16kB.
111 Bugzilla 879. Patch from Paul Fisher.
113 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
114 Heavily based on revision 40f9a89a from Simon Arlott's tree.
117 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
119 PP/05 Variable $av_failed, true if the AV scanner deferred.
120 Bugzilla 1078. Patch from John Horne.
122 PP/06 Stop make process more reliably on build failure.
123 Bugzilla 1087. Patch from Heiko Schlittermann.
125 PP/07 Make maildir_use_size_file an _expandable_ boolean.
126 Bugzilla 1089. Patch from Heiko Schlittermann.
128 PP/08 Handle ${run} returning more data than OS pipe buffer size.
129 Bugzilla 1131. Patch from Holger Weiß.
131 PP/09 Handle IPv6 addresses with SPF.
132 Bugzilla 860. Patch from Wolfgang Breyha.
134 PP/10 GnuTLS: support TLS 1.2 & 1.1.
136 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
139 PP/11 match_* no longer expand right-hand-side by default.
140 New compile-time build option, EXPAND_LISTMATCH_RHS.
141 New expansion conditions, "inlist", "inlisti".
143 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
145 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
147 PP/14 fix log_write() format string regression from TF/03.
148 Bugzilla 1152. Patch from Dmitry Isaikin.
154 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
156 PP/02 Harmonised TLS library version reporting; only show if debugging.
157 Layout now matches that introduced for other libraries in 4.74 PP/03.
159 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
161 PP/04 New "dns_use_edns0" global option.
163 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
166 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
167 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
169 TK/01 Updated PolarSSL code to 0.14.2.
170 Bugzilla 1097. Patch from Andreas Metzler.
172 PP/07 Catch divide-by-zero in ${eval:...}.
175 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
178 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
179 format-string attack -- SECURITY: remote arbitrary code execution.
181 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
182 time unintentionally subject to list matching rules, letting the header
183 cause arbitrary Exim lookups (of items which can occur in lists, *not*
184 arbitrary string expansion). This allowed for information disclosure.
186 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
187 INT_MIN/-1 -- value coerced to INT_MAX.
193 NM/01 Workround for PCRE version dependency in version reporting
196 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
197 This fixes portability to compilers other than gcc, notably
198 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
200 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
201 makefiles for portability to HP-UX and POSIX correctness.
203 PP/01 Permit LOOKUP_foo enabling on the make command-line.
204 Also via indented variable definition in the Makefile.
205 (Debugging by Oliver Heesakkers).
207 PP/02 Restore caching of spamd results with expanded spamd_address.
208 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
210 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
211 Improves build reliability. Fix from: Frank Elsner
213 NM/02 Fix wide character breakage in the rfc2047 coding
214 Fixes bug 1064. Patch from Andrey N. Oktyabrski
216 NM/03 Allow underscore in dnslist lookups
217 Fixes bug 1026. Patch from Graeme Fowler
219 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
220 Code patches from Adam Ciarcinski of NetBSD.
222 NM/04 Fixed exiqgrep to cope with mailq missing size issue
225 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
226 is logged, to avoid truncation. Patch from John Horne.
228 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
229 Patch from Jakob Hirsch.
231 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
232 SQL string expansion failure details.
233 Patch from Andrey Oktyabrski.
235 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
236 Patch from Simon Arlott.
238 PP/09 New lookups functionality failed to compile on old gcc which rejects
239 extern declarations in function scope.
240 Patch from Oliver Fleischmann
242 PP/10 Use sig_atomic_t for flags set from signal handlers.
243 Check getgroups() return and improve debugging.
244 Fixed developed for diagnosis in bug 927 (which turned out to be
247 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
248 Patch from Mark Zealey.
250 PP/12 Bugzilla 1056: Improved spamd server selection.
251 Patch from Mark Zealey.
253 PP/13 Bugzilla 1086: Deal with maildir quota file races.
254 Based on patch from Heiko Schlittermann.
256 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
257 Patch from Uwe Doering, sign-off by Michael Haardt.
259 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
260 variable declaration deep within a block. Bug and patch from
263 PP/15 lookups-Makefile IRIX compatibilty coercion.
265 PP/16 Make DISABLE_DKIM build knob functional.
267 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
268 Patch by Simon Arlott
270 TF/03 Fix valgrind.h portability to C89 compilers that do not support
271 variable argument macros. Our copy now differs from upstream.
277 TF/01 Failure to get a lock on a hints database can have serious
278 consequences so log it to the panic log.
280 TF/02 Log LMTP confirmation messages in the same way as SMTP,
281 controlled using the smtp_confirmation log selector.
283 TF/03 Include the error message when we fail to unlink a spool file.
285 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
286 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
287 for maintaining out-of-tree patches for some time.
289 PP/01 Bugzilla 139: Documentation and portability issues.
290 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
291 Handle per-OS dynamic-module compilation flags.
293 PP/02 Let /dev/null have normal permissions.
294 The 4.73 fixes were a little too stringent and complained about the
295 permissions on /dev/null. Exempt it from some checks.
296 Reported by Andreas M. Kirchwitz.
298 PP/03 Report version information for many libraries, including
299 Exim version information for dynamically loaded libraries. Created
300 version.h, now support a version extension string for distributors
301 who patch heavily. Dynamic module ABI change.
303 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
304 privilege escalation vulnerability whereby the Exim run-time user
305 can cause root to append content of the attacker's choosing to
308 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
311 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
312 If dropping privileges for untrusted macros, we disabled normal logging
313 on the basis that it would fail; for the Exim run-time user, this is not
314 the case, and it resulted in successful deliveries going unlogged.
315 Fixed. Reported by Andreas Metzler.
321 PP/01 Date: & Message-Id: revert to normally being appended to a message,
322 only prepend for the Resent-* case. Fixes regression introduced in
323 Exim 4.70 by NM/22 for Bugzilla 607.
325 PP/02 Include check_rfc2047_length in configure.default because we're seeing
326 increasing numbers of administrators be bitten by this.
328 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
330 PP/03 Bugzilla 994: added openssl_options main configuration option.
332 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
334 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
336 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
338 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
339 without a peer certificate, leading to a segfault because of an
340 assumption that peers always have certificates. Be a little more
341 paranoid. Problem reported by Martin Tscholak.
343 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
344 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
345 NB: ClamAV planning to remove STREAM in "middle of 2010".
346 CL also introduces -bmalware, various -d+acl logging additions and
347 more caution in buffer sizes.
349 PP/09 Implemented reverse_ip expansion operator.
351 PP/10 Bugzilla 937: provide a "debug" ACL control.
353 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
355 PP/12 Bugzilla 973: Implement --version.
357 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
359 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
361 PP/15 Bugzilla 816: support multiple condition rules on Routers.
363 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
364 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
365 ignore trailing whitespace.
367 JJ/02 prevent non-panic DKIM error from being sent to paniclog
369 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
372 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
373 Notification from Dr Andrew Aitchison.
375 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
376 ExtendedDetectionInfo response format.
377 Notification from John Horne.
379 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
382 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
383 XSL and documented dependency on system catalogs, with examples of how
386 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
389 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
390 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
391 configuration file which is writeable by the Exim user or group.
393 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
394 of configuration files to cover files specified with the -C option if
395 they are going to be used with root privileges, not just the default
398 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
399 option (effectively making it always true).
401 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
402 files to be used while preserving root privileges.
404 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
405 that rogue child processes cannot use them.
407 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
408 run-time user, instead of root.
410 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
411 Exim run-time user without dropping privileges.
413 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
414 result string, instead of calling string_vformat() twice with the same
417 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
418 for other users. Others should always drop root privileges if they use
419 -C on the command line, even for a whitelisted configure file.
421 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
423 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
429 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
430 $data_path, and $header_path variables; fixed documentation bugs and
433 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
434 exipick to access non-standard spools, including the "frozen" queue
437 NM/01 Bugzilla 965: Support mysql stored procedures.
438 Patch from Alain Williams
440 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
442 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
443 Patch from Andreas Metzler
445 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
446 Patch from Kirill Miazine
448 NM/05 Bugzilla 671: Added umask to procmail example.
450 JJ/03 installed exipick 20100323.0, fixing doc bug
452 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
453 directory. Notification and patch from Dan Rosenberg.
455 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
457 TK/02 Improve log output when DKIM signing operation fails.
459 MH/01 Treat the transport option dkim_domain as a colon separated
460 list, not as a single string, and sign the message with each element,
461 omitting multiple occurences of the same signer.
463 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
464 Bugzilla 985, 986. Patch by Simon Arlott
466 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
467 Patch by Simon Arlott
469 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
470 MBX locking. Notification from Dan Rosenberg.
476 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
478 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
480 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
482 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
484 NM/04 Bugzilla 915: Flag broken perl installation during build.
490 TK/01 Added patch by Johannes Berg that expands the main option
491 "spamd_address" if it starts with a dollar sign.
493 TK/02 Write list of recipients to X-Envelope-Sender header when building
494 the mbox-format spool file for content scanning (suggested by Jakob
497 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
498 (http://www.dcc-servers.net/) support via dccifd. Activated by
499 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
501 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
502 by Mark Daniel Reidel <mr@df.eu>.
504 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
505 When building exim an external PCRE library is now needed -
506 PCRE is a system library on the majority of modern systems.
507 See entry on PCRE_LIBS in EDITME file.
509 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
510 conversation. Added nologin parameter to request.
511 Patch contributed by Kirill Miazine.
513 TF/01 Do not log submission mode rewrites if they do not change the address.
515 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
517 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
518 log files in place. Contributed by Roberto Lima.
520 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
522 TF/03 Bugzilla 615: When checking the local_parts router precondition
523 after a local_part_suffix or local_part_prefix option, Exim now
524 does not use the address's named list lookup cache, since this
525 contains cached lookups for the whole local part.
527 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
528 Robert Millan. Documentation is in experimental-spec.txt.
530 TF/04 Bugzilla 668: Fix parallel build (make -j).
532 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
534 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
535 Patch provided by Jan Srzednicki.
537 TF/05 Leading white space used to be stripped from $spam_report which
538 wrecked the formatting. Now it is preserved.
540 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
541 that they are available at delivery time.
543 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
545 TF/08 TLS error reporting now respects the incoming_interface and
546 incoming_port log selectors.
548 TF/09 Produce a more useful error message if an SMTP transport's hosts
549 setting expands to an empty string.
551 NM/06 Bugzilla 744: EXPN did not work under TLS.
552 Patch provided by Phil Pennock.
554 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
555 Patch provided by Richard Godbee.
557 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
558 acl_smtp_notquit, added index entry.
560 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
561 Patch provided by Eugene Bujak.
563 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
564 accept(). Patch provided by Maxim Dounin.
566 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
567 Patch provided by Phil Pennock.
569 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
571 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
572 Patch provided by Brad "anomie" Jorsch.
574 NM/14 Bugzilla 622: Added timeout setting to transport filter.
575 Patch provided by Dean Brooks.
577 TK/05 Add native DKIM support (does not depend on external libraries).
579 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
580 Patch provided by Graeme Fowler.
582 NM/16 Bugzilla 851: Documentation example syntax fix.
584 NM/17 Changed NOTICE file to remove references to embedded PCRE.
586 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
589 NM/19 Bugzilla 745: TLS version reporting.
590 Patch provided by Phil Pennock.
592 NM/20 Bugzilla 167: bool: condition support.
593 Patch provided by Phil Pennock.
595 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
596 clients. Patch provided by Phil Pennock.
598 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
599 Patch provided by Brad "anomie" Jorsch.
601 NM/23 Bugzilla 687: Fix misparses in eximstats.
602 Patch provided by Heiko Schlittermann.
604 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
605 Patch provided by Heiko Schlittermann.
607 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
608 plus update to original patch.
610 NM/26 Bugzilla 799: Documentation correction for ratelimit.
612 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
613 Patch provided by David Brownlee.
615 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
617 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
619 NM/30 Bugzilla 888: TLS documentation bugfixes.
621 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
623 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
624 Unlike the original bugzilla I have changed all shell scripts in src tree.
626 NM/33 Bugzilla 898: Transport filter timeout fix.
627 Patch by Todd Rinaldo.
629 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
630 Patch by Serge Demonchaux.
632 NM/35 Bugzilla 39: Base64 decode bug fixes.
633 Patch by Jakob Hirsch.
635 NM/36 Bugzilla 909: Correct connect() call in dcc code.
637 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
639 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
641 NM/39 Bugzilla 911: Fixed MakeLinks build script.
647 TK/01 Add preliminary DKIM support. Currently requires a forked version of
648 ALT-N's libdkim that I have put here:
649 http://duncanthrax.net/exim-experimental/
651 Note to Michael Haardt: I had to rename some vars in sieve.c. They
652 were called 'true' and it seems that C99 defines that as a reserved
653 keyword to be used with 'bool' variable types. That means you could
654 not include C99-style headers which use bools without triggering
655 build errors in sieve.c.
657 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
658 as mailq or other aliases. Changed the --help handling significantly
659 to do whats expected. exim_usage() emits usage/help information.
661 SC/01 Added the -bylocaldomain option to eximstats.
663 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
665 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
667 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
673 PH/01 Another patch from the Sieve maintainer.
675 PH/02 When an IPv6 address is converted to a string for single-key lookup
676 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
677 dots are used instead of colons so that keys in lsearch files need not
678 contain colons. This was done some time before quoting was made available
679 in lsearch files. However, iplsearch files do require colons in IPv6 keys
680 (notated using the quote facility) so as to distinguish them from IPv4
681 keys. This meant that lookups for IP addresses in host lists did not work
682 for iplsearch lookups.
684 This has been fixed by arranging for IPv6 addresses to be expressed with
685 colons if the lookup type is iplsearch. This is not incompatible, because
686 previously such lookups could never work.
688 The situation is now rather anomolous, since one *can* have colons in
689 ordinary lsearch keys. However, making the change in all cases is
690 incompatible and would probably break a number of configurations.
692 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
695 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
696 conversion specification without a maximum field width, thereby enabling
697 a rogue spamd server to cause a buffer overflow. While nobody in their
698 right mind would setup Exim to query an untrusted spamd server, an
699 attacker that gains access to a server running spamd could potentially
700 exploit this vulnerability to run arbitrary code as the Exim user.
702 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
703 $primary_hostname instead of what libspf2 thinks the hosts name is.
705 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
706 a directory entry by the name of the lookup key. Previously, if a
707 symlink pointed to a non-existing file or a file in a directory that
708 Exim lacked permissions to read, a lookup for a key matching that
709 symlink would fail. Now it is enough that a matching directory entry
710 exists, symlink or not. (Bugzilla 503.)
712 PH/03 The body_linecount and body_zerocount variables are now exported in the
715 PH/04 Added the $dnslist_matched variable.
717 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
718 This means they are set thereafter only if the connection becomes
721 PH/06 Added the client_condition to authenticators so that some can be skipped
722 by clients under certain conditions.
724 PH/07 The error message for a badly-placed control=no_multiline_responses left
725 "_responses" off the end of the name.
727 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
729 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
730 (without spaces) instead of just copying the configuration text.
732 PH/10 Added the /noupdate option to the ratelimit ACL condition.
734 PH/11 Added $max_received_linelength.
736 PH/12 Added +ignore_defer and +include_defer to host lists.
738 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
739 way in which PCRE > 7.0 is built.
741 PH/14 Implemented queue_only_load_latch.
743 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
744 MAIL command. The effect was to mangle the value on 64-bit systems.
746 PH/16 Another patch from the Sieve maintainer.
748 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
750 PH/18 If a system quota error occurred while trying to create the file for
751 a maildir delivery, the message "Mailbox is full" was not appended to the
752 bounce if the delivery eventually timed out. Change 4.67/27 below applied
753 only to a quota excession during the actual writing of the file.
755 PH/19 It seems that peer DN values may contain newlines (and other non-printing
756 characters?) which causes problems in log lines. The DN values are now
757 passed through string_printing() before being added to log lines.
759 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
760 and InterBase are left for another time.)
762 PH/21 Added message_body_newlines option.
764 PH/22 Guard against possible overflow in moan_check_errorcopy().
766 PH/23 POSIX allows open() to be a macro; guard against that.
768 PH/24 If the recipient of an error message contained an @ in the local part
769 (suitably quoted, of course), incorrect values were put in $domain and
770 $local_part during the evaluation of errors_copy.
776 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
777 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
780 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
781 issue a MAIL command.
783 PH/02 In an ACL statement such as
785 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
787 if a client was not listed at all, or was listed with a value other than
788 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
789 the condition was not true (as it should be), so access was not denied.
790 The bug was that the ! inversion was incorrectly passed on to the second
791 item. This has been fixed.
793 PH/03 Added additional dnslists conditions == and =& which are different from
794 = and & when the dns lookup returns more than one IP address.
796 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
797 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
799 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
800 FSYNC, which compiles an option called disable_fsync that allows for
801 bypassing fsync(). The documentation is heavily laced with warnings.
803 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
805 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
806 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
807 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
808 including adding "make clean"; (3) Added -fPIC when compiling the test
809 dynamically loaded module, to get rid of a warning.
811 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
812 message fails, move_frozen_messages = true and ignore_bounce_errors_after
813 = 0s. The bug is otherwise harmless.
815 PH/07 There was a bug in the dovecot authenticator such that the value of
816 $auth1 could be overwritten, and so not correctly preserved, after a
817 successful authentication. This usually meant that the value preserved by
818 the server_setid option was incorrect.
820 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
822 PH/09 Installed PCRE release 7.0.
824 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
825 run for batched SMTP input. It is now run at the start of every message
826 in the batch. While fixing this I discovered that the process information
827 (output by running exiwhat) was not always getting set for -bs and -bS
828 input. This is fixed, and it now also says "batched" for BSMTP.
830 PH/11 Added control=no_pipelining.
832 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
833 patch, slightly modified), and move the expansion of helo_data till after
834 the connection is made in the smtp transport (so it can use these
837 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
839 PH/14 Added log_selector = +pid.
841 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
843 PH/16 Add ${if forany and ${if forall.
845 PH/17 Added dsn_from option to vary the From: line in DSNs.
847 PH/18 Flush SMTP output before performing a callout, unless control =
848 no_callout_flush is set.
850 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
851 was true (the default) a successful delivery failed to delete the retry
852 item, thus causing premature timeout of the address. The bug is now
855 PH/20 Added hosts_avoid_pipelining to the smtp transport.
857 PH/21 Long custom messages for fakedefer and fakereject are now split up
858 into multiline reponses in the same way that messages for "deny" and
859 other ACL rejections are.
861 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
862 with slight modification.
864 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
865 draft, changing the syntax and factoring some duplicate code".
867 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
868 for deliveries of the second and subsequent messages over the same SMTP
871 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
872 ${reduce, with only minor "tidies".
874 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
876 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
877 expansion side effects.
879 PH/27 When a message times out after an over-quota error from an Exim-imposed
880 quota, the bounce message says "mailbox is full". This message was not
881 being given when it was a system quota that was exceeded. It now should
884 MH/03 Made $recipients available in local_scan(). local_scan() already has
885 better access to the recipient list through recipients_list[], but
886 $recipients can be useful in postmaster-provided expansion strings.
888 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
889 in the case of a MAIL command with additional options following the
890 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
891 were accidentally chopped off.
893 PH/29 SMTP synchronization checks are implemented when a command is read -
894 there is a check that no more input is waiting when there shouldn't be
895 any. However, for some commands, a delay in an ACL can mean that it is
896 some time before the response is written. In this time, more input might
897 arrive, invalidly. So now there are extra checks after an ACL has run for
898 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
899 pipelining has not been advertised.
901 PH/30 MH's patch to allow iscntrl() characters to be list separators.
903 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
904 returned in the SMTP response when smtp_return_error_details was false.
907 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
908 instead of the C I/O that was originally supplied, because problems were
911 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
912 Exim which did not show up earlier: it was assuming that a call to
913 SSL_CTX_set_info_callback() might give an error value. In fact, there is
914 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
915 was a macro that became an assignment, so it seemed to work. This has
916 changed to a proper function call with a void return, hence the compile
917 error. Exim's code has been fixed.
919 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
922 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
924 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
926 PH/37 If a message is not accepted after it has had an id assigned (e.g.
927 because it turns out to be too big or there is a timeout) there is no
928 "Completed" line in the log. When some messages of this type were
929 selected by exigrep, they were listed as "not completed". Others were
930 picked up by some special patterns. I have improved the selection
931 criteria to be more general.
933 PH/38 The host_find_failed option in the manualroute router can now be set
934 to "ignore", to completely ignore a host whose IP address cannot be
935 found. If all hosts are ignored, the behaviour is controlled by the new
936 host_all_ignored option.
938 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
939 homing or because of multiple MX records with /mx) generated more than
940 one IP address, and the following item turned out to be the local host,
941 all the secondary addresses of the first item were incorrectly removed
942 from the list, along with the local host and any following hosts (which
943 is what is supposed to happen).
945 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
946 whoever called Exim into the -H file. In the case of the daemon it was
947 behaving confusingly. When first started, it used values for whoever
948 started the daemon, but after a SIGHUP it used the Exim user (because it
949 calls itself on a restart). I have changed the code so that it now always
952 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
953 message are rejected with the same error (e.g. no authentication or bad
954 sender address), and a DATA command is nevertheless sent (as can happen
955 with PIPELINING or a stupid MUA), the error message that was given to the
956 RCPT commands is included in the rejection of the DATA command. This is
957 intended to be helpful for MUAs that show only the final error to their
960 PH/42 Another patch from the Sieve maintainer.
962 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
963 Eximstats - Fixed some broken HTML links and added missing column headers
965 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
966 columns for Rejects, Temp Rejects, Ham, and Spam rows.
968 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
970 PH/43 Yet another patch from the Sieve maintainer.
972 PH/44 I found a way to check for a TCP/IP connection going away before sending
973 the response to the final '.' that terminates a message, but only in the
974 case where the client has not sent further data following the '.'
975 (unfortunately, this is allowed). However, in many cases there won't be
976 any further data because there won't be any more messages to send. A call
977 to select() can be used: if it shows that the input is "ready", there is
978 either input waiting, or the socket has been closed. An attempt to read
979 the next input character can distinguish the two cases. Previously, Exim
980 would have sent an OK response which the client would never have see.
981 This could lead to message repetition. This fix should cure that, at
982 least in a lot of common cases.
984 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
985 advertised in response to EHLO.
991 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
992 fixed by 4.65/MH/01 (is this a record?) are fixed:
994 (i) An empty string was always treated as zero by the numeric comparison
995 operators. This behaviour has been restored.
997 (ii) It is documented that the numeric comparison operators always treat
998 their arguments as decimal numbers. This was broken in that numbers
999 starting with 0 were being interpreted as octal.
1001 While fixing these problems I realized that there was another issue that
1002 hadn't been noticed. Values of message_size_limit (both the global option
1003 and the transport option) were treated as octal if they started with 0.
1004 The documentation was vague. These values are now always treated as
1005 decimal, and I will make that clear in the documentation.
1011 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1012 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1015 MH/01 Don't check that the operands of numeric comparison operators are
1016 integers when their expansion is in "skipping" mode (fixes bug
1017 introduced by 4.64-PH/07).
1019 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1020 child addresses, Exim now panics and dies. Previously, because the count
1021 is held in a short int, deliveries were likely to be lost. As such a
1022 large number of recipients for a single message is ridiculous
1023 (performance will be very, very poor), I have chosen to impose a limit
1024 rather than extend the field.
1030 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1031 leftover -K file (the existence of which was triggered by #402).
1032 While we were at it, introduced process PID as part of the -K
1033 filename. This should rule out race conditions when creating
1036 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1037 processing considerably. Previous code took too long for large mails,
1038 triggering a timeout which in turn triggers #401.
1040 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1041 in the DK code in transports.c. sendfile() is not really portable,
1042 hence the _LINUX specificness.
1044 TF/01 In the add_headers option to the mail command in an Exim filter,
1045 there was a bug that Exim would claim a syntax error in any
1046 header after the first one which had an odd number of characters
1049 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1050 callout verification, Exim cached a "reject" for the entire domain. This
1051 is correct for most verifications, but it is not correct for a recipient
1052 verification with use_sender or use_postmaster set, because in that case
1053 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1054 case of MAIL FROM:<> rejection from other early rejections (e.g.
1055 rejection of HELO). When verifying a recipient using a non-null MAIL
1056 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1057 Whatever the result of the callout, the value of the domain cache is
1058 left unchanged (for any other kind of callout, getting as far as trying
1059 RCPT means that the domain itself is ok).
1061 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1064 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1065 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1066 session, the daemon ignores ECONNECT errors and logs others; it now
1067 ignores EPIPE as well.
1069 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1070 (quoted-printable decoding).
1072 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1073 later the small subsequent patch to fix an introduced bug.
1075 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1077 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1079 PH/08 An error is now given if message_size_limit is specified negative.
1081 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1082 to be given (somewhat) arbitrary names.
1084 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1087 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1088 miscellaneous code fixes
1090 PH/10 Added the log_reject_target ACL modifier to specify where to log
1093 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1094 hostname. This is wrong, because it relates to the incoming message (and
1095 probably the interface on which it is arriving) and not to the outgoing
1096 callout (which could be using a different interface). This has been
1097 changed to use the value of the helo_data option from the smtp transport
1098 instead - this is what is used when a message is actually being sent. If
1099 there is no remote transport (possible with a router that sets up host
1100 addresses), $smtp_active_hostname is used.
1102 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1103 tweaks were necessary in order to get it to work (see also 21 below):
1104 (a) The code assumed that strncpy() returns a negative number on buffer
1105 overflow, which isn't the case. Replaced with Exim's string_format()
1107 (b) There were several signed/unsigned issues. I just did the minimum
1108 hacking in of casts. There is scope for a larger refactoring.
1109 (c) The code used strcasecmp() which is not a standard C function.
1110 Replaced with Exim's strcmpic() function.
1111 (d) The code set only $1; it now sets $auth1 as well.
1112 (e) A simple test gave the error "authentication client didn't specify
1113 service in request". It would seem that Dovecot has changed its
1114 interface. Fortunately there's a specification; I followed it and
1115 changed what the client sends and it appears to be working now.
1117 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1120 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1121 address A is aliased to B and C, where B exists and C does not. Without
1122 -v the output is "A verified" because verification stops after a
1123 successful redirection if more than one address is generated. However,
1124 with -v the child addresses are also verified. Exim was outputting "A
1125 failed to verify" and then showing the successful verification for C,
1126 with its parentage. It now outputs "B failed to verify", showing B's
1127 parentage before showing the successful verification of C.
1129 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1130 look up a TXT record in a specific list after matching in a combined
1133 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1134 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1135 they consult the DNS. I had assumed they would set it the way they
1136 wanted; and indeed my experiments on Linux seem to show that in some
1137 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1138 To be on the safe side, however, I have now made the interface to
1139 host_find_byname() similar to host_find_bydns(), with an argument
1140 containing the DNS resolver options. The host_find_byname() function now
1141 sets these options at its start, just as host_find_bydns() does. The smtp
1142 transport options dns_qualify_single and dns_search_parents are passed to
1143 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1144 of host_find_byname() use the default settings of RES_DEFNAMES
1145 (qualify_single) but not RES_DNSRCH (search_parents).
1147 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1148 spool_read_header() do less string testing, by means of a preliminary
1149 switch on the second character of optional "-foo" lines. (This is
1150 overdue, caused by the large number of possibilities that now exist.
1151 Originally there were few.) While I was there, I also converted the
1152 str(n)cmp tests so they don't re-test the leading "-" and the first
1153 character, in the hope this might squeeze out yet more improvement.
1155 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1156 flag allowing group syntax was set by the header_syntax check but not
1157 turned off, possible causing trouble later; (2) The flag was not being
1158 set at all for the header_verify test, causing "group"-style headers to
1159 be rejected. I have now set it in this case, and also caused header_
1160 verify to ignore an empty address taken from a group. While doing this, I
1161 came across some other cases where the code for allowing group syntax
1162 while scanning a header line wasn't quite right (mostly, not resetting
1163 the flag correctly in the right place). These bugs could have caused
1164 trouble for malformed header lines. I hope it is now all correct.
1166 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1167 with the "reply" argument non-NULL. The code, however (which originally
1168 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1169 but it didn't always do it. This confused somebody who was copying the
1170 code for some other use. I have removed all the tests.
1172 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1173 feature that was used to support insecure browsers during the U.S. crypto
1174 embargo. It requires special client support, and Exim is probably the
1175 only MTA that supported it -- and would never use it because real RSA is
1176 always available. This code has been removed, because it had the bad
1177 effect of slowing Exim down by computing (never used) parameters for the
1178 RSA_EXPORT functionality.
1180 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1181 authenticator to fail if there's a tab character in the incoming data
1182 (there should never be unless someone is messing about, as it's supposed
1183 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1184 if the connection is using TLS or if the remote IP is the same as the
1185 local IP, and the "valid-client-cert option" if a client certificate has
1188 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1189 authenticators. This can be used for authorization after authentication
1190 succeeds. (In the case of plaintext, it servers for both authentication
1193 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1194 if any retry times were supplied.
1196 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1197 connection, where there is no client IP address to check. In this
1198 situation, the verify now always succeeds.
1200 PH/25 Applied John Jetmore's -Mset patch.
1202 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1204 PH/27 In a string expansion for a processed (not raw) header when multiple
1205 headers of the same name were present, leading whitespace was being
1206 removed from all of them, but trailing whitespace was being removed only
1207 from the last one. Now trailing whitespace is removed from each header
1208 before concatenation. Completely empty headers in a concatenation (as
1209 before) are ignored.
1211 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1212 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1214 PH/29 [Removed. This was a change that I later backed out, and forgot to
1215 correct the ChangeLog entry (that I had efficiently created) before
1216 committing the later change.]
1218 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1219 address errors (4xx response to RCPT) over the same connection as other
1220 messages routed to the same hosts. Such deliveries are always "forced",
1221 so retry times are not inspected. This resulted in far too many retries
1222 for the affected addresses. The effect occurred only when there were more
1223 hosts than the hosts_max_try setting in the smtp transport when it had
1224 the 4xx errors. Those hosts that it had tried were not added to the list
1225 of hosts for which the message was waiting, so if all were tried, there
1226 was no problem. Two fixes have been applied:
1228 (i) If there are any address or message errors in an SMTP delivery, none
1229 of the hosts (tried or untried) are now added to the list of hosts
1230 for which the message is waiting, so the message should not be a
1231 candidate for sending over the same connection that was used for a
1232 successful delivery of some other message. This seems entirely
1233 reasonable: after all the message is NOT "waiting for some host".
1234 This is so "obvious" that I'm not sure why it wasn't done
1235 previously. Hope I haven't missed anything, but it can't do any
1236 harm, as the worst effect is to miss an optimization.
1238 (ii) If, despite (i), such a delivery is accidentally attempted, the
1239 routing retry time is respected, so at least it doesn't keep
1240 hammering the server.
1242 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1243 in ${readsocket because some servers need this prod.
1245 PH/32 Added some extra debug output when updating a wait-xxx database.
1247 PH/33 The hint "could be header name not terminated by colon", which has been
1248 given for certain expansion errors for a long time, was not being given
1249 for the ${if def:h_colon_omitted{... case.
1251 PH/34 The spec says: "With one important exception, whenever a domain list is
1252 being scanned, $domain contains the subject domain." There was at least
1253 one case where this was not true.
1255 PH/35 The error "getsockname() failed: connection reset by peer" was being
1256 written to the panic log as well as the main log, but it isn't really
1257 panic-worthy as it just means the connection died rather early on. I have
1258 removed the panic log writing for the ECONNRESET error when getsockname()
1261 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1262 runs only) independently of the message's sender address. This meant
1263 that, if the 4xx error was in fact related to the sender, a different
1264 message to the same recipient with a different sender could confuse
1265 things. In particualar, this can happen when sending to a greylisting
1266 server, but other circumstances could also provoke similar problems.
1267 I have changed the default so that the retry time for these errors is now
1268 based a combination of the sender and recipient addresses. This change
1269 can be overridden by setting address_retry_include_sender=false in the
1272 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1273 remote server are returned as part of bounce messages. This was not
1274 happening for LMTP over a pipe (the lmtp transport), but now it is the
1275 same for both kinds of LMTP.
1277 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1278 in header lines that were in fact CNAMEs. This is no longer the case.
1280 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1281 and queue runs started by the daemon processed all messages. This has
1282 been fixed so that -R and -S can now usefully be given with -q<time>.
1284 PH/40 Import PCRE release 6.7 (fixes some bugs).
1286 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1288 PH/42 Give an error if -q is specified more than once.
1290 PH/43 Renamed the variables $interface_address and $interface_port as
1291 $received_ip_address and $received_port, to make it clear that these
1292 values apply to message reception, and not to the outgoing interface when
1293 a message is delivered. (The old names remain recognized, of course.)
1295 PH/44 There was no timeout on the connect() call when using a Unix domain
1296 socket in the ${readsocket expansion. There now is.
1298 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1299 be meaningful with "accept".
1301 SC/01 Eximstats V1.43
1302 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1304 SC/02 Eximstats V1.44
1305 Use a glob alias rather than an array ref in the generated
1306 parser. This improves both readability and performance.
1308 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1309 Collect SpamAssassin and rejection statistics.
1310 Don't display local sender or destination tables unless
1311 there is data to show.
1312 Added average volumes into the top table text output.
1314 SC/04 Eximstats V1.46
1315 Collect data on the number of addresses (recipients)
1316 as well as the number of messages.
1318 SC/05 Eximstats V1.47
1319 Added 'Message too big' to the list of mail rejection
1320 reasons (thanks to Marco Gaiarin).
1322 SC/06 Eximstats V1.48
1323 Mainlog lines which have GMT offsets and are too short to
1324 have a flag are now skipped.
1326 SC/07 Eximstats V1.49 (Alain Williams)
1327 Added the -emptyok flag.
1329 SC/08 Eximstats V1.50
1330 Fixes for obtaining the IP address from reject messages.
1332 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1333 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1334 whitesspace changes from 4.64-PH/27
1336 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1339 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1340 are found, allow negative numbers in numeric criteria)
1342 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1344 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1345 to match changes made in 4.64-PH/43
1347 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1349 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1350 "domain" is actually an IP address, and force a failure. This locks out
1351 those revolvers/nameservers that support "A-for-A" lookups, in
1352 contravention of the specifications.
1354 PH/48 When a host name was looked up from an IP address, and the subsequent
1355 forward lookup of the name timed out, the host name was left in
1356 $sender_host_name, contrary to the specification.
1358 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1359 restricted to single-key lookups, Exim was not diagnosing an error if
1360 * or *@ was used with a query-style lookup.
1362 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1364 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1365 long ago, but noone interested enough thought of it. Let's just say that
1366 the "1.1" means that there are some new functions that weren't there at
1367 some point in the past.
1369 PH/51 Error processing for expansion failure of helo_data from an smtp
1370 transport during callout processing was broken.
1372 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1373 tested/used via the -bh/-bhc/-bs options.
1375 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1376 bug, fixed in subsequent PCRE releases).
1378 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1379 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1381 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1387 SC/01 Use a glob alias rather than an array ref in eximstats generated
1388 parser. This improves both readability and performance.
1390 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1391 Don't display local sender or destination tables in eximstats unless
1392 there is data to show.
1393 Added average volumes into the eximstats top table text output.
1395 SC/03 Collect data on the number of addresses (recipients) as well
1396 as the number of messages in eximstats.
1398 TF/01 Correct an error in the documentation for the redirect router. Exim
1399 does (usually) call initgroups() when daemonizing.
1401 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1402 with consistent privilege compared to when running as a daemon.
1404 TF/03 Note in the spec that $authenticated_id is not set for local
1405 submissions from trusted users.
1407 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1408 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1410 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1411 by adding some example configuration directives to the default
1412 configuration file. A little bit of work is required to uncomment the
1413 directives and define how usernames and passwords are checked, but
1414 there is now a framework to start from.
1416 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1417 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1418 without this. I don't know how relevant this is to other LDAP libraries.
1420 PH/02 Add the verb name to the "unknown ACL verb" error.
1422 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1424 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1426 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1427 directory not expanded when it should be if an expanded home directory
1428 was set for the address (which is overridden by the transport).
1430 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1433 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1434 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1435 because it is too late at that time, and has no effect.
1437 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1438 security issue with \' (bugzilla #107). I could not use the
1439 PQescapeStringConn() function, because it needs a PGconn value as one of
1442 PH/08 When testing addresses using -bt, indicate those final addresses that
1443 are duplicates that would not cause an additional delivery. At least one
1444 person was confused, thinking that -bt output corresponded to deliveries.
1445 (Suppressing duplicates isn't a good idea as you lose the information
1446 about possibly different redirections that led to the duplicates.)
1448 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1449 systems where poll() doesn't work, in particular OS X.
1451 PH/10 Added more information to debugging output for retry time not reached.
1453 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1454 operations in malware.c.
1456 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1459 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1460 syslog_duplication set false, log lines that would normally be written
1461 both the the main log and to the reject log were not written to syslog at
1464 PH/14 In the default configuration, change the use of "message" in ACL warn
1465 statements to "add_header".
1467 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1468 not followed by a command (e.g. "seen endif").
1470 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1471 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1474 PH/17 Added extra conditions to the default value of delay_warning_condition
1478 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1479 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1480 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1483 The Auto-Submitted: and various List- headers are standardised, whereas I
1484 don't think Precedence: ever was.
1486 PH/18 Refactored debugging code in route_finduser() to show more information,
1487 in particular, the error code if getpwnam() issues one.
1489 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1490 This is apparently needed in addition to the PH/07 change above to avoid
1491 any possible encoding problems.
1493 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1494 but not after initializing Perl.
1496 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1497 output them only if debugging. By default they are written stderr,
1498 apparently, which is not desirable.
1500 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1503 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1506 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1508 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1509 authenticated or an ident call has been made. Suppress the default
1510 values for $authenticated_id and $authenticated_sender (but permit -oMai
1511 and -oMas) when testing with -bh.
1513 PH/24 Re-jigged the order of the tests in the default configuration so that the
1514 tests for valid domains and recipients precede the DNS black list and CSA
1515 tests, on the grounds that those ones are more expensive.
1517 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1518 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1519 This bug exists in every version of Exim that I still have, right back to
1522 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1523 However, an attempt to turn on case-sensitivity in a regex key by
1524 including (?-i) didn't work because the subject string was already
1525 lowercased, and the effects were non-intuitive. It turns out that a
1526 one-line patch can be used to allow (?-i) to work as expected.
1532 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1533 other effects) broke the use of negated acl sub-conditions.
1535 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1538 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1539 "Deny" causes Exim to reject the incoming connection with a 554 error.
1540 Unfortunately, if there is a major crisis, such as a disk failure,
1541 tcp-wrappers gives "deny", whereas what one would like would be some
1542 kind of temporary error. A kludge has been added to help with this.
1543 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1544 554 error is used if errno is still zero or contains ENOENT (which occurs
1545 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1548 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1550 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1551 errors. Otherwise a message that provokes a temporary error (when other
1552 messages do not) can cause a whole host to time out.
1554 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1555 addresses were routed directly to files or pipes from a redirect router.
1556 File deliveries just didn't batch; pipe deliveries might have suffered
1559 PH/06 A failure to get a lock for a hints database would erroneously always say
1560 "Failed to get write lock", even when it was really a read lock.
1562 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1563 of 0600. This has been changed to use the value of the lockfile_mode
1564 option (which defaults to 0600).
1566 PH/08 Applied small patch from the Sieve maintainer.
1568 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1569 folder from quota calculations, a direct delivery into this folder messed
1570 up the contents of the maildirsize file. This was because the regex was
1571 used only to exclude .Trash (or whatever) when the size of the mailbox
1572 was calculated. There was no check that a delivery was happening into an
1573 excluded directory. This bug has been fixed by ignoring all quota
1574 processing for deliveries into excluded directories.
1576 PH/10 Added the maildirfolder_create_regex option to appendfile.
1582 PH/01 The code for finding all the local interface addresses on a FreeBSD
1583 system running IPv6 was broken. This may well have applied to all BSD
1584 systems, as well as to others that have similar system calls. The broken
1585 code found IPv4 interfaces correctly, but gave incorrect values for the
1586 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1587 that it would not match correctly against @[] and not recognize the IPv6
1590 PH/02 The ipliteral router was not recognizing addresses of the form user@
1591 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1593 PH/03 Added disable_ipv6.
1595 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1596 decoded form, because it is most often used to construct To: headers
1597 lines in autoreplies, and the decoded form may well be syntactically
1598 invalid. However, $reply_address has leading white space removed, and all
1599 newlines turned into spaces so that the autoreply transport does not
1602 PH/05 If group was specified without a user on a router, and no group or user
1603 was specified on a transport, the group from the router was ignored.
1605 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1606 for visible compile-time settings that can be used to change these
1607 numbers, for those that want even more. Backwards compatibility with old
1608 spool files has been maintained. However, going back to a previous Exim
1609 release will lost any variables that are in spool files.
1611 PH/07 Two small changes when running in the test harness: increase delay when
1612 passing a TCP/IP connection to a new process, in case the original
1613 process has to generate a bounce, and remove special handling of
1614 127.0.0.2 (sic), which is no longer necessary.
1616 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1617 be the same on different OS.
1619 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1622 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1623 whether --show-vars was specified or not
1625 JJ/02 exipick: Added support for new ACL variable spool format introduced
1628 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1629 syntactically invalid From: or Reply-to: line, and a filter used this to
1630 generate an autoreply, and therefore failed to obtain an address for the
1631 autoreply, Exim could try to deliver to a non-existent relative file
1632 name, causing unrelated and misleading errors. What now happens is that
1633 it logs this as a hard delivery error, but does not attempt to create a
1636 PH/11 The exinext utility has a -C option for testing purposes, but although
1637 the given file was scanned by exinext itself; it wasn't being passed on
1638 when Exim was called.
1640 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1641 an end-of-file indication when reading a command response.
1643 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1644 compiled. In many other places in Exim, IPv6 addresses are always
1645 recognized, so I have changed this. It also means that IPv4 domain
1646 literals of the form [IPV4:n.n.n.n] are now always recognized.
1648 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1649 used if the router is not running as root, for example, when verifying at
1650 ACL time, or when using -bh. The debugging output from this situation was
1651 non-existent - all you got was a failure to exec. I have made two
1654 (a) Failures to set uid/gid, the current directory, or a process leader
1655 in a subprocess such as that created by queryprogram now generate
1656 suitable debugging ouput when -d is set.
1658 (b) The queryprogram router detects when it is not running as root,
1659 outputs suitable debugging information if -d is set, and then runs
1660 the subprocess without attempting to change uid/gid.
1662 PH/15 Minor change to Makefile for building test_host (undocumented testing
1665 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1666 additional section of a DNS packet that returns MX or SRV records.
1667 Instead, it always explicitly searches for A/AAAA records. This avoids
1668 major problems that occur when a DNS server includes only records of one
1669 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1670 fixed another bug: if SRV records were looked up and the corresponding
1671 address records were *not* found in the additional section, the port
1672 values from the SRV records were lost.
1674 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1675 using the correct key (the original address) when searching the retry
1676 rules in order to find which one to use for generating the retry hint.
1678 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1679 adding the default one. Similarly, if it contains a Reply-To: header, the
1680 errors_reply_to option, if set, is not used.
1682 PH/19 When calculating a retry time, Exim used to measure the "time since
1683 failure" by looking at the "first failed" field in the retry record. Now
1684 it does not use this if it is later than than the arrival time of the
1685 message. Instead it uses the arrival time. This makes for better
1686 behaviour in cases where some deliveries succeed, thus re-setting the
1687 "first failed" field. An example is a quota failure for a huge message
1688 when small messages continue to be delivered. Without this change, the
1689 "time since failure" will always be short, possible causing more frequent
1690 delivery attempts for the huge message than are intended.
1691 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1693 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1694 $1, $2, $3) because the numerical variables can be reset during some
1695 expansion items (e.g. "match"), thereby losing the authentication data.
1697 PH/21 Make -bV show the size of off_t variables so that the test suite can
1698 decide whether to run tests for quotas > 2G.
1700 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1701 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1702 filecount value is greater than 2G or if a quota value is greater than 2G
1703 on a system where the size of off_t is not greater than 4, a panic error
1706 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1707 never match. The debug and -bh output now contains an explicit error
1708 message indicating a malformed IPv4 address or mask.
1710 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1711 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1712 PH/23 above applies.
1714 PH/25 Do not write to syslog when running in the test harness. The only
1715 occasion when this arises is a failure to open the main or panic logs
1716 (for which there is an explicit test).
1718 PH/26 Added the /no_tell option to "control=freeze".
1720 PH/27 If a host name lookup failed very early in a connection, for example, if
1721 the IP address matched host_lookup and the reverse lookup yielded a name
1722 that did not have a forward lookup, an error message of the form "no IP
1723 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1724 could be logged. Now it outputs the IP address instead of "NULL".
1726 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1727 allows the sender and the authenticated sender to be set when
1728 submitting a message from within Exim. Since child_open_exim() is
1729 documented for local_scan(), the new function should be too.
1731 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1732 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1733 results in an empty string is now treated as unset.
1735 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1737 PH/31 Added sender_verify_fail logging option.
1739 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1740 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1741 tidied the source and removed it altogether.
1743 PH/33 When a queue run was abandoned because the load average was too high, a
1744 log line was always written; now it is written only if the queue_run log
1745 selector is set. In addition, the log line for abandonment now contains
1746 information about the queue run such as the pid. This is always present
1747 in "start" and "stop" lines but was omitted from the "abandon" line.
1749 PH/34 Omit spaces between a header name and the colon in the error message that
1750 is given when verify = headers_syntax fails (if there are lots of them,
1751 the message gets confusing).
1753 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1754 names, as there are now some PTR records that contain slashes. This check
1755 is only to protect against broken name servers that fall over on strange
1756 characters, so the fact that it applies to all lookups doesn't matter.
1758 PH/36 Now that the new test suite is complete, we can remove some of the
1759 special code in Exim that was needed for the old test suite. For example,
1760 sorting DNS records because real resolvers return them in an arbitrary
1761 order. The new test suite's fake resolver always returns records in the
1764 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1765 bounces) except when queue_only is set, to avoid logging races between
1766 the different processes.
1768 PH/38 Panic-die if .include specifies a non-absolute path.
1770 PH/39 A tweak to the "H" retry rule from its user.
1772 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1773 a label. They prevented compilation on older perls.
1775 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1776 a warning to be raised on newish perls.
1778 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1779 on queue. Changes to match documented behaviour of showing count of
1780 messages matching specified criteria.
1782 PH/40 Changed the default ident timeout from 30s to 5s.
1784 PH/41 Added support for the use of login_cap features, on those BSD systems
1785 that have them, for controlling the resources used by pipe deliveries.
1787 PH/42 The content-scanning code uses fopen() to create files in which to put
1788 message data. Previously it was not paying any attention to the mode of
1789 the files. Exim runs with umask(0) because the rest of the code creates
1790 files with open(), and sets the required mode explicitly. Thus, these
1791 files were ending up world-writeable. This was not a big issue, because,
1792 being within the spool directory, they were not world-accessible. I have
1793 created a function called modefopen, which takes an additional mode
1794 argument. It sets umask(777), creates the file, chmods it to the required
1795 mode, then resets the umask. All the relevant calls to fopen() in the
1796 content scanning code have been changed to use this function.
1798 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1799 to 24 hours. This avoids potential overflow problems when processing G
1800 and H retry rules. I suspect nobody ever tinkers with this value.
1802 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1804 PH/45 When the plaintext authenticator is running as a client, the server's
1805 challenges are checked to ensure they are valid base64 strings. By
1806 default, the authentication attempt is cancelled if an invalid string is
1807 received. Setting client_ignore_invalid_base64 true ignores these errors.
1808 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1809 they are received. Thus, the responses can be made to depend on the
1810 challenges. If an invalid string is ignored, an empty string is placed in
1813 PH/46 Messages that are created by the autoreply transport now contains a
1814 References: header, in accordance with RFCs 2822 and 3834.
1816 PH/47 Added authenticated_sender_force to the smtp transport.
1818 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1820 PH/49 Installed latest patch from the Sieve maintainer.
1822 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1823 was also set, the appendfile transport was unnecessarily scanning a
1824 directory of message files (e.g. for maildir delivery) to find the count
1825 of files (along with the size), even though it did not need this
1826 information. It now does the scan only if it needs to find either the
1827 size of the count of files.
1829 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1831 PH/52 Two bugs concerned with error handling when the smtp transport is
1834 (i) Exim was not creating retry information for temporary errors given
1835 for individual recipients after the DATA command when the smtp transport
1836 was used in LMTP mode. This meant that they could be retried too
1837 frequently, and not timed out correctly.
1839 (ii) Exim was setting the flag that allows error details to be returned
1840 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1841 recipients that were returned after the DATA command.
1843 PH/53 This is related to PH/52, but is more general: for any failing address,
1844 when detailed error information was permitted to be returned to the
1845 sender, but the error was temporary, then after the final timeout, only
1846 "retry timeout exceeded" was returned. Now it returns the full error as
1847 well as "retry timeout exceeded".
1849 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1850 do this, and (what is worse) MTAs that accept it.
1852 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1853 will now be deprecated.
1855 PH/56 New os.c-cygwin from the Cygwin maintainer.
1857 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1858 formats (previously only available in exim formats via -bpr, -bpru,
1859 and -bpra. Now also available in native and exiqgrep formats)
1861 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1862 with very large, slow to parse queues
1864 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1866 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1868 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1869 responses to authentication challenges, though it was showing the
1870 challenges; (ii) I've removed the CR characters from the debug output for
1873 PH/58 Allow for the insertion of a newline as well as a space when a string
1874 is turned into more than one encoded-word during RFC 2047 encoding. The
1875 Sieve code now uses this.
1877 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1878 data_4xx, lost_connection, tls_required.
1880 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1881 message was being sent as an SMTP response.
1883 PH/61 Add -l and -k options to exicyclog.
1885 PH/62 When verifying, if an address was redirected to one new address, so that
1886 verification continued, and the new address failed or deferred after
1887 having set something in $address_data, the value of $address_data was not
1888 passed back to the ACL. This was different to the case when no
1889 redirection occurred. The value is now passed back in both cases.
1891 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1892 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1893 use login_cap.h, so on its own it isn't the distinguishing feature. The
1894 new name refers directly to the setclassresources() function.
1896 PH/65 Added configuration files for NetBSD3.
1898 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1900 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1901 is preferred over IPv4.
1903 PH/68 The bounce_return_message and bounce_return_body options were not being
1904 honoured for bounces generated during the reception of non-SMTP messages.
1905 In particular, this applied to messages rejected by the ACL. This bug has
1906 been fixed. However, if bounce_return_message is true and bounce_return_
1907 body is false, the headers that are returned for a non-SMTP message
1908 include only those that have been read before the error was detected.
1909 (In the case of an ACL rejection, they have all been read.)
1911 PH/69 The HTML version of the specification is now built in a directory called
1912 spec_html instead of spec.html, because the latter looks like a path with
1913 a MIME-type, and this confuses some software.
1915 PH/70 Catch two compiler warnings in sieve.c.
1917 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1918 function verify_get_ident() calls ip_connect() to connect a socket, but
1919 if the "connect()" function timed out, ip_connect() used to close the
1920 socket. However, verify_get_ident() also closes the socket later, and in
1921 between Exim writes to the log, which may get opened at this point. When
1922 the socket was closed in ip_connect(), the log could get the same file
1923 descriptor number as the socket. This naturally causes chaos. The fix is
1924 not to close the socket in ip_connect(); the socket should be closed by
1925 the function that creates it. There was only one place in the code where
1926 this was missing, in the iplookup router, which I don't think anybody now
1927 uses, but I've fixed it anyway.
1929 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1930 well as to direct DNS lookups. Otherwise the handling of names in host
1931 lists is inconsistent and therefore confusing.
1937 PH/01 Two changes to the default runtime configuration:
1939 (1) Move the checks for relay_from_hosts and authenticated clients from
1940 after to before the (commented out) DNS black list checks.
1942 (2) Add control=submission to the relay_from_hosts and authenticated
1943 clients checks, on the grounds that messages accepted by these
1944 statements are most likely to be submissions.
1946 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1948 (1) Generate an error if the third argument for the ${prvs expansion is
1951 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1954 (3) Reset the variables that are obtained from the first argument of
1955 ${prvscheck and used in the second argument before leaving the code,
1956 because their memory is reclaimed, so using them afterwards may do
1959 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1960 one (it's much easier than Tom thought :-).
1962 (5) Because of (4), we can now allow for the use of $prvscheck_result
1963 inside the third argument.
1965 PH/03 For some reason, the default setting of PATH when running a command from
1966 a pipe transport was just "/usr/bin". I have changed it to
1969 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1970 anything to be listed in the output from -bV.
1972 PH/05 When a filter generated an autoreply, the entire To: header line was
1973 quoted in the delivery log line, like this:
1975 => >A.N.Other <ano@some.domain> <original@ddress> ...
1977 This has been changed so that it extracts the operative address. There
1978 may be more than one such address. If so, they are comma-separated, like
1981 => >ano@some.domain,ona@other.domain <original@ddress> ...
1983 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1984 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1985 not being looked up in the rDNS to get a host name, Exim was showing the
1986 IP address twice in Received: lines, even though the IP addresses were
1987 identical. For example:
1989 Received: from [1.2.3.4] (helo=[1.2.3.4])
1991 However, if the real host name was known, it was omitting the HELO data
1992 if it matched the actual IP address. This has been tidied up so that it
1993 doesn't show the same IP address twice.
1995 PH/07 When both +timestamp and +memory debugging was on, the value given by
1996 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1997 called by the debug printing, thereby overwriting the timestamp buffer.
1998 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2000 PH/08 When the original message was included in an autoreply transport, it
2001 always said "this is a copy of the message, including all the headers",
2002 even if body_only or headers_only was set. It now gives an appropriate
2005 PH/09 Applied a patch from the Sieve maintainer which:
2007 o fixes some comments
2008 o adds the (disabled) notify extension core
2009 o adds some debug output for the result of if/elsif tests
2010 o points to the current vacation draft in the documentation
2011 and documents the missing references header update
2015 o fixes a bug in processing the envelope test (when testing
2016 multiple envelope elements, the last element determinted the
2019 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2020 Electronic Mail") by including:
2022 Auto-submitted: auto-generated
2024 in the messages that it generates (bounce messages and others, such as
2025 warnings). In the case of bounce messages for non-SMTP mesages, there was
2026 also a typo: it was using "Auto_submitted" (underscore instead of
2027 hyphen). Since every message generated by Exim is necessarily in response
2028 to another message, thes have all been changed to:
2030 Auto-Submitted: auto-replied
2032 in accordance with these statements in the RFC:
2034 The auto-replied keyword:
2036 - SHOULD be used on messages sent in direct response to another
2037 message by an automatic process,
2039 - MUST NOT be used on manually-generated messages,
2041 - MAY be used on Delivery Status Notifications (DSNs) and Message
2042 Disposition Notifications (MDNs),
2044 - MUST NOT be used on messages generated by automatic or periodic
2045 processes, except for messages which are automatic responses to
2048 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2049 to the default Received: header definition.
2051 PH/12 Added log selector acl_warn_skipped (default on).
2053 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2054 variables because (a) they are in the wrong storage pool and (b) even if
2055 they were copied, it wouldn't work properly because of the caching.
2057 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2058 checking when decoding. Apparently there are clients that generate
2059 overlong encoded strings. Why am I not surprised?
2061 PH/15 If the first argument of "${if match_address" was not empty, but did not
2062 contain an "@" character, Exim crashed. Now it writes a panic log message
2063 and treats the condition as false.
2065 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2067 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2068 extension "envelope-auth". The code is finished and in agreement with
2069 other implementations, but there is no documentation so far and in fact,
2070 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2071 not changing the active code.
2073 Print executed "if" and "elsif" statements when debugging is used. This
2074 helps a great deal to understand what a filter does.
2076 Document more things not specified clearly in RFC3028. I had all this
2077 sorted out, when out of a sudden new issues came to my mind. Oops."
2079 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2082 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2083 canonical form (as documented). However, after a host name lookup from
2084 the IP address, check_host() was doing a simple string comparison with
2085 addresses acquired from the DNS when checking that the found name did
2086 have the original IP as one of its addresses. Since any found IPv6
2087 addresses are likely to be in abbreviated form, the comparison could
2088 fail. Luckily, there already exists a function for doing the comparison
2089 by converting both addresses to binary, so now that is used instead of
2090 the text comparison.
2092 PH/20 There was another similar case to PH/19, when a complete host name was
2093 given in a host list; looking up its IP address could give an abbreviated
2094 form, whereas the current host's name might or might not be abbreviated.
2095 The same fix has been applied.
2101 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2102 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2105 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2106 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2108 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2110 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2111 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2112 Also, for "Auto-Submitted", treat anything other than "no" as
2113 non-personal, in accordance with RFC 3834. (Previously it treated
2114 anything starting "auto-" as non-personal.)
2116 TF/01 The control=submission/name=... option had a problem with syntax
2117 errors if the name included a slash character. The /name= option
2118 now slurps the rest of the string, so it can include any characters
2119 but it must come last in the list of options (after /sender_retain
2122 PH/05 Some modifications to the interface to the fake nameserver for the new
2130 TK/01 Added the "success_on_redirect" address verification option. See
2131 NewStuff for rationale and an example.
2133 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2135 PH/02 Patch to exigrep to allow it to work on syslog lines.
2137 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2138 fread() to skip over the body file's header line, because in Cygwin the
2139 header line is locked and is inaccessible.
2141 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2142 co-exist for some time) to make it clear that it is the Exim ID that is
2143 referenced, not the Message-ID: header line.
2145 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2146 string_format() function, because snprintf() does not exist on all
2149 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2150 ${stat: expansion item.
2152 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2153 protocol synchronization error", to keep the pedants happy.
2155 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2156 well as for IRIX systems, when gcc is being used. See the host.c source
2159 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2161 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2164 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2165 transport and to the smtp transport in LMTP mode.
2167 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2169 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2170 run a filter in a subprocess. This could lead to confusion in subsequent
2171 lookups in the parent process. There should also be a search_tidyup() at
2172 the end of the subprocess.
2174 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2175 only if the host matched helo_try_verify_hosts, which caused the
2176 verification to occur when the EHLO/HELO command was issued. The ACL just
2177 tested the remembered result. Now, if a previous verification attempt has
2178 not happened, "verify = helo" does it there and then.
2180 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2182 TK/03 Fix log output including CR from clamd.
2184 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2185 exist provoked a memory error which could cause a segfault.
2187 PH/15 Installed PCRE 6.2
2189 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2191 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2192 of the problem. Specifically, suggested +O2 rather than +O1 for the
2195 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2197 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2198 record, the port was not correctly logged when the outgoing_port log
2199 selector was set (it logged the transort's default port).
2201 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2202 fallback_hosts, and "hosts" in the smtp transport.
2204 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2205 host errors such as "Connection refused".
2207 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2208 authentication with radiusclient 0.4.9:
2210 - Error returned from rc_read_config was caught wrongly
2211 - Username/password not passed on to radius server due to wrong length.
2213 The presumption is that some radiusclient API changes for 4.51/PH/17
2214 were not taken care of correctly. The code is still untested by me (my
2215 Linux distribution still has 0.3.2 of radiusclient), but it was
2216 contributed by a Radius user.
2218 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2219 expanding the "port" option of the smtp transport.
2221 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2222 while reading a MIME header. Thanks to Tom Hughes for a patch.
2224 PH/24 Include config.h inside local_scan.h so that configuration settings are
2227 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2228 that in an ACL for RCPT (for example), you can examine exactly what was
2231 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2232 commands, but it was not correctly comparing the address with the actual
2233 client host address. Thus, it would show the EHLO address in Received:
2234 header lines when this was not necessary.
2236 PH/27 Added the % operator to ${eval:}.
2238 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2239 it should be ignoring failures (because with -C, for example, it has lost
2240 privilege). It wasn't ignoring creation failures other than "already
2243 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2246 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2247 because an input line was too long, either on its own, or by virtue of
2248 too many continuations, the temporary file was not being removed, and the
2249 return code was incorrect.
2251 PH/31 Missing "BOOL" in function definition in filtertest.c.
2253 PH/32 Applied Sieve patches from the maintainer.
2255 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2257 PH/33 Added "verify = not_blind".
2259 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2260 Local/Makefile (with some defaults set). These are used in built scripts
2261 such as exicyclog, but they have never been used in the exim_install
2262 script (though there are many overriding facilities there). I have
2263 arranged that the exim_install script now takes note of these two
2266 PH/35 Installed configuration files for Dragonfly.
2268 PH/36 When a locally submitted message by a trusted user did not contain a
2269 From: header, and the sender address was obtained from -f or from an SMTP
2270 MAIL command, and the trusted user did not use -F to supply a sender
2271 name, $originator_name was incorrectly used when constructing a From:
2272 header. Furthermore, $originator_name was used for submission mode
2273 messages from external hosts without From: headers in a similar way,
2274 which is clearly wrong.
2276 PH/37 Added control=suppress_local_fixups.
2278 PH/38 When log_selector = +received_sender was set, and the addition of the
2279 sender made the log line's construction buffer exactly full, or one byte
2280 less than full, an overflow happened when the terminating "\n" was
2283 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2284 when the result of a list match is failure because a DNS lookup failed.
2286 PH/40 RM_COMMAND is now used in the building process.
2288 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2289 the "build-* directories that it finds.
2291 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2292 address in a domain literal was a prefix of an interface address.
2294 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2295 when verifying a sender address, unless rewrite_headers is false.
2297 PH/44 Wrote a long comment about why errors_to addresses are verified as
2298 recipients, not senders.
2300 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2301 the ratelimit ACL was added.
2303 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2305 PH/46 Added extra information about PostgreSQL errors to the error string.
2307 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2308 avoiding the need to install special zones in a real server. This is
2309 backwards compatible; if it can't find the fake resolver, it drops back.
2310 Thus, both old and new test suites can be run.
2312 TF/02 Added util/ratelimit.pl
2314 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2315 clock is set back in time.
2317 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2318 Candler <B.Candler@pobox.com>.
2320 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2321 OK for addresses that are the result of redirections.
2323 PH/48 A number of further additions for the benefit of the new test suite,
2324 including a fake gethostbyname() that interfaces to the fake DNS resolver
2327 TF/06 The fix for widen_domains has also been applied to qualify_single and
2328 search_parents which are the other dnslookup options that can cause
2331 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2334 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2336 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2337 service" if it encountered an MX record with an empty target hostname.
2338 The message is now "an MX or SRV record indicated no SMTP service".
2340 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2341 if the DNS of the sending site is misconfigured. This is quite a
2342 common situation. This change restores the behaviour of treating a
2343 helo verification defer as a failure.
2345 PH/51 If self=fail was set on a router, the bounce message did not include the
2346 actual error message.
2352 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2354 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2355 command itself ended in error, the underlying message about the transport
2356 filter timeout was being overwritten with the pipe command error. Now the
2357 underlying error message should be appended to the second error message.
2359 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2361 PH/02 Reduce the amount of output that "make" produces by default. Full output
2362 can still be requested.
2364 PH/03 The warning log line about a condition test deferring for a "warn" verb
2365 was being output only once per connection, rather than after each
2366 occurrence (because it was using the same function as for successful
2367 "warn" verbs). This seems wrong, so I have changed it.
2369 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2370 it should not have, which might have caused a crash in the right
2371 circumstances, but probably never did.
2373 PH/04 Installed a modified version of Tony Finch's patch to make submission
2374 mode fix the return path as well as the Sender: header line, and to
2375 add a /name= option so that you can make the user's friendly name appear
2378 TF/03 Added the control = fakedefer ACL modifier.
2380 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2381 Mark Lowes for thorough testing.
2383 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2385 TK/03 Merged latest SRS patch from Miles Wilton.
2387 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2388 with the definition in sysexits.h (which is #included earlier).
2389 Fortunately, Exim does not actually use EX_OK. The code used to try to
2390 preserve the sysexits.h value, by assumimg that macro definitions were
2391 scanned for macro replacements. I have been disabused of this notion,
2392 so now the code just undefines EX_OK before #including unistd.h.
2394 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2395 in the smtp transport. When a block could not be written in a single
2396 write() function, the timeout was being re-applied to each part-write.
2397 This seems wrong - if the receiver was accepting one byte at a time it
2398 would take for ever. The timeout is now adjusted when this happens. It
2399 doesn't have to be particularly precise.
2401 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2402 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2404 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2405 without a domain if the check to <postmaster@domain> fails.
2407 SC/01 Eximstats: added -xls and the ability to specify output files
2408 (patch written by Frank Heydlauf).
2410 SC/02 Eximstats: use FileHandles for outputing results.
2412 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2414 SC/04 Eximstats: fixed display of large numbers with -nvr option
2416 SC/05 Eximstats: fixed merging of reports with empty tables.
2418 SC/06 Eximstats: added the -include_original_destination flag
2420 SC/07 Eximstats: removed tabs and trailing whitespace.
2422 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2424 TK/06 MBOX spool code: Add real "From " MBOX separator line
2425 so the .eml file is really in mbox format (even though
2426 most programs do not really care). Patch from Alex Miller.
2428 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2429 The latter is generated from $received_to and is only set if the
2430 message has one envelope recipient. SA can use these headers,
2431 obviously out-of-the-box. Patch from Alex Miller.
2433 PH/08 The ${def test on a variable was returning false if the variable's
2434 value was "0", contrary to what the specification has always said!
2435 The result should be true unless the variable is empty.
2437 PH/09 The syntax error of a character other than { following "${if
2438 def:variable_name" (after optional whitespace) was not being diagnosed.
2439 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2440 accidental colon was present, for example, could give incorrect results.
2442 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2443 result is used (not including appendfile, where other changes are about
2446 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2447 This involved changing a lot of size variables from int to off_t. It
2448 should work with maildirs and everything.
2450 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2451 spamd dying while we are connected to it.
2453 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2456 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2457 function for BDB 4.3.
2459 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2461 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2462 the sources for PCRE have been re-arranged and more files are now
2465 PH/15 The code I had for printing potentially long long variables in PH/11
2466 above was not the best (it lost precision). The length of off_t variables
2467 is now inspected at build time, and an appropriate printing format (%ld
2468 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2469 to be "long long int" or "long int". This is needed for the internal
2470 formatting function string_vformat().
2472 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2473 the configuration file to be ":syslog", then the script "guesses" where
2474 the logs files are, rather than using the compiled in default. In our
2475 case the guess is not the same as the compiled default, so the script
2476 suddenly stopped working when I started to use syslog. The patch checks
2477 to see if log_file_path is "". If so, it attempts to read it from exim
2478 with no configuration file to get the compiled in version, before it
2479 falls back to the previous guessing code."
2481 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2482 implementing BATV in an Exim configuration. See NewStuff for the gory
2485 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2486 Makefile that are specific to HP-UX.
2488 PH/18 If the "use_postmaster" option was set for a recipient callout together
2489 with the "random" option, the postmaster address was used as the MAIL
2490 FROM address for the random test, but not for the subsequent recipient
2491 test. It is now used for both.
2493 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2494 patch removes a few documentation additions to RFC 3028, because the
2495 latest draft now contains them. It adds the new en;ascii-case comparator
2496 and a new error check for 8bit text in MIME parts. Comparator and
2497 require names are now matched exactly. I enabled the subaddress
2498 extension, but it is not well tested yet (read: it works for me)."
2500 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2501 rework some of the code of TK/09 above to avoid the hardwired use of
2502 "%lld" and "long long". Replaced the call to snprintf() with a call to
2505 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2506 records point to non-existent hosts", "retry timeout exceeded", and
2507 "retry time not reached for any host after a long failure period".
2509 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2510 experimental DomainKeys support:
2512 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2513 (2) On an error such as an illegally used "control", the wrong name for
2514 the control was given.
2516 These problems did NOT occur unless DomainKeys support was compiled.
2518 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2520 PH/24 Added ${if match_ip condition.
2522 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2523 databases so that it will be absolutely obvious if a crash occurs in the
2524 DB library. This is a regular occurrence (often caused by mis-matched
2527 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2528 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2529 picked up on a user's system that detects such things. There doesn't seem
2530 to be a gcc warning option for this - only an attribute that has to be
2531 put on the function's prototype. It seems that in Fedora Core 4 they have
2532 set this on a number of new functions. No doubt there will be more in due
2535 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2536 specify a transport. However, if an address that was verified by such a
2537 router was the subject of a callout, Exim crashed because it tried to
2538 read the rcpt_include_affixes from the non-existent transport. Now it
2539 just assumes that the setting of that option is false. This bug was
2540 introduced by 4.51/PH/31.
2542 PH/28 Changed -d+all to exclude +memory, because that information is very
2543 rarely of interest, but it makes the output a lot bigger. People tend to
2544 do -d+all out of habit.
2546 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2547 code in os-type was giving problems when libc.so lives in lib64, like on
2550 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2551 aren't the modern standard, and it seems that some systems' include files
2552 don't always have them. Exim was already checking for some of the newer
2553 ones like T_AAAA, and defining it itself. I've added checks for all the
2554 record types that Exim uses.
2556 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2557 not automatically generating a new one, as it is supposed to. This
2558 prevented TLS from working. If the file did exist, but contained invalid
2559 data, a new version was generated, as expected. It was only the case of a
2560 non-existent file that was broken.
2562 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2563 with a change in libdomainkeys > 0.64.
2565 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2566 from DNS. If the selector record carries the flag, it now has
2567 precedence over the domain-wide flag.
2569 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2571 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2572 the use of a port name that isn't defined in /etc/services) occurred, the
2573 message was deferred as in a normal delivery, and thus remained on the
2574 spool, instead of being failed because of the mua_wrapper setting. This
2575 is now fixed, and I tidied up some of the mua_wrapper messages at the
2578 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2579 the messages in a hash of arrays rather than using individual hashes.
2580 This is a bit cleaner and results in dramatic memory savings, albeit
2581 at a slight CPU cost.
2583 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2584 as requested by Marc Sherman.
2586 SC/10 Eximstats: added histograms for user specified patterns as requested
2589 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2591 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2592 fopen() in the content-scanning modules that did not already have it.
2598 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2599 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2601 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2603 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2605 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2606 to test Sieve filters that use "vacation".
2608 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2609 that changes the way the GnuTLS parameters are stored in the cache file.
2610 The new format can be generated externally. For backward compatibility,
2611 if the data in the cache doesn't make sense, Exim assumes it has read an
2612 old-format file, and it generates new data and writes a new file. This
2613 means that you can't go back to an older release without removing the
2616 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2617 work if there are any delivery delays because "one_time" forces the
2618 parent to be marked "delivered", so its unseen clone is never tried
2619 again. For this reason, Exim now forbids the simultaneous setting of
2622 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2623 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2624 entry if you want to know the details. The fix, however, neglected to
2625 consider the case when local delivery batching is involved. The test for
2626 "previously delivered" was not happening when checking to see if an
2627 address could be batched with a previous (undelivered) one; under
2628 certain circumstances this could lead to multiple deliveries to the same
2631 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2632 in its include files, and this causes problems building Exim.
2634 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2635 header_syntax) but Exim was just ignoring anything given after a slash.
2636 In particular, this caused confusion with an attempt to use "verify =
2637 reverse_host_lookup/defer_ok". An error is now given when options are
2638 supplied for verify items that do not have them. (Maybe reverse_host_
2639 lookup should have a defer_ok option, but that's a different point.)
2641 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2642 defined by RFC 821) to 2048, because there were problems with some AUTH
2643 commands, and RFC 1869 says the size should be increased for extended
2644 SMTP commands that take arguments.
2646 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2649 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2650 "unknown" error; now it says that the functionality isn't in the binary.
2652 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2653 an address' error message when a string expansion fails (syntax or
2654 whatever). Otherwise the password may appear in the log. Following change
2655 PH/42 below, there is no longer a chance of it appearing in a bounce
2658 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2660 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2661 of its addresses was ever tried. (Bugzilla bug #2.)
2663 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2664 the result incorrectly in the debug output. (It correctly added a newline
2665 to what was transported.)
2667 TF/01 Added $received_time.
2669 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2670 commented out examples of how to interface to a virus scanner and to
2671 SpamAssassin. Also added commented examples of av_scanner and
2672 spamd_address settings.
2674 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2675 and controls are allowed in which ACLs. There were a couple of minor
2676 errors. Some of the entries in the conditions table (which is a table of
2677 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2678 negation of where the condition IS allowed.
2680 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2682 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2683 header file does not have a version number, so I've had to invent a new
2684 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2685 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2686 radiusclient), but it was contributed by a Radius user.
2688 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2689 files or directories, for OpenSSL.
2691 PH/19 When an Exim process that is running as root has to create an Exim log
2692 file, it does so in a subprocess that runs as exim:exim so as to get the
2693 ownership right at creation (otherwise, other Exim processes might see
2694 the file with the wrong ownership). There was no test for failure of this
2695 fork() call, which would lead to the process getting stuck as it waited
2696 for a non-existent subprocess. Forks do occasionally fail when resources
2697 run out. I reviewed all the other calls to fork(); they all seem to check
2700 PH/20 When checking for unexpected SMTP input at connect time (before writing
2701 the banner), Exim was not dealing correctly with a non-positive return
2702 from the read() function. If the client had disconnected by this time,
2703 the result was a log entry for a synchronization error with an empty
2704 string after "input=" when read() returned zero. If read() returned -1
2705 (an event I could not check), uninitialized data bytes were printed.
2706 There were reports of junk text (parts of files, etc) appearing after
2709 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2711 PH/22 Added support for macro redefinition, and (re)definition in between
2712 driver and ACL definitions.
2714 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2715 forgetting to use the resulting value; it was using the unexpanded value.
2717 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2718 hadn't been configured. The fix is from Juergen Kreileder, who
2719 understands it better than I do:
2721 "Here's what I see happening with three configured cyrus_sasl
2722 authenticators configured (plain, login, cram-md5):
2724 On startup auth_cyrus_sasl_init() gets called for each of these.
2725 This means three calls to sasl_listmech() without a specified mech_list.
2726 => SASL tests which mechs of all available mechs actually work
2727 => three warnings about OTP not working
2728 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2730 With the patch, sasl_listmech() also gets called three times. But now
2731 SASL's mech_list option is set to the server_mech specified in the the
2732 authenticator. Or in other words, the answer from sasl_listmech()
2733 gets limited to just the mech you're testing for (which is different
2735 => the return list contains just 'plain' or 'login', 'cram-md5' or
2736 nothing depending on the value of ob->server_mech.
2738 I've just tested the patch: Authentication still works fine,
2739 unavailable mechs specified in the exim configuration are still
2740 caught, and the auth.log warnings about OTP are gone."
2742 PH/25 When debugging is enabled, the contents of the command line are added
2743 to the debugging output, even when log_selector=+arguments is not
2746 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2747 answer is "GNU", and only if the return is "GNU/something" is the answer
2750 PH/27 $acl_verify_message is now set immediately after the failure of a
2751 verification in an ACL, and so is available in subsequent modifiers. In
2752 particular, the message can be preserved by coding like this:
2754 warn !verify = sender
2755 set acl_m0 = $acl_verify_message
2757 Previously, $acl_verify_message was set only while expanding "message"
2758 and "log_message" when a very denied access.
2760 PH/28 Modified OS/os.c-Linux with
2762 -#ifndef OS_LOAD_AVERAGE
2763 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2765 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2766 nomenclature these days.)
2768 PH/29 Installed patch from the Sieve maintainer that adds the options
2769 sieve_useraddress and sieve_subaddress to the redirect router.
2771 PH/30 In these circumstances:
2772 . Two addresses routed to the same list of hosts;
2773 . First host does not offer TLS;
2774 . First host accepts first address;
2775 . First host gives temporary error to second address;
2776 . Second host offers TLS and a TLS session is established;
2777 . Second host accepts second address.
2778 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2779 and peerdn, if requested) that were in fact used only for the second
2782 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2783 attention to any local part prefix or suffix that was matched by the
2784 router that accepted the address. It now behaves in the same way as it
2785 does for delivery: the affixes are removed from the local part unless
2786 rcpt_include_affixes is set on the transport.
2788 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2789 timeout during the DATA phase of an incoming message.
2791 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2792 applied a patch sanctioned by the Sieve maintainer.
2794 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2795 the uid or gid is negative. A case of a negative gid caused this to be
2796 noticed. The fix allows for either to be negative.
2798 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2799 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2802 PH/36 The change PH/12 above was broken. Fixed it.
2804 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2805 the grounds that routing the same address twice would always produce the
2806 same answer. This might have been true once, but it is certainly no
2807 longer true now. Routing a child address may depend on the previous
2808 routing that produced that child. Some complicated redirection strategies
2809 went wrong when messages had multiple recipients, and made Exim's
2810 behaviour dependent on the order in which the addresses were given.
2812 I have moved the duplicate checking until after the routing is complete.
2813 Exim scans the addresses that are assigned to local and remote
2814 transports, and removes any duplicates. This means that more work will be
2815 done, as duplicates will always all be routed, but duplicates are
2816 presumably rare, so I don't expect this is of any significance.
2818 For deliveries to pipes, files, and autoreplies, the duplicate checking
2819 still happens during the routing process, since they are not going to be
2822 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2823 It corrects a timeout issue with spamd. This is Ian's comment: "The
2824 background is that sometimes spamd either never reads data from a
2825 connection it has accepted, or it never writes response data. The exiscan
2826 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2827 blindly assumes that writes won't block so it may never time out."
2829 PH/39 Allow G after quota size as well as K and M.
2831 PH/40 The value set for $authenticated_id in an authenticator may not contain
2832 binary zeroes or newlines because the value is written to log lines and
2833 to spool files. There was no check on this. Now the value is run through
2834 the string_printing() function so that such characters are converted to
2835 printable escape sequences.
2837 PH/41 $message_linecount is a new variable that contains the total number of
2838 lines in the message. Compare $body_linecount, which is the count for the
2841 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2842 bounce and delay warning messages, except in certain special cases, which
2845 (a) An SMTP error message from a remote host;
2846 (b) A message specified in a :fail: redirection;
2847 (c) A message specified in a "fail" command in a system filter;
2848 (d) A message specified in a FAIL return from the queryprogram router;
2849 (e) A message specified by the cannot_route_message router option.
2851 In these cases only, Exim does include the error details in bounce and
2852 warning messages. There are also a few cases where bland messages such
2853 as "unrouteable address" or "local delivery error" are given.
2855 PH/43 $value is now also set for the "else" part of a ${run expansion.
2857 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2858 being worked on, but at least Exim now implements the latest version to
2861 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2862 process to complete was treated as a delivery failure, a timeout while
2863 writing the message to the pipe was logged, but erroneously treated as a
2864 successful delivery. Such timeouts include transport filter timeouts. For
2865 consistency with the overall process timeout, these timeouts are now
2866 treated as errors, giving rise to delivery failures by default. However,
2867 there is now a new Boolean option for the pipe transport called
2868 timeout_defer, which, if set TRUE, converts the failures into defers for
2869 both kinds of timeout. A transport filter timeout is now identified in
2872 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2873 systems where "make" and "gmake" are different, calling "gmake" at top
2874 level broke things. I've arranged for the value of $(MAKE) to be passed
2875 from the Makefile to this script so that it can call the same version of
2879 A note about Exim versions 4.44 and 4.50
2880 ----------------------------------------
2882 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2883 changes of various kinds. As a consequence, a big documentation update was
2884 needed. This delayed the release for rather longer than seemed good, especially
2885 in the light of a couple of (minor) security issues. Therefore, the changes
2886 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2887 release. So 4.44 and 4.50 are in effect two different branches that both start
2890 I have left the 4.50 change log unchanged; it contains all the changes since
2891 4.43. The change log for 4.44 is below; many of its items are identical to
2892 those for 4.50. This seems to be the most sensible way to preserve the
2893 historical information.
2899 1. Minor wording change to the doc/README.SIEVE file.
2901 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2902 computation of the current number of files was incorrect.
2904 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2905 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2908 4. Give more explanation in the error message when the command for a transport
2909 filter fails to execute.
2911 5. There are several places where Exim runs a non-Exim command in a
2912 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2913 was being done only for the command run by the queryprogram router. It is
2914 now done for all such subprocesses. The other cases are: ${run, transport
2915 filters, and the commands run by the lmtp and pipe transports.
2917 6. Added CONFIGURE_GROUP build-time option.
2919 7. Some older OS have a limit of 256 on the maximum number of file
2920 descriptors. Exim was using setrlimit() to set 1000 as a large value
2921 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2922 systems. I've change it so that if it can't get 1000, it tries for 256.
2924 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2925 was an oversight, and furthermore, ever since the addition of extra
2926 controls (e.g. 4.43/32), the checks on when to allow different forms of
2927 "control" were broken. There should now be diagnostics for all cases when a
2928 control that does not make sense is encountered.
2930 9. Added the /retain_sender option to "control=submission".
2932 10. $recipients is now available in the predata ACL (oversight).
2934 11. Tidy the search cache before the fork to do a delivery from a message
2935 received from the command line. Otherwise the child will trigger a lookup
2936 failure and thereby defer the delivery if it tries to use (for example) a
2937 cached ldap connection that the parent has called unbind on.
2939 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2940 of $address_data from the recipient verification was clobbered by the
2941 sender verification.
2943 13. The value of address_data from a sender verification is now available in
2944 $sender_address_data in subsequent conditions in the ACL statement.
2946 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2948 15. Added a new option "connect=<time>" to callout options, to set a different
2951 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2952 was its contents. (It was OK if the option was not defined at all.)
2954 17. A "Completed" log line is now written for messages that are removed from
2955 the spool by the -Mrm option.
2957 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2958 information about exactly what failed.
2960 19. Added -dd to debug only the daemon process.
2962 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2963 handles timeouts, both on the server side and network timeouts. Renamed the
2964 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2966 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2967 It is now set to "smtps".
2969 22. $host_address is now set to the target address during the checking of
2970 ignore_target_hosts.
2972 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2973 being passed; this would have caused $sender_host_name to have been used if
2974 matching the list had actually called for a host name (not very likely,
2975 since this list is usually IP addresses). A host name is now passed as
2978 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2979 code that specifies a non-restarting handler (typically sigaction() in
2980 modern systems) in an attempt to fix a rare and obscure crash bug.
2982 25. Narrowed the window for a race in the daemon that could cause it to ignore
2983 SIGCHLD signals. This is not a major problem, because they are used only to
2984 wake it up if nothing else does.
2986 26. A malformed maildirsize file could cause Exim to calculate negative values
2987 for the mailbox size or file count. Odd effects could occur as a result.
2988 The maildirsize information is now recalculated if the size or filecount
2991 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2992 support for a long time. Removed HAVE_SYS_VFS_H.
2994 28. Installed the latest version of exipick from John Jetmore.
2996 29. In an address list, if the pattern was not a regular expression, an empty
2997 subject address (from a bounce message) matched only if the pattern was an
2998 empty string. Non-empty patterns were not even tested. This was the wrong
2999 because it is perfectly reasonable to use an empty address as part of a
3000 database query. An empty address is now tested by patterns that are
3001 lookups. However, all the other forms of pattern expect the subject to
3002 contain a local part and a domain, and therefore, for them, an empty
3003 address still always fails if the pattern is not itself empty.
3005 30. Exim went into a mad DNS loop when attempting to do a callout where the
3006 host was specified on an smtp transport, and looking it up yielded more
3007 than one IP address.
3009 31. Re-factored the code for checking spool and log partition space into a
3010 function that finds that data and another that does the check. The former
3011 is then used to implement four new variables: $spool_space, $log_space,
3012 $spool_inodes, and $log_inodes.
3014 32. The RFC2047 encoding function was originally intended for short strings
3015 such as real names; it was not keeping to the 75-character limit for
3016 encoded words that the RFC imposes. It now respects the limit, and
3017 generates multiple encoded words if necessary. To be on the safe side, I
3018 have increased the buffer size for the ${rfc2047: expansion operator from
3021 33. It is now permitted to omit both strings after an "if" condition; if the
3022 condition is true, the result is "true". As before, when the second string
3023 is omitted, a false condition yields an empty string. This makes it less
3024 cumbersome to write custom ACL and router conditions.
3026 34. Failure to deliver a bounce message always caused it to be frozen, even if
3027 there was an errors_to setting on the router. The errors_to setting is now
3030 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3031 canonical form (fully expanded) before being placed in
3032 $sender_host_address.
3034 36. The table in the code that translates DNS record types into text (T_A to
3035 "A" for instance) was missing entries for NS and CNAME. It is just possible
3036 that this could have caused confusion if both these types were looked up
3037 for the same domain, because the text type is used as part of Exim's
3038 per-process caching. But the chance of anyone hitting this buglet seems
3041 37. The dnsdb lookup has been extended in a number of ways.
3043 (1) There is a new type, "zns", which walks up the domain tree until it
3044 finds some nameserver records. It should be used with care.
3046 (2) There is a new type, "mxh", which is like "mx" except that it returns
3047 just the host names, not the priorities.
3049 (3) It is now possible to give a list of domains (or IP addresses) to be
3050 looked up. The behaviour when one of the lookups defers can be
3051 controlled by a keyword.
3053 (4) It is now possible to specify the separator character for use when
3054 multiple records are returned.
3056 38. The dnslists ACL condition has been extended: it is now possible to supply
3057 a list of IP addresses and/or domains to be looked up in a particular DNS
3060 39. Added log_selector=+queue_time_overall.
3062 40. When running the queue in the test harness, wait just a tad after forking a
3063 delivery process, to get repeatability of debugging output.
3065 41. Include certificate and key file names in error message when GnuTLS fails
3066 to set them up, because the GnuTLS error message doesn't include the name
3067 of the failing file when there is a problem reading it.
3069 42. Allow both -bf and -bF in the same test run.
3071 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3073 44. The "Exiscan patch" is now merged into the mainline Exim source.
3075 45. Sometimes the final signoff response after QUIT could fail to get
3076 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3077 < 0 before doing a fflush(). This bug looks as though it goes back to the
3078 introduction of TLS in release 3.20, but "sometimes" must have been rare
3079 because the tests only now provoked it.
3081 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3082 (this can affect the format of dates).
3084 47. exim_tidydb, when checking for the continued existence of a message for
3085 which it has found a message-specific retry record, was not finding
3086 messages that were in split spool directories. Consequently, it was
3087 deleting retry records that should have stayed in existence.
3089 48. Steve fixed some bugs in eximstats.
3091 49. The SPA authentication driver was not abandoning authentication and moving
3092 on to the next authenticator when an expansion was forced to fail,
3093 contradicting the general specification for all authenticators. Instead it
3094 was generating a temporary error. It now behaves as specified.
3096 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3097 (the order specifies the preference for clients). The order is now AES256,
3098 AES128, 3DES, ARCFOUR128.
3100 51. Small patch to Sieve code - explicitly set From: when generating an
3103 52. Exim crashed if a remote delivery caused a very long error message to be
3104 recorded - for instance if somebody sent an entire SpamAssassin report back
3105 as a large number of 550 error lines. This bug was coincidentally fixed by
3106 increasing the size of one of Exim's internal buffers (big_buffer) that
3107 happened as part of the Exiscan merge. However, to be on the safe side, I
3108 have made the code more robust (and fixed the comments that describe what
3111 53. Now that there can be additional text after "Completed" in log lines (if
3112 the queue_time_overall log selector is set), a one-byte patch to exigrep
3113 was needed to allow it to recognize "Completed" as not the last thing in
3116 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3117 patch that reportedly fixes this has been added. I am not expert enough to
3118 create a test for it. This is what the patch creator wrote:
3120 "I found a little strange behaviour of ldap code when working with
3121 Windows 2003 AD Domain, where users was placed in more than one
3122 Organization Units. When I tried to give exim partial DN, the exit code
3123 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3124 But simultaneously result of request was absolutely normal ldap result,
3125 so I produce this patch..."
3127 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3128 REFERENCE, so I have modified the code to exclude the patch when that macro
3131 55. Some experimental protocols are using DNS PTR records for new purposes. The
3132 keys for these records are domain names, not reversed IP addresses. The
3133 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3134 leaves it alone. Component reversal etc. now happens only for IP addresses.
3137 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3139 57. Double the size of the debug message buffer (to 2048) so that more of very
3140 long debug lines gets shown.
3142 58. The exicyclog utility now does better if the number of log files to keep
3143 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3145 59. Two changes related to the smtp_active_hostname option:
3147 (1) $smtp_active_hostname is now available as a variable.
3148 (2) The default for smtp_banner uses $smtp_active_hostname instead
3149 of $primary_hostname.
3151 60. The host_aton() function is supposed to be passed a string that is known
3152 to be a valid IP address. However, in the case of IPv6 addresses, it was
3153 not checking this. This is a hostage to fortune. Exim now panics and dies
3154 if the condition is not met. A case was found where this could be provoked
3155 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3156 components; fortuitously, this particular loophole had already been fixed
3157 by change 4.50/55 above.
3159 If there are any other similar loopholes, the new check in host_aton()
3160 itself should stop them being exploited. The report I received stated that
3161 data on the command line could provoke the exploit when Exim was running as
3162 exim, but did not say which command line option was involved. All I could
3163 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3164 running as the user.
3167 61. There was a buffer overflow vulnerability in the SPA authentication code
3168 (which came originally from the Samba project). I have added a test to the
3169 spa_base64_to_bits() function which I hope fixes it.
3172 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3173 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3175 63. The daemon start-up calls getloadavg() while still root for those OS that
3176 need the first call to be done as root, but it missed one case: when
3177 deliver_queue_load_max is set with deliver_drop_privilege. This is
3178 necessary for the benefit of the queue runner, because there is no re-exec
3179 when deliver_drop_privilege is set.
3181 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3182 This has been fixed.
3184 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3185 was in use, was not putting the data itself into the right store pool;
3186 consequently, it could be overwritten for a subsequent message in the same
3187 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3190 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3192 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3193 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3194 the function were treating the return as a boolean value, which happened to
3195 work because 0=false and not-0=true, but is not correct code.
3197 68. The host_aton() function was not handling scoped IPv6 addresses (those
3198 with, for example, "%eth0" on the end) correctly.
3200 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3201 negated items (that is, ~something) in unsigned ints. Some compilers
3202 apparently mutter when there is no cast.
3204 70. If an address verification called from an ACL failed, and did not produce a
3205 user-specific message (i.e. there was only a "system" message), nothing was
3206 put in $acl_verify_message. In this situation, it now puts the system
3209 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3210 session; change 4.31/43 added the unwanted input to the log line - except
3211 that it did not do this in the start of session case. It now does.
3213 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3214 This is wrong and can cause the other end to generate a synchronization
3215 error if it is another Exim or anything else that does the synchronization
3216 check. A QUIT command is no longer sent after a timeout.
3218 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3219 during host lookups.
3221 74. The defer_ok option of callout verification was not working if it was used
3222 when verifying addresses in header lines, that is, for this case:
3224 verify = header_sender/callout=defer_ok
3226 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3227 those file descriptors could be used for SMTP connections. If anything
3228 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3229 could be sent to the SMTP client, causing chaos. The daemon now opens
3230 stdin, stdout, and stderr to /dev/null when it puts itself into the
3233 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3234 log by default. The user can override this with suitable Perl magic.
3236 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3237 the log message when discard triggers, was not working for the DATA ACL or
3238 for the non-SMTP ACL.
3240 78. Error message wording change in sieve.c.
3242 79. If smtp_accept_max_per_host was set, the number of connections could be
3243 restricted to fewer than expected, because the daemon was trying to set up
3244 a new connection before checking whether the processes handling previous
3245 connections had finished. The check for completed processes is now done
3246 earlier. On busy systems, this bug wouldn't be noticed because something
3247 else would have woken the daemon, and it would have reaped the completed
3250 80. If a message was submitted locally by a user whose login name contained one
3251 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3252 It caused a spool format error. I have fixed the spool reading code. A
3253 related problem was that the "from" clause in the Received: line became
3254 illegal because of the space(s). It is now covered by ${quote_local_part.
3256 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3259 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3261 83. Patch from Sieve maintainer to fix unterminated string problem in
3262 "vacation" handling.
3264 84. Some minor changes to the Linux configuration files to help with other
3265 OS variants using glibc.
3267 85. One more patch for Sieve to update vacation handling to latest spec.
3270 ----------------------------------------------------
3271 See the note above about the 4.44 and 4.50 releases.
3272 ----------------------------------------------------
3278 1. Change 4.43/35 introduced a bug that caused file counts to be
3279 incorrectly computed when quota_filecount was set in an appendfile
3282 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3283 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3286 3. Give more explanation in the error message when the command for a transport
3287 filter fails to execute.
3289 4. There are several places where Exim runs a non-Exim command in a
3290 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3291 was being done only for the command run by the queryprogram router. It is
3292 now done for all such subprocesses. The other cases are: ${run, transport
3293 filters, and the commands run by the lmtp and pipe transports.
3295 5. Some older OS have a limit of 256 on the maximum number of file
3296 descriptors. Exim was using setrlimit() to set 1000 as a large value
3297 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3298 systems. I've change it so that if it can't get 1000, it tries for 256.
3300 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3301 was an oversight, and furthermore, ever since the addition of extra
3302 controls (e.g. 4.43/32), the checks on when to allow different forms of
3303 "control" were broken. There should now be diagnostics for all cases when a
3304 control that does not make sense is encountered.
3306 7. $recipients is now available in the predata ACL (oversight).
3308 8. Tidy the search cache before the fork to do a delivery from a message
3309 received from the command line. Otherwise the child will trigger a lookup
3310 failure and thereby defer the delivery if it tries to use (for example) a
3311 cached ldap connection that the parent has called unbind on.
3313 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3314 of $address_data from the recipient verification was clobbered by the
3315 sender verification.
3317 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3318 was its contents. (It was OK if the option was not defined at all.)
3320 11. A "Completed" log line is now written for messages that are removed from
3321 the spool by the -Mrm option.
3323 12. $host_address is now set to the target address during the checking of
3324 ignore_target_hosts.
3326 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3327 being passed; this would have caused $sender_host_name to have been used if
3328 matching the list had actually called for a host name (not very likely,
3329 since this list is usually IP addresses). A host name is now passed as
3332 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3333 code that specifies a non-restarting handler (typically sigaction() in
3334 modern systems) in an attempt to fix a rare and obscure crash bug.
3336 15. Narrowed the window for a race in the daemon that could cause it to ignore
3337 SIGCHLD signals. This is not a major problem, because they are used only to
3338 wake it up if nothing else does.
3340 16. A malformed maildirsize file could cause Exim to calculate negative values
3341 for the mailbox size or file count. Odd effects could occur as a result.
3342 The maildirsize information is now recalculated if the size or filecount
3345 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3346 support for a long time. Removed HAVE_SYS_VFS_H.
3348 18. Updated exipick to current release from John Jetmore.
3350 19. Allow an empty sender to be matched against a lookup in an address list.
3351 Previously the only cases considered were a regular expression, or an
3354 20. Exim went into a mad DNS lookup loop when doing a callout where the
3355 host was specified on the transport, if the DNS lookup yielded more than
3358 21. The RFC2047 encoding function was originally intended for short strings
3359 such as real names; it was not keeping to the 75-character limit for
3360 encoded words that the RFC imposes. It now respects the limit, and
3361 generates multiple encoded words if necessary. To be on the safe side, I
3362 have increased the buffer size for the ${rfc2047: expansion operator from
3365 22. Failure to deliver a bounce message always caused it to be frozen, even if
3366 there was an errors_to setting on the router. The errors_to setting is now
3369 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3370 canonical form (fully expanded) before being placed in
3371 $sender_host_address.
3373 24. Updated eximstats to version 1.33
3375 25. Include certificate and key file names in error message when GnuTLS fails
3376 to set them up, because the GnuTLS error message doesn't include the name
3377 of the failing file when there is a problem reading it.
3379 26. Expand error message when OpenSSL has problems setting up cert/key files.
3382 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3383 (this can affect the format of dates).
3385 28. exim_tidydb, when checking for the continued existence of a message for
3386 which it has found a message-specific retry record, was not finding
3387 messages that were in split spool directories. Consequently, it was
3388 deleting retry records that should have stayed in existence.
3390 29. eximstats updated to version 1.35
3391 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3392 1.35 - bugfix such that pie charts by volume are generated correctly
3394 30. The SPA authentication driver was not abandoning authentication and moving
3395 on to the next authenticator when an expansion was forced to fail,
3396 contradicting the general specification for all authenticators. Instead it
3397 was generating a temporary error. It now behaves as specified.
3399 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3400 (the order specifies the preference for clients). The order is now AES256,
3401 AES128, 3DES, ARCFOUR128.
3403 31. Small patch to Sieve code - explicitly set From: when generating an
3406 32. Exim crashed if a remote delivery caused a very long error message to be
3407 recorded - for instance if somebody sent an entire SpamAssassin report back
3408 as a large number of 550 error lines. This bug was coincidentally fixed by
3409 increasing the size of one of Exim's internal buffers (big_buffer) that
3410 happened as part of the Exiscan merge. However, to be on the safe side, I
3411 have made the code more robust (and fixed the comments that describe what
3414 33. Some experimental protocols are using DNS PTR records for new purposes. The
3415 keys for these records are domain names, not reversed IP addresses. The
3416 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3417 leaves it alone. Component reversal etc. now happens only for IP addresses.
3420 34. The host_aton() function is supposed to be passed a string that is known
3421 to be a valid IP address. However, in the case of IPv6 addresses, it was
3422 not checking this. This is a hostage to fortune. Exim now panics and dies
3423 if the condition is not met. A case was found where this could be provoked
3424 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3425 components; fortuitously, this particular loophole had already been fixed
3426 by change 4.50/55 or 4.44/33 above.
3428 If there are any other similar loopholes, the new check in host_aton()
3429 itself should stop them being exploited. The report I received stated that
3430 data on the command line could provoke the exploit when Exim was running as
3431 exim, but did not say which command line option was involved. All I could
3432 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3433 running as the user.
3436 35. There was a buffer overflow vulnerability in the SPA authentication code
3437 (which came originally from the Samba project). I have added a test to the
3438 spa_base64_to_bits() function which I hope fixes it.
3441 36. The daemon start-up calls getloadavg() while still root for those OS that
3442 need the first call to be done as root, but it missed one case: when
3443 deliver_queue_load_max is set with deliver_drop_privilege. This is
3444 necessary for the benefit of the queue runner, because there is no re-exec
3445 when deliver_drop_privilege is set.
3447 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3448 was in use, was not putting the data itself into the right store pool;
3449 consequently, it could be overwritten for a subsequent message in the same
3450 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3453 38. Sometimes the final signoff response after QUIT could fail to get
3454 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3455 < 0 before doing a fflush(). This bug looks as though it goes back to the
3456 introduction of TLS in release 3.20, but "sometimes" must have been rare
3457 because the tests only now provoked it.
3463 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3464 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3465 Now it returns an integer. A number of calls were still expecting a T/F
3466 return. Fortuitously, in all cases, the tests worked in OK situations,
3467 which is the norm. However, things would have gone wrong on any write
3468 failures on the smtp file descriptor. This function is used when sending
3469 messages over SMTP and also when doing verify callouts.
3471 2. When Exim is called to do synchronous delivery of a locally submitted
3472 message (the -odf or -odi options), it no longer closes stderr before doing
3475 3. Implemented the mua_wrapper option.
3477 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3479 5. Implemented the functions header_remove(), header_testname(),
3480 header_add_at_position(), and receive_remove_recipient(), and exported them
3483 6. If an ACL "warn" statement specified the addition of headers, Exim already
3484 inserted X-ACL-Warn: at the start if there was no header name. However, it
3485 was not making this test for the second and subsequent header lines if
3486 there were newlines in the string. This meant that an invalid header could
3487 be inserted if Exim was badly configured.
3489 7. Allow an ACL "warn" statement to add header lines at the start or after all
3490 the Received: headers, as well as at the end.
3492 8. Added the rcpt_4xx retry error code.
3494 9. Added postmaster_mailfrom=xxx to callout verification option.
3496 10. Added mailfrom=xxxx to the callout verification option, for verify=
3499 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3500 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3502 12. Inserted some casts to stop certain compilers warning when using pointer
3503 differences as field lengths or precisions in printf-type calls (mostly
3504 affecting debugging statements).
3506 13. Added optional readline() support for -be (dynamically loaded).
3508 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3509 same clock tick as a message's arrival, so that its received time was the
3510 same as the "first fail" time on the retry record, and that message
3511 remained on the queue past the ultimate address timeout, every queue runner
3512 would try a delivery (because it was past the ultimate address timeout) but
3513 after another failure, the ultimate address timeout, which should have then
3514 bounced the address, did not kick in. This was a "< instead of <=" error;
3515 in most cases the first failure would have been in the next clock tick
3516 after the received time, and all would be well.
3518 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3519 being recognized when the domain list was tested by the match_domain
3520 condition in an expansion string.
3522 16. Added the ${str2b64: operator.
3524 17. Exim was always calling setrlimit() to set a large limit for the number of
3525 processes, without checking whether the existing limit was already
3526 adequate. (It did check for the limit on file descriptors.) Furthermore,
3527 errors from getrlimit() and setrlimit() were being ignored. Now they are
3528 logged to the main and panic logs, but Exim does carry on, to try to do its
3529 job under whatever limits there are.
3531 18. Imported PCRE 5.0.
3533 19. Trivial typo in log message " temporarily refused connection" (the leading
3536 20. If the log selector return_path_on_delivery was set and an address was
3537 redirected to /dev/null, the delivery process crashed because it assumed
3538 that a return path would always be set for a "successful" delivery. In this
3539 case, the whole delivery is bypassed as an optimization, and therefore no
3542 21. Internal re-arrangement: the function for sending a challenge and reading
3543 a response while authentication was assuming a zero-terminated challenge
3544 string. It's now changed to take a pointer and a length, to allow for
3545 binary data in such strings.
3547 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3549 23. Exim was not respecting finduser_retries when seeking the login of the
3550 uid under which it was called; it was always trying 10 times. (The default
3551 setting of finduser_retries is zero.) Also, it was sleeping after the final
3552 failure, which is pointless.
3554 24. Implemented tls_on_connect_ports.
3556 25. Implemented acl_smtp_predata.
3558 26. If the domain in control=submission is set empty, Exim assumes that the
3559 authenticated id is a complete email address when it generates From: or
3560 Sender: header lines.
3562 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3563 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3564 chown and chgrp in /bin and hostname in /usr/bin.
3566 28. Exim was keeping the "process log" file open after each use, just as it
3567 does for the main log. This opens the possibility of it remaining open for
3568 long periods when the USR1 signal hits a daemon. Occasional processlog
3569 errors were reported, that could have been caused by this. Anyway, it seems
3570 much more sensible not to leave this file open at all, so that is what now
3573 29. The long-running daemon process does not normally write to the log once it
3574 has entered its main loop, and it closes the log before doing so. This is
3575 so that log files can straightforwardly be renamed and moved. However,
3576 there are a couple of unusual error situations where the daemon does write
3577 log entries, and I had neglected to close the log afterwards.
3579 30. The text of an SMTP error response that was received during a remote
3580 delivery was being truncated at 512 bytes. This is too short for some of
3581 the long messages that one sometimes sees. I've increased the limit to
3584 31. It is now possible to make retry rules that apply only when a message has a
3585 specific sender, in particular, an empty sender.
3587 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3588 it possible to be selective about when SMTP synchronization is enforced.
3590 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3592 32. Implemented hosts_connection_nolog.
3594 33. Added an ACL for QUIT.
3596 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3599 35. Added mailbox_size and mailbox_filecount to appendfile.
3601 36. Added control = no_multiline_responses to ACLs.
3603 37. There was a bug in the logic of the code that waits for the clock to tick
3604 in the case where the clock went backwards by a substantial amount such
3605 that the microsecond fraction of "now" was more than the microsecond
3606 fraction of "then" (but the whole seconds number was less).
3608 38. Added support for the libradius Radius client library this is found on
3609 FreeBSD (previously only the radiusclient library was supported).
3615 1. When certain lookups returned multiple values in the form name=value, the
3616 quoting of the values was not always being done properly. Specifically:
3617 (a) If the value started with a double quote, but contained no whitespace,
3619 (b) If the value contained whitespace other than a space character (i.e.
3620 tabs or newlines or carriage returns) it was not quoted.
3621 This fix has been applied to the mysql and pgsql lookups by writing a
3622 separate quoting function and calling it from the lookup code. The fix
3623 should probably also be applied to nisplus, ibase and oracle lookups, but
3624 since I cannot test any of those, I have not disturbed their existing code.
3626 2. A hit in the callout cache for a specific address caused a log line with no
3627 reason for rejecting RCPT. Now it says "Previous (cached) callout
3628 verification failure".
3630 3. There was an off-by-one bug in the queryprogram router. An over-long
3631 return line was truncated at 256 instead of 255 characters, thereby
3632 overflowing its buffer with the terminating zero. As well as fixing this, I
3633 have increased the buffer size to 1024 (and made a note to document this).
3635 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3636 when Exim is waiting for an SMTP response from a remote server, Exim
3637 restarts its select() call on the socket, thereby resetting its timeout.
3638 This is not a problem when such interrupts are rare. Somebody set up a cron
3639 job to run exiwhat every 2 minutes, which is less than the normal select()
3640 timeout (5 or 10 minutes). This meant that the select() timeout never
3641 kicked in because it was always reset. I have fixed this by comparing the
3642 time when an interrupt arrives with the time at the start of the first call
3643 to select(). If more time than the timeout has elapsed, the interrupt is
3644 treated as a timeout.
3646 5. Some internal re-factoring in preparation for the addition of Sieve
3647 extensions (by MH). In particular, the "personal" test is moved to a
3648 separate function, and given an option for scanning Cc: and Bcc: (which is
3649 not set for Exim filters).
3651 6. When Exim created an email address using the login of the caller as the
3652 local part (e.g. when creating a From: or Sender: header line), it was not
3653 quoting the local part when it contained special characters such as @.
3655 7. Installed new OpenBSD configuration files.
3657 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3658 try to make them clearer.
3660 9. Callout options, other than the timeout value, were being ignored when
3661 verifying sender addresses in header lines. For example, when using
3663 verify = header_sender/callout=no_cache
3665 the cache was (incorrectly) being used.
3667 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3668 only the Cygwin environment.
3670 11. When return_path_on_delivery was set as a log selector, if different remote
3671 addresses in the same message used different return paths and parallel
3672 remote delivery occurred, the wrong values would sometimes be logged.
3673 (Whenever a remote delivery process finished, the return path value from
3674 the most recently started remote delivery process was logged.)
3676 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3677 lines when AUTH and/or TLS are in use. This is the "received protocol"
3678 field. Exim used to use "asmtp" for authenticated SMTP, without any
3679 indication (in the protocol name) for TLS use. Now it follows the RFC and
3680 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3681 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3682 names appear in log lines as well as in Received: header lines.
3684 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3685 extensions, and comparison tests, and to fix some bugs.
3687 14. Changes to the "personal" filter test:
3689 (1) The test was buggy in that it was just doing the equivalent of
3690 "contains" tests on header lines. For example, if a user's address was
3691 anne@some.where, the "personal" test would incorrectly be true for
3693 To: susanne@some.where
3695 This test is now done by extracting each address from the header in turn,
3696 and checking the entire address. Other tests that are part of "personal"
3697 are now done using regular expressions (for example, to check local parts
3698 of addresses in From: header lines).
3700 (2) The list of non-personal local parts in From: addresses has been
3701 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3702 taken from the Sieve specification recommendations.
3704 (3) If the message contains any header line starting with "List-" it is
3705 treated as non-personal.
3707 (4) The test for "circular" in the Subject: header line has been removed
3708 because it now seems ill-conceived.
3710 15. Minor typos in src/EDITME comments corrected.
3712 16. Installed latest exipick from John Jetmore.
3714 17. If headers_add on a router specified a text string that was too long for
3715 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3716 of string_sprintf() is now avoided.
3718 18. $message_body_size was not set (it was always zero) when running the DATA
3719 ACL and the local_scan() function.
3721 19. For the "mail" command in an Exim filter, no default was being set for
3722 the once_repeat time, causing a random time value to be used if "once" was
3723 specified. (If the value happened to be <= 0, no repeat happened.) The
3724 default is now 0s, meaning "never repeat". The "vacation" command was OK
3725 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3726 (I found it when inspecting the code).
3728 20. There is now an overall timeout for performing a callout verification. It
3729 defaults to 4 times the callout timeout, which applies to individual SMTP
3730 commands during the callout. The overall timeout applies when there is more
3731 than one host that can be tried. The timeout is checked before trying the
3732 next host. This prevents very long delays if there are a large number of
3733 hosts and all are timing out (e.g. when the network connections are timing
3734 out). The value of the overall timeout can be changed by specifying an
3735 additional sub-option for "callout", called "maxwait". For example:
3737 verify = sender/callout=5s,maxwait=20s
3739 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3740 to the end before writing, but this should make it even safer).
3742 22. Exim was forgetting that it had advertised PIPELINING for the second and
3743 subsequent messages on an SMTP connection. It was also not resetting its
3744 memory on STARTTLS and an internal HELO.
3746 23. When Exim logs an SMTP synchronization error within a session, it now
3747 records whether PIPELINING has been advertised or not.
3749 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3750 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3751 rather than long int.
3753 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3755 26. Added the never_mail option to autoreply.
3761 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3762 crash if the getsockname() call failed; this can happen if a connection is
3763 closed very soon after it is established. The problem was simply in the
3764 order in which certain operations were done, causing Exim to try to write
3765 to the SMTP stream before it had set up the file descriptor. The bug has
3766 been fixed by making things happen in the correct order.
3772 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3773 before the connection was closed, thus losing the rejection response.
3775 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3776 some early Solaris releases, but causes trouble in current releases where
3777 socklen_t is defined.
3779 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3782 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3785 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3786 that releasing the top of it at the end releases what was used for sub-
3787 expansions (unless the block got too big). However, discard this block if
3788 the first thing is a variable or header, so that we can use its block when
3789 it is dynamic (useful for very large $message_headers, for example).
3791 6. Lookups now cache *every* query, not just the most recent. A new, separate
3792 store pool is used for this. It can be recovered when all lookup caches are
3793 flushed. Lookups now release memory at the end of their result strings.
3794 This has involved some general refactoring of the lookup sources.
3796 7. Some code has been added to the store_xxx() functions to reduce the amount
3797 of flapping under certain conditions.
3799 8. log_incoming_interface used to affect only the <= reception log lines. Now
3800 it causes the local interface and port to be added to several more SMTP log
3801 lines, for example "SMTP connection from", and rejection lines.
3803 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3805 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3807 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3808 could be overwritten at the end of the current message (or the start of a
3809 new message if it was set in a HELO ACL). The value is now preserved for
3810 the duration of the SMTP connection.
3812 12. If a transport had a headers_rewrite setting, and a matching header line
3813 contained an unqualified address, that address was qualified, even if it
3814 did not match any rewriting rules. The underlying bug was that the values
3815 of the flags that permit the existence of unqualified sender and recipient
3816 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3817 non-local messages, and by -bnq for local messages) were not being
3818 preserved with the message after it was received.
3820 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3821 "next input=" as part of the text comprising the host identity instead of
3822 the correct text. The code was using the same buffer for two different
3823 strings. However, depending on which order the printing function evaluated
3824 its arguments, the bug did not always show up. Under Linux, for example, my
3825 test suite worked just fine.
3827 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3828 doesn't work with some older versions of Perl. It has been changed to "my",
3829 which in any case is probably the better facility to use.
3831 15. A really picky compiler found some instances of statements for creating
3832 error messages that either had too many or two few arguments for the format
3835 16. The size of the buffer for calls to the DNS resolver has been increased
3836 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3837 for addresses that have a lot of PTR records. This alleviates a problem; it
3838 does not fully solve it.
3840 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3841 buffer now truncates the list and logs the incident, which is the same
3842 action as happens when Exim is looking up a host name and its aliases.
3843 Previously in this situation something unpredictable would happen;
3844 sometimes it was "internal error: store_reset failed".
3846 18. If a server dropped the connection unexpectedly when an Exim client was
3847 using GnuTLS and trying to read a response, the client delivery process
3848 crashed while trying to generate an error log message.
3850 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3851 string, for example:
3853 warn message = H1: something\nH2: something
3855 the text was added as a single header line from Exim's point of view
3856 though it ended up OK in the delivered message. However, searching for the
3857 second and subsequent header lines using $h_h2: did not work. This has been
3858 fixed. Similarly, if a system filter added multiple headers in this way,
3859 the routers could not see them.
3861 20. Expanded the error message when iplsearch is called with an invalid key to
3862 suggest using net-iplsearch in a host list.
3864 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3865 ACLs are no longer actually imposed (and a message to that effect is
3868 22. If a "gecos" field in a passwd entry contained escaped characters, in
3869 particular, if it contained a \" sequence, Exim got it wrong when building
3870 a From: or a Sender: header from that name. A second bug also caused
3871 incorrect handling when an unquoted " was present following a character
3872 that needed quoting.
3874 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3875 was not being matched caselessly.
3877 24. Arranged for all hyphens in the exim.8 source to be escaped with
3880 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3881 properly. Recipient callout cache records were still being keyed to include
3882 the sender, even when use_sender was set false. This led to far more
3883 callouts that were necessary. The sender is no longer included in the key
3884 when use_sender is false.
3886 26. Added "control = submission" modifier to ACLs.
3888 27. Added the ${base62d: operator to decode base 62 numbers.
3890 28. dnsdb lookups can now access SRV records.
3892 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3893 the configuration file.
3895 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3896 (-v) mode. This makes the output for a verbose queue run more intelligible.
3898 31. Added a use_postmaster feature to recipient callouts.
3900 32. Added the $body_zerocount variable, containing the number of binary zero
3901 bytes in the message body.
3903 33. The time of last modification of the "new" subdirectory is now used as the
3904 "mailbox time last read" when there is a quota error for a maildir
3907 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3909 35. Added +ignore_unknown as a special item in host lists.
3911 36. Code for decoding IPv6 addresses in host lists is now included, even if
3912 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3913 address was recognized as an IP address, but was then not correctly decoded
3914 into binary, causing unexpected and incorrect effects when compared with
3921 1. Very minor rewording of debugging text in manualroute to say "list of
3922 hosts" instead of "hostlist".
3924 2. If verify=header_syntax was set, and a header line with an unqualified
3925 address (no domain) and a large number of spaces between the end of the
3926 name and the colon was received, the reception process suffered a buffer
3927 overflow, and (when I tested it) crashed. This was caused by some obsolete
3928 code that should have been removed. The fix is to remove it!
3930 3. When running in the test harness, delay a bit after writing a bounce
3931 message to get a bit more predictability in the log output.
3933 4. Added a call to search_tidyup() just before forking a reception process. In
3934 theory, someone could use a lookup in the expansion of smtp_accept_max_
3935 per_host which, without the tidyup, could leave open a database connection.
3937 5. Added the variables $recipient_data and $sender_data which get set from a
3938 lookup success in an ACL "recipients" or "senders" condition, or a router
3939 "senders" option, similar to $domain_data and $local_part_data.
3941 6. Moved the writing of debug_print from before to after the "senders" test
3944 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3945 problems for message scanning, either using a data ACL, or using
3946 local_scan() because the Received: header was not generated till after they
3947 were called (in order to set the time as the time of reception completion).
3948 I have revised the way this works. The header is now generated after the
3949 body is received, but before the ACL or local_scan() are called. After they
3950 are run, the timestamp in the header is updated.
3956 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3957 before starting a queue runner without re-exec. This happened only when
3958 deliver_drop_privilege was set or when the Exim user was set to root. The
3959 effect of the bug was that timeouts during subsequent deliveries caused
3960 crashes instead of being properly handled. The handler is now left at its
3961 default (and expected) setting.
3963 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3964 message, again when deliver_drop_privilege is set or Exim is run as root.
3965 The bug described in (1) was not present in this case, but the tidying up
3966 of the other signals was missing. I have made the two cases consistent.
3968 3. The ignore_target_hosts setting on a manualroute router was being ignored
3969 for hosts that were looked up using the /MX notation.
3971 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3974 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3975 operated on the sender address. After changing the $sender_address to <>
3976 for the sender address verify, Exim was re-instated it as the original
3977 (before rewriting) address, but remembering that it had rewritten it, so it
3978 wasn't rewriting it again. This bug also had the effect of breaking the
3979 sender address verification caching when the sender address was rewritten.
3981 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3982 This has been changed so that if the ip literal address matches
3983 ignore_target_hosts, the router declines.
3985 7. Added expansion conditions match_domain, match_address, and match_local_
3986 part (NOT match_host).
3988 8. The placeholder for the Received: header didn't have a length field set.
3990 9. Added code to Exim itself and to exim_lock to test for a specific race
3991 condition that could lead to file corruption when using MBX delivery. The
3992 issue is with the lockfile that is created in /tmp. If this file is removed
3993 after a process has opened it but before that process has acquired a lock,
3994 there is the potential for a second process to recreate the file and also
3995 acquire a lock. This could lead to two Exim processes writing to the file
3996 at the same time. The added code performs the same test as UW imapd; it
3997 checks after acquiring the lock that its file descriptor still refers to
3998 the same named file.
4000 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4001 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4004 11. Added the smtp_active_hostname option. If used, this will typically be made
4005 to depend on the incoming interface address. Because $interface_address is
4006 not set up until the daemon has forked a reception process, error responses
4007 that can happen earlier (such as "too many connections") no longer contain
4010 12. If an expansion in a condition on a "warn" statement fails because a lookup
4011 defers, the "warn" statement is abandoned, and the next ACL statement is
4012 processed. Previously this caused the whole ACL to be aborted.
4014 13. Added the iplsearch lookup type.
4016 14. Added ident_timeout as a log selector.
4018 15. Added tls_certificate_verified as a log selector.
4020 16. Added a global option tls_require_ciphers (compare the smtp transport
4021 option of the same name). This controls incoming TLS connections.
4023 17. I finally figured out how to make tls_require_ciphers do a similar thing
4024 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4025 before starting the TLS session.
4027 18. Tabs are now shown as \t in -bP output.
4029 19. If the log selector return_path_on_delivery was set, Exim crashed when
4030 bouncing a message because it had too many Received: header lines.
4032 20. If two routers both had headers_remove settings, and the first one included
4033 a superfluous trailing colon, the final name in the first list and the
4034 first name in the second list were incorrectly joined into one item (with a
4035 colon in the middle).
4041 1. Added -C and -D options to the exinext utility, mainly to make it easier
4042 to include in the automated testing, but these could be helpful when
4043 multiple configurations are in use.
4045 2. The exinext utility was not formatting the output nicely when there was
4046 an alternate port involved in the retry record key, nor when there was a
4047 message id as well (for retries that were specific to a specific message
4048 and a specific host). It was also confused by IPv6 addresses, because of
4049 the additional colons they contain. I have fixed the IPv4 problem, and
4050 patched it up to do a reasonable job for IPv6.
4052 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4053 delivery, the log line now contains "pipelined" if PIPELINING was used.
4055 4. An SMTP transport process used to panic and die if the bind() call to set
4056 an explicit outgoing interface failed. This has been changed; it is now
4057 treated in the same way as a connect() failure.
4059 5. A reference to $sender_host_name in the part of a conditional expansion
4060 that was being skipped was still causing a DNS lookup. This no longer
4063 6. The def: expansion condition was not recognizing references to header lines
4064 that used bh_ and bheader_.
4066 7. Added the _cache feature to named lists.
4068 8. The code for checking quota_filecount in the appendfile transport was
4069 allowing one more file than it should have been.
4071 9. For compatibility with Sendmail, the command line option
4079 and sets the incoming protocol and host name (for trusted callers). The
4080 host name and its colon can be omitted when only the protocol is to be set.
4081 Note the Exim already has two private options, -pd and -ps, that refer to
4082 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4083 "s", but I don't think that's a major issue.
4085 10. A number of refactoring changes to the code, none of which should affect
4088 (a) The number of logging options was getting close to filling up the
4089 32-bit word that was used as a bit map. I have split them into two classes:
4090 those that are passed in the argument to log_write(), and those that are
4091 only ever tested independently outside of that function. These are now in
4092 separate 32-bit words, so there is plenty of room for expansion again.
4093 There is no change in the user interface or the logging behaviour.
4095 (b) When building, for example, log lines, the code previously used a
4096 macro that called string_cat() twice, in order to add two strings. This is
4097 not really sufficiently general. Furthermore, there was one instance where
4098 it was actually wrong because one of the argument was used twice, and in
4099 one call a function was used. (As it happened, calling the function twice
4100 did not affect the overall behaviour.) The macro has been replaced by a
4101 function that can join an arbitrary number of extra strings onto a growing
4104 (c) The code for expansion conditions now uses a table and a binary chop
4105 instead of a serial search (which was left over from when there were very
4106 few conditions). Also, it now recognizes conditions like "pam" even when
4107 the relevant support is not compiled in: a suitably worded error message is
4108 given if an attempt is made to use such a condition.
4110 11. Added ${time_interval:xxxxx}.
4112 12. A bug was causing one of the ddress fields not to be passed back correctly
4113 from remote delivery subprocesses. The field in question was not being
4114 subsequently used, so this caused to problems in practice.
4116 13. Added new log selectors queue_time and deliver_time.
4118 14. Might have fixed a bug in maildirsizefile handling that threw up
4119 "unexpected character" debug warnings, and recalculated the data
4120 unnecessarily. In any case, I expanded the warning message to give more
4123 15. Added the message "Restricted characters in address" to the statements in
4124 the default ACL that block characters like @ and % in local parts.
4126 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4127 Three changes have been made:
4129 (a) There was a serious bug; a negative response to MAIL caused the whole
4130 recipient domain to be cached as invalid, thereby blocking all messages
4131 to all local parts at the same domain, from all senders. This bug has
4132 been fixed. The domain is no longer cached after a negative response to
4133 MAIL if the sender used is not empty.
4135 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4138 (c) A new callout option, "use_sender" has been added for people who want
4139 the modified behaviour.
4145 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4148 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4149 indeed breaks things for older releases.
4151 3. Added additional logging to the case where there is a problem reading data
4152 from a filter that is running in a subprocess using a pipe, in order to
4153 try to track down a specific problem.
4155 4. Testing facility fudge: when running in the test harness and attempting
4156 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4157 getting "No route to host". Convert this to a timeout.
4159 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4162 6. Some OS don't have socklen_t but use size_t instead. This affects the
4163 fifth argument of getsockopt() amongst other things. This is now
4164 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4165 can be set for individual OS. I have set it for SunOS5, OSF1, and
4166 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4167 some earlier ones do not.
4169 7. Change 4.30/15 was not doing the test caselessly.
4171 8. The standard form for an IPv6 address literal was being rejected by address
4172 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4173 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4174 this, as well as the form without the "IPv6" on the front (but only when
4175 address literals are enabled, of course).
4177 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4179 10. Exim crashed if a message with an empty sender address specified by -f
4180 encountered a router with an errors_to setting. This could be provoked only
4181 by a command such as
4185 where an empty string was supplied; "<>" did not hit this bug.
4187 11. Installed PCRE release 4.5.
4189 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4190 remained set. It is now erased.
4192 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4193 times from message ids (which are base 36 rather than the normal 62).
4195 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4196 were being counted as actual protocol errors, and logged if the log
4197 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4198 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4199 and DATA following a set of rejected RCPTs do not count as protocol errors.
4200 In other words, Exim assumes they were pipelined, though this may not
4201 actually be the case. Of course, in all cases the client gets an
4202 appropriate error code.
4204 15. If a lookup fails in an ACL condition, a message about the failure may
4205 be available; it is used if testing the ACL cannot continue, because most
4206 such messages specify what the cause of the deferral is. However, some
4207 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4208 that caused an old message to be retained and used if a later statement
4209 caused a defer, replacing the real cause of the deferral.
4211 16. If an IP address had so many PTR records that the DNS lookup buffer
4212 was not large enough to hold them, Exim could crash while trying to process
4213 the truncated data. It now detects and logs this case.
4215 17. Further to 4.21/58, another change has been made: if (and only if) the
4216 first line of a message (the first header line) ends with CRLF, a bare LF
4217 in a subsequent header line has a space inserted after it, so as not to
4218 terminate the header.
4220 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4221 unnecessarily, used atoi() instead of strtol(), and didn't check the
4222 termination when getting file sizes from file names by regex.
4224 19. Completely re-implemented the support for maildirsize files, in the light
4225 of a number of problems with the previous contributed implementation
4226 (4.30/29). In particular:
4228 . If the quota is zero, the maildirsize file is maintained, but no quota is
4231 . If the maildir directory does not exist, it is created before any attempt
4232 to write a maildirsize file.
4234 . The quota value in the file is just a cache; if the quota is changed in
4235 the transport, the new value overrides.
4237 . A regular expression is available for excluding directories from the
4240 20. The autoreply transport checks the characters in options that define the
4241 message's headers; it allows continued headers, but it was checking with
4242 isspace() after an embedded newline instead of explicitly looking for a
4245 21. If all the "regular" hosts to which an address was routed had passed their
4246 expiry times, and had not reached their retry times, the address was
4247 bounced, even if fallback hosts were defined. Now Exim should go on to try
4250 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4251 equivalent code in the SMTP transport. Some hosts send humungous responses
4252 to HELO/EHLO, more than 1024 it seems.
4254 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4255 gives compiler warnings in some environments. I've now done it "properly",
4258 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4259 (because of problems with the built-in one) was declared to return uschar *
4260 instead of char *, causing compiler failure.
4262 25. Fixed a file descriptor leak when processing alias/forward files.
4264 26. Fixed a minor format string issue in dbfn.c.
4266 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4268 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4269 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4272 29. When a sender address is verified, it is cached, to save repeating the test
4273 when there is more than one recipient in a message. However, when the
4274 verification involves a callout, it is possible for different callout
4275 options to be set for different recipients. It is too complicated to keep
4276 track of this in the cache, so now Exim always runs a verification when a
4277 callout is required, relying on the callout cache for the optimization.
4278 The overhead is duplication of the address routing, but this should not be
4281 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4282 to be verified with callout=postmaster, and the main callout worked but the
4283 postmaster check failed, the verification correctly failed. However, if a
4284 subsequent RCPT command asked for sender verification *without* the
4285 postmaster check, incorrect caching caused this verification also to fail,
4288 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4289 it was not caching the DNS options (qualify_single, search_parents) that
4290 were used when the lookup failed. A subsequent lookup with different
4291 options therefore always gave the same answer, though there were cases
4292 where it should not have. (Example: a "domains = !$mx_any" option on a
4293 dnslookup router: the "domains" option is always processed without any
4294 widening, but the router might have qualify_single set.) Now Exim uses the
4295 cached value only when the same options are set.
4297 32. Added John Jetmore's "exipick" utility to the distribution.
4299 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4300 than a timeout (e.g. a certificate is required, and is not provided), an
4301 Exim server now closes the connection immediately. Previously it waited for
4302 the client to close - but if the client is SSL, it seems that they each
4303 wait for each other, leading to a delay before one of them times out.
4305 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4306 maintained 0.8.x compatibility because I don't think many are using it, and
4307 it is clearly obsolete.
4309 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4312 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4313 client certificate was expired. A simple patch fixes this, though I don't
4314 understand the full logic of why the verify callback is called multiple
4317 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4318 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4319 which causes problems with some clients (such as the Certicom SSL Plus
4320 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4321 disables the coutermeasure allowing Eudora to connect."
4323 38. Exim was not checking that a write() to a log file succeeded. This could
4324 lead to Bad Things if a log got too big, in particular if it hit a file
4325 size limit. Exim now panics and dies if it cannot write to a log file, just
4326 as it does if it cannot open a log file.
4328 39. Modified OS/Makefile-Linux so that it now contains
4330 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4332 The two -D definitions ensure that Exim is compiled with large file
4333 support, which makes it possible to handle log files that are bigger than
4336 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4337 instance) a domain was checked against a named list that involved a lookup,
4338 causing $domain_data to be set, then another domain was checked against the
4339 same list, then the first domain was re-checked, the value of $domain_data
4340 after the final check could be wrong. In particular, if the second check
4341 failed, it could be set empty. This bug probably also applied to
4344 41. The strip_trailing_dot option was not being applied to the address given
4345 with the -f command-line option.
4347 42. The code for reading a message's header from the spool was incrementing
4348 $received_count, but never initializing it. This meant that the value was
4349 incorrect (doubled) while delivering a message in the same process in which
4350 it was received. In the most common configuration of Exim, this never
4351 happens - a fresh exec is done - but it can happen when
4352 deliver_drop_privilege is set.
4354 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4355 it now includes up to 150 characters of the unexpected data in the log
4358 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4359 and building data strings. The size of both of these buffers was 10 000
4360 bytes - far larger than anybody would *ever* want, thought I. Needless to
4361 say, somebody hit the limit. I have increased the maximum line length to
4362 20 000 and the maximum data length of concatenated lines to 100 000. I have
4363 also fixed two bugs, because there was no checking on these buffers. Tsk,
4364 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4365 buffer is too small.
4367 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4368 lsearch lookups. Now it does.
4370 46. When parsing a route_list item in a manualroute router, a fixed-length
4371 buffer was used for the list of hosts. I made this 1024 bytes long,
4372 thinking that nobody would ever have a list of hosts that long. Wrong.
4373 Somebody had a whole pile of complicated expansion conditions, and the
4374 string was silently truncated, leading to an expansion error. It turns out
4375 that it is easier to change to an unlimited length (owing to other changes
4376 that have happened since this code was originally written) than to build
4377 structure for giving a limitation error. The length of the item that
4378 expands into the list of hosts is now unlimited.
4380 47. The lsearch lookup could not handle data where the length of text line was
4381 more than 4095 characters. Such lines were truncated, leading to shortened
4382 data being returned. It should now handle lines of any length.
4384 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4385 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4388 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4389 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4390 no longer happens in comment lines. A list of replacements is now placed
4391 at the head of all of the source files, except those whose only change is
4392 to replace PERL_COMMAND in the very first #! line.
4394 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4395 messages on the queue, with a bottom-up merge sort, using code contributed
4396 by Michael Haardt. This should make operations like -bp somewhat faster on
4397 large queues. It won't affect queue runners, except when queue_run_in_order
4400 51. Installed eximstats 1.31 in the distribution.
4402 52. Added support for SRV lookups to the dnslookup router.
4404 53. If an ACL referred to $message_body or $message_body_end, the value was not
4405 reset for any messages that followed in the same SMTP session.
4407 54. The store-handling optimization for building very long strings was not
4408 differentiating between the different store pools. I don't think this
4409 actually made any difference in practice, but I've tidied it.
4411 55. While running the routers to verify a sender address, $sender_address
4412 was still set to the sender address. This is wrong, because when routing to
4413 send a bounce to the sender, it would be empty. Therefore, I have changed
4414 it so that, while verifying a sender address, $sender_address is set to <>.
4415 (There is no change to what happens when verifying a recipient address.)
4417 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4418 target A or AAAA records (if not already returned) without resetting the
4419 qualify_single or search_parents options of the DNS resolver. These are
4420 inappropriate in this case because the targets of MX and SRV records must
4421 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4422 target that, when qualified, matched something in the local domain. These
4423 two options are now turned off when doing these lookups.
4425 57. It seems that at least some releases of Reiserfs (which does not have the
4426 concept of a fixed number of inodes) returns zero and not -1 for the
4427 number of available inodes. This interacted badly with check_spool_inodes,
4428 which assumed that -1 was the "no such thing" setting. What I have done is
4429 to check that the total number of inodes is greater than zero before doing
4430 the test of how many are available.
4432 58. When a "warn" ACL statement has a log_message modifier, the message is
4433 remembered, and not repeated. This is to avoid a lot of repetition when a
4434 message has many recipients that cause the same warning to be written.
4435 Howewer, Exim was preserving the list of already written lines for an
4436 entire SMTP session, which doesn't seem right. The memory is now reset if a
4437 new message is started.
4439 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4440 debugging output unless log_rewrite was also set.
4442 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4443 of (int)(handle) when we know that handle contains (void *)(-1).
4445 61. The Exim daemon panic-logs an error return when it closes the incoming
4446 connection. However "connection reset by peer" seems to be common, and
4447 isn't really an error worthy of noting specially, so that particular error
4450 62. When Exim is trying to find all the local interfaces, it used to panic and
4451 die if the ioctl to get the interface flags failed. However, it seems that
4452 on at least one OS (Solaris 9) it is possible to have an interface that is
4453 included in the list of interfaces, but for which you get a failure error
4454 for this call. This happens when the interface is not "plumbed" into a
4455 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4456 failure of the "get flags" call assumes that the interface is down.
4458 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4459 makes life easier for people who are doing arithmetic on fields extracted
4460 from dates, where you often get leading zeros that should not be
4461 interpreted as octal.
4463 64. Added qualify_domain to the redirect router, to override the global
4466 65. If a pathologically long header line contained very many addresses (the
4467 report of this problem mentioned 10 000) and each of them was rewritten,
4468 Exim could use up a very large amount of memory. (It kept on making new
4469 copies of the header line as it rewrote, and never released the old ones.)
4470 At the expense of a bit more processing, the header rewriting function has
4471 been changed so that it no longer eats memory in this way.
4473 66. The generation of the Received: header has been moved from the time that a
4474 message starts to be received, to the time that it finishes. The timestamp
4475 in the Received: header should now be very close to that of the <= log
4476 line. There are two side-effects of this change:
4478 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4479 logged header lines no longer include the local Received: line, because
4480 it has not yet been created. The same applies to a copy of the message
4481 that is returned to a non-SMTP sender when a message is rejected.
4483 (b) When a filter file is tested using -bf, no additional Received: header
4484 is added to the test message. After some thought, I decided that this
4487 This change does not affect the value of $received_for. It is still set
4488 after address rewriting, but before local_scan() is called.
4490 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4492 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4493 gave an unhelpful panic error message, and a defer error. I have managed to
4494 change this behaviour so that it now rejects any supplied certificate,
4495 which seems right, as the list of acceptable certificates is empty.
4497 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4498 gave an unhelpful defer error. I have not managed to make this reject any
4499 supplied certificates, but the error message it gives is "no certificate
4500 supplied", which is not helpful.
4502 70. exigrep's output now also includes lines that are not associated with any
4503 message, but which match the given pattern. Implemented by a patch from
4504 Martin Sluka, which also tidied up the Perl a bit.
4506 71. Recipient callout verification, like sender verification, was using <> in
4507 the MAIL FROM command. This isn't really the right thing, since the actual
4508 sender may affect whether the remote host accepts the recipient or not. I
4509 have changed it to use the actual sender in the callout; this means that
4510 the cache record is now keyed on a recipient/sender pair, not just the
4511 recipient address. There doesn't seem to be a real danger of callout loops,
4512 since a callout by the remote host to check the sender would use <>.
4513 [SEE ABOVE: changed after hitting problems.]
4515 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4516 temporary errors. However, in the case of such a code being given after
4517 the end of a data transmission (i.e. after ".") Exim was failing to write
4518 a retry record for the message. (Yes, there was some broken host that was
4519 actually sending 8xx at this point.)
4521 73. An unknown lookup type in a host list could cause Exim to panic-die when
4522 the list was checked. (An example that provoked this was putting <; in the
4523 middle of a list instead of at the start.) If this happened during a DATA
4524 ACL check, a -D file could be left lying around. This kind of configuration
4525 error no longer causes Exim to die; instead it causes a defer errror. The
4526 incident is still logged to the main and panic logs.
4528 74. Buglet left over from Exim 3 conversion. The message "too many messages
4529 in one connection" was written to the rejectlog but not the mainlog, except
4530 when address rewriting (yes!) was being logged.
4532 75. Added write_rejectlog option.
4534 76. When a system filter was run not as root (that is, when system_filter_user
4535 was set), the values of the $n variables were not being returned to the
4536 main process; thus, they were not subsequently available in the $sn
4539 77. Added +return_path_on_delivery log selector.
4541 78. A connection timeout was being treated differently from recipients deferred
4542 when testing hosts_max_try with a message that was older than the host's
4543 retry timeout. (The host should not be counted, thus allowing all hosts to
4544 be tried at least once before bouncing.) This may have been the cause of an
4545 occasionally reported bug whereby a message would remain on the queue
4546 longer than the retry timeout, but would be bounced if a delivery was
4547 forced. I say "may" because I never totally pinned down the problem;
4548 setting up timeout/retry tests is difficult. See also the next item.
4550 79. The ultimate address timeout was not being applied to errors that involved
4551 a combination of host plus message (for example, a timeout on a MAIL
4552 command). When an address resolved to a number of possible hosts, and they
4553 were not all tried for each delivery (e.g. because of hosts_max_try), a
4554 message could remain on the queue longer than the retry timeout.
4556 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4559 81. Fixed an obscure SMTP outgoing bug which required at least the following
4560 conditions: (a) there was another message waiting for the same server;
4561 (b) the server returned 5xx to all RCPT commands in the first message so
4562 that the message was not completed; (c) the server dropped the connection
4563 or gave a negative response to the RSET that Exim sends to abort the
4564 transaction. The observed case was a dropped connection after DATA that had
4565 been sent in pipelining mode. That is, the server had advertised PIPELINING
4566 but was not implementing it correctly. The effect of the bug was incorrect
4567 behaviour, such as trying another host, and this could lead to a crash.
4573 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4574 and daemon.c were passed as pointers to ints; they should have been
4575 pointers to socklen_t variables (which are typically unsigned ints).
4577 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4580 3. Fixed a really odd bug that affected only the testing scheme; patching a
4581 certain fixed string in the binary changed the value of another string that
4582 happened to be identical to the end of the original first string.
4584 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4585 name", it returns that address as the IP address. On some operating
4586 systems (e.g. Solaris), it also passes back the IP address string as the
4587 "host name". However, on others (e.g. Linux), it passes back an empty
4588 string. Exim wasn't checking for this, and was changing the host name to an
4589 empty string, assuming it had been canonicized.
4591 5. Although rare, it is permitted to have more than one PTR record for a given
4592 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4593 all the names associated with an address, because they do in Solaris.
4594 However, it seems that they do not in Linux for data that comes from the
4595 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4596 I found this out when I moved to a new Linux workstation and tried to run
4597 the Exim test suite.
4599 To get round this problem I have changed the code so that it now does its
4600 own call to the DNS to look up PTR records when searching for a host name.
4601 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4602 addresses that are only in /etc/hosts are still found.
4604 This behaviour is, however, controlled by an option called host_lookup_
4605 order, which defaults to "bydns:byaddr". If people want to use the other
4606 order, or indeed, just use one or the other means of lookup, they can
4607 specify it in this variable.
4609 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4610 some operating systems, this comes back from gethostbyaddr() as an empty
4611 string, and this is what Exim used to test for. However, it seems that in
4612 other systems, "." is yielded. Exim now tests for this case too.
4614 7. The values of check_spool_space and check_log_space are now held internally
4615 as a number of kilobytes instead of an absolute number of bytes. If a
4616 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4617 kilobyte. This means that much larger values can be stored.
4619 8. Exim monitor: an attempt to get the action menu when not actually pointing
4620 at a message produces an empty menu entitled "No message selected". This
4621 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4622 no entries in it ("Shell widget menu has zero width and/or height"). So I
4623 have added a single, blank menu entry in this case.
4625 9. Added ${quote_local_part.
4627 10. MIME decoding is now applied to the contents of Subject: header lines when
4630 11. Now that a reference to $sender_host_address automatically causes a reverse
4631 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4632 host lookup before query-style lookups in lists that might use this
4633 variable. This has therefore been abolished, and the "net-" prefix is no
4634 longer necessary for query-style lookups.
4636 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4637 have been a typo for LFLAGS, so it has been changed.
4639 13. The install script calls Exim with "-C /dev/null" in order to find the
4640 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4641 to be output. Howeve, since Exim outputs its version number before the
4642 error, it didn't break the script. It just looked ugly. I fixed this by
4643 always allowing "-C /dev/null" if the caller is root.
4645 14. Ignore overlarge ACL variable number when reading spool file - insurance
4646 against a later release with more variables having written the file.
4648 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4649 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4650 this, as well as the form without the "IPv6" on the front.
4652 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4653 OS/Makefile-Darwin file.
4655 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4656 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4658 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4659 variable. However, some OS use ints for inodes. I've added cast to long int
4660 to get rid of the compiler warning.
4662 19. I had forgotten to lock out "/../" in configuration file names when
4663 ALT_CONFIG_PREFIX was set.
4665 20. Routers used for verification do not need to specify transports. However,
4666 if such a router generated a host list, and callout was configured, Exim
4667 crashed, because it could not find a port number from the (non-existent)
4668 transport. It now assumes port 25 in this circumstance.
4670 21. Added the -t option to exigrep.
4672 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4673 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4674 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4675 them is set, LOOKUP_LSEARCH is forced.
4677 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4679 24. Added sender and host information to the "rejected by local_scan()" log
4680 line; previously there was no indication of these.
4682 25. Added .include_if_exists.
4684 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4685 Linux. It turns out that not all file systems support this. Apparently some
4686 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4687 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4688 not supported on the file descriptor, is now ignored when Exim is trying to
4689 sync a directory. This applies only to Linux.
4691 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4693 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4694 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4697 29. Added support for maildirsize files from supplied patch (modified a bit).
4699 30. The use of :fail: followed by an empty string could lead Exim to respond to
4700 sender verification failures with (e.g.):
4702 550 Verification failed for <xxx>
4703 550 Sender verify failed
4705 where the first response line was missing the '-' that indicates it is not
4706 the final line of the response.
4708 31. The loop for finding the name of the user that called Exim had a hardwired
4709 limit of 10; it now uses the value of finduser_retries, which is used for
4710 all other user lookups.
4712 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4715 33. Exim was neglecting to zero errno before one call of strtol() when
4716 expanding a string and expecting an integer value. On some systems this
4717 resulted in spurious "integer overflow" errors. Also, it was casting the
4718 result into an int without checking.
4720 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4721 did not work. The code looks as if it has *never* worked, though it appears
4722 to have been documented since at least releast 1.62. I have made it work.
4724 35. The "timeout_DNS" error in retry rules, also documented since at least
4725 1.62, also never worked. As it isn't clear exactly what this means, and
4726 clearly it isn't a major issue, I have abolished the feature by treating it
4727 as "timeout", and writing a warning to the main and panic logs.
4729 36. The display of retry rules for -brt wasn't always showing the error code
4732 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4733 timeout_connect_A, timeout_connect_MX.
4735 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4736 to the empty sender.
4738 39. The daemon was not analysing the content of -oX till after it had closed
4739 stderr and disconnected from the controlling terminal. This meant that any
4740 syntax errors were only noted on the panic log, and the return code from
4741 the command was 0. By re-arranging the code a little, I've made the
4742 decoding happen first, so such errors now appear on stderr, and the return
4743 code is 1. However, the actual setting up of the sockets still happens in
4744 the disconnected process, so errors there are still only recorded on the
4747 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4748 connections (as happens on some IP stacks) was logged at start up time as
4749 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4750 it from "IPv6 and IPv4", which means that two separate sockets are being
4753 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4754 says whether AF_INET or AF_INET6 was passed as an argument.
4756 42. Exiwhat output was messed up when time zones were included in log
4759 43. Exiwhat now gives more information about the daemon's listening ports,
4760 and whether -tls-on-connect was used.
4762 44. The "port" option of the smtp transport is now expanded.
4764 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4765 silently ignored. Now an error message is written to the main and panic
4768 46. There's a new ACL modifier called "logwrite" which writes to a log file
4769 as soon as it is encountered.
4771 47. Added $local_user_uid and $local_user_gid at routing time.
4773 48. Exim crashed when trying to verify a sender address that was being
4776 49. Exim was recognizing only a space character after ".include". It now also
4777 recognizes a tab character.
4779 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4780 extracting the relevant information from the specification. The man page no
4781 longer contains scrambled data for the -d option, and I've added a section
4782 at the front about calling Exim under different names.
4784 51. Added "extra_headers" argument to the "mail" command in filter files.
4786 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4789 53. Installed eximstats 1.29.
4791 54. Added transport_filter_timeout as a generic transport option.
4793 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4794 Cc: header lines. This was required by RFC 822, but it not required by RFC
4797 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4798 incoming messages that did not have them. Now it does so only if the
4799 message originates locally, that is, if there is no associated remote host
4800 address. When Resent- header lines are present, this applies to the Resent-
4801 lines rather than the non-Resent- lines.
4803 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4804 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4806 58. Messages for configuration errors now include the name of the main
4807 configuration file - useful now that there may be more than one file in a
4808 list (.included file names were always shown).
4810 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4811 for those rare installations that do not start the daemon as root or run it
4812 setuid root. I've cut out the call to initgroups() if the daemon is not
4815 60. The Exim user and group can now be bound into the binary as text strings
4816 that are looked up at the start of Exim's processing.
4818 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4820 62. Added $mailstore_basename variable.
4822 63. Installed patch to sieve.c from Michael Haardt.
4824 64. When Exim failed to open the panic log after failing to open the main log,
4825 the original message it was trying to log was written to stderr and debug
4826 output, but if they were not available (the usual case in production), it
4827 was lost. Now it is written to syslog before the two lines that record the
4828 failures to open the logs.
4830 65. Users' Exim filters run in subprocesses under the user's uid. It is
4831 possible for a "deliver" command or an alias in a "personal" command to
4832 provoke an address rewrite. If logging of address rewriting is configured,
4833 this fails because the process is not running as root or exim. There may be
4834 a better way of dealing with this, but for the moment (because 4.30 needs
4835 to be released), I have disabled address rewrite logging when running a
4836 filter in a non-root, non-exim process.
4842 1. The buildconfig auxiliary program wasn't quoting the value set for
4843 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4844 not defined. This bug was masked in 4.22 by the effect that was fixed in
4847 2. Some messages that were rejected after a message id was allocated were
4848 shown as "incomplete" by exigrep. It no longer does this for messages that
4849 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4851 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4852 have allow_domain_literals set, the ID did not get logged in the <= line.
4853 Domain literals are now always recognized in Message-ID: header lines.
4855 4. The first argument for a ${extract expansion item is the key name or field
4856 number. Leading and trailing spaces in this item were not being ignored,
4857 causing some misleading effects.
4859 5. When deliver_drop_privilege was set, single queue runner processes started
4860 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4861 same command in the process it spins off) were not dropping privilege.
4863 6. When the daemon running as "exim" started a queue runner, it always
4864 re-executed Exim in the spun-off process. This is a waste of effort when
4865 deliver_drop_privilege is set. The new process now just calls the
4866 queue-runner function directly.
4872 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4875 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4876 ignored. Though the use of -oP was forcing the writing of a pid file, it
4877 was always written to the default place.
4879 3. If the message "no IP address found for host xxxx" is generated during
4880 incoming verification, it is now followed by identification of the incoming
4881 connection (so you can more easily find what provoked it).
4883 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4885 5. Added some features to "harden" Exim a bit more against certain attacks:
4887 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4888 be put in Local/Makefile. This is like the never_users runtime option,
4889 but it cannot be overridden. The default setting is "root".
4891 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4892 prefix string with which any file named in a -C command line option
4895 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4896 is retained for -C and -D only if the caller of Exim is root. Without
4897 it, the exim user may also use -C and -D and retain privilege.
4899 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4900 command line option is disabled.
4902 6. Macro names set by the -D option must start with an upper case letter, just
4903 like macro names defined in the configuration file.
4905 7. Added "dereference=" facility to LDAP.
4907 8. Two instances of the typo "uknown" in the source files are fixed.
4909 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4910 the Configure-Makefile script screwed up while processing it.
4912 10. Incorporated PCRE 4.4.
4914 11. The SMTP synchronization check was not operating right at the start of an
4915 SMTP session. For example, it could not catch a HELO sent before the client
4916 waited for the greeting. There is now a check for outstanding input at the
4917 point when the greeting is written. Because of the duplex, asynchronous
4918 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4919 way, but not yet received, when the check is performed.
4921 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4922 on TCP/IP sockets, because this apparently causes some broken clients to
4925 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4926 unchanged) from the Cygwin maintainer.
4928 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4929 was supported instead of testing for mbx. Effectively a typo.
4931 15. The spa authenticator server code was not checking that the input it
4932 received was valid base64.
4934 16. The debug output line for the "set" modifier in ACLs was not showing the
4935 name of the variable that was being set.
4937 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4939 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4940 lookup on its own. Something else was needed to trigger the lookup. For
4941 example, a match in host_lookup or the need for a host name in a host list.
4942 Now, if $sender_host_name is referenced and the host name has not yet been
4943 looked up, a lookup is performed. If the lookup fails, the variable remains
4944 empty, and $host_lookup_failed is set to "1".
4946 19. Added "eqi" as a case-independent comparison operator.
4948 20. The saslauthd authentication condition could segfault if neither service
4949 nor realm was specified.
4951 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4952 error message that was logged was misleading, and incoming SMTP
4953 connections were dropped. The message is now more accurate, and temporary
4954 errors are given to SMTP connections.
4956 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4957 (except RSET) with a 421 error, until QUIT is received. However, it was
4958 failing to send a response to QUIT.
4960 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4961 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4962 ACL is now run after the helo_try_verify_hosts code.
4964 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4965 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4966 case-independent, so other case variants are also recognized). Apparently
4967 some systems use these upper case variants.
4969 25. If more than two messages were waiting for the same host, and a transport
4970 filter was specified for the transport, Exim sent two messages over the
4971 same TCP/IP connection, and then failed with "socket operation on non-
4972 socket" when it tried to send the third.
4974 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4976 27. The extern definition of crypt16() in expand.c was not being excluded when
4977 the OS had its own crypt16() function.
4979 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4980 as a preferred synonym for return_size_limit, both as an option and as an
4983 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4985 30. Changed the default configuration ACL to relax the local part checking rule
4986 for addresses that are not in any local domains. For these addresses,
4987 slashes and pipe symbols are allowed within local parts, but the sequence
4988 /../ is explicitly forbidden.
4990 31. SPA server authentication was not clearing the challenge buffer before
4993 32. log_message in a "warn" ACL statement was writing to the reject log as
4994 well as to the main log, which contradicts the documentation and doesn't
4995 seem right (because no rejection is happening). So I have stopped it.
4997 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4998 However, I am unable to do any testing of this.
5000 34. Fixed an infelicity in the appendfile transport. When checking directories
5001 for a mailbox, to see if any needed to be created, it was accidentally
5002 using path names with one or more superfluous leading slashes; tracing
5003 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5005 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5006 contents are added to the log line that is written for every discarded
5007 recipient. (Previously a log_message setting was ignored.)
5009 36. The ${quote: operator now quotes the string if it is empty.
5011 37. The install script runs exim in order to find its version number. If for
5012 some reason other than non-existence or emptiness, which it checks, it
5013 could not run './exim', it was installing it with an empty version number,
5014 i.e. as "exim-". This error state is now caught, and the installation is
5017 38. An argument was missing from the function that creates an error message
5018 when Exim fails to connect to the socket for saslauthd authentication.
5019 This could cause Exim to crash, or give a corrupted message.
5021 39. Added isip, isip4, and isip6 to ${if conditions.
5023 40. The ACL variables $acl_xx are now saved with the message, and can be
5024 accessed later in routers, transports, and filters.
5026 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5027 strings in the file are not string-expanded.
5029 42. If a MAIL command specified a SIZE value that was too large to fit into an
5030 int variable, the check against message_size_limit failed. Such values are
5031 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5032 day this will have to be increased, but I don't think I want to be around
5033 when emails are that large.
5040 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5041 iconv() is not standard in FreeBSD.
5043 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5044 IPv6 enabled. The observed symptom was a segmentation fault on return from
5045 the function os_common_find_running_interfaces() in src/os.c.
5047 3. In the check_special_case() function in daemon.c I had used "errno" as an
5048 argument name, which causes warnings on some systems. This was basically a
5049 typo, since it was named "eno" in the comments!
5051 4. The code that waits for the clock to tick (at a resolution of some fraction
5052 of a second) so as to ensure message-id uniqueness was always waiting for
5053 at least one whole tick, when it could have waited for less. [This is
5054 almost certainly not relevant at current processor speeds, where it is
5055 unlikely to ever wait at all. But we try to future-proof.]
5057 5. The function that sleeps for a time interval that includes fractions of a
5058 second contained a race. It did not block SIGALRM between setting the
5059 timer, and suspending (a couple of lines later). If the interval was short
5060 and the sigsuspend() was delayed until after it had expired, the suspension
5061 never ended. On busy systems this could lead to processes getting stuck for
5064 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5065 process, before it forks any delivery processes. The open lookup caching
5066 mechanism meant that the open file or database connection was passed into
5067 the delivery process. The problem was that delivery processes always tidy
5068 up cached lookup data. This could cause a problem for the next delivery
5069 process started by the queue runner, because the external queue runner
5070 process does not know about the closure. So the next delivery process
5071 still has data in the lookup cache. In the case of a file lookup, there was
5072 no problem because closing a file descriptor in a subprocess doesn't affect
5073 the parent. However, if the lookup was caching a connection to a database,
5074 the connection was closed, and the second delivery process was likely to
5075 see errors such as "PGSQL: query failed: server closed the connection
5076 unexpectedly". The problem has been fixed by closing all cached lookups
5077 in a queue runner before running a delivery process.
5079 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5080 seem to have the "const" qualifier which it has on other OS. I've
5083 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5084 *of the same type* (client or server) with the same public name that an
5085 error should be diagnosed.
5087 9. When Exim looked up a host name for an IP address, but failed to find the
5088 original IP address when looking up the host name (a safety check), it
5089 output the message "<ip address> does not match any IP for NULL", which was
5090 confusing, to say the least. The bug was that the host name should have
5091 appeared instead of "NULL".
5093 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5094 user that is built into the binary, and the -C or -D options is used, root
5095 privilege is dropped before the configuration file is read. In addition,
5096 logging is switched to stderr instead of the normal log files. If the
5097 configuration then re-defines the Exim user, the unprivileged environment
5098 is probably not what is expected, so Exim logs a panic warning message (but
5101 However, if deliver_drop_privilege is set, the unprivileged state may well
5102 be exactly what is intended, so the warning has been cut out in that case,
5103 and Exim is allowed to try to write to its normal log files.
5109 1. smtp_return_error_details was not giving details for temporary sender
5110 or receiver verification errors.
5112 2. Diagnose a configuration error if two authenticators have the same public
5115 3. Exim used not to create the message log file for a message until the first
5116 delivery attempt. This could be confusing when incoming messages were held
5117 for policy or load reasons. The message log file is now created at the time
5118 the message is received, and an initial "Received" line is written to it.
5120 4. The automatically generated man page for command line options had a minor
5121 bug that caused no ill effects; however, a more serious problem was that
5122 the procedure for building the man page automatically didn't always
5123 operate. Consequently, release 4.20 contains an out-of-date version. This
5124 shouldn't happen again.
5126 5. When building Exim with embedded Perl support, the script that builds the
5127 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5128 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5130 6. The freeze_tell option was not being used for messages that were frozen on
5131 arrival, either by an ACL or by local_scan().
5133 7. Added the smtp_incomplete_transaction log selector.
5135 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5136 was accepting AUTH without a new EHLO.
5138 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5139 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5142 10. Logging of TCP/IP connections (when configured) now happens in the main
5143 daemon process instead of the child process, so that the TCP/IP connection
5144 count is more accurate (but it can never be perfect).
5146 11. The use of "drop" in a nested ACL was not being handled correctly in the
5147 outer ACL. Now, if condition failure induced by the nested "drop" causes
5148 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5149 or "require"), the connection is dropped.
5151 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5152 that yield "discard" can now be used with an "accept" or a "discard" verb,
5153 but an error is generated for any others (because I can't see a useful way
5154 to define what should happen).
5156 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5157 are now processed in the same way as lines in the Exim configuration file.
5158 In particular, continuation lines are supported.
5160 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5162 15. Added -ti meaning -t -i.
5164 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5165 domains, and warn by logging if others are found.
5167 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5168 returned from GIFCONF when Exim is trying to find the list of interfaces on
5169 a host. The code in os.c has been modified to copy each ifreq to an aligned
5170 structure in all cases.
5172 Also, in some cases, the returned ifreq's were being copied to a 'struct
5173 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5174 means the last couple of bytes of an IPv6 address could be chopped if the
5175 ifreq contained only a normal sockaddr (14 bytes storage).
5177 18. Named domain lists were not supported in the hosts_treat_as_local option.
5178 An entry such as +xxxx was not recognized, and was treated as a literal
5181 19. Ensure that header lines added by a DATA ACL are included in the reject log
5182 if the ACL subsequently rejects the message.
5184 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5185 MD5 (which is deprecated).
5187 21. When testing a filter file using -bf, Exim was writing a message when it
5188 took the sender from a "From " line in the message, but it was not doing so
5189 when it took $return_path from a Return-Path: header line. It now does.
5191 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5192 with a valid header line field name (a series of printing characters
5193 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5195 23. Changed "disc" in the source to "disk" to conform to the documentation and
5196 the book and for uniformity.
5198 24. Ignore Sendmail's -Ooption=value command line item.
5200 25. When execve() failed while trying to run a command in a pipe transport,
5201 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5202 could be confused with a return value of 69 from the command itself. This
5203 has been changed to 127, the value the shell returns if it is asked to run
5204 a non-existent command. The wording for the related log line suggests a
5205 non-existent command as the problem.
5207 26. If received_header_text expands to an empty string, do not add a Received:
5208 header line to the message. (Well, it adds a token one on the spool, but
5209 marks it "old" so that it doesn't get used or transmitted.)
5211 27. Installed eximstats 1.28 (addition of -nt option).
5213 28. There was no check for failure on the call to getsockname() in the daemon
5214 code. This can fail if there is a shortage of resources on the system, with
5215 ENOMEM, for example. A temporary error is now given on failure.
5217 29. Contrary to the C standard, it seems that in some environments, the
5218 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5219 program. Exim now does this explicitly; it affects the formatting of
5220 timestamps using strftime().
5222 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5223 complaints. I've now initialized all the variables, to avoid this.
5225 32. Header lines added by a system filter were not being "seen" during
5226 transport-time rewrites.
5228 33. The info_callback() function passed to OpenSSL is set up with type void
5229 (*)(SSL *, int, int), as described somewhere. However, when calling the
5230 function (actually a macro) that sets it up, the type void(*)() is
5231 expected. I've put in a cast to prevent warnings from picky compilers.
5233 34. If a DNS black list lookup found a CNAME record, but there were no A
5234 records associated with the domain it pointed at, Exim crashed.
5236 35. If a DNS black list lookup returned more than one A record, Exim ignored
5237 all but the first. It now scans all returned addresses if a particular IP
5238 value is being sought. In this situation, the contents of the
5239 $dnslist_value variable are a list of all the addresses, separated by a
5242 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5243 to accept a host name and all its aliases if the forward lookup for any of
5244 them yielded the IP address of the incoming connection. Now it accepts only
5245 those names whose forward lookup yields the correct IP address. Any other
5246 names are discarded. This closes a loophole whereby a rogue DNS
5247 administrator could create reverse DNS records to break through a
5248 wildcarded host restriction in an ACL.
5250 37. If a user filter or a system filter that ran in a subprocess used any of
5251 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5252 the wrong values were passed to the pipe command ($thisaddress had the
5253 value of $0, $0 had the value of $1, etc). This bug was introduced by
5254 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5256 38. Improved the line breaking for long SMTP error messages from ACLs.
5257 Previously, if there was no break point between 40 and 75 characters, Exim
5258 left the rest of the message alone. Two changes have been made: (a) I've
5259 reduced the minimum length to 35 characters; (b) if it can't find a break
5260 point between 35 and 75 characters, it looks ahead and uses the first one
5261 that it finds. This may give the occasional overlong line, but at least the
5262 remaining text gets split now.
5264 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5265 file descriptors might be low, and that setting 1000 would always raise it.
5266 It turns out that in some environments, the limit is already over 1000 and
5267 that lowering it causes trouble. So now Exim takes care not to decrease it.
5269 40. When delivering a message, the value of $return_path is set to $sender_
5270 address at the start of routing (routers may change the value). By an
5271 oversight, this default was not being set up when an address was tested by
5272 -bt or -bv, which affected the outcome if any router or filter referred to
5275 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5276 log only when "message" or "log_message" are set. However, if one of the
5277 conditions was an address verification, or a call to a nested ACL, the
5278 messages generated by the underlying test were being passed through. This
5279 no longer happens. The underlying message is available in $acl_verify_
5280 message for both "message" and "log_message" expansions, so it can be
5281 passed through if needed.
5283 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5284 new expansion $bh_ to give the encoded byte string without charset
5285 translation. Translation happens only if iconv() is available; HAVE_ICONV
5286 indicates this at build time. HEADERS_CHARSET gives the charset to
5287 translate to; headers_charset can change it in the configuration, and
5288 "headers charset" can change it in an individual filter file.
5290 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5291 that creates RFC 2047 encoded "words" labels them as that charset instead
5292 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5293 expansion operator; (ii) when Exim creates a From: line for a local
5294 message; (iii) when a header line is rewritten to include a "phrase" part.
5296 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5297 buggy, causing it to skip the first lines of messages whose message ID
5298 ended in 'D'. This would not have bitten before Exim release 4.14, because
5299 message IDs were unlikely to end in 'D' before then. The effect was to have
5300 incorrect size information for certain domains.
5302 45. #include "config.h" was missing at the start of the crypt16.c module. This
5303 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5306 46. If there was a timeout during a "random" callout check, Exim treated it as
5307 a failure of the random address, and carried on sending RSET and the real
5308 address. If the delay was just some slowness somewhere, the response to the
5309 original RCPT would be taken as a response to RSET and so on, causing
5310 mayhem of various kinds.
5312 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5313 when I implemented it. It didn't allow for the fact that some option values
5314 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5315 the right test for positive values.
5317 48. Domain names in DNS records are case-independent. Exim always looks them up
5318 in lower case. Some resolvers return domain names in exactly the case they
5319 appear in the zone file, that is, they may contain uppercase letters. Not
5320 all resolvers do this - some return always lower case. Exim was treating a
5321 change of case by a resolver as a change of domain, similar to a widening
5322 of a domain abbreviation. This triggered its re-routing code and so it was
5323 trying to route what was effectively the same domain again. This normally
5324 caused routing to fail (because the router wouldn't handle the domain
5325 twice). Now Exim checks for this case specially, and just changes the
5326 casing of the domain that it ultimately uses when it transmits the message
5329 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5332 50. If a filter generated a file delivery with a non-absolute name (possible if
5333 no home directory exists for the router), the forbid_file option was not
5336 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5337 the existing equality matching.
5339 52. Exim was using ints instead of ino_t variables in some places where it was
5340 dealing with inode numbers.
5342 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5343 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5344 called TMPDIR, and if it finds it is different, it changes its value.
5346 54. The smtp_printf() function is now made available to local_scan() so
5347 additional output lines can be written before returning. There is also an
5348 smtp_fflush() function to enable the detection of a dropped connection.
5349 The variables smtp_input and smtp_batched_input are exported to
5352 55. Changed the default runtime configuration: the message "Unknown user"
5353 has been removed from the ACL, and instead placed on the localuser router,
5354 using the cannot_route_message feature. This means that any verification
5355 failures that generate their own messages won't get overridden. Similarly,
5356 the "Unrouteable address" message that was in the ACL for unverifiable
5357 relay addresses has also been removed.
5359 56. Added hosts_avoid_esmtp to the smtp transport.
5361 57. The exicyclog script was not checking for the esoteric option
5362 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5363 will work only if exicyclog is run under the appropriate euid.
5365 58. Following a discussion on the list, the rules by which Exim recognises line
5366 endings on incoming messages have been changed. The -dropcr and drop_cr
5367 options are now no-ops, retained only for backwards compatibility. The
5368 following line terminators are recognized: LF CRLF CR. However, special
5369 processing applies to CR:
5371 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5372 nor a local message in the state where . is a terminator.
5374 (ii) If a bare CR is encountered in a header line, an extra space is added
5375 after the line terminator so as not to end the header. The reasoning
5376 behind this is that bare CRs in header lines are most likely either
5377 to be mistakes, or people trying to play silly games.
5379 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5380 was being incorrectly given as 18 bytes larger than it should have been.
5381 This is a VOB (very old bug).
5383 60. This may never have affected anything current, but just in case it has:
5384 When the local host is found other than at the start of a list of hosts,
5385 the local host, those with the same MX, and any that follow, are discarded.
5386 When the list in question was part of a longer list of hosts, the following
5387 hosts (not currently being processed) were also being discarded. This no
5388 longer happens. I'm not sure if this situation could ever has previously
5391 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5394 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5395 header. This is something that is recommended in a new Internet Draft, and
5396 is something that is documented as being done by Sendmail. There are two
5397 possible values. For messages generated by the autoreply transport, Exim
5400 Auto-Submitted: auto-replied
5402 whereas for all other generated messages (e.g. bounces) it adds
5404 Auto-Submitted: auto-generated
5406 63. The "personal" condition in filters now includes a test for the
5407 Auto-Submitted: header. If it contains the string "auto-" the message it
5408 not considered personal.
5410 64. Added rcpt_include_affixes as a generic transport option.
5412 65. Added queue_only_override (default true).
5414 66. Added the syslog_duplication option.
5416 67. If what should have been the first header line of a message consisted of
5417 a space followed by a colon, Exim was mis-interpreting it as a header line.
5418 It isn't of course - it is syntactically invalid and should therefore be
5419 treated as the start of the message body. The misbehaviour could have
5420 caused a number of strange effects, including loss of data in subsequent
5421 header lines, and spool format errors.
5423 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5424 client host had authenticated. This control can now be exercised by an ACL
5425 for more flexibility.
5427 69. By default, callouts do not happen when testing with -bh. There is now a
5428 variant, -bhc, which does actually run the callout code, including
5429 consulting and updating the callout cache.
5431 70. Added support for saslauthd authentication, courtesy of Alexander
5434 71. If statvfs() failed on the spool or log directories while checking their
5435 size for availability, Exim confusingly gave the error "space shortage".
5436 Furthermore, in debugging mode it crashed with a floating point exception.
5437 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5438 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5439 serious problem, Exim now writes to the main and panic logs when this
5440 happens, with details of the failure. It then refuses to accept the
5441 incoming message, giving the message "spool directory problem" or "log
5442 directory problem" with a 421 code for SMTP messages.
5444 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5445 0, 1, and 2 exist, because some OS complain for execs without them (see
5446 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5447 descriptors. However, the code omitted to check that the open succeeded,
5448 causing mysterious errors if for some reason the permissions on /dev/null
5449 got screwed. Now Exim writes a message to the main and panic logs, and
5450 bombs out if it can't open /dev/null.
5452 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5453 interact so that it is all more flexible. It is supposed to remain
5454 backwards compatible. Also added extra_local_interfaces.
5456 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5457 to bomb out with an assertion failure - to the client this appears as a
5458 connection drop. This problem occurs in the part of the code that was taken
5459 from the Samba project. Fortunately, the assertion is in a very simple
5460 function, so I have fixed this by reproducing the function inline in the
5461 one place where it is called, and arranging for authentication to fail
5462 instead of killing the process with assert().
5464 75. The SPA client code was not working when the server requested OEM rather
5465 than Unicode encoding.
5467 76. Added code to make require_files with a specific uid setting more usable in
5468 the case where statting the file as root fails - usually a non-root-mounted
5469 NFS file system. When this happens and the failure is EACCES, Exim now
5470 forks a subprocess and does the per-uid checking as the relevant uid.
5472 77. Added process_log_path.
5474 78. If log_file_path was not explicitly set, a setting of check_log_space or
5475 check_log_inodes was ignored.
5477 79. If a space check for the spool or log partitions fails, the incident is now
5478 logged. Of course, in the latter case the data may get lost...
5480 80. Added the %p formatting code to string_format() so that it can be used to
5481 print addresses in debug_print(). Adjusted all the address printing in the
5482 debugging in store.c to use %p rather than %d.
5484 81. There was a concern that a line of code in smtp_in.c could overflow a
5485 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5486 initially expressed, the concern was not well-founded, because trailing
5487 spaces are removed early. However, if the trailing spaces were followed by
5488 a NULL, they did not get removed, so the overflow was possible. Two fixes
5491 (a) I re-wrote the offending code in a cleaner fashion.
5492 (b) If an incoming SMTP command contains a NULL character, it is rejected
5495 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5496 runs initgroups(), so that if the Exim user is in any additional groups,
5497 they will be used during message reception.
5503 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob