1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
48 PP/01 Solaris build fix for Oracle's LDAP libraries.
49 Bugzilla 1109, patch from Stephen Usher.
51 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
53 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
56 TF/02 Fix a couple more cases where we did not log the error message
57 when unlink() failed. See also change 4.74-TF/03.
59 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
60 lock up or crash if it happened to be inside a call to libc when it
61 got a SIGUSR1 from exiwhat.
63 The SIGUSR1 handler appends the current process status to the process
64 log which is later printed by exiwhat. It used to use the general
65 purpose logging code to do this, but several functions it calls are
68 The new output code in the SIGUSR1 handler is specific to the process
69 log, and simple enough that it's easy to inspect for signal safety.
70 Removing some special cases also simplifies the general logging code.
71 Removing the spurious timestamps from the process log simplifies
74 TF/04 Improved ratelimit ACL condition.
76 The /noupdate option has been deprecated in favour of /readonly which
77 has clearer semantics. The /leaky, /strict, and /readonly update modes
78 are mutually exclusive. The update mode is no longer included in the
79 database key; it just determines when the database is updated. (This
80 means that when you upgrde Exim will forget old rate measurements.)
82 Exim now checks that the per_* options are used with an update mode that
83 makes sense for the current ACL. For example, when Exim is processing a
84 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
85 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
86 must specify per_mail/readonly. If you omit the update mode it defaults to
87 /leaky where that makes sense (as before) or /readonly where required.
89 The /noupdate option is now undocumented but still supported for
90 backwards compatibility. It is equivalent to /readonly except that in
91 ACLs where /readonly is required you may specify /leaky/noupdate or
92 /strict/noupdate which are treated the same as /readonly.
94 A useful new feature is the /count= option. This is a generalization
95 of the per_byte option, so that you can measure the throughput of other
96 aggregate values. For example, the per_byte option is now equivalent
97 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
99 The per_rcpt option has been generalized using the /count= mechanism
100 (though it's more complicated than the per_byte equivalence). When it is
101 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
102 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
103 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
104 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
105 Note that using per_rcpt with a non-readonly update mode in more than
106 one ACL will cause the recipients to be double-counted. (The per_mail
107 and per_byte options don't have this problem.)
109 The handling of very low rates has changed slightly. If the computed rate
110 is less than the event's count (usually one) then this event is the first
111 after a long gap. In this case the rate is set to the same as this event's
112 count, so that the first message of a spam run is counted properly.
114 The major new feature is a mechanism for counting the rate of unique
115 events. The new per_addr option counts the number of different
116 recipients that someone has sent messages to in the last time period. It
117 behaves like per_rcpt if all the recipient addresses are different, but
118 duplicate recipient addresses do not increase the measured rate. Like
119 the /count= option this is a general mechanism, so the per_addr option
120 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
121 example, measure the rate that a client uses different sender addresses
122 with the options per_mail/unique=$sender_address. There are further
123 details in the main documentation.
125 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
127 TF/06 Removed a few PCRE remnants.
129 TF/07 Automatically extract Exim's version number from tags in the git
130 repository when doing development or release builds.
132 PP/02 Raise smtp_cmd_buffer_size to 16kB.
133 Bugzilla 879. Patch from Paul Fisher.
135 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
136 Heavily based on revision 40f9a89a from Simon Arlott's tree.
139 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
141 PP/05 Variable $av_failed, true if the AV scanner deferred.
142 Bugzilla 1078. Patch from John Horne.
144 PP/06 Stop make process more reliably on build failure.
145 Bugzilla 1087. Patch from Heiko Schlittermann.
147 PP/07 Make maildir_use_size_file an _expandable_ boolean.
148 Bugzilla 1089. Patch from Heiko Schlittermann.
150 PP/08 Handle ${run} returning more data than OS pipe buffer size.
151 Bugzilla 1131. Patch from Holger Weiß.
153 PP/09 Handle IPv6 addresses with SPF.
154 Bugzilla 860. Patch from Wolfgang Breyha.
156 PP/10 GnuTLS: support TLS 1.2 & 1.1.
158 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
161 PP/11 match_* no longer expand right-hand-side by default.
162 New compile-time build option, EXPAND_LISTMATCH_RHS.
163 New expansion conditions, "inlist", "inlisti".
165 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
167 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
169 PP/14 fix log_write() format string regression from TF/03.
170 Bugzilla 1152. Patch from Dmitry Isaikin.
176 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
178 PP/02 Harmonised TLS library version reporting; only show if debugging.
179 Layout now matches that introduced for other libraries in 4.74 PP/03.
181 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
183 PP/04 New "dns_use_edns0" global option.
185 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
188 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
189 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
191 TK/01 Updated PolarSSL code to 0.14.2.
192 Bugzilla 1097. Patch from Andreas Metzler.
194 PP/07 Catch divide-by-zero in ${eval:...}.
197 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
200 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
201 format-string attack -- SECURITY: remote arbitrary code execution.
203 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
204 time unintentionally subject to list matching rules, letting the header
205 cause arbitrary Exim lookups (of items which can occur in lists, *not*
206 arbitrary string expansion). This allowed for information disclosure.
208 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
209 INT_MIN/-1 -- value coerced to INT_MAX.
215 NM/01 Workround for PCRE version dependency in version reporting
218 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
219 This fixes portability to compilers other than gcc, notably
220 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
222 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
223 makefiles for portability to HP-UX and POSIX correctness.
225 PP/01 Permit LOOKUP_foo enabling on the make command-line.
226 Also via indented variable definition in the Makefile.
227 (Debugging by Oliver Heesakkers).
229 PP/02 Restore caching of spamd results with expanded spamd_address.
230 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
232 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
233 Improves build reliability. Fix from: Frank Elsner
235 NM/02 Fix wide character breakage in the rfc2047 coding
236 Fixes bug 1064. Patch from Andrey N. Oktyabrski
238 NM/03 Allow underscore in dnslist lookups
239 Fixes bug 1026. Patch from Graeme Fowler
241 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
242 Code patches from Adam Ciarcinski of NetBSD.
244 NM/04 Fixed exiqgrep to cope with mailq missing size issue
247 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
248 is logged, to avoid truncation. Patch from John Horne.
250 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
251 Patch from Jakob Hirsch.
253 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
254 SQL string expansion failure details.
255 Patch from Andrey Oktyabrski.
257 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
258 Patch from Simon Arlott.
260 PP/09 New lookups functionality failed to compile on old gcc which rejects
261 extern declarations in function scope.
262 Patch from Oliver Fleischmann
264 PP/10 Use sig_atomic_t for flags set from signal handlers.
265 Check getgroups() return and improve debugging.
266 Fixed developed for diagnosis in bug 927 (which turned out to be
269 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
270 Patch from Mark Zealey.
272 PP/12 Bugzilla 1056: Improved spamd server selection.
273 Patch from Mark Zealey.
275 PP/13 Bugzilla 1086: Deal with maildir quota file races.
276 Based on patch from Heiko Schlittermann.
278 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
279 Patch from Uwe Doering, sign-off by Michael Haardt.
281 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
282 variable declaration deep within a block. Bug and patch from
285 PP/15 lookups-Makefile IRIX compatibilty coercion.
287 PP/16 Make DISABLE_DKIM build knob functional.
289 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
290 Patch by Simon Arlott
292 TF/03 Fix valgrind.h portability to C89 compilers that do not support
293 variable argument macros. Our copy now differs from upstream.
299 TF/01 Failure to get a lock on a hints database can have serious
300 consequences so log it to the panic log.
302 TF/02 Log LMTP confirmation messages in the same way as SMTP,
303 controlled using the smtp_confirmation log selector.
305 TF/03 Include the error message when we fail to unlink a spool file.
307 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
308 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
309 for maintaining out-of-tree patches for some time.
311 PP/01 Bugzilla 139: Documentation and portability issues.
312 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
313 Handle per-OS dynamic-module compilation flags.
315 PP/02 Let /dev/null have normal permissions.
316 The 4.73 fixes were a little too stringent and complained about the
317 permissions on /dev/null. Exempt it from some checks.
318 Reported by Andreas M. Kirchwitz.
320 PP/03 Report version information for many libraries, including
321 Exim version information for dynamically loaded libraries. Created
322 version.h, now support a version extension string for distributors
323 who patch heavily. Dynamic module ABI change.
325 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
326 privilege escalation vulnerability whereby the Exim run-time user
327 can cause root to append content of the attacker's choosing to
330 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
333 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
334 If dropping privileges for untrusted macros, we disabled normal logging
335 on the basis that it would fail; for the Exim run-time user, this is not
336 the case, and it resulted in successful deliveries going unlogged.
337 Fixed. Reported by Andreas Metzler.
343 PP/01 Date: & Message-Id: revert to normally being appended to a message,
344 only prepend for the Resent-* case. Fixes regression introduced in
345 Exim 4.70 by NM/22 for Bugzilla 607.
347 PP/02 Include check_rfc2047_length in configure.default because we're seeing
348 increasing numbers of administrators be bitten by this.
350 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
352 PP/03 Bugzilla 994: added openssl_options main configuration option.
354 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
356 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
358 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
360 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
361 without a peer certificate, leading to a segfault because of an
362 assumption that peers always have certificates. Be a little more
363 paranoid. Problem reported by Martin Tscholak.
365 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
366 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
367 NB: ClamAV planning to remove STREAM in "middle of 2010".
368 CL also introduces -bmalware, various -d+acl logging additions and
369 more caution in buffer sizes.
371 PP/09 Implemented reverse_ip expansion operator.
373 PP/10 Bugzilla 937: provide a "debug" ACL control.
375 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
377 PP/12 Bugzilla 973: Implement --version.
379 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
381 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
383 PP/15 Bugzilla 816: support multiple condition rules on Routers.
385 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
386 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
387 ignore trailing whitespace.
389 JJ/02 prevent non-panic DKIM error from being sent to paniclog
391 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
394 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
395 Notification from Dr Andrew Aitchison.
397 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
398 ExtendedDetectionInfo response format.
399 Notification from John Horne.
401 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
404 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
405 XSL and documented dependency on system catalogs, with examples of how
408 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
411 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
412 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
413 configuration file which is writeable by the Exim user or group.
415 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
416 of configuration files to cover files specified with the -C option if
417 they are going to be used with root privileges, not just the default
420 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
421 option (effectively making it always true).
423 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
424 files to be used while preserving root privileges.
426 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
427 that rogue child processes cannot use them.
429 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
430 run-time user, instead of root.
432 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
433 Exim run-time user without dropping privileges.
435 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
436 result string, instead of calling string_vformat() twice with the same
439 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
440 for other users. Others should always drop root privileges if they use
441 -C on the command line, even for a whitelisted configure file.
443 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
445 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
451 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
452 $data_path, and $header_path variables; fixed documentation bugs and
455 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
456 exipick to access non-standard spools, including the "frozen" queue
459 NM/01 Bugzilla 965: Support mysql stored procedures.
460 Patch from Alain Williams
462 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
464 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
465 Patch from Andreas Metzler
467 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
468 Patch from Kirill Miazine
470 NM/05 Bugzilla 671: Added umask to procmail example.
472 JJ/03 installed exipick 20100323.0, fixing doc bug
474 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
475 directory. Notification and patch from Dan Rosenberg.
477 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
479 TK/02 Improve log output when DKIM signing operation fails.
481 MH/01 Treat the transport option dkim_domain as a colon separated
482 list, not as a single string, and sign the message with each element,
483 omitting multiple occurences of the same signer.
485 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
486 Bugzilla 985, 986. Patch by Simon Arlott
488 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
489 Patch by Simon Arlott
491 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
492 MBX locking. Notification from Dan Rosenberg.
498 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
500 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
502 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
504 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
506 NM/04 Bugzilla 915: Flag broken perl installation during build.
512 TK/01 Added patch by Johannes Berg that expands the main option
513 "spamd_address" if it starts with a dollar sign.
515 TK/02 Write list of recipients to X-Envelope-Sender header when building
516 the mbox-format spool file for content scanning (suggested by Jakob
519 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
520 (http://www.dcc-servers.net/) support via dccifd. Activated by
521 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
523 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
524 by Mark Daniel Reidel <mr@df.eu>.
526 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
527 When building exim an external PCRE library is now needed -
528 PCRE is a system library on the majority of modern systems.
529 See entry on PCRE_LIBS in EDITME file.
531 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
532 conversation. Added nologin parameter to request.
533 Patch contributed by Kirill Miazine.
535 TF/01 Do not log submission mode rewrites if they do not change the address.
537 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
539 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
540 log files in place. Contributed by Roberto Lima.
542 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
544 TF/03 Bugzilla 615: When checking the local_parts router precondition
545 after a local_part_suffix or local_part_prefix option, Exim now
546 does not use the address's named list lookup cache, since this
547 contains cached lookups for the whole local part.
549 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
550 Robert Millan. Documentation is in experimental-spec.txt.
552 TF/04 Bugzilla 668: Fix parallel build (make -j).
554 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
556 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
557 Patch provided by Jan Srzednicki.
559 TF/05 Leading white space used to be stripped from $spam_report which
560 wrecked the formatting. Now it is preserved.
562 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
563 that they are available at delivery time.
565 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
567 TF/08 TLS error reporting now respects the incoming_interface and
568 incoming_port log selectors.
570 TF/09 Produce a more useful error message if an SMTP transport's hosts
571 setting expands to an empty string.
573 NM/06 Bugzilla 744: EXPN did not work under TLS.
574 Patch provided by Phil Pennock.
576 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
577 Patch provided by Richard Godbee.
579 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
580 acl_smtp_notquit, added index entry.
582 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
583 Patch provided by Eugene Bujak.
585 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
586 accept(). Patch provided by Maxim Dounin.
588 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
589 Patch provided by Phil Pennock.
591 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
593 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
594 Patch provided by Brad "anomie" Jorsch.
596 NM/14 Bugzilla 622: Added timeout setting to transport filter.
597 Patch provided by Dean Brooks.
599 TK/05 Add native DKIM support (does not depend on external libraries).
601 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
602 Patch provided by Graeme Fowler.
604 NM/16 Bugzilla 851: Documentation example syntax fix.
606 NM/17 Changed NOTICE file to remove references to embedded PCRE.
608 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
611 NM/19 Bugzilla 745: TLS version reporting.
612 Patch provided by Phil Pennock.
614 NM/20 Bugzilla 167: bool: condition support.
615 Patch provided by Phil Pennock.
617 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
618 clients. Patch provided by Phil Pennock.
620 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
621 Patch provided by Brad "anomie" Jorsch.
623 NM/23 Bugzilla 687: Fix misparses in eximstats.
624 Patch provided by Heiko Schlittermann.
626 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
627 Patch provided by Heiko Schlittermann.
629 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
630 plus update to original patch.
632 NM/26 Bugzilla 799: Documentation correction for ratelimit.
634 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
635 Patch provided by David Brownlee.
637 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
639 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
641 NM/30 Bugzilla 888: TLS documentation bugfixes.
643 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
645 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
646 Unlike the original bugzilla I have changed all shell scripts in src tree.
648 NM/33 Bugzilla 898: Transport filter timeout fix.
649 Patch by Todd Rinaldo.
651 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
652 Patch by Serge Demonchaux.
654 NM/35 Bugzilla 39: Base64 decode bug fixes.
655 Patch by Jakob Hirsch.
657 NM/36 Bugzilla 909: Correct connect() call in dcc code.
659 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
661 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
663 NM/39 Bugzilla 911: Fixed MakeLinks build script.
669 TK/01 Add preliminary DKIM support. Currently requires a forked version of
670 ALT-N's libdkim that I have put here:
671 http://duncanthrax.net/exim-experimental/
673 Note to Michael Haardt: I had to rename some vars in sieve.c. They
674 were called 'true' and it seems that C99 defines that as a reserved
675 keyword to be used with 'bool' variable types. That means you could
676 not include C99-style headers which use bools without triggering
677 build errors in sieve.c.
679 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
680 as mailq or other aliases. Changed the --help handling significantly
681 to do whats expected. exim_usage() emits usage/help information.
683 SC/01 Added the -bylocaldomain option to eximstats.
685 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
687 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
689 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
695 PH/01 Another patch from the Sieve maintainer.
697 PH/02 When an IPv6 address is converted to a string for single-key lookup
698 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
699 dots are used instead of colons so that keys in lsearch files need not
700 contain colons. This was done some time before quoting was made available
701 in lsearch files. However, iplsearch files do require colons in IPv6 keys
702 (notated using the quote facility) so as to distinguish them from IPv4
703 keys. This meant that lookups for IP addresses in host lists did not work
704 for iplsearch lookups.
706 This has been fixed by arranging for IPv6 addresses to be expressed with
707 colons if the lookup type is iplsearch. This is not incompatible, because
708 previously such lookups could never work.
710 The situation is now rather anomolous, since one *can* have colons in
711 ordinary lsearch keys. However, making the change in all cases is
712 incompatible and would probably break a number of configurations.
714 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
717 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
718 conversion specification without a maximum field width, thereby enabling
719 a rogue spamd server to cause a buffer overflow. While nobody in their
720 right mind would setup Exim to query an untrusted spamd server, an
721 attacker that gains access to a server running spamd could potentially
722 exploit this vulnerability to run arbitrary code as the Exim user.
724 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
725 $primary_hostname instead of what libspf2 thinks the hosts name is.
727 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
728 a directory entry by the name of the lookup key. Previously, if a
729 symlink pointed to a non-existing file or a file in a directory that
730 Exim lacked permissions to read, a lookup for a key matching that
731 symlink would fail. Now it is enough that a matching directory entry
732 exists, symlink or not. (Bugzilla 503.)
734 PH/03 The body_linecount and body_zerocount variables are now exported in the
737 PH/04 Added the $dnslist_matched variable.
739 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
740 This means they are set thereafter only if the connection becomes
743 PH/06 Added the client_condition to authenticators so that some can be skipped
744 by clients under certain conditions.
746 PH/07 The error message for a badly-placed control=no_multiline_responses left
747 "_responses" off the end of the name.
749 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
751 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
752 (without spaces) instead of just copying the configuration text.
754 PH/10 Added the /noupdate option to the ratelimit ACL condition.
756 PH/11 Added $max_received_linelength.
758 PH/12 Added +ignore_defer and +include_defer to host lists.
760 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
761 way in which PCRE > 7.0 is built.
763 PH/14 Implemented queue_only_load_latch.
765 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
766 MAIL command. The effect was to mangle the value on 64-bit systems.
768 PH/16 Another patch from the Sieve maintainer.
770 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
772 PH/18 If a system quota error occurred while trying to create the file for
773 a maildir delivery, the message "Mailbox is full" was not appended to the
774 bounce if the delivery eventually timed out. Change 4.67/27 below applied
775 only to a quota excession during the actual writing of the file.
777 PH/19 It seems that peer DN values may contain newlines (and other non-printing
778 characters?) which causes problems in log lines. The DN values are now
779 passed through string_printing() before being added to log lines.
781 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
782 and InterBase are left for another time.)
784 PH/21 Added message_body_newlines option.
786 PH/22 Guard against possible overflow in moan_check_errorcopy().
788 PH/23 POSIX allows open() to be a macro; guard against that.
790 PH/24 If the recipient of an error message contained an @ in the local part
791 (suitably quoted, of course), incorrect values were put in $domain and
792 $local_part during the evaluation of errors_copy.
798 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
799 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
802 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
803 issue a MAIL command.
805 PH/02 In an ACL statement such as
807 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
809 if a client was not listed at all, or was listed with a value other than
810 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
811 the condition was not true (as it should be), so access was not denied.
812 The bug was that the ! inversion was incorrectly passed on to the second
813 item. This has been fixed.
815 PH/03 Added additional dnslists conditions == and =& which are different from
816 = and & when the dns lookup returns more than one IP address.
818 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
819 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
821 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
822 FSYNC, which compiles an option called disable_fsync that allows for
823 bypassing fsync(). The documentation is heavily laced with warnings.
825 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
827 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
828 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
829 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
830 including adding "make clean"; (3) Added -fPIC when compiling the test
831 dynamically loaded module, to get rid of a warning.
833 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
834 message fails, move_frozen_messages = true and ignore_bounce_errors_after
835 = 0s. The bug is otherwise harmless.
837 PH/07 There was a bug in the dovecot authenticator such that the value of
838 $auth1 could be overwritten, and so not correctly preserved, after a
839 successful authentication. This usually meant that the value preserved by
840 the server_setid option was incorrect.
842 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
844 PH/09 Installed PCRE release 7.0.
846 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
847 run for batched SMTP input. It is now run at the start of every message
848 in the batch. While fixing this I discovered that the process information
849 (output by running exiwhat) was not always getting set for -bs and -bS
850 input. This is fixed, and it now also says "batched" for BSMTP.
852 PH/11 Added control=no_pipelining.
854 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
855 patch, slightly modified), and move the expansion of helo_data till after
856 the connection is made in the smtp transport (so it can use these
859 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
861 PH/14 Added log_selector = +pid.
863 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
865 PH/16 Add ${if forany and ${if forall.
867 PH/17 Added dsn_from option to vary the From: line in DSNs.
869 PH/18 Flush SMTP output before performing a callout, unless control =
870 no_callout_flush is set.
872 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
873 was true (the default) a successful delivery failed to delete the retry
874 item, thus causing premature timeout of the address. The bug is now
877 PH/20 Added hosts_avoid_pipelining to the smtp transport.
879 PH/21 Long custom messages for fakedefer and fakereject are now split up
880 into multiline reponses in the same way that messages for "deny" and
881 other ACL rejections are.
883 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
884 with slight modification.
886 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
887 draft, changing the syntax and factoring some duplicate code".
889 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
890 for deliveries of the second and subsequent messages over the same SMTP
893 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
894 ${reduce, with only minor "tidies".
896 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
898 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
899 expansion side effects.
901 PH/27 When a message times out after an over-quota error from an Exim-imposed
902 quota, the bounce message says "mailbox is full". This message was not
903 being given when it was a system quota that was exceeded. It now should
906 MH/03 Made $recipients available in local_scan(). local_scan() already has
907 better access to the recipient list through recipients_list[], but
908 $recipients can be useful in postmaster-provided expansion strings.
910 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
911 in the case of a MAIL command with additional options following the
912 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
913 were accidentally chopped off.
915 PH/29 SMTP synchronization checks are implemented when a command is read -
916 there is a check that no more input is waiting when there shouldn't be
917 any. However, for some commands, a delay in an ACL can mean that it is
918 some time before the response is written. In this time, more input might
919 arrive, invalidly. So now there are extra checks after an ACL has run for
920 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
921 pipelining has not been advertised.
923 PH/30 MH's patch to allow iscntrl() characters to be list separators.
925 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
926 returned in the SMTP response when smtp_return_error_details was false.
929 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
930 instead of the C I/O that was originally supplied, because problems were
933 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
934 Exim which did not show up earlier: it was assuming that a call to
935 SSL_CTX_set_info_callback() might give an error value. In fact, there is
936 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
937 was a macro that became an assignment, so it seemed to work. This has
938 changed to a proper function call with a void return, hence the compile
939 error. Exim's code has been fixed.
941 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
944 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
946 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
948 PH/37 If a message is not accepted after it has had an id assigned (e.g.
949 because it turns out to be too big or there is a timeout) there is no
950 "Completed" line in the log. When some messages of this type were
951 selected by exigrep, they were listed as "not completed". Others were
952 picked up by some special patterns. I have improved the selection
953 criteria to be more general.
955 PH/38 The host_find_failed option in the manualroute router can now be set
956 to "ignore", to completely ignore a host whose IP address cannot be
957 found. If all hosts are ignored, the behaviour is controlled by the new
958 host_all_ignored option.
960 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
961 homing or because of multiple MX records with /mx) generated more than
962 one IP address, and the following item turned out to be the local host,
963 all the secondary addresses of the first item were incorrectly removed
964 from the list, along with the local host and any following hosts (which
965 is what is supposed to happen).
967 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
968 whoever called Exim into the -H file. In the case of the daemon it was
969 behaving confusingly. When first started, it used values for whoever
970 started the daemon, but after a SIGHUP it used the Exim user (because it
971 calls itself on a restart). I have changed the code so that it now always
974 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
975 message are rejected with the same error (e.g. no authentication or bad
976 sender address), and a DATA command is nevertheless sent (as can happen
977 with PIPELINING or a stupid MUA), the error message that was given to the
978 RCPT commands is included in the rejection of the DATA command. This is
979 intended to be helpful for MUAs that show only the final error to their
982 PH/42 Another patch from the Sieve maintainer.
984 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
985 Eximstats - Fixed some broken HTML links and added missing column headers
987 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
988 columns for Rejects, Temp Rejects, Ham, and Spam rows.
990 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
992 PH/43 Yet another patch from the Sieve maintainer.
994 PH/44 I found a way to check for a TCP/IP connection going away before sending
995 the response to the final '.' that terminates a message, but only in the
996 case where the client has not sent further data following the '.'
997 (unfortunately, this is allowed). However, in many cases there won't be
998 any further data because there won't be any more messages to send. A call
999 to select() can be used: if it shows that the input is "ready", there is
1000 either input waiting, or the socket has been closed. An attempt to read
1001 the next input character can distinguish the two cases. Previously, Exim
1002 would have sent an OK response which the client would never have see.
1003 This could lead to message repetition. This fix should cure that, at
1004 least in a lot of common cases.
1006 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1007 advertised in response to EHLO.
1013 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1014 fixed by 4.65/MH/01 (is this a record?) are fixed:
1016 (i) An empty string was always treated as zero by the numeric comparison
1017 operators. This behaviour has been restored.
1019 (ii) It is documented that the numeric comparison operators always treat
1020 their arguments as decimal numbers. This was broken in that numbers
1021 starting with 0 were being interpreted as octal.
1023 While fixing these problems I realized that there was another issue that
1024 hadn't been noticed. Values of message_size_limit (both the global option
1025 and the transport option) were treated as octal if they started with 0.
1026 The documentation was vague. These values are now always treated as
1027 decimal, and I will make that clear in the documentation.
1033 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1034 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1037 MH/01 Don't check that the operands of numeric comparison operators are
1038 integers when their expansion is in "skipping" mode (fixes bug
1039 introduced by 4.64-PH/07).
1041 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1042 child addresses, Exim now panics and dies. Previously, because the count
1043 is held in a short int, deliveries were likely to be lost. As such a
1044 large number of recipients for a single message is ridiculous
1045 (performance will be very, very poor), I have chosen to impose a limit
1046 rather than extend the field.
1052 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1053 leftover -K file (the existence of which was triggered by #402).
1054 While we were at it, introduced process PID as part of the -K
1055 filename. This should rule out race conditions when creating
1058 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1059 processing considerably. Previous code took too long for large mails,
1060 triggering a timeout which in turn triggers #401.
1062 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1063 in the DK code in transports.c. sendfile() is not really portable,
1064 hence the _LINUX specificness.
1066 TF/01 In the add_headers option to the mail command in an Exim filter,
1067 there was a bug that Exim would claim a syntax error in any
1068 header after the first one which had an odd number of characters
1071 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1072 callout verification, Exim cached a "reject" for the entire domain. This
1073 is correct for most verifications, but it is not correct for a recipient
1074 verification with use_sender or use_postmaster set, because in that case
1075 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1076 case of MAIL FROM:<> rejection from other early rejections (e.g.
1077 rejection of HELO). When verifying a recipient using a non-null MAIL
1078 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1079 Whatever the result of the callout, the value of the domain cache is
1080 left unchanged (for any other kind of callout, getting as far as trying
1081 RCPT means that the domain itself is ok).
1083 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1086 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1087 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1088 session, the daemon ignores ECONNECT errors and logs others; it now
1089 ignores EPIPE as well.
1091 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1092 (quoted-printable decoding).
1094 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1095 later the small subsequent patch to fix an introduced bug.
1097 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1099 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1101 PH/08 An error is now given if message_size_limit is specified negative.
1103 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1104 to be given (somewhat) arbitrary names.
1106 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1109 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1110 miscellaneous code fixes
1112 PH/10 Added the log_reject_target ACL modifier to specify where to log
1115 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1116 hostname. This is wrong, because it relates to the incoming message (and
1117 probably the interface on which it is arriving) and not to the outgoing
1118 callout (which could be using a different interface). This has been
1119 changed to use the value of the helo_data option from the smtp transport
1120 instead - this is what is used when a message is actually being sent. If
1121 there is no remote transport (possible with a router that sets up host
1122 addresses), $smtp_active_hostname is used.
1124 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1125 tweaks were necessary in order to get it to work (see also 21 below):
1126 (a) The code assumed that strncpy() returns a negative number on buffer
1127 overflow, which isn't the case. Replaced with Exim's string_format()
1129 (b) There were several signed/unsigned issues. I just did the minimum
1130 hacking in of casts. There is scope for a larger refactoring.
1131 (c) The code used strcasecmp() which is not a standard C function.
1132 Replaced with Exim's strcmpic() function.
1133 (d) The code set only $1; it now sets $auth1 as well.
1134 (e) A simple test gave the error "authentication client didn't specify
1135 service in request". It would seem that Dovecot has changed its
1136 interface. Fortunately there's a specification; I followed it and
1137 changed what the client sends and it appears to be working now.
1139 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1142 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1143 address A is aliased to B and C, where B exists and C does not. Without
1144 -v the output is "A verified" because verification stops after a
1145 successful redirection if more than one address is generated. However,
1146 with -v the child addresses are also verified. Exim was outputting "A
1147 failed to verify" and then showing the successful verification for C,
1148 with its parentage. It now outputs "B failed to verify", showing B's
1149 parentage before showing the successful verification of C.
1151 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1152 look up a TXT record in a specific list after matching in a combined
1155 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1156 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1157 they consult the DNS. I had assumed they would set it the way they
1158 wanted; and indeed my experiments on Linux seem to show that in some
1159 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1160 To be on the safe side, however, I have now made the interface to
1161 host_find_byname() similar to host_find_bydns(), with an argument
1162 containing the DNS resolver options. The host_find_byname() function now
1163 sets these options at its start, just as host_find_bydns() does. The smtp
1164 transport options dns_qualify_single and dns_search_parents are passed to
1165 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1166 of host_find_byname() use the default settings of RES_DEFNAMES
1167 (qualify_single) but not RES_DNSRCH (search_parents).
1169 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1170 spool_read_header() do less string testing, by means of a preliminary
1171 switch on the second character of optional "-foo" lines. (This is
1172 overdue, caused by the large number of possibilities that now exist.
1173 Originally there were few.) While I was there, I also converted the
1174 str(n)cmp tests so they don't re-test the leading "-" and the first
1175 character, in the hope this might squeeze out yet more improvement.
1177 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1178 flag allowing group syntax was set by the header_syntax check but not
1179 turned off, possible causing trouble later; (2) The flag was not being
1180 set at all for the header_verify test, causing "group"-style headers to
1181 be rejected. I have now set it in this case, and also caused header_
1182 verify to ignore an empty address taken from a group. While doing this, I
1183 came across some other cases where the code for allowing group syntax
1184 while scanning a header line wasn't quite right (mostly, not resetting
1185 the flag correctly in the right place). These bugs could have caused
1186 trouble for malformed header lines. I hope it is now all correct.
1188 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1189 with the "reply" argument non-NULL. The code, however (which originally
1190 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1191 but it didn't always do it. This confused somebody who was copying the
1192 code for some other use. I have removed all the tests.
1194 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1195 feature that was used to support insecure browsers during the U.S. crypto
1196 embargo. It requires special client support, and Exim is probably the
1197 only MTA that supported it -- and would never use it because real RSA is
1198 always available. This code has been removed, because it had the bad
1199 effect of slowing Exim down by computing (never used) parameters for the
1200 RSA_EXPORT functionality.
1202 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1203 authenticator to fail if there's a tab character in the incoming data
1204 (there should never be unless someone is messing about, as it's supposed
1205 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1206 if the connection is using TLS or if the remote IP is the same as the
1207 local IP, and the "valid-client-cert option" if a client certificate has
1210 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1211 authenticators. This can be used for authorization after authentication
1212 succeeds. (In the case of plaintext, it servers for both authentication
1215 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1216 if any retry times were supplied.
1218 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1219 connection, where there is no client IP address to check. In this
1220 situation, the verify now always succeeds.
1222 PH/25 Applied John Jetmore's -Mset patch.
1224 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1226 PH/27 In a string expansion for a processed (not raw) header when multiple
1227 headers of the same name were present, leading whitespace was being
1228 removed from all of them, but trailing whitespace was being removed only
1229 from the last one. Now trailing whitespace is removed from each header
1230 before concatenation. Completely empty headers in a concatenation (as
1231 before) are ignored.
1233 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1234 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1236 PH/29 [Removed. This was a change that I later backed out, and forgot to
1237 correct the ChangeLog entry (that I had efficiently created) before
1238 committing the later change.]
1240 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1241 address errors (4xx response to RCPT) over the same connection as other
1242 messages routed to the same hosts. Such deliveries are always "forced",
1243 so retry times are not inspected. This resulted in far too many retries
1244 for the affected addresses. The effect occurred only when there were more
1245 hosts than the hosts_max_try setting in the smtp transport when it had
1246 the 4xx errors. Those hosts that it had tried were not added to the list
1247 of hosts for which the message was waiting, so if all were tried, there
1248 was no problem. Two fixes have been applied:
1250 (i) If there are any address or message errors in an SMTP delivery, none
1251 of the hosts (tried or untried) are now added to the list of hosts
1252 for which the message is waiting, so the message should not be a
1253 candidate for sending over the same connection that was used for a
1254 successful delivery of some other message. This seems entirely
1255 reasonable: after all the message is NOT "waiting for some host".
1256 This is so "obvious" that I'm not sure why it wasn't done
1257 previously. Hope I haven't missed anything, but it can't do any
1258 harm, as the worst effect is to miss an optimization.
1260 (ii) If, despite (i), such a delivery is accidentally attempted, the
1261 routing retry time is respected, so at least it doesn't keep
1262 hammering the server.
1264 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1265 in ${readsocket because some servers need this prod.
1267 PH/32 Added some extra debug output when updating a wait-xxx database.
1269 PH/33 The hint "could be header name not terminated by colon", which has been
1270 given for certain expansion errors for a long time, was not being given
1271 for the ${if def:h_colon_omitted{... case.
1273 PH/34 The spec says: "With one important exception, whenever a domain list is
1274 being scanned, $domain contains the subject domain." There was at least
1275 one case where this was not true.
1277 PH/35 The error "getsockname() failed: connection reset by peer" was being
1278 written to the panic log as well as the main log, but it isn't really
1279 panic-worthy as it just means the connection died rather early on. I have
1280 removed the panic log writing for the ECONNRESET error when getsockname()
1283 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1284 runs only) independently of the message's sender address. This meant
1285 that, if the 4xx error was in fact related to the sender, a different
1286 message to the same recipient with a different sender could confuse
1287 things. In particualar, this can happen when sending to a greylisting
1288 server, but other circumstances could also provoke similar problems.
1289 I have changed the default so that the retry time for these errors is now
1290 based a combination of the sender and recipient addresses. This change
1291 can be overridden by setting address_retry_include_sender=false in the
1294 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1295 remote server are returned as part of bounce messages. This was not
1296 happening for LMTP over a pipe (the lmtp transport), but now it is the
1297 same for both kinds of LMTP.
1299 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1300 in header lines that were in fact CNAMEs. This is no longer the case.
1302 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1303 and queue runs started by the daemon processed all messages. This has
1304 been fixed so that -R and -S can now usefully be given with -q<time>.
1306 PH/40 Import PCRE release 6.7 (fixes some bugs).
1308 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1310 PH/42 Give an error if -q is specified more than once.
1312 PH/43 Renamed the variables $interface_address and $interface_port as
1313 $received_ip_address and $received_port, to make it clear that these
1314 values apply to message reception, and not to the outgoing interface when
1315 a message is delivered. (The old names remain recognized, of course.)
1317 PH/44 There was no timeout on the connect() call when using a Unix domain
1318 socket in the ${readsocket expansion. There now is.
1320 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1321 be meaningful with "accept".
1323 SC/01 Eximstats V1.43
1324 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1326 SC/02 Eximstats V1.44
1327 Use a glob alias rather than an array ref in the generated
1328 parser. This improves both readability and performance.
1330 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1331 Collect SpamAssassin and rejection statistics.
1332 Don't display local sender or destination tables unless
1333 there is data to show.
1334 Added average volumes into the top table text output.
1336 SC/04 Eximstats V1.46
1337 Collect data on the number of addresses (recipients)
1338 as well as the number of messages.
1340 SC/05 Eximstats V1.47
1341 Added 'Message too big' to the list of mail rejection
1342 reasons (thanks to Marco Gaiarin).
1344 SC/06 Eximstats V1.48
1345 Mainlog lines which have GMT offsets and are too short to
1346 have a flag are now skipped.
1348 SC/07 Eximstats V1.49 (Alain Williams)
1349 Added the -emptyok flag.
1351 SC/08 Eximstats V1.50
1352 Fixes for obtaining the IP address from reject messages.
1354 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1355 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1356 whitesspace changes from 4.64-PH/27
1358 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1361 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1362 are found, allow negative numbers in numeric criteria)
1364 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1366 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1367 to match changes made in 4.64-PH/43
1369 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1371 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1372 "domain" is actually an IP address, and force a failure. This locks out
1373 those revolvers/nameservers that support "A-for-A" lookups, in
1374 contravention of the specifications.
1376 PH/48 When a host name was looked up from an IP address, and the subsequent
1377 forward lookup of the name timed out, the host name was left in
1378 $sender_host_name, contrary to the specification.
1380 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1381 restricted to single-key lookups, Exim was not diagnosing an error if
1382 * or *@ was used with a query-style lookup.
1384 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1386 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1387 long ago, but noone interested enough thought of it. Let's just say that
1388 the "1.1" means that there are some new functions that weren't there at
1389 some point in the past.
1391 PH/51 Error processing for expansion failure of helo_data from an smtp
1392 transport during callout processing was broken.
1394 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1395 tested/used via the -bh/-bhc/-bs options.
1397 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1398 bug, fixed in subsequent PCRE releases).
1400 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1401 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1403 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1409 SC/01 Use a glob alias rather than an array ref in eximstats generated
1410 parser. This improves both readability and performance.
1412 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1413 Don't display local sender or destination tables in eximstats unless
1414 there is data to show.
1415 Added average volumes into the eximstats top table text output.
1417 SC/03 Collect data on the number of addresses (recipients) as well
1418 as the number of messages in eximstats.
1420 TF/01 Correct an error in the documentation for the redirect router. Exim
1421 does (usually) call initgroups() when daemonizing.
1423 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1424 with consistent privilege compared to when running as a daemon.
1426 TF/03 Note in the spec that $authenticated_id is not set for local
1427 submissions from trusted users.
1429 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1430 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1432 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1433 by adding some example configuration directives to the default
1434 configuration file. A little bit of work is required to uncomment the
1435 directives and define how usernames and passwords are checked, but
1436 there is now a framework to start from.
1438 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1439 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1440 without this. I don't know how relevant this is to other LDAP libraries.
1442 PH/02 Add the verb name to the "unknown ACL verb" error.
1444 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1446 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1448 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1449 directory not expanded when it should be if an expanded home directory
1450 was set for the address (which is overridden by the transport).
1452 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1455 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1456 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1457 because it is too late at that time, and has no effect.
1459 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1460 security issue with \' (bugzilla #107). I could not use the
1461 PQescapeStringConn() function, because it needs a PGconn value as one of
1464 PH/08 When testing addresses using -bt, indicate those final addresses that
1465 are duplicates that would not cause an additional delivery. At least one
1466 person was confused, thinking that -bt output corresponded to deliveries.
1467 (Suppressing duplicates isn't a good idea as you lose the information
1468 about possibly different redirections that led to the duplicates.)
1470 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1471 systems where poll() doesn't work, in particular OS X.
1473 PH/10 Added more information to debugging output for retry time not reached.
1475 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1476 operations in malware.c.
1478 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1481 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1482 syslog_duplication set false, log lines that would normally be written
1483 both the the main log and to the reject log were not written to syslog at
1486 PH/14 In the default configuration, change the use of "message" in ACL warn
1487 statements to "add_header".
1489 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1490 not followed by a command (e.g. "seen endif").
1492 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1493 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1496 PH/17 Added extra conditions to the default value of delay_warning_condition
1500 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1501 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1502 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1505 The Auto-Submitted: and various List- headers are standardised, whereas I
1506 don't think Precedence: ever was.
1508 PH/18 Refactored debugging code in route_finduser() to show more information,
1509 in particular, the error code if getpwnam() issues one.
1511 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1512 This is apparently needed in addition to the PH/07 change above to avoid
1513 any possible encoding problems.
1515 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1516 but not after initializing Perl.
1518 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1519 output them only if debugging. By default they are written stderr,
1520 apparently, which is not desirable.
1522 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1525 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1528 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1530 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1531 authenticated or an ident call has been made. Suppress the default
1532 values for $authenticated_id and $authenticated_sender (but permit -oMai
1533 and -oMas) when testing with -bh.
1535 PH/24 Re-jigged the order of the tests in the default configuration so that the
1536 tests for valid domains and recipients precede the DNS black list and CSA
1537 tests, on the grounds that those ones are more expensive.
1539 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1540 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1541 This bug exists in every version of Exim that I still have, right back to
1544 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1545 However, an attempt to turn on case-sensitivity in a regex key by
1546 including (?-i) didn't work because the subject string was already
1547 lowercased, and the effects were non-intuitive. It turns out that a
1548 one-line patch can be used to allow (?-i) to work as expected.
1554 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1555 other effects) broke the use of negated acl sub-conditions.
1557 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1560 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1561 "Deny" causes Exim to reject the incoming connection with a 554 error.
1562 Unfortunately, if there is a major crisis, such as a disk failure,
1563 tcp-wrappers gives "deny", whereas what one would like would be some
1564 kind of temporary error. A kludge has been added to help with this.
1565 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1566 554 error is used if errno is still zero or contains ENOENT (which occurs
1567 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1570 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1572 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1573 errors. Otherwise a message that provokes a temporary error (when other
1574 messages do not) can cause a whole host to time out.
1576 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1577 addresses were routed directly to files or pipes from a redirect router.
1578 File deliveries just didn't batch; pipe deliveries might have suffered
1581 PH/06 A failure to get a lock for a hints database would erroneously always say
1582 "Failed to get write lock", even when it was really a read lock.
1584 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1585 of 0600. This has been changed to use the value of the lockfile_mode
1586 option (which defaults to 0600).
1588 PH/08 Applied small patch from the Sieve maintainer.
1590 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1591 folder from quota calculations, a direct delivery into this folder messed
1592 up the contents of the maildirsize file. This was because the regex was
1593 used only to exclude .Trash (or whatever) when the size of the mailbox
1594 was calculated. There was no check that a delivery was happening into an
1595 excluded directory. This bug has been fixed by ignoring all quota
1596 processing for deliveries into excluded directories.
1598 PH/10 Added the maildirfolder_create_regex option to appendfile.
1604 PH/01 The code for finding all the local interface addresses on a FreeBSD
1605 system running IPv6 was broken. This may well have applied to all BSD
1606 systems, as well as to others that have similar system calls. The broken
1607 code found IPv4 interfaces correctly, but gave incorrect values for the
1608 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1609 that it would not match correctly against @[] and not recognize the IPv6
1612 PH/02 The ipliteral router was not recognizing addresses of the form user@
1613 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1615 PH/03 Added disable_ipv6.
1617 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1618 decoded form, because it is most often used to construct To: headers
1619 lines in autoreplies, and the decoded form may well be syntactically
1620 invalid. However, $reply_address has leading white space removed, and all
1621 newlines turned into spaces so that the autoreply transport does not
1624 PH/05 If group was specified without a user on a router, and no group or user
1625 was specified on a transport, the group from the router was ignored.
1627 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1628 for visible compile-time settings that can be used to change these
1629 numbers, for those that want even more. Backwards compatibility with old
1630 spool files has been maintained. However, going back to a previous Exim
1631 release will lost any variables that are in spool files.
1633 PH/07 Two small changes when running in the test harness: increase delay when
1634 passing a TCP/IP connection to a new process, in case the original
1635 process has to generate a bounce, and remove special handling of
1636 127.0.0.2 (sic), which is no longer necessary.
1638 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1639 be the same on different OS.
1641 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1644 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1645 whether --show-vars was specified or not
1647 JJ/02 exipick: Added support for new ACL variable spool format introduced
1650 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1651 syntactically invalid From: or Reply-to: line, and a filter used this to
1652 generate an autoreply, and therefore failed to obtain an address for the
1653 autoreply, Exim could try to deliver to a non-existent relative file
1654 name, causing unrelated and misleading errors. What now happens is that
1655 it logs this as a hard delivery error, but does not attempt to create a
1658 PH/11 The exinext utility has a -C option for testing purposes, but although
1659 the given file was scanned by exinext itself; it wasn't being passed on
1660 when Exim was called.
1662 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1663 an end-of-file indication when reading a command response.
1665 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1666 compiled. In many other places in Exim, IPv6 addresses are always
1667 recognized, so I have changed this. It also means that IPv4 domain
1668 literals of the form [IPV4:n.n.n.n] are now always recognized.
1670 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1671 used if the router is not running as root, for example, when verifying at
1672 ACL time, or when using -bh. The debugging output from this situation was
1673 non-existent - all you got was a failure to exec. I have made two
1676 (a) Failures to set uid/gid, the current directory, or a process leader
1677 in a subprocess such as that created by queryprogram now generate
1678 suitable debugging ouput when -d is set.
1680 (b) The queryprogram router detects when it is not running as root,
1681 outputs suitable debugging information if -d is set, and then runs
1682 the subprocess without attempting to change uid/gid.
1684 PH/15 Minor change to Makefile for building test_host (undocumented testing
1687 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1688 additional section of a DNS packet that returns MX or SRV records.
1689 Instead, it always explicitly searches for A/AAAA records. This avoids
1690 major problems that occur when a DNS server includes only records of one
1691 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1692 fixed another bug: if SRV records were looked up and the corresponding
1693 address records were *not* found in the additional section, the port
1694 values from the SRV records were lost.
1696 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1697 using the correct key (the original address) when searching the retry
1698 rules in order to find which one to use for generating the retry hint.
1700 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1701 adding the default one. Similarly, if it contains a Reply-To: header, the
1702 errors_reply_to option, if set, is not used.
1704 PH/19 When calculating a retry time, Exim used to measure the "time since
1705 failure" by looking at the "first failed" field in the retry record. Now
1706 it does not use this if it is later than than the arrival time of the
1707 message. Instead it uses the arrival time. This makes for better
1708 behaviour in cases where some deliveries succeed, thus re-setting the
1709 "first failed" field. An example is a quota failure for a huge message
1710 when small messages continue to be delivered. Without this change, the
1711 "time since failure" will always be short, possible causing more frequent
1712 delivery attempts for the huge message than are intended.
1713 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1715 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1716 $1, $2, $3) because the numerical variables can be reset during some
1717 expansion items (e.g. "match"), thereby losing the authentication data.
1719 PH/21 Make -bV show the size of off_t variables so that the test suite can
1720 decide whether to run tests for quotas > 2G.
1722 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1723 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1724 filecount value is greater than 2G or if a quota value is greater than 2G
1725 on a system where the size of off_t is not greater than 4, a panic error
1728 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1729 never match. The debug and -bh output now contains an explicit error
1730 message indicating a malformed IPv4 address or mask.
1732 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1733 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1734 PH/23 above applies.
1736 PH/25 Do not write to syslog when running in the test harness. The only
1737 occasion when this arises is a failure to open the main or panic logs
1738 (for which there is an explicit test).
1740 PH/26 Added the /no_tell option to "control=freeze".
1742 PH/27 If a host name lookup failed very early in a connection, for example, if
1743 the IP address matched host_lookup and the reverse lookup yielded a name
1744 that did not have a forward lookup, an error message of the form "no IP
1745 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1746 could be logged. Now it outputs the IP address instead of "NULL".
1748 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1749 allows the sender and the authenticated sender to be set when
1750 submitting a message from within Exim. Since child_open_exim() is
1751 documented for local_scan(), the new function should be too.
1753 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1754 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1755 results in an empty string is now treated as unset.
1757 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1759 PH/31 Added sender_verify_fail logging option.
1761 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1762 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1763 tidied the source and removed it altogether.
1765 PH/33 When a queue run was abandoned because the load average was too high, a
1766 log line was always written; now it is written only if the queue_run log
1767 selector is set. In addition, the log line for abandonment now contains
1768 information about the queue run such as the pid. This is always present
1769 in "start" and "stop" lines but was omitted from the "abandon" line.
1771 PH/34 Omit spaces between a header name and the colon in the error message that
1772 is given when verify = headers_syntax fails (if there are lots of them,
1773 the message gets confusing).
1775 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1776 names, as there are now some PTR records that contain slashes. This check
1777 is only to protect against broken name servers that fall over on strange
1778 characters, so the fact that it applies to all lookups doesn't matter.
1780 PH/36 Now that the new test suite is complete, we can remove some of the
1781 special code in Exim that was needed for the old test suite. For example,
1782 sorting DNS records because real resolvers return them in an arbitrary
1783 order. The new test suite's fake resolver always returns records in the
1786 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1787 bounces) except when queue_only is set, to avoid logging races between
1788 the different processes.
1790 PH/38 Panic-die if .include specifies a non-absolute path.
1792 PH/39 A tweak to the "H" retry rule from its user.
1794 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1795 a label. They prevented compilation on older perls.
1797 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1798 a warning to be raised on newish perls.
1800 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1801 on queue. Changes to match documented behaviour of showing count of
1802 messages matching specified criteria.
1804 PH/40 Changed the default ident timeout from 30s to 5s.
1806 PH/41 Added support for the use of login_cap features, on those BSD systems
1807 that have them, for controlling the resources used by pipe deliveries.
1809 PH/42 The content-scanning code uses fopen() to create files in which to put
1810 message data. Previously it was not paying any attention to the mode of
1811 the files. Exim runs with umask(0) because the rest of the code creates
1812 files with open(), and sets the required mode explicitly. Thus, these
1813 files were ending up world-writeable. This was not a big issue, because,
1814 being within the spool directory, they were not world-accessible. I have
1815 created a function called modefopen, which takes an additional mode
1816 argument. It sets umask(777), creates the file, chmods it to the required
1817 mode, then resets the umask. All the relevant calls to fopen() in the
1818 content scanning code have been changed to use this function.
1820 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1821 to 24 hours. This avoids potential overflow problems when processing G
1822 and H retry rules. I suspect nobody ever tinkers with this value.
1824 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1826 PH/45 When the plaintext authenticator is running as a client, the server's
1827 challenges are checked to ensure they are valid base64 strings. By
1828 default, the authentication attempt is cancelled if an invalid string is
1829 received. Setting client_ignore_invalid_base64 true ignores these errors.
1830 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1831 they are received. Thus, the responses can be made to depend on the
1832 challenges. If an invalid string is ignored, an empty string is placed in
1835 PH/46 Messages that are created by the autoreply transport now contains a
1836 References: header, in accordance with RFCs 2822 and 3834.
1838 PH/47 Added authenticated_sender_force to the smtp transport.
1840 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1842 PH/49 Installed latest patch from the Sieve maintainer.
1844 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1845 was also set, the appendfile transport was unnecessarily scanning a
1846 directory of message files (e.g. for maildir delivery) to find the count
1847 of files (along with the size), even though it did not need this
1848 information. It now does the scan only if it needs to find either the
1849 size of the count of files.
1851 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1853 PH/52 Two bugs concerned with error handling when the smtp transport is
1856 (i) Exim was not creating retry information for temporary errors given
1857 for individual recipients after the DATA command when the smtp transport
1858 was used in LMTP mode. This meant that they could be retried too
1859 frequently, and not timed out correctly.
1861 (ii) Exim was setting the flag that allows error details to be returned
1862 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1863 recipients that were returned after the DATA command.
1865 PH/53 This is related to PH/52, but is more general: for any failing address,
1866 when detailed error information was permitted to be returned to the
1867 sender, but the error was temporary, then after the final timeout, only
1868 "retry timeout exceeded" was returned. Now it returns the full error as
1869 well as "retry timeout exceeded".
1871 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1872 do this, and (what is worse) MTAs that accept it.
1874 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1875 will now be deprecated.
1877 PH/56 New os.c-cygwin from the Cygwin maintainer.
1879 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1880 formats (previously only available in exim formats via -bpr, -bpru,
1881 and -bpra. Now also available in native and exiqgrep formats)
1883 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1884 with very large, slow to parse queues
1886 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1888 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1890 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1891 responses to authentication challenges, though it was showing the
1892 challenges; (ii) I've removed the CR characters from the debug output for
1895 PH/58 Allow for the insertion of a newline as well as a space when a string
1896 is turned into more than one encoded-word during RFC 2047 encoding. The
1897 Sieve code now uses this.
1899 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1900 data_4xx, lost_connection, tls_required.
1902 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1903 message was being sent as an SMTP response.
1905 PH/61 Add -l and -k options to exicyclog.
1907 PH/62 When verifying, if an address was redirected to one new address, so that
1908 verification continued, and the new address failed or deferred after
1909 having set something in $address_data, the value of $address_data was not
1910 passed back to the ACL. This was different to the case when no
1911 redirection occurred. The value is now passed back in both cases.
1913 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1914 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1915 use login_cap.h, so on its own it isn't the distinguishing feature. The
1916 new name refers directly to the setclassresources() function.
1918 PH/65 Added configuration files for NetBSD3.
1920 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1922 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1923 is preferred over IPv4.
1925 PH/68 The bounce_return_message and bounce_return_body options were not being
1926 honoured for bounces generated during the reception of non-SMTP messages.
1927 In particular, this applied to messages rejected by the ACL. This bug has
1928 been fixed. However, if bounce_return_message is true and bounce_return_
1929 body is false, the headers that are returned for a non-SMTP message
1930 include only those that have been read before the error was detected.
1931 (In the case of an ACL rejection, they have all been read.)
1933 PH/69 The HTML version of the specification is now built in a directory called
1934 spec_html instead of spec.html, because the latter looks like a path with
1935 a MIME-type, and this confuses some software.
1937 PH/70 Catch two compiler warnings in sieve.c.
1939 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1940 function verify_get_ident() calls ip_connect() to connect a socket, but
1941 if the "connect()" function timed out, ip_connect() used to close the
1942 socket. However, verify_get_ident() also closes the socket later, and in
1943 between Exim writes to the log, which may get opened at this point. When
1944 the socket was closed in ip_connect(), the log could get the same file
1945 descriptor number as the socket. This naturally causes chaos. The fix is
1946 not to close the socket in ip_connect(); the socket should be closed by
1947 the function that creates it. There was only one place in the code where
1948 this was missing, in the iplookup router, which I don't think anybody now
1949 uses, but I've fixed it anyway.
1951 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1952 well as to direct DNS lookups. Otherwise the handling of names in host
1953 lists is inconsistent and therefore confusing.
1959 PH/01 Two changes to the default runtime configuration:
1961 (1) Move the checks for relay_from_hosts and authenticated clients from
1962 after to before the (commented out) DNS black list checks.
1964 (2) Add control=submission to the relay_from_hosts and authenticated
1965 clients checks, on the grounds that messages accepted by these
1966 statements are most likely to be submissions.
1968 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1970 (1) Generate an error if the third argument for the ${prvs expansion is
1973 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1976 (3) Reset the variables that are obtained from the first argument of
1977 ${prvscheck and used in the second argument before leaving the code,
1978 because their memory is reclaimed, so using them afterwards may do
1981 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1982 one (it's much easier than Tom thought :-).
1984 (5) Because of (4), we can now allow for the use of $prvscheck_result
1985 inside the third argument.
1987 PH/03 For some reason, the default setting of PATH when running a command from
1988 a pipe transport was just "/usr/bin". I have changed it to
1991 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1992 anything to be listed in the output from -bV.
1994 PH/05 When a filter generated an autoreply, the entire To: header line was
1995 quoted in the delivery log line, like this:
1997 => >A.N.Other <ano@some.domain> <original@ddress> ...
1999 This has been changed so that it extracts the operative address. There
2000 may be more than one such address. If so, they are comma-separated, like
2003 => >ano@some.domain,ona@other.domain <original@ddress> ...
2005 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2006 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2007 not being looked up in the rDNS to get a host name, Exim was showing the
2008 IP address twice in Received: lines, even though the IP addresses were
2009 identical. For example:
2011 Received: from [1.2.3.4] (helo=[1.2.3.4])
2013 However, if the real host name was known, it was omitting the HELO data
2014 if it matched the actual IP address. This has been tidied up so that it
2015 doesn't show the same IP address twice.
2017 PH/07 When both +timestamp and +memory debugging was on, the value given by
2018 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2019 called by the debug printing, thereby overwriting the timestamp buffer.
2020 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2022 PH/08 When the original message was included in an autoreply transport, it
2023 always said "this is a copy of the message, including all the headers",
2024 even if body_only or headers_only was set. It now gives an appropriate
2027 PH/09 Applied a patch from the Sieve maintainer which:
2029 o fixes some comments
2030 o adds the (disabled) notify extension core
2031 o adds some debug output for the result of if/elsif tests
2032 o points to the current vacation draft in the documentation
2033 and documents the missing references header update
2037 o fixes a bug in processing the envelope test (when testing
2038 multiple envelope elements, the last element determinted the
2041 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2042 Electronic Mail") by including:
2044 Auto-submitted: auto-generated
2046 in the messages that it generates (bounce messages and others, such as
2047 warnings). In the case of bounce messages for non-SMTP mesages, there was
2048 also a typo: it was using "Auto_submitted" (underscore instead of
2049 hyphen). Since every message generated by Exim is necessarily in response
2050 to another message, thes have all been changed to:
2052 Auto-Submitted: auto-replied
2054 in accordance with these statements in the RFC:
2056 The auto-replied keyword:
2058 - SHOULD be used on messages sent in direct response to another
2059 message by an automatic process,
2061 - MUST NOT be used on manually-generated messages,
2063 - MAY be used on Delivery Status Notifications (DSNs) and Message
2064 Disposition Notifications (MDNs),
2066 - MUST NOT be used on messages generated by automatic or periodic
2067 processes, except for messages which are automatic responses to
2070 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2071 to the default Received: header definition.
2073 PH/12 Added log selector acl_warn_skipped (default on).
2075 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2076 variables because (a) they are in the wrong storage pool and (b) even if
2077 they were copied, it wouldn't work properly because of the caching.
2079 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2080 checking when decoding. Apparently there are clients that generate
2081 overlong encoded strings. Why am I not surprised?
2083 PH/15 If the first argument of "${if match_address" was not empty, but did not
2084 contain an "@" character, Exim crashed. Now it writes a panic log message
2085 and treats the condition as false.
2087 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2089 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2090 extension "envelope-auth". The code is finished and in agreement with
2091 other implementations, but there is no documentation so far and in fact,
2092 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2093 not changing the active code.
2095 Print executed "if" and "elsif" statements when debugging is used. This
2096 helps a great deal to understand what a filter does.
2098 Document more things not specified clearly in RFC3028. I had all this
2099 sorted out, when out of a sudden new issues came to my mind. Oops."
2101 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2104 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2105 canonical form (as documented). However, after a host name lookup from
2106 the IP address, check_host() was doing a simple string comparison with
2107 addresses acquired from the DNS when checking that the found name did
2108 have the original IP as one of its addresses. Since any found IPv6
2109 addresses are likely to be in abbreviated form, the comparison could
2110 fail. Luckily, there already exists a function for doing the comparison
2111 by converting both addresses to binary, so now that is used instead of
2112 the text comparison.
2114 PH/20 There was another similar case to PH/19, when a complete host name was
2115 given in a host list; looking up its IP address could give an abbreviated
2116 form, whereas the current host's name might or might not be abbreviated.
2117 The same fix has been applied.
2123 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2124 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2127 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2128 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2130 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2132 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2133 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2134 Also, for "Auto-Submitted", treat anything other than "no" as
2135 non-personal, in accordance with RFC 3834. (Previously it treated
2136 anything starting "auto-" as non-personal.)
2138 TF/01 The control=submission/name=... option had a problem with syntax
2139 errors if the name included a slash character. The /name= option
2140 now slurps the rest of the string, so it can include any characters
2141 but it must come last in the list of options (after /sender_retain
2144 PH/05 Some modifications to the interface to the fake nameserver for the new
2152 TK/01 Added the "success_on_redirect" address verification option. See
2153 NewStuff for rationale and an example.
2155 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2157 PH/02 Patch to exigrep to allow it to work on syslog lines.
2159 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2160 fread() to skip over the body file's header line, because in Cygwin the
2161 header line is locked and is inaccessible.
2163 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2164 co-exist for some time) to make it clear that it is the Exim ID that is
2165 referenced, not the Message-ID: header line.
2167 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2168 string_format() function, because snprintf() does not exist on all
2171 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2172 ${stat: expansion item.
2174 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2175 protocol synchronization error", to keep the pedants happy.
2177 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2178 well as for IRIX systems, when gcc is being used. See the host.c source
2181 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2183 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2186 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2187 transport and to the smtp transport in LMTP mode.
2189 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2191 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2192 run a filter in a subprocess. This could lead to confusion in subsequent
2193 lookups in the parent process. There should also be a search_tidyup() at
2194 the end of the subprocess.
2196 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2197 only if the host matched helo_try_verify_hosts, which caused the
2198 verification to occur when the EHLO/HELO command was issued. The ACL just
2199 tested the remembered result. Now, if a previous verification attempt has
2200 not happened, "verify = helo" does it there and then.
2202 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2204 TK/03 Fix log output including CR from clamd.
2206 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2207 exist provoked a memory error which could cause a segfault.
2209 PH/15 Installed PCRE 6.2
2211 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2213 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2214 of the problem. Specifically, suggested +O2 rather than +O1 for the
2217 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2219 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2220 record, the port was not correctly logged when the outgoing_port log
2221 selector was set (it logged the transort's default port).
2223 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2224 fallback_hosts, and "hosts" in the smtp transport.
2226 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2227 host errors such as "Connection refused".
2229 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2230 authentication with radiusclient 0.4.9:
2232 - Error returned from rc_read_config was caught wrongly
2233 - Username/password not passed on to radius server due to wrong length.
2235 The presumption is that some radiusclient API changes for 4.51/PH/17
2236 were not taken care of correctly. The code is still untested by me (my
2237 Linux distribution still has 0.3.2 of radiusclient), but it was
2238 contributed by a Radius user.
2240 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2241 expanding the "port" option of the smtp transport.
2243 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2244 while reading a MIME header. Thanks to Tom Hughes for a patch.
2246 PH/24 Include config.h inside local_scan.h so that configuration settings are
2249 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2250 that in an ACL for RCPT (for example), you can examine exactly what was
2253 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2254 commands, but it was not correctly comparing the address with the actual
2255 client host address. Thus, it would show the EHLO address in Received:
2256 header lines when this was not necessary.
2258 PH/27 Added the % operator to ${eval:}.
2260 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2261 it should be ignoring failures (because with -C, for example, it has lost
2262 privilege). It wasn't ignoring creation failures other than "already
2265 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2268 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2269 because an input line was too long, either on its own, or by virtue of
2270 too many continuations, the temporary file was not being removed, and the
2271 return code was incorrect.
2273 PH/31 Missing "BOOL" in function definition in filtertest.c.
2275 PH/32 Applied Sieve patches from the maintainer.
2277 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2279 PH/33 Added "verify = not_blind".
2281 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2282 Local/Makefile (with some defaults set). These are used in built scripts
2283 such as exicyclog, but they have never been used in the exim_install
2284 script (though there are many overriding facilities there). I have
2285 arranged that the exim_install script now takes note of these two
2288 PH/35 Installed configuration files for Dragonfly.
2290 PH/36 When a locally submitted message by a trusted user did not contain a
2291 From: header, and the sender address was obtained from -f or from an SMTP
2292 MAIL command, and the trusted user did not use -F to supply a sender
2293 name, $originator_name was incorrectly used when constructing a From:
2294 header. Furthermore, $originator_name was used for submission mode
2295 messages from external hosts without From: headers in a similar way,
2296 which is clearly wrong.
2298 PH/37 Added control=suppress_local_fixups.
2300 PH/38 When log_selector = +received_sender was set, and the addition of the
2301 sender made the log line's construction buffer exactly full, or one byte
2302 less than full, an overflow happened when the terminating "\n" was
2305 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2306 when the result of a list match is failure because a DNS lookup failed.
2308 PH/40 RM_COMMAND is now used in the building process.
2310 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2311 the "build-* directories that it finds.
2313 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2314 address in a domain literal was a prefix of an interface address.
2316 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2317 when verifying a sender address, unless rewrite_headers is false.
2319 PH/44 Wrote a long comment about why errors_to addresses are verified as
2320 recipients, not senders.
2322 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2323 the ratelimit ACL was added.
2325 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2327 PH/46 Added extra information about PostgreSQL errors to the error string.
2329 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2330 avoiding the need to install special zones in a real server. This is
2331 backwards compatible; if it can't find the fake resolver, it drops back.
2332 Thus, both old and new test suites can be run.
2334 TF/02 Added util/ratelimit.pl
2336 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2337 clock is set back in time.
2339 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2340 Candler <B.Candler@pobox.com>.
2342 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2343 OK for addresses that are the result of redirections.
2345 PH/48 A number of further additions for the benefit of the new test suite,
2346 including a fake gethostbyname() that interfaces to the fake DNS resolver
2349 TF/06 The fix for widen_domains has also been applied to qualify_single and
2350 search_parents which are the other dnslookup options that can cause
2353 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2356 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2358 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2359 service" if it encountered an MX record with an empty target hostname.
2360 The message is now "an MX or SRV record indicated no SMTP service".
2362 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2363 if the DNS of the sending site is misconfigured. This is quite a
2364 common situation. This change restores the behaviour of treating a
2365 helo verification defer as a failure.
2367 PH/51 If self=fail was set on a router, the bounce message did not include the
2368 actual error message.
2374 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2376 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2377 command itself ended in error, the underlying message about the transport
2378 filter timeout was being overwritten with the pipe command error. Now the
2379 underlying error message should be appended to the second error message.
2381 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2383 PH/02 Reduce the amount of output that "make" produces by default. Full output
2384 can still be requested.
2386 PH/03 The warning log line about a condition test deferring for a "warn" verb
2387 was being output only once per connection, rather than after each
2388 occurrence (because it was using the same function as for successful
2389 "warn" verbs). This seems wrong, so I have changed it.
2391 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2392 it should not have, which might have caused a crash in the right
2393 circumstances, but probably never did.
2395 PH/04 Installed a modified version of Tony Finch's patch to make submission
2396 mode fix the return path as well as the Sender: header line, and to
2397 add a /name= option so that you can make the user's friendly name appear
2400 TF/03 Added the control = fakedefer ACL modifier.
2402 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2403 Mark Lowes for thorough testing.
2405 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2407 TK/03 Merged latest SRS patch from Miles Wilton.
2409 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2410 with the definition in sysexits.h (which is #included earlier).
2411 Fortunately, Exim does not actually use EX_OK. The code used to try to
2412 preserve the sysexits.h value, by assumimg that macro definitions were
2413 scanned for macro replacements. I have been disabused of this notion,
2414 so now the code just undefines EX_OK before #including unistd.h.
2416 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2417 in the smtp transport. When a block could not be written in a single
2418 write() function, the timeout was being re-applied to each part-write.
2419 This seems wrong - if the receiver was accepting one byte at a time it
2420 would take for ever. The timeout is now adjusted when this happens. It
2421 doesn't have to be particularly precise.
2423 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2424 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2426 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2427 without a domain if the check to <postmaster@domain> fails.
2429 SC/01 Eximstats: added -xls and the ability to specify output files
2430 (patch written by Frank Heydlauf).
2432 SC/02 Eximstats: use FileHandles for outputing results.
2434 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2436 SC/04 Eximstats: fixed display of large numbers with -nvr option
2438 SC/05 Eximstats: fixed merging of reports with empty tables.
2440 SC/06 Eximstats: added the -include_original_destination flag
2442 SC/07 Eximstats: removed tabs and trailing whitespace.
2444 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2446 TK/06 MBOX spool code: Add real "From " MBOX separator line
2447 so the .eml file is really in mbox format (even though
2448 most programs do not really care). Patch from Alex Miller.
2450 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2451 The latter is generated from $received_to and is only set if the
2452 message has one envelope recipient. SA can use these headers,
2453 obviously out-of-the-box. Patch from Alex Miller.
2455 PH/08 The ${def test on a variable was returning false if the variable's
2456 value was "0", contrary to what the specification has always said!
2457 The result should be true unless the variable is empty.
2459 PH/09 The syntax error of a character other than { following "${if
2460 def:variable_name" (after optional whitespace) was not being diagnosed.
2461 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2462 accidental colon was present, for example, could give incorrect results.
2464 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2465 result is used (not including appendfile, where other changes are about
2468 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2469 This involved changing a lot of size variables from int to off_t. It
2470 should work with maildirs and everything.
2472 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2473 spamd dying while we are connected to it.
2475 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2478 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2479 function for BDB 4.3.
2481 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2483 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2484 the sources for PCRE have been re-arranged and more files are now
2487 PH/15 The code I had for printing potentially long long variables in PH/11
2488 above was not the best (it lost precision). The length of off_t variables
2489 is now inspected at build time, and an appropriate printing format (%ld
2490 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2491 to be "long long int" or "long int". This is needed for the internal
2492 formatting function string_vformat().
2494 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2495 the configuration file to be ":syslog", then the script "guesses" where
2496 the logs files are, rather than using the compiled in default. In our
2497 case the guess is not the same as the compiled default, so the script
2498 suddenly stopped working when I started to use syslog. The patch checks
2499 to see if log_file_path is "". If so, it attempts to read it from exim
2500 with no configuration file to get the compiled in version, before it
2501 falls back to the previous guessing code."
2503 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2504 implementing BATV in an Exim configuration. See NewStuff for the gory
2507 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2508 Makefile that are specific to HP-UX.
2510 PH/18 If the "use_postmaster" option was set for a recipient callout together
2511 with the "random" option, the postmaster address was used as the MAIL
2512 FROM address for the random test, but not for the subsequent recipient
2513 test. It is now used for both.
2515 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2516 patch removes a few documentation additions to RFC 3028, because the
2517 latest draft now contains them. It adds the new en;ascii-case comparator
2518 and a new error check for 8bit text in MIME parts. Comparator and
2519 require names are now matched exactly. I enabled the subaddress
2520 extension, but it is not well tested yet (read: it works for me)."
2522 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2523 rework some of the code of TK/09 above to avoid the hardwired use of
2524 "%lld" and "long long". Replaced the call to snprintf() with a call to
2527 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2528 records point to non-existent hosts", "retry timeout exceeded", and
2529 "retry time not reached for any host after a long failure period".
2531 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2532 experimental DomainKeys support:
2534 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2535 (2) On an error such as an illegally used "control", the wrong name for
2536 the control was given.
2538 These problems did NOT occur unless DomainKeys support was compiled.
2540 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2542 PH/24 Added ${if match_ip condition.
2544 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2545 databases so that it will be absolutely obvious if a crash occurs in the
2546 DB library. This is a regular occurrence (often caused by mis-matched
2549 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2550 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2551 picked up on a user's system that detects such things. There doesn't seem
2552 to be a gcc warning option for this - only an attribute that has to be
2553 put on the function's prototype. It seems that in Fedora Core 4 they have
2554 set this on a number of new functions. No doubt there will be more in due
2557 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2558 specify a transport. However, if an address that was verified by such a
2559 router was the subject of a callout, Exim crashed because it tried to
2560 read the rcpt_include_affixes from the non-existent transport. Now it
2561 just assumes that the setting of that option is false. This bug was
2562 introduced by 4.51/PH/31.
2564 PH/28 Changed -d+all to exclude +memory, because that information is very
2565 rarely of interest, but it makes the output a lot bigger. People tend to
2566 do -d+all out of habit.
2568 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2569 code in os-type was giving problems when libc.so lives in lib64, like on
2572 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2573 aren't the modern standard, and it seems that some systems' include files
2574 don't always have them. Exim was already checking for some of the newer
2575 ones like T_AAAA, and defining it itself. I've added checks for all the
2576 record types that Exim uses.
2578 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2579 not automatically generating a new one, as it is supposed to. This
2580 prevented TLS from working. If the file did exist, but contained invalid
2581 data, a new version was generated, as expected. It was only the case of a
2582 non-existent file that was broken.
2584 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2585 with a change in libdomainkeys > 0.64.
2587 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2588 from DNS. If the selector record carries the flag, it now has
2589 precedence over the domain-wide flag.
2591 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2593 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2594 the use of a port name that isn't defined in /etc/services) occurred, the
2595 message was deferred as in a normal delivery, and thus remained on the
2596 spool, instead of being failed because of the mua_wrapper setting. This
2597 is now fixed, and I tidied up some of the mua_wrapper messages at the
2600 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2601 the messages in a hash of arrays rather than using individual hashes.
2602 This is a bit cleaner and results in dramatic memory savings, albeit
2603 at a slight CPU cost.
2605 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2606 as requested by Marc Sherman.
2608 SC/10 Eximstats: added histograms for user specified patterns as requested
2611 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2613 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2614 fopen() in the content-scanning modules that did not already have it.
2620 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2621 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2623 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2625 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2627 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2628 to test Sieve filters that use "vacation".
2630 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2631 that changes the way the GnuTLS parameters are stored in the cache file.
2632 The new format can be generated externally. For backward compatibility,
2633 if the data in the cache doesn't make sense, Exim assumes it has read an
2634 old-format file, and it generates new data and writes a new file. This
2635 means that you can't go back to an older release without removing the
2638 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2639 work if there are any delivery delays because "one_time" forces the
2640 parent to be marked "delivered", so its unseen clone is never tried
2641 again. For this reason, Exim now forbids the simultaneous setting of
2644 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2645 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2646 entry if you want to know the details. The fix, however, neglected to
2647 consider the case when local delivery batching is involved. The test for
2648 "previously delivered" was not happening when checking to see if an
2649 address could be batched with a previous (undelivered) one; under
2650 certain circumstances this could lead to multiple deliveries to the same
2653 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2654 in its include files, and this causes problems building Exim.
2656 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2657 header_syntax) but Exim was just ignoring anything given after a slash.
2658 In particular, this caused confusion with an attempt to use "verify =
2659 reverse_host_lookup/defer_ok". An error is now given when options are
2660 supplied for verify items that do not have them. (Maybe reverse_host_
2661 lookup should have a defer_ok option, but that's a different point.)
2663 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2664 defined by RFC 821) to 2048, because there were problems with some AUTH
2665 commands, and RFC 1869 says the size should be increased for extended
2666 SMTP commands that take arguments.
2668 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2671 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2672 "unknown" error; now it says that the functionality isn't in the binary.
2674 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2675 an address' error message when a string expansion fails (syntax or
2676 whatever). Otherwise the password may appear in the log. Following change
2677 PH/42 below, there is no longer a chance of it appearing in a bounce
2680 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2682 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2683 of its addresses was ever tried. (Bugzilla bug #2.)
2685 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2686 the result incorrectly in the debug output. (It correctly added a newline
2687 to what was transported.)
2689 TF/01 Added $received_time.
2691 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2692 commented out examples of how to interface to a virus scanner and to
2693 SpamAssassin. Also added commented examples of av_scanner and
2694 spamd_address settings.
2696 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2697 and controls are allowed in which ACLs. There were a couple of minor
2698 errors. Some of the entries in the conditions table (which is a table of
2699 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2700 negation of where the condition IS allowed.
2702 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2704 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2705 header file does not have a version number, so I've had to invent a new
2706 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2707 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2708 radiusclient), but it was contributed by a Radius user.
2710 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2711 files or directories, for OpenSSL.
2713 PH/19 When an Exim process that is running as root has to create an Exim log
2714 file, it does so in a subprocess that runs as exim:exim so as to get the
2715 ownership right at creation (otherwise, other Exim processes might see
2716 the file with the wrong ownership). There was no test for failure of this
2717 fork() call, which would lead to the process getting stuck as it waited
2718 for a non-existent subprocess. Forks do occasionally fail when resources
2719 run out. I reviewed all the other calls to fork(); they all seem to check
2722 PH/20 When checking for unexpected SMTP input at connect time (before writing
2723 the banner), Exim was not dealing correctly with a non-positive return
2724 from the read() function. If the client had disconnected by this time,
2725 the result was a log entry for a synchronization error with an empty
2726 string after "input=" when read() returned zero. If read() returned -1
2727 (an event I could not check), uninitialized data bytes were printed.
2728 There were reports of junk text (parts of files, etc) appearing after
2731 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2733 PH/22 Added support for macro redefinition, and (re)definition in between
2734 driver and ACL definitions.
2736 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2737 forgetting to use the resulting value; it was using the unexpanded value.
2739 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2740 hadn't been configured. The fix is from Juergen Kreileder, who
2741 understands it better than I do:
2743 "Here's what I see happening with three configured cyrus_sasl
2744 authenticators configured (plain, login, cram-md5):
2746 On startup auth_cyrus_sasl_init() gets called for each of these.
2747 This means three calls to sasl_listmech() without a specified mech_list.
2748 => SASL tests which mechs of all available mechs actually work
2749 => three warnings about OTP not working
2750 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2752 With the patch, sasl_listmech() also gets called three times. But now
2753 SASL's mech_list option is set to the server_mech specified in the the
2754 authenticator. Or in other words, the answer from sasl_listmech()
2755 gets limited to just the mech you're testing for (which is different
2757 => the return list contains just 'plain' or 'login', 'cram-md5' or
2758 nothing depending on the value of ob->server_mech.
2760 I've just tested the patch: Authentication still works fine,
2761 unavailable mechs specified in the exim configuration are still
2762 caught, and the auth.log warnings about OTP are gone."
2764 PH/25 When debugging is enabled, the contents of the command line are added
2765 to the debugging output, even when log_selector=+arguments is not
2768 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2769 answer is "GNU", and only if the return is "GNU/something" is the answer
2772 PH/27 $acl_verify_message is now set immediately after the failure of a
2773 verification in an ACL, and so is available in subsequent modifiers. In
2774 particular, the message can be preserved by coding like this:
2776 warn !verify = sender
2777 set acl_m0 = $acl_verify_message
2779 Previously, $acl_verify_message was set only while expanding "message"
2780 and "log_message" when a very denied access.
2782 PH/28 Modified OS/os.c-Linux with
2784 -#ifndef OS_LOAD_AVERAGE
2785 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2787 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2788 nomenclature these days.)
2790 PH/29 Installed patch from the Sieve maintainer that adds the options
2791 sieve_useraddress and sieve_subaddress to the redirect router.
2793 PH/30 In these circumstances:
2794 . Two addresses routed to the same list of hosts;
2795 . First host does not offer TLS;
2796 . First host accepts first address;
2797 . First host gives temporary error to second address;
2798 . Second host offers TLS and a TLS session is established;
2799 . Second host accepts second address.
2800 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2801 and peerdn, if requested) that were in fact used only for the second
2804 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2805 attention to any local part prefix or suffix that was matched by the
2806 router that accepted the address. It now behaves in the same way as it
2807 does for delivery: the affixes are removed from the local part unless
2808 rcpt_include_affixes is set on the transport.
2810 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2811 timeout during the DATA phase of an incoming message.
2813 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2814 applied a patch sanctioned by the Sieve maintainer.
2816 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2817 the uid or gid is negative. A case of a negative gid caused this to be
2818 noticed. The fix allows for either to be negative.
2820 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2821 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2824 PH/36 The change PH/12 above was broken. Fixed it.
2826 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2827 the grounds that routing the same address twice would always produce the
2828 same answer. This might have been true once, but it is certainly no
2829 longer true now. Routing a child address may depend on the previous
2830 routing that produced that child. Some complicated redirection strategies
2831 went wrong when messages had multiple recipients, and made Exim's
2832 behaviour dependent on the order in which the addresses were given.
2834 I have moved the duplicate checking until after the routing is complete.
2835 Exim scans the addresses that are assigned to local and remote
2836 transports, and removes any duplicates. This means that more work will be
2837 done, as duplicates will always all be routed, but duplicates are
2838 presumably rare, so I don't expect this is of any significance.
2840 For deliveries to pipes, files, and autoreplies, the duplicate checking
2841 still happens during the routing process, since they are not going to be
2844 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2845 It corrects a timeout issue with spamd. This is Ian's comment: "The
2846 background is that sometimes spamd either never reads data from a
2847 connection it has accepted, or it never writes response data. The exiscan
2848 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2849 blindly assumes that writes won't block so it may never time out."
2851 PH/39 Allow G after quota size as well as K and M.
2853 PH/40 The value set for $authenticated_id in an authenticator may not contain
2854 binary zeroes or newlines because the value is written to log lines and
2855 to spool files. There was no check on this. Now the value is run through
2856 the string_printing() function so that such characters are converted to
2857 printable escape sequences.
2859 PH/41 $message_linecount is a new variable that contains the total number of
2860 lines in the message. Compare $body_linecount, which is the count for the
2863 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2864 bounce and delay warning messages, except in certain special cases, which
2867 (a) An SMTP error message from a remote host;
2868 (b) A message specified in a :fail: redirection;
2869 (c) A message specified in a "fail" command in a system filter;
2870 (d) A message specified in a FAIL return from the queryprogram router;
2871 (e) A message specified by the cannot_route_message router option.
2873 In these cases only, Exim does include the error details in bounce and
2874 warning messages. There are also a few cases where bland messages such
2875 as "unrouteable address" or "local delivery error" are given.
2877 PH/43 $value is now also set for the "else" part of a ${run expansion.
2879 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2880 being worked on, but at least Exim now implements the latest version to
2883 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2884 process to complete was treated as a delivery failure, a timeout while
2885 writing the message to the pipe was logged, but erroneously treated as a
2886 successful delivery. Such timeouts include transport filter timeouts. For
2887 consistency with the overall process timeout, these timeouts are now
2888 treated as errors, giving rise to delivery failures by default. However,
2889 there is now a new Boolean option for the pipe transport called
2890 timeout_defer, which, if set TRUE, converts the failures into defers for
2891 both kinds of timeout. A transport filter timeout is now identified in
2894 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2895 systems where "make" and "gmake" are different, calling "gmake" at top
2896 level broke things. I've arranged for the value of $(MAKE) to be passed
2897 from the Makefile to this script so that it can call the same version of
2901 A note about Exim versions 4.44 and 4.50
2902 ----------------------------------------
2904 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2905 changes of various kinds. As a consequence, a big documentation update was
2906 needed. This delayed the release for rather longer than seemed good, especially
2907 in the light of a couple of (minor) security issues. Therefore, the changes
2908 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2909 release. So 4.44 and 4.50 are in effect two different branches that both start
2912 I have left the 4.50 change log unchanged; it contains all the changes since
2913 4.43. The change log for 4.44 is below; many of its items are identical to
2914 those for 4.50. This seems to be the most sensible way to preserve the
2915 historical information.
2921 1. Minor wording change to the doc/README.SIEVE file.
2923 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2924 computation of the current number of files was incorrect.
2926 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2927 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2930 4. Give more explanation in the error message when the command for a transport
2931 filter fails to execute.
2933 5. There are several places where Exim runs a non-Exim command in a
2934 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2935 was being done only for the command run by the queryprogram router. It is
2936 now done for all such subprocesses. The other cases are: ${run, transport
2937 filters, and the commands run by the lmtp and pipe transports.
2939 6. Added CONFIGURE_GROUP build-time option.
2941 7. Some older OS have a limit of 256 on the maximum number of file
2942 descriptors. Exim was using setrlimit() to set 1000 as a large value
2943 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2944 systems. I've change it so that if it can't get 1000, it tries for 256.
2946 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2947 was an oversight, and furthermore, ever since the addition of extra
2948 controls (e.g. 4.43/32), the checks on when to allow different forms of
2949 "control" were broken. There should now be diagnostics for all cases when a
2950 control that does not make sense is encountered.
2952 9. Added the /retain_sender option to "control=submission".
2954 10. $recipients is now available in the predata ACL (oversight).
2956 11. Tidy the search cache before the fork to do a delivery from a message
2957 received from the command line. Otherwise the child will trigger a lookup
2958 failure and thereby defer the delivery if it tries to use (for example) a
2959 cached ldap connection that the parent has called unbind on.
2961 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2962 of $address_data from the recipient verification was clobbered by the
2963 sender verification.
2965 13. The value of address_data from a sender verification is now available in
2966 $sender_address_data in subsequent conditions in the ACL statement.
2968 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2970 15. Added a new option "connect=<time>" to callout options, to set a different
2973 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2974 was its contents. (It was OK if the option was not defined at all.)
2976 17. A "Completed" log line is now written for messages that are removed from
2977 the spool by the -Mrm option.
2979 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2980 information about exactly what failed.
2982 19. Added -dd to debug only the daemon process.
2984 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2985 handles timeouts, both on the server side and network timeouts. Renamed the
2986 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2988 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2989 It is now set to "smtps".
2991 22. $host_address is now set to the target address during the checking of
2992 ignore_target_hosts.
2994 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2995 being passed; this would have caused $sender_host_name to have been used if
2996 matching the list had actually called for a host name (not very likely,
2997 since this list is usually IP addresses). A host name is now passed as
3000 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3001 code that specifies a non-restarting handler (typically sigaction() in
3002 modern systems) in an attempt to fix a rare and obscure crash bug.
3004 25. Narrowed the window for a race in the daemon that could cause it to ignore
3005 SIGCHLD signals. This is not a major problem, because they are used only to
3006 wake it up if nothing else does.
3008 26. A malformed maildirsize file could cause Exim to calculate negative values
3009 for the mailbox size or file count. Odd effects could occur as a result.
3010 The maildirsize information is now recalculated if the size or filecount
3013 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3014 support for a long time. Removed HAVE_SYS_VFS_H.
3016 28. Installed the latest version of exipick from John Jetmore.
3018 29. In an address list, if the pattern was not a regular expression, an empty
3019 subject address (from a bounce message) matched only if the pattern was an
3020 empty string. Non-empty patterns were not even tested. This was the wrong
3021 because it is perfectly reasonable to use an empty address as part of a
3022 database query. An empty address is now tested by patterns that are
3023 lookups. However, all the other forms of pattern expect the subject to
3024 contain a local part and a domain, and therefore, for them, an empty
3025 address still always fails if the pattern is not itself empty.
3027 30. Exim went into a mad DNS loop when attempting to do a callout where the
3028 host was specified on an smtp transport, and looking it up yielded more
3029 than one IP address.
3031 31. Re-factored the code for checking spool and log partition space into a
3032 function that finds that data and another that does the check. The former
3033 is then used to implement four new variables: $spool_space, $log_space,
3034 $spool_inodes, and $log_inodes.
3036 32. The RFC2047 encoding function was originally intended for short strings
3037 such as real names; it was not keeping to the 75-character limit for
3038 encoded words that the RFC imposes. It now respects the limit, and
3039 generates multiple encoded words if necessary. To be on the safe side, I
3040 have increased the buffer size for the ${rfc2047: expansion operator from
3043 33. It is now permitted to omit both strings after an "if" condition; if the
3044 condition is true, the result is "true". As before, when the second string
3045 is omitted, a false condition yields an empty string. This makes it less
3046 cumbersome to write custom ACL and router conditions.
3048 34. Failure to deliver a bounce message always caused it to be frozen, even if
3049 there was an errors_to setting on the router. The errors_to setting is now
3052 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3053 canonical form (fully expanded) before being placed in
3054 $sender_host_address.
3056 36. The table in the code that translates DNS record types into text (T_A to
3057 "A" for instance) was missing entries for NS and CNAME. It is just possible
3058 that this could have caused confusion if both these types were looked up
3059 for the same domain, because the text type is used as part of Exim's
3060 per-process caching. But the chance of anyone hitting this buglet seems
3063 37. The dnsdb lookup has been extended in a number of ways.
3065 (1) There is a new type, "zns", which walks up the domain tree until it
3066 finds some nameserver records. It should be used with care.
3068 (2) There is a new type, "mxh", which is like "mx" except that it returns
3069 just the host names, not the priorities.
3071 (3) It is now possible to give a list of domains (or IP addresses) to be
3072 looked up. The behaviour when one of the lookups defers can be
3073 controlled by a keyword.
3075 (4) It is now possible to specify the separator character for use when
3076 multiple records are returned.
3078 38. The dnslists ACL condition has been extended: it is now possible to supply
3079 a list of IP addresses and/or domains to be looked up in a particular DNS
3082 39. Added log_selector=+queue_time_overall.
3084 40. When running the queue in the test harness, wait just a tad after forking a
3085 delivery process, to get repeatability of debugging output.
3087 41. Include certificate and key file names in error message when GnuTLS fails
3088 to set them up, because the GnuTLS error message doesn't include the name
3089 of the failing file when there is a problem reading it.
3091 42. Allow both -bf and -bF in the same test run.
3093 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3095 44. The "Exiscan patch" is now merged into the mainline Exim source.
3097 45. Sometimes the final signoff response after QUIT could fail to get
3098 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3099 < 0 before doing a fflush(). This bug looks as though it goes back to the
3100 introduction of TLS in release 3.20, but "sometimes" must have been rare
3101 because the tests only now provoked it.
3103 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3104 (this can affect the format of dates).
3106 47. exim_tidydb, when checking for the continued existence of a message for
3107 which it has found a message-specific retry record, was not finding
3108 messages that were in split spool directories. Consequently, it was
3109 deleting retry records that should have stayed in existence.
3111 48. Steve fixed some bugs in eximstats.
3113 49. The SPA authentication driver was not abandoning authentication and moving
3114 on to the next authenticator when an expansion was forced to fail,
3115 contradicting the general specification for all authenticators. Instead it
3116 was generating a temporary error. It now behaves as specified.
3118 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3119 (the order specifies the preference for clients). The order is now AES256,
3120 AES128, 3DES, ARCFOUR128.
3122 51. Small patch to Sieve code - explicitly set From: when generating an
3125 52. Exim crashed if a remote delivery caused a very long error message to be
3126 recorded - for instance if somebody sent an entire SpamAssassin report back
3127 as a large number of 550 error lines. This bug was coincidentally fixed by
3128 increasing the size of one of Exim's internal buffers (big_buffer) that
3129 happened as part of the Exiscan merge. However, to be on the safe side, I
3130 have made the code more robust (and fixed the comments that describe what
3133 53. Now that there can be additional text after "Completed" in log lines (if
3134 the queue_time_overall log selector is set), a one-byte patch to exigrep
3135 was needed to allow it to recognize "Completed" as not the last thing in
3138 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3139 patch that reportedly fixes this has been added. I am not expert enough to
3140 create a test for it. This is what the patch creator wrote:
3142 "I found a little strange behaviour of ldap code when working with
3143 Windows 2003 AD Domain, where users was placed in more than one
3144 Organization Units. When I tried to give exim partial DN, the exit code
3145 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3146 But simultaneously result of request was absolutely normal ldap result,
3147 so I produce this patch..."
3149 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3150 REFERENCE, so I have modified the code to exclude the patch when that macro
3153 55. Some experimental protocols are using DNS PTR records for new purposes. The
3154 keys for these records are domain names, not reversed IP addresses. The
3155 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3156 leaves it alone. Component reversal etc. now happens only for IP addresses.
3159 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3161 57. Double the size of the debug message buffer (to 2048) so that more of very
3162 long debug lines gets shown.
3164 58. The exicyclog utility now does better if the number of log files to keep
3165 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3167 59. Two changes related to the smtp_active_hostname option:
3169 (1) $smtp_active_hostname is now available as a variable.
3170 (2) The default for smtp_banner uses $smtp_active_hostname instead
3171 of $primary_hostname.
3173 60. The host_aton() function is supposed to be passed a string that is known
3174 to be a valid IP address. However, in the case of IPv6 addresses, it was
3175 not checking this. This is a hostage to fortune. Exim now panics and dies
3176 if the condition is not met. A case was found where this could be provoked
3177 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3178 components; fortuitously, this particular loophole had already been fixed
3179 by change 4.50/55 above.
3181 If there are any other similar loopholes, the new check in host_aton()
3182 itself should stop them being exploited. The report I received stated that
3183 data on the command line could provoke the exploit when Exim was running as
3184 exim, but did not say which command line option was involved. All I could
3185 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3186 running as the user.
3189 61. There was a buffer overflow vulnerability in the SPA authentication code
3190 (which came originally from the Samba project). I have added a test to the
3191 spa_base64_to_bits() function which I hope fixes it.
3194 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3195 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3197 63. The daemon start-up calls getloadavg() while still root for those OS that
3198 need the first call to be done as root, but it missed one case: when
3199 deliver_queue_load_max is set with deliver_drop_privilege. This is
3200 necessary for the benefit of the queue runner, because there is no re-exec
3201 when deliver_drop_privilege is set.
3203 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3204 This has been fixed.
3206 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3207 was in use, was not putting the data itself into the right store pool;
3208 consequently, it could be overwritten for a subsequent message in the same
3209 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3212 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3214 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3215 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3216 the function were treating the return as a boolean value, which happened to
3217 work because 0=false and not-0=true, but is not correct code.
3219 68. The host_aton() function was not handling scoped IPv6 addresses (those
3220 with, for example, "%eth0" on the end) correctly.
3222 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3223 negated items (that is, ~something) in unsigned ints. Some compilers
3224 apparently mutter when there is no cast.
3226 70. If an address verification called from an ACL failed, and did not produce a
3227 user-specific message (i.e. there was only a "system" message), nothing was
3228 put in $acl_verify_message. In this situation, it now puts the system
3231 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3232 session; change 4.31/43 added the unwanted input to the log line - except
3233 that it did not do this in the start of session case. It now does.
3235 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3236 This is wrong and can cause the other end to generate a synchronization
3237 error if it is another Exim or anything else that does the synchronization
3238 check. A QUIT command is no longer sent after a timeout.
3240 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3241 during host lookups.
3243 74. The defer_ok option of callout verification was not working if it was used
3244 when verifying addresses in header lines, that is, for this case:
3246 verify = header_sender/callout=defer_ok
3248 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3249 those file descriptors could be used for SMTP connections. If anything
3250 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3251 could be sent to the SMTP client, causing chaos. The daemon now opens
3252 stdin, stdout, and stderr to /dev/null when it puts itself into the
3255 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3256 log by default. The user can override this with suitable Perl magic.
3258 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3259 the log message when discard triggers, was not working for the DATA ACL or
3260 for the non-SMTP ACL.
3262 78. Error message wording change in sieve.c.
3264 79. If smtp_accept_max_per_host was set, the number of connections could be
3265 restricted to fewer than expected, because the daemon was trying to set up
3266 a new connection before checking whether the processes handling previous
3267 connections had finished. The check for completed processes is now done
3268 earlier. On busy systems, this bug wouldn't be noticed because something
3269 else would have woken the daemon, and it would have reaped the completed
3272 80. If a message was submitted locally by a user whose login name contained one
3273 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3274 It caused a spool format error. I have fixed the spool reading code. A
3275 related problem was that the "from" clause in the Received: line became
3276 illegal because of the space(s). It is now covered by ${quote_local_part.
3278 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3281 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3283 83. Patch from Sieve maintainer to fix unterminated string problem in
3284 "vacation" handling.
3286 84. Some minor changes to the Linux configuration files to help with other
3287 OS variants using glibc.
3289 85. One more patch for Sieve to update vacation handling to latest spec.
3292 ----------------------------------------------------
3293 See the note above about the 4.44 and 4.50 releases.
3294 ----------------------------------------------------
3300 1. Change 4.43/35 introduced a bug that caused file counts to be
3301 incorrectly computed when quota_filecount was set in an appendfile
3304 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3305 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3308 3. Give more explanation in the error message when the command for a transport
3309 filter fails to execute.
3311 4. There are several places where Exim runs a non-Exim command in a
3312 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3313 was being done only for the command run by the queryprogram router. It is
3314 now done for all such subprocesses. The other cases are: ${run, transport
3315 filters, and the commands run by the lmtp and pipe transports.
3317 5. Some older OS have a limit of 256 on the maximum number of file
3318 descriptors. Exim was using setrlimit() to set 1000 as a large value
3319 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3320 systems. I've change it so that if it can't get 1000, it tries for 256.
3322 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3323 was an oversight, and furthermore, ever since the addition of extra
3324 controls (e.g. 4.43/32), the checks on when to allow different forms of
3325 "control" were broken. There should now be diagnostics for all cases when a
3326 control that does not make sense is encountered.
3328 7. $recipients is now available in the predata ACL (oversight).
3330 8. Tidy the search cache before the fork to do a delivery from a message
3331 received from the command line. Otherwise the child will trigger a lookup
3332 failure and thereby defer the delivery if it tries to use (for example) a
3333 cached ldap connection that the parent has called unbind on.
3335 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3336 of $address_data from the recipient verification was clobbered by the
3337 sender verification.
3339 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3340 was its contents. (It was OK if the option was not defined at all.)
3342 11. A "Completed" log line is now written for messages that are removed from
3343 the spool by the -Mrm option.
3345 12. $host_address is now set to the target address during the checking of
3346 ignore_target_hosts.
3348 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3349 being passed; this would have caused $sender_host_name to have been used if
3350 matching the list had actually called for a host name (not very likely,
3351 since this list is usually IP addresses). A host name is now passed as
3354 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3355 code that specifies a non-restarting handler (typically sigaction() in
3356 modern systems) in an attempt to fix a rare and obscure crash bug.
3358 15. Narrowed the window for a race in the daemon that could cause it to ignore
3359 SIGCHLD signals. This is not a major problem, because they are used only to
3360 wake it up if nothing else does.
3362 16. A malformed maildirsize file could cause Exim to calculate negative values
3363 for the mailbox size or file count. Odd effects could occur as a result.
3364 The maildirsize information is now recalculated if the size or filecount
3367 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3368 support for a long time. Removed HAVE_SYS_VFS_H.
3370 18. Updated exipick to current release from John Jetmore.
3372 19. Allow an empty sender to be matched against a lookup in an address list.
3373 Previously the only cases considered were a regular expression, or an
3376 20. Exim went into a mad DNS lookup loop when doing a callout where the
3377 host was specified on the transport, if the DNS lookup yielded more than
3380 21. The RFC2047 encoding function was originally intended for short strings
3381 such as real names; it was not keeping to the 75-character limit for
3382 encoded words that the RFC imposes. It now respects the limit, and
3383 generates multiple encoded words if necessary. To be on the safe side, I
3384 have increased the buffer size for the ${rfc2047: expansion operator from
3387 22. Failure to deliver a bounce message always caused it to be frozen, even if
3388 there was an errors_to setting on the router. The errors_to setting is now
3391 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3392 canonical form (fully expanded) before being placed in
3393 $sender_host_address.
3395 24. Updated eximstats to version 1.33
3397 25. Include certificate and key file names in error message when GnuTLS fails
3398 to set them up, because the GnuTLS error message doesn't include the name
3399 of the failing file when there is a problem reading it.
3401 26. Expand error message when OpenSSL has problems setting up cert/key files.
3404 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3405 (this can affect the format of dates).
3407 28. exim_tidydb, when checking for the continued existence of a message for
3408 which it has found a message-specific retry record, was not finding
3409 messages that were in split spool directories. Consequently, it was
3410 deleting retry records that should have stayed in existence.
3412 29. eximstats updated to version 1.35
3413 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3414 1.35 - bugfix such that pie charts by volume are generated correctly
3416 30. The SPA authentication driver was not abandoning authentication and moving
3417 on to the next authenticator when an expansion was forced to fail,
3418 contradicting the general specification for all authenticators. Instead it
3419 was generating a temporary error. It now behaves as specified.
3421 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3422 (the order specifies the preference for clients). The order is now AES256,
3423 AES128, 3DES, ARCFOUR128.
3425 31. Small patch to Sieve code - explicitly set From: when generating an
3428 32. Exim crashed if a remote delivery caused a very long error message to be
3429 recorded - for instance if somebody sent an entire SpamAssassin report back
3430 as a large number of 550 error lines. This bug was coincidentally fixed by
3431 increasing the size of one of Exim's internal buffers (big_buffer) that
3432 happened as part of the Exiscan merge. However, to be on the safe side, I
3433 have made the code more robust (and fixed the comments that describe what
3436 33. Some experimental protocols are using DNS PTR records for new purposes. The
3437 keys for these records are domain names, not reversed IP addresses. The
3438 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3439 leaves it alone. Component reversal etc. now happens only for IP addresses.
3442 34. The host_aton() function is supposed to be passed a string that is known
3443 to be a valid IP address. However, in the case of IPv6 addresses, it was
3444 not checking this. This is a hostage to fortune. Exim now panics and dies
3445 if the condition is not met. A case was found where this could be provoked
3446 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3447 components; fortuitously, this particular loophole had already been fixed
3448 by change 4.50/55 or 4.44/33 above.
3450 If there are any other similar loopholes, the new check in host_aton()
3451 itself should stop them being exploited. The report I received stated that
3452 data on the command line could provoke the exploit when Exim was running as
3453 exim, but did not say which command line option was involved. All I could
3454 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3455 running as the user.
3458 35. There was a buffer overflow vulnerability in the SPA authentication code
3459 (which came originally from the Samba project). I have added a test to the
3460 spa_base64_to_bits() function which I hope fixes it.
3463 36. The daemon start-up calls getloadavg() while still root for those OS that
3464 need the first call to be done as root, but it missed one case: when
3465 deliver_queue_load_max is set with deliver_drop_privilege. This is
3466 necessary for the benefit of the queue runner, because there is no re-exec
3467 when deliver_drop_privilege is set.
3469 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3470 was in use, was not putting the data itself into the right store pool;
3471 consequently, it could be overwritten for a subsequent message in the same
3472 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3475 38. Sometimes the final signoff response after QUIT could fail to get
3476 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3477 < 0 before doing a fflush(). This bug looks as though it goes back to the
3478 introduction of TLS in release 3.20, but "sometimes" must have been rare
3479 because the tests only now provoked it.
3485 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3486 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3487 Now it returns an integer. A number of calls were still expecting a T/F
3488 return. Fortuitously, in all cases, the tests worked in OK situations,
3489 which is the norm. However, things would have gone wrong on any write
3490 failures on the smtp file descriptor. This function is used when sending
3491 messages over SMTP and also when doing verify callouts.
3493 2. When Exim is called to do synchronous delivery of a locally submitted
3494 message (the -odf or -odi options), it no longer closes stderr before doing
3497 3. Implemented the mua_wrapper option.
3499 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3501 5. Implemented the functions header_remove(), header_testname(),
3502 header_add_at_position(), and receive_remove_recipient(), and exported them
3505 6. If an ACL "warn" statement specified the addition of headers, Exim already
3506 inserted X-ACL-Warn: at the start if there was no header name. However, it
3507 was not making this test for the second and subsequent header lines if
3508 there were newlines in the string. This meant that an invalid header could
3509 be inserted if Exim was badly configured.
3511 7. Allow an ACL "warn" statement to add header lines at the start or after all
3512 the Received: headers, as well as at the end.
3514 8. Added the rcpt_4xx retry error code.
3516 9. Added postmaster_mailfrom=xxx to callout verification option.
3518 10. Added mailfrom=xxxx to the callout verification option, for verify=
3521 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3522 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3524 12. Inserted some casts to stop certain compilers warning when using pointer
3525 differences as field lengths or precisions in printf-type calls (mostly
3526 affecting debugging statements).
3528 13. Added optional readline() support for -be (dynamically loaded).
3530 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3531 same clock tick as a message's arrival, so that its received time was the
3532 same as the "first fail" time on the retry record, and that message
3533 remained on the queue past the ultimate address timeout, every queue runner
3534 would try a delivery (because it was past the ultimate address timeout) but
3535 after another failure, the ultimate address timeout, which should have then
3536 bounced the address, did not kick in. This was a "< instead of <=" error;
3537 in most cases the first failure would have been in the next clock tick
3538 after the received time, and all would be well.
3540 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3541 being recognized when the domain list was tested by the match_domain
3542 condition in an expansion string.
3544 16. Added the ${str2b64: operator.
3546 17. Exim was always calling setrlimit() to set a large limit for the number of
3547 processes, without checking whether the existing limit was already
3548 adequate. (It did check for the limit on file descriptors.) Furthermore,
3549 errors from getrlimit() and setrlimit() were being ignored. Now they are
3550 logged to the main and panic logs, but Exim does carry on, to try to do its
3551 job under whatever limits there are.
3553 18. Imported PCRE 5.0.
3555 19. Trivial typo in log message " temporarily refused connection" (the leading
3558 20. If the log selector return_path_on_delivery was set and an address was
3559 redirected to /dev/null, the delivery process crashed because it assumed
3560 that a return path would always be set for a "successful" delivery. In this
3561 case, the whole delivery is bypassed as an optimization, and therefore no
3564 21. Internal re-arrangement: the function for sending a challenge and reading
3565 a response while authentication was assuming a zero-terminated challenge
3566 string. It's now changed to take a pointer and a length, to allow for
3567 binary data in such strings.
3569 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3571 23. Exim was not respecting finduser_retries when seeking the login of the
3572 uid under which it was called; it was always trying 10 times. (The default
3573 setting of finduser_retries is zero.) Also, it was sleeping after the final
3574 failure, which is pointless.
3576 24. Implemented tls_on_connect_ports.
3578 25. Implemented acl_smtp_predata.
3580 26. If the domain in control=submission is set empty, Exim assumes that the
3581 authenticated id is a complete email address when it generates From: or
3582 Sender: header lines.
3584 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3585 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3586 chown and chgrp in /bin and hostname in /usr/bin.
3588 28. Exim was keeping the "process log" file open after each use, just as it
3589 does for the main log. This opens the possibility of it remaining open for
3590 long periods when the USR1 signal hits a daemon. Occasional processlog
3591 errors were reported, that could have been caused by this. Anyway, it seems
3592 much more sensible not to leave this file open at all, so that is what now
3595 29. The long-running daemon process does not normally write to the log once it
3596 has entered its main loop, and it closes the log before doing so. This is
3597 so that log files can straightforwardly be renamed and moved. However,
3598 there are a couple of unusual error situations where the daemon does write
3599 log entries, and I had neglected to close the log afterwards.
3601 30. The text of an SMTP error response that was received during a remote
3602 delivery was being truncated at 512 bytes. This is too short for some of
3603 the long messages that one sometimes sees. I've increased the limit to
3606 31. It is now possible to make retry rules that apply only when a message has a
3607 specific sender, in particular, an empty sender.
3609 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3610 it possible to be selective about when SMTP synchronization is enforced.
3612 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3614 32. Implemented hosts_connection_nolog.
3616 33. Added an ACL for QUIT.
3618 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3621 35. Added mailbox_size and mailbox_filecount to appendfile.
3623 36. Added control = no_multiline_responses to ACLs.
3625 37. There was a bug in the logic of the code that waits for the clock to tick
3626 in the case where the clock went backwards by a substantial amount such
3627 that the microsecond fraction of "now" was more than the microsecond
3628 fraction of "then" (but the whole seconds number was less).
3630 38. Added support for the libradius Radius client library this is found on
3631 FreeBSD (previously only the radiusclient library was supported).
3637 1. When certain lookups returned multiple values in the form name=value, the
3638 quoting of the values was not always being done properly. Specifically:
3639 (a) If the value started with a double quote, but contained no whitespace,
3641 (b) If the value contained whitespace other than a space character (i.e.
3642 tabs or newlines or carriage returns) it was not quoted.
3643 This fix has been applied to the mysql and pgsql lookups by writing a
3644 separate quoting function and calling it from the lookup code. The fix
3645 should probably also be applied to nisplus, ibase and oracle lookups, but
3646 since I cannot test any of those, I have not disturbed their existing code.
3648 2. A hit in the callout cache for a specific address caused a log line with no
3649 reason for rejecting RCPT. Now it says "Previous (cached) callout
3650 verification failure".
3652 3. There was an off-by-one bug in the queryprogram router. An over-long
3653 return line was truncated at 256 instead of 255 characters, thereby
3654 overflowing its buffer with the terminating zero. As well as fixing this, I
3655 have increased the buffer size to 1024 (and made a note to document this).
3657 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3658 when Exim is waiting for an SMTP response from a remote server, Exim
3659 restarts its select() call on the socket, thereby resetting its timeout.
3660 This is not a problem when such interrupts are rare. Somebody set up a cron
3661 job to run exiwhat every 2 minutes, which is less than the normal select()
3662 timeout (5 or 10 minutes). This meant that the select() timeout never
3663 kicked in because it was always reset. I have fixed this by comparing the
3664 time when an interrupt arrives with the time at the start of the first call
3665 to select(). If more time than the timeout has elapsed, the interrupt is
3666 treated as a timeout.
3668 5. Some internal re-factoring in preparation for the addition of Sieve
3669 extensions (by MH). In particular, the "personal" test is moved to a
3670 separate function, and given an option for scanning Cc: and Bcc: (which is
3671 not set for Exim filters).
3673 6. When Exim created an email address using the login of the caller as the
3674 local part (e.g. when creating a From: or Sender: header line), it was not
3675 quoting the local part when it contained special characters such as @.
3677 7. Installed new OpenBSD configuration files.
3679 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3680 try to make them clearer.
3682 9. Callout options, other than the timeout value, were being ignored when
3683 verifying sender addresses in header lines. For example, when using
3685 verify = header_sender/callout=no_cache
3687 the cache was (incorrectly) being used.
3689 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3690 only the Cygwin environment.
3692 11. When return_path_on_delivery was set as a log selector, if different remote
3693 addresses in the same message used different return paths and parallel
3694 remote delivery occurred, the wrong values would sometimes be logged.
3695 (Whenever a remote delivery process finished, the return path value from
3696 the most recently started remote delivery process was logged.)
3698 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3699 lines when AUTH and/or TLS are in use. This is the "received protocol"
3700 field. Exim used to use "asmtp" for authenticated SMTP, without any
3701 indication (in the protocol name) for TLS use. Now it follows the RFC and
3702 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3703 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3704 names appear in log lines as well as in Received: header lines.
3706 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3707 extensions, and comparison tests, and to fix some bugs.
3709 14. Changes to the "personal" filter test:
3711 (1) The test was buggy in that it was just doing the equivalent of
3712 "contains" tests on header lines. For example, if a user's address was
3713 anne@some.where, the "personal" test would incorrectly be true for
3715 To: susanne@some.where
3717 This test is now done by extracting each address from the header in turn,
3718 and checking the entire address. Other tests that are part of "personal"
3719 are now done using regular expressions (for example, to check local parts
3720 of addresses in From: header lines).
3722 (2) The list of non-personal local parts in From: addresses has been
3723 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3724 taken from the Sieve specification recommendations.
3726 (3) If the message contains any header line starting with "List-" it is
3727 treated as non-personal.
3729 (4) The test for "circular" in the Subject: header line has been removed
3730 because it now seems ill-conceived.
3732 15. Minor typos in src/EDITME comments corrected.
3734 16. Installed latest exipick from John Jetmore.
3736 17. If headers_add on a router specified a text string that was too long for
3737 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3738 of string_sprintf() is now avoided.
3740 18. $message_body_size was not set (it was always zero) when running the DATA
3741 ACL and the local_scan() function.
3743 19. For the "mail" command in an Exim filter, no default was being set for
3744 the once_repeat time, causing a random time value to be used if "once" was
3745 specified. (If the value happened to be <= 0, no repeat happened.) The
3746 default is now 0s, meaning "never repeat". The "vacation" command was OK
3747 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3748 (I found it when inspecting the code).
3750 20. There is now an overall timeout for performing a callout verification. It
3751 defaults to 4 times the callout timeout, which applies to individual SMTP
3752 commands during the callout. The overall timeout applies when there is more
3753 than one host that can be tried. The timeout is checked before trying the
3754 next host. This prevents very long delays if there are a large number of
3755 hosts and all are timing out (e.g. when the network connections are timing
3756 out). The value of the overall timeout can be changed by specifying an
3757 additional sub-option for "callout", called "maxwait". For example:
3759 verify = sender/callout=5s,maxwait=20s
3761 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3762 to the end before writing, but this should make it even safer).
3764 22. Exim was forgetting that it had advertised PIPELINING for the second and
3765 subsequent messages on an SMTP connection. It was also not resetting its
3766 memory on STARTTLS and an internal HELO.
3768 23. When Exim logs an SMTP synchronization error within a session, it now
3769 records whether PIPELINING has been advertised or not.
3771 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3772 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3773 rather than long int.
3775 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3777 26. Added the never_mail option to autoreply.
3783 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3784 crash if the getsockname() call failed; this can happen if a connection is
3785 closed very soon after it is established. The problem was simply in the
3786 order in which certain operations were done, causing Exim to try to write
3787 to the SMTP stream before it had set up the file descriptor. The bug has
3788 been fixed by making things happen in the correct order.
3794 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3795 before the connection was closed, thus losing the rejection response.
3797 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3798 some early Solaris releases, but causes trouble in current releases where
3799 socklen_t is defined.
3801 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3804 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3807 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3808 that releasing the top of it at the end releases what was used for sub-
3809 expansions (unless the block got too big). However, discard this block if
3810 the first thing is a variable or header, so that we can use its block when
3811 it is dynamic (useful for very large $message_headers, for example).
3813 6. Lookups now cache *every* query, not just the most recent. A new, separate
3814 store pool is used for this. It can be recovered when all lookup caches are
3815 flushed. Lookups now release memory at the end of their result strings.
3816 This has involved some general refactoring of the lookup sources.
3818 7. Some code has been added to the store_xxx() functions to reduce the amount
3819 of flapping under certain conditions.
3821 8. log_incoming_interface used to affect only the <= reception log lines. Now
3822 it causes the local interface and port to be added to several more SMTP log
3823 lines, for example "SMTP connection from", and rejection lines.
3825 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3827 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3829 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3830 could be overwritten at the end of the current message (or the start of a
3831 new message if it was set in a HELO ACL). The value is now preserved for
3832 the duration of the SMTP connection.
3834 12. If a transport had a headers_rewrite setting, and a matching header line
3835 contained an unqualified address, that address was qualified, even if it
3836 did not match any rewriting rules. The underlying bug was that the values
3837 of the flags that permit the existence of unqualified sender and recipient
3838 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3839 non-local messages, and by -bnq for local messages) were not being
3840 preserved with the message after it was received.
3842 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3843 "next input=" as part of the text comprising the host identity instead of
3844 the correct text. The code was using the same buffer for two different
3845 strings. However, depending on which order the printing function evaluated
3846 its arguments, the bug did not always show up. Under Linux, for example, my
3847 test suite worked just fine.
3849 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3850 doesn't work with some older versions of Perl. It has been changed to "my",
3851 which in any case is probably the better facility to use.
3853 15. A really picky compiler found some instances of statements for creating
3854 error messages that either had too many or two few arguments for the format
3857 16. The size of the buffer for calls to the DNS resolver has been increased
3858 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3859 for addresses that have a lot of PTR records. This alleviates a problem; it
3860 does not fully solve it.
3862 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3863 buffer now truncates the list and logs the incident, which is the same
3864 action as happens when Exim is looking up a host name and its aliases.
3865 Previously in this situation something unpredictable would happen;
3866 sometimes it was "internal error: store_reset failed".
3868 18. If a server dropped the connection unexpectedly when an Exim client was
3869 using GnuTLS and trying to read a response, the client delivery process
3870 crashed while trying to generate an error log message.
3872 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3873 string, for example:
3875 warn message = H1: something\nH2: something
3877 the text was added as a single header line from Exim's point of view
3878 though it ended up OK in the delivered message. However, searching for the
3879 second and subsequent header lines using $h_h2: did not work. This has been
3880 fixed. Similarly, if a system filter added multiple headers in this way,
3881 the routers could not see them.
3883 20. Expanded the error message when iplsearch is called with an invalid key to
3884 suggest using net-iplsearch in a host list.
3886 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3887 ACLs are no longer actually imposed (and a message to that effect is
3890 22. If a "gecos" field in a passwd entry contained escaped characters, in
3891 particular, if it contained a \" sequence, Exim got it wrong when building
3892 a From: or a Sender: header from that name. A second bug also caused
3893 incorrect handling when an unquoted " was present following a character
3894 that needed quoting.
3896 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3897 was not being matched caselessly.
3899 24. Arranged for all hyphens in the exim.8 source to be escaped with
3902 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3903 properly. Recipient callout cache records were still being keyed to include
3904 the sender, even when use_sender was set false. This led to far more
3905 callouts that were necessary. The sender is no longer included in the key
3906 when use_sender is false.
3908 26. Added "control = submission" modifier to ACLs.
3910 27. Added the ${base62d: operator to decode base 62 numbers.
3912 28. dnsdb lookups can now access SRV records.
3914 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3915 the configuration file.
3917 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3918 (-v) mode. This makes the output for a verbose queue run more intelligible.
3920 31. Added a use_postmaster feature to recipient callouts.
3922 32. Added the $body_zerocount variable, containing the number of binary zero
3923 bytes in the message body.
3925 33. The time of last modification of the "new" subdirectory is now used as the
3926 "mailbox time last read" when there is a quota error for a maildir
3929 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3931 35. Added +ignore_unknown as a special item in host lists.
3933 36. Code for decoding IPv6 addresses in host lists is now included, even if
3934 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3935 address was recognized as an IP address, but was then not correctly decoded
3936 into binary, causing unexpected and incorrect effects when compared with
3943 1. Very minor rewording of debugging text in manualroute to say "list of
3944 hosts" instead of "hostlist".
3946 2. If verify=header_syntax was set, and a header line with an unqualified
3947 address (no domain) and a large number of spaces between the end of the
3948 name and the colon was received, the reception process suffered a buffer
3949 overflow, and (when I tested it) crashed. This was caused by some obsolete
3950 code that should have been removed. The fix is to remove it!
3952 3. When running in the test harness, delay a bit after writing a bounce
3953 message to get a bit more predictability in the log output.
3955 4. Added a call to search_tidyup() just before forking a reception process. In
3956 theory, someone could use a lookup in the expansion of smtp_accept_max_
3957 per_host which, without the tidyup, could leave open a database connection.
3959 5. Added the variables $recipient_data and $sender_data which get set from a
3960 lookup success in an ACL "recipients" or "senders" condition, or a router
3961 "senders" option, similar to $domain_data and $local_part_data.
3963 6. Moved the writing of debug_print from before to after the "senders" test
3966 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3967 problems for message scanning, either using a data ACL, or using
3968 local_scan() because the Received: header was not generated till after they
3969 were called (in order to set the time as the time of reception completion).
3970 I have revised the way this works. The header is now generated after the
3971 body is received, but before the ACL or local_scan() are called. After they
3972 are run, the timestamp in the header is updated.
3978 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3979 before starting a queue runner without re-exec. This happened only when
3980 deliver_drop_privilege was set or when the Exim user was set to root. The
3981 effect of the bug was that timeouts during subsequent deliveries caused
3982 crashes instead of being properly handled. The handler is now left at its
3983 default (and expected) setting.
3985 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3986 message, again when deliver_drop_privilege is set or Exim is run as root.
3987 The bug described in (1) was not present in this case, but the tidying up
3988 of the other signals was missing. I have made the two cases consistent.
3990 3. The ignore_target_hosts setting on a manualroute router was being ignored
3991 for hosts that were looked up using the /MX notation.
3993 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3996 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3997 operated on the sender address. After changing the $sender_address to <>
3998 for the sender address verify, Exim was re-instated it as the original
3999 (before rewriting) address, but remembering that it had rewritten it, so it
4000 wasn't rewriting it again. This bug also had the effect of breaking the
4001 sender address verification caching when the sender address was rewritten.
4003 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4004 This has been changed so that if the ip literal address matches
4005 ignore_target_hosts, the router declines.
4007 7. Added expansion conditions match_domain, match_address, and match_local_
4008 part (NOT match_host).
4010 8. The placeholder for the Received: header didn't have a length field set.
4012 9. Added code to Exim itself and to exim_lock to test for a specific race
4013 condition that could lead to file corruption when using MBX delivery. The
4014 issue is with the lockfile that is created in /tmp. If this file is removed
4015 after a process has opened it but before that process has acquired a lock,
4016 there is the potential for a second process to recreate the file and also
4017 acquire a lock. This could lead to two Exim processes writing to the file
4018 at the same time. The added code performs the same test as UW imapd; it
4019 checks after acquiring the lock that its file descriptor still refers to
4020 the same named file.
4022 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4023 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4026 11. Added the smtp_active_hostname option. If used, this will typically be made
4027 to depend on the incoming interface address. Because $interface_address is
4028 not set up until the daemon has forked a reception process, error responses
4029 that can happen earlier (such as "too many connections") no longer contain
4032 12. If an expansion in a condition on a "warn" statement fails because a lookup
4033 defers, the "warn" statement is abandoned, and the next ACL statement is
4034 processed. Previously this caused the whole ACL to be aborted.
4036 13. Added the iplsearch lookup type.
4038 14. Added ident_timeout as a log selector.
4040 15. Added tls_certificate_verified as a log selector.
4042 16. Added a global option tls_require_ciphers (compare the smtp transport
4043 option of the same name). This controls incoming TLS connections.
4045 17. I finally figured out how to make tls_require_ciphers do a similar thing
4046 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4047 before starting the TLS session.
4049 18. Tabs are now shown as \t in -bP output.
4051 19. If the log selector return_path_on_delivery was set, Exim crashed when
4052 bouncing a message because it had too many Received: header lines.
4054 20. If two routers both had headers_remove settings, and the first one included
4055 a superfluous trailing colon, the final name in the first list and the
4056 first name in the second list were incorrectly joined into one item (with a
4057 colon in the middle).
4063 1. Added -C and -D options to the exinext utility, mainly to make it easier
4064 to include in the automated testing, but these could be helpful when
4065 multiple configurations are in use.
4067 2. The exinext utility was not formatting the output nicely when there was
4068 an alternate port involved in the retry record key, nor when there was a
4069 message id as well (for retries that were specific to a specific message
4070 and a specific host). It was also confused by IPv6 addresses, because of
4071 the additional colons they contain. I have fixed the IPv4 problem, and
4072 patched it up to do a reasonable job for IPv6.
4074 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4075 delivery, the log line now contains "pipelined" if PIPELINING was used.
4077 4. An SMTP transport process used to panic and die if the bind() call to set
4078 an explicit outgoing interface failed. This has been changed; it is now
4079 treated in the same way as a connect() failure.
4081 5. A reference to $sender_host_name in the part of a conditional expansion
4082 that was being skipped was still causing a DNS lookup. This no longer
4085 6. The def: expansion condition was not recognizing references to header lines
4086 that used bh_ and bheader_.
4088 7. Added the _cache feature to named lists.
4090 8. The code for checking quota_filecount in the appendfile transport was
4091 allowing one more file than it should have been.
4093 9. For compatibility with Sendmail, the command line option
4101 and sets the incoming protocol and host name (for trusted callers). The
4102 host name and its colon can be omitted when only the protocol is to be set.
4103 Note the Exim already has two private options, -pd and -ps, that refer to
4104 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4105 "s", but I don't think that's a major issue.
4107 10. A number of refactoring changes to the code, none of which should affect
4110 (a) The number of logging options was getting close to filling up the
4111 32-bit word that was used as a bit map. I have split them into two classes:
4112 those that are passed in the argument to log_write(), and those that are
4113 only ever tested independently outside of that function. These are now in
4114 separate 32-bit words, so there is plenty of room for expansion again.
4115 There is no change in the user interface or the logging behaviour.
4117 (b) When building, for example, log lines, the code previously used a
4118 macro that called string_cat() twice, in order to add two strings. This is
4119 not really sufficiently general. Furthermore, there was one instance where
4120 it was actually wrong because one of the argument was used twice, and in
4121 one call a function was used. (As it happened, calling the function twice
4122 did not affect the overall behaviour.) The macro has been replaced by a
4123 function that can join an arbitrary number of extra strings onto a growing
4126 (c) The code for expansion conditions now uses a table and a binary chop
4127 instead of a serial search (which was left over from when there were very
4128 few conditions). Also, it now recognizes conditions like "pam" even when
4129 the relevant support is not compiled in: a suitably worded error message is
4130 given if an attempt is made to use such a condition.
4132 11. Added ${time_interval:xxxxx}.
4134 12. A bug was causing one of the ddress fields not to be passed back correctly
4135 from remote delivery subprocesses. The field in question was not being
4136 subsequently used, so this caused to problems in practice.
4138 13. Added new log selectors queue_time and deliver_time.
4140 14. Might have fixed a bug in maildirsizefile handling that threw up
4141 "unexpected character" debug warnings, and recalculated the data
4142 unnecessarily. In any case, I expanded the warning message to give more
4145 15. Added the message "Restricted characters in address" to the statements in
4146 the default ACL that block characters like @ and % in local parts.
4148 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4149 Three changes have been made:
4151 (a) There was a serious bug; a negative response to MAIL caused the whole
4152 recipient domain to be cached as invalid, thereby blocking all messages
4153 to all local parts at the same domain, from all senders. This bug has
4154 been fixed. The domain is no longer cached after a negative response to
4155 MAIL if the sender used is not empty.
4157 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4160 (c) A new callout option, "use_sender" has been added for people who want
4161 the modified behaviour.
4167 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4170 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4171 indeed breaks things for older releases.
4173 3. Added additional logging to the case where there is a problem reading data
4174 from a filter that is running in a subprocess using a pipe, in order to
4175 try to track down a specific problem.
4177 4. Testing facility fudge: when running in the test harness and attempting
4178 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4179 getting "No route to host". Convert this to a timeout.
4181 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4184 6. Some OS don't have socklen_t but use size_t instead. This affects the
4185 fifth argument of getsockopt() amongst other things. This is now
4186 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4187 can be set for individual OS. I have set it for SunOS5, OSF1, and
4188 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4189 some earlier ones do not.
4191 7. Change 4.30/15 was not doing the test caselessly.
4193 8. The standard form for an IPv6 address literal was being rejected by address
4194 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4195 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4196 this, as well as the form without the "IPv6" on the front (but only when
4197 address literals are enabled, of course).
4199 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4201 10. Exim crashed if a message with an empty sender address specified by -f
4202 encountered a router with an errors_to setting. This could be provoked only
4203 by a command such as
4207 where an empty string was supplied; "<>" did not hit this bug.
4209 11. Installed PCRE release 4.5.
4211 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4212 remained set. It is now erased.
4214 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4215 times from message ids (which are base 36 rather than the normal 62).
4217 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4218 were being counted as actual protocol errors, and logged if the log
4219 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4220 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4221 and DATA following a set of rejected RCPTs do not count as protocol errors.
4222 In other words, Exim assumes they were pipelined, though this may not
4223 actually be the case. Of course, in all cases the client gets an
4224 appropriate error code.
4226 15. If a lookup fails in an ACL condition, a message about the failure may
4227 be available; it is used if testing the ACL cannot continue, because most
4228 such messages specify what the cause of the deferral is. However, some
4229 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4230 that caused an old message to be retained and used if a later statement
4231 caused a defer, replacing the real cause of the deferral.
4233 16. If an IP address had so many PTR records that the DNS lookup buffer
4234 was not large enough to hold them, Exim could crash while trying to process
4235 the truncated data. It now detects and logs this case.
4237 17. Further to 4.21/58, another change has been made: if (and only if) the
4238 first line of a message (the first header line) ends with CRLF, a bare LF
4239 in a subsequent header line has a space inserted after it, so as not to
4240 terminate the header.
4242 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4243 unnecessarily, used atoi() instead of strtol(), and didn't check the
4244 termination when getting file sizes from file names by regex.
4246 19. Completely re-implemented the support for maildirsize files, in the light
4247 of a number of problems with the previous contributed implementation
4248 (4.30/29). In particular:
4250 . If the quota is zero, the maildirsize file is maintained, but no quota is
4253 . If the maildir directory does not exist, it is created before any attempt
4254 to write a maildirsize file.
4256 . The quota value in the file is just a cache; if the quota is changed in
4257 the transport, the new value overrides.
4259 . A regular expression is available for excluding directories from the
4262 20. The autoreply transport checks the characters in options that define the
4263 message's headers; it allows continued headers, but it was checking with
4264 isspace() after an embedded newline instead of explicitly looking for a
4267 21. If all the "regular" hosts to which an address was routed had passed their
4268 expiry times, and had not reached their retry times, the address was
4269 bounced, even if fallback hosts were defined. Now Exim should go on to try
4272 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4273 equivalent code in the SMTP transport. Some hosts send humungous responses
4274 to HELO/EHLO, more than 1024 it seems.
4276 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4277 gives compiler warnings in some environments. I've now done it "properly",
4280 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4281 (because of problems with the built-in one) was declared to return uschar *
4282 instead of char *, causing compiler failure.
4284 25. Fixed a file descriptor leak when processing alias/forward files.
4286 26. Fixed a minor format string issue in dbfn.c.
4288 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4290 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4291 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4294 29. When a sender address is verified, it is cached, to save repeating the test
4295 when there is more than one recipient in a message. However, when the
4296 verification involves a callout, it is possible for different callout
4297 options to be set for different recipients. It is too complicated to keep
4298 track of this in the cache, so now Exim always runs a verification when a
4299 callout is required, relying on the callout cache for the optimization.
4300 The overhead is duplication of the address routing, but this should not be
4303 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4304 to be verified with callout=postmaster, and the main callout worked but the
4305 postmaster check failed, the verification correctly failed. However, if a
4306 subsequent RCPT command asked for sender verification *without* the
4307 postmaster check, incorrect caching caused this verification also to fail,
4310 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4311 it was not caching the DNS options (qualify_single, search_parents) that
4312 were used when the lookup failed. A subsequent lookup with different
4313 options therefore always gave the same answer, though there were cases
4314 where it should not have. (Example: a "domains = !$mx_any" option on a
4315 dnslookup router: the "domains" option is always processed without any
4316 widening, but the router might have qualify_single set.) Now Exim uses the
4317 cached value only when the same options are set.
4319 32. Added John Jetmore's "exipick" utility to the distribution.
4321 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4322 than a timeout (e.g. a certificate is required, and is not provided), an
4323 Exim server now closes the connection immediately. Previously it waited for
4324 the client to close - but if the client is SSL, it seems that they each
4325 wait for each other, leading to a delay before one of them times out.
4327 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4328 maintained 0.8.x compatibility because I don't think many are using it, and
4329 it is clearly obsolete.
4331 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4334 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4335 client certificate was expired. A simple patch fixes this, though I don't
4336 understand the full logic of why the verify callback is called multiple
4339 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4340 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4341 which causes problems with some clients (such as the Certicom SSL Plus
4342 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4343 disables the coutermeasure allowing Eudora to connect."
4345 38. Exim was not checking that a write() to a log file succeeded. This could
4346 lead to Bad Things if a log got too big, in particular if it hit a file
4347 size limit. Exim now panics and dies if it cannot write to a log file, just
4348 as it does if it cannot open a log file.
4350 39. Modified OS/Makefile-Linux so that it now contains
4352 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4354 The two -D definitions ensure that Exim is compiled with large file
4355 support, which makes it possible to handle log files that are bigger than
4358 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4359 instance) a domain was checked against a named list that involved a lookup,
4360 causing $domain_data to be set, then another domain was checked against the
4361 same list, then the first domain was re-checked, the value of $domain_data
4362 after the final check could be wrong. In particular, if the second check
4363 failed, it could be set empty. This bug probably also applied to
4366 41. The strip_trailing_dot option was not being applied to the address given
4367 with the -f command-line option.
4369 42. The code for reading a message's header from the spool was incrementing
4370 $received_count, but never initializing it. This meant that the value was
4371 incorrect (doubled) while delivering a message in the same process in which
4372 it was received. In the most common configuration of Exim, this never
4373 happens - a fresh exec is done - but it can happen when
4374 deliver_drop_privilege is set.
4376 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4377 it now includes up to 150 characters of the unexpected data in the log
4380 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4381 and building data strings. The size of both of these buffers was 10 000
4382 bytes - far larger than anybody would *ever* want, thought I. Needless to
4383 say, somebody hit the limit. I have increased the maximum line length to
4384 20 000 and the maximum data length of concatenated lines to 100 000. I have
4385 also fixed two bugs, because there was no checking on these buffers. Tsk,
4386 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4387 buffer is too small.
4389 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4390 lsearch lookups. Now it does.
4392 46. When parsing a route_list item in a manualroute router, a fixed-length
4393 buffer was used for the list of hosts. I made this 1024 bytes long,
4394 thinking that nobody would ever have a list of hosts that long. Wrong.
4395 Somebody had a whole pile of complicated expansion conditions, and the
4396 string was silently truncated, leading to an expansion error. It turns out
4397 that it is easier to change to an unlimited length (owing to other changes
4398 that have happened since this code was originally written) than to build
4399 structure for giving a limitation error. The length of the item that
4400 expands into the list of hosts is now unlimited.
4402 47. The lsearch lookup could not handle data where the length of text line was
4403 more than 4095 characters. Such lines were truncated, leading to shortened
4404 data being returned. It should now handle lines of any length.
4406 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4407 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4410 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4411 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4412 no longer happens in comment lines. A list of replacements is now placed
4413 at the head of all of the source files, except those whose only change is
4414 to replace PERL_COMMAND in the very first #! line.
4416 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4417 messages on the queue, with a bottom-up merge sort, using code contributed
4418 by Michael Haardt. This should make operations like -bp somewhat faster on
4419 large queues. It won't affect queue runners, except when queue_run_in_order
4422 51. Installed eximstats 1.31 in the distribution.
4424 52. Added support for SRV lookups to the dnslookup router.
4426 53. If an ACL referred to $message_body or $message_body_end, the value was not
4427 reset for any messages that followed in the same SMTP session.
4429 54. The store-handling optimization for building very long strings was not
4430 differentiating between the different store pools. I don't think this
4431 actually made any difference in practice, but I've tidied it.
4433 55. While running the routers to verify a sender address, $sender_address
4434 was still set to the sender address. This is wrong, because when routing to
4435 send a bounce to the sender, it would be empty. Therefore, I have changed
4436 it so that, while verifying a sender address, $sender_address is set to <>.
4437 (There is no change to what happens when verifying a recipient address.)
4439 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4440 target A or AAAA records (if not already returned) without resetting the
4441 qualify_single or search_parents options of the DNS resolver. These are
4442 inappropriate in this case because the targets of MX and SRV records must
4443 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4444 target that, when qualified, matched something in the local domain. These
4445 two options are now turned off when doing these lookups.
4447 57. It seems that at least some releases of Reiserfs (which does not have the
4448 concept of a fixed number of inodes) returns zero and not -1 for the
4449 number of available inodes. This interacted badly with check_spool_inodes,
4450 which assumed that -1 was the "no such thing" setting. What I have done is
4451 to check that the total number of inodes is greater than zero before doing
4452 the test of how many are available.
4454 58. When a "warn" ACL statement has a log_message modifier, the message is
4455 remembered, and not repeated. This is to avoid a lot of repetition when a
4456 message has many recipients that cause the same warning to be written.
4457 Howewer, Exim was preserving the list of already written lines for an
4458 entire SMTP session, which doesn't seem right. The memory is now reset if a
4459 new message is started.
4461 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4462 debugging output unless log_rewrite was also set.
4464 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4465 of (int)(handle) when we know that handle contains (void *)(-1).
4467 61. The Exim daemon panic-logs an error return when it closes the incoming
4468 connection. However "connection reset by peer" seems to be common, and
4469 isn't really an error worthy of noting specially, so that particular error
4472 62. When Exim is trying to find all the local interfaces, it used to panic and
4473 die if the ioctl to get the interface flags failed. However, it seems that
4474 on at least one OS (Solaris 9) it is possible to have an interface that is
4475 included in the list of interfaces, but for which you get a failure error
4476 for this call. This happens when the interface is not "plumbed" into a
4477 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4478 failure of the "get flags" call assumes that the interface is down.
4480 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4481 makes life easier for people who are doing arithmetic on fields extracted
4482 from dates, where you often get leading zeros that should not be
4483 interpreted as octal.
4485 64. Added qualify_domain to the redirect router, to override the global
4488 65. If a pathologically long header line contained very many addresses (the
4489 report of this problem mentioned 10 000) and each of them was rewritten,
4490 Exim could use up a very large amount of memory. (It kept on making new
4491 copies of the header line as it rewrote, and never released the old ones.)
4492 At the expense of a bit more processing, the header rewriting function has
4493 been changed so that it no longer eats memory in this way.
4495 66. The generation of the Received: header has been moved from the time that a
4496 message starts to be received, to the time that it finishes. The timestamp
4497 in the Received: header should now be very close to that of the <= log
4498 line. There are two side-effects of this change:
4500 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4501 logged header lines no longer include the local Received: line, because
4502 it has not yet been created. The same applies to a copy of the message
4503 that is returned to a non-SMTP sender when a message is rejected.
4505 (b) When a filter file is tested using -bf, no additional Received: header
4506 is added to the test message. After some thought, I decided that this
4509 This change does not affect the value of $received_for. It is still set
4510 after address rewriting, but before local_scan() is called.
4512 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4514 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4515 gave an unhelpful panic error message, and a defer error. I have managed to
4516 change this behaviour so that it now rejects any supplied certificate,
4517 which seems right, as the list of acceptable certificates is empty.
4519 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4520 gave an unhelpful defer error. I have not managed to make this reject any
4521 supplied certificates, but the error message it gives is "no certificate
4522 supplied", which is not helpful.
4524 70. exigrep's output now also includes lines that are not associated with any
4525 message, but which match the given pattern. Implemented by a patch from
4526 Martin Sluka, which also tidied up the Perl a bit.
4528 71. Recipient callout verification, like sender verification, was using <> in
4529 the MAIL FROM command. This isn't really the right thing, since the actual
4530 sender may affect whether the remote host accepts the recipient or not. I
4531 have changed it to use the actual sender in the callout; this means that
4532 the cache record is now keyed on a recipient/sender pair, not just the
4533 recipient address. There doesn't seem to be a real danger of callout loops,
4534 since a callout by the remote host to check the sender would use <>.
4535 [SEE ABOVE: changed after hitting problems.]
4537 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4538 temporary errors. However, in the case of such a code being given after
4539 the end of a data transmission (i.e. after ".") Exim was failing to write
4540 a retry record for the message. (Yes, there was some broken host that was
4541 actually sending 8xx at this point.)
4543 73. An unknown lookup type in a host list could cause Exim to panic-die when
4544 the list was checked. (An example that provoked this was putting <; in the
4545 middle of a list instead of at the start.) If this happened during a DATA
4546 ACL check, a -D file could be left lying around. This kind of configuration
4547 error no longer causes Exim to die; instead it causes a defer errror. The
4548 incident is still logged to the main and panic logs.
4550 74. Buglet left over from Exim 3 conversion. The message "too many messages
4551 in one connection" was written to the rejectlog but not the mainlog, except
4552 when address rewriting (yes!) was being logged.
4554 75. Added write_rejectlog option.
4556 76. When a system filter was run not as root (that is, when system_filter_user
4557 was set), the values of the $n variables were not being returned to the
4558 main process; thus, they were not subsequently available in the $sn
4561 77. Added +return_path_on_delivery log selector.
4563 78. A connection timeout was being treated differently from recipients deferred
4564 when testing hosts_max_try with a message that was older than the host's
4565 retry timeout. (The host should not be counted, thus allowing all hosts to
4566 be tried at least once before bouncing.) This may have been the cause of an
4567 occasionally reported bug whereby a message would remain on the queue
4568 longer than the retry timeout, but would be bounced if a delivery was
4569 forced. I say "may" because I never totally pinned down the problem;
4570 setting up timeout/retry tests is difficult. See also the next item.
4572 79. The ultimate address timeout was not being applied to errors that involved
4573 a combination of host plus message (for example, a timeout on a MAIL
4574 command). When an address resolved to a number of possible hosts, and they
4575 were not all tried for each delivery (e.g. because of hosts_max_try), a
4576 message could remain on the queue longer than the retry timeout.
4578 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4581 81. Fixed an obscure SMTP outgoing bug which required at least the following
4582 conditions: (a) there was another message waiting for the same server;
4583 (b) the server returned 5xx to all RCPT commands in the first message so
4584 that the message was not completed; (c) the server dropped the connection
4585 or gave a negative response to the RSET that Exim sends to abort the
4586 transaction. The observed case was a dropped connection after DATA that had
4587 been sent in pipelining mode. That is, the server had advertised PIPELINING
4588 but was not implementing it correctly. The effect of the bug was incorrect
4589 behaviour, such as trying another host, and this could lead to a crash.
4595 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4596 and daemon.c were passed as pointers to ints; they should have been
4597 pointers to socklen_t variables (which are typically unsigned ints).
4599 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4602 3. Fixed a really odd bug that affected only the testing scheme; patching a
4603 certain fixed string in the binary changed the value of another string that
4604 happened to be identical to the end of the original first string.
4606 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4607 name", it returns that address as the IP address. On some operating
4608 systems (e.g. Solaris), it also passes back the IP address string as the
4609 "host name". However, on others (e.g. Linux), it passes back an empty
4610 string. Exim wasn't checking for this, and was changing the host name to an
4611 empty string, assuming it had been canonicized.
4613 5. Although rare, it is permitted to have more than one PTR record for a given
4614 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4615 all the names associated with an address, because they do in Solaris.
4616 However, it seems that they do not in Linux for data that comes from the
4617 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4618 I found this out when I moved to a new Linux workstation and tried to run
4619 the Exim test suite.
4621 To get round this problem I have changed the code so that it now does its
4622 own call to the DNS to look up PTR records when searching for a host name.
4623 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4624 addresses that are only in /etc/hosts are still found.
4626 This behaviour is, however, controlled by an option called host_lookup_
4627 order, which defaults to "bydns:byaddr". If people want to use the other
4628 order, or indeed, just use one or the other means of lookup, they can
4629 specify it in this variable.
4631 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4632 some operating systems, this comes back from gethostbyaddr() as an empty
4633 string, and this is what Exim used to test for. However, it seems that in
4634 other systems, "." is yielded. Exim now tests for this case too.
4636 7. The values of check_spool_space and check_log_space are now held internally
4637 as a number of kilobytes instead of an absolute number of bytes. If a
4638 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4639 kilobyte. This means that much larger values can be stored.
4641 8. Exim monitor: an attempt to get the action menu when not actually pointing
4642 at a message produces an empty menu entitled "No message selected". This
4643 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4644 no entries in it ("Shell widget menu has zero width and/or height"). So I
4645 have added a single, blank menu entry in this case.
4647 9. Added ${quote_local_part.
4649 10. MIME decoding is now applied to the contents of Subject: header lines when
4652 11. Now that a reference to $sender_host_address automatically causes a reverse
4653 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4654 host lookup before query-style lookups in lists that might use this
4655 variable. This has therefore been abolished, and the "net-" prefix is no
4656 longer necessary for query-style lookups.
4658 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4659 have been a typo for LFLAGS, so it has been changed.
4661 13. The install script calls Exim with "-C /dev/null" in order to find the
4662 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4663 to be output. Howeve, since Exim outputs its version number before the
4664 error, it didn't break the script. It just looked ugly. I fixed this by
4665 always allowing "-C /dev/null" if the caller is root.
4667 14. Ignore overlarge ACL variable number when reading spool file - insurance
4668 against a later release with more variables having written the file.
4670 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4671 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4672 this, as well as the form without the "IPv6" on the front.
4674 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4675 OS/Makefile-Darwin file.
4677 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4678 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4680 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4681 variable. However, some OS use ints for inodes. I've added cast to long int
4682 to get rid of the compiler warning.
4684 19. I had forgotten to lock out "/../" in configuration file names when
4685 ALT_CONFIG_PREFIX was set.
4687 20. Routers used for verification do not need to specify transports. However,
4688 if such a router generated a host list, and callout was configured, Exim
4689 crashed, because it could not find a port number from the (non-existent)
4690 transport. It now assumes port 25 in this circumstance.
4692 21. Added the -t option to exigrep.
4694 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4695 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4696 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4697 them is set, LOOKUP_LSEARCH is forced.
4699 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4701 24. Added sender and host information to the "rejected by local_scan()" log
4702 line; previously there was no indication of these.
4704 25. Added .include_if_exists.
4706 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4707 Linux. It turns out that not all file systems support this. Apparently some
4708 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4709 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4710 not supported on the file descriptor, is now ignored when Exim is trying to
4711 sync a directory. This applies only to Linux.
4713 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4715 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4716 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4719 29. Added support for maildirsize files from supplied patch (modified a bit).
4721 30. The use of :fail: followed by an empty string could lead Exim to respond to
4722 sender verification failures with (e.g.):
4724 550 Verification failed for <xxx>
4725 550 Sender verify failed
4727 where the first response line was missing the '-' that indicates it is not
4728 the final line of the response.
4730 31. The loop for finding the name of the user that called Exim had a hardwired
4731 limit of 10; it now uses the value of finduser_retries, which is used for
4732 all other user lookups.
4734 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4737 33. Exim was neglecting to zero errno before one call of strtol() when
4738 expanding a string and expecting an integer value. On some systems this
4739 resulted in spurious "integer overflow" errors. Also, it was casting the
4740 result into an int without checking.
4742 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4743 did not work. The code looks as if it has *never* worked, though it appears
4744 to have been documented since at least releast 1.62. I have made it work.
4746 35. The "timeout_DNS" error in retry rules, also documented since at least
4747 1.62, also never worked. As it isn't clear exactly what this means, and
4748 clearly it isn't a major issue, I have abolished the feature by treating it
4749 as "timeout", and writing a warning to the main and panic logs.
4751 36. The display of retry rules for -brt wasn't always showing the error code
4754 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4755 timeout_connect_A, timeout_connect_MX.
4757 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4758 to the empty sender.
4760 39. The daemon was not analysing the content of -oX till after it had closed
4761 stderr and disconnected from the controlling terminal. This meant that any
4762 syntax errors were only noted on the panic log, and the return code from
4763 the command was 0. By re-arranging the code a little, I've made the
4764 decoding happen first, so such errors now appear on stderr, and the return
4765 code is 1. However, the actual setting up of the sockets still happens in
4766 the disconnected process, so errors there are still only recorded on the
4769 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4770 connections (as happens on some IP stacks) was logged at start up time as
4771 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4772 it from "IPv6 and IPv4", which means that two separate sockets are being
4775 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4776 says whether AF_INET or AF_INET6 was passed as an argument.
4778 42. Exiwhat output was messed up when time zones were included in log
4781 43. Exiwhat now gives more information about the daemon's listening ports,
4782 and whether -tls-on-connect was used.
4784 44. The "port" option of the smtp transport is now expanded.
4786 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4787 silently ignored. Now an error message is written to the main and panic
4790 46. There's a new ACL modifier called "logwrite" which writes to a log file
4791 as soon as it is encountered.
4793 47. Added $local_user_uid and $local_user_gid at routing time.
4795 48. Exim crashed when trying to verify a sender address that was being
4798 49. Exim was recognizing only a space character after ".include". It now also
4799 recognizes a tab character.
4801 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4802 extracting the relevant information from the specification. The man page no
4803 longer contains scrambled data for the -d option, and I've added a section
4804 at the front about calling Exim under different names.
4806 51. Added "extra_headers" argument to the "mail" command in filter files.
4808 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4811 53. Installed eximstats 1.29.
4813 54. Added transport_filter_timeout as a generic transport option.
4815 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4816 Cc: header lines. This was required by RFC 822, but it not required by RFC
4819 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4820 incoming messages that did not have them. Now it does so only if the
4821 message originates locally, that is, if there is no associated remote host
4822 address. When Resent- header lines are present, this applies to the Resent-
4823 lines rather than the non-Resent- lines.
4825 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4826 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4828 58. Messages for configuration errors now include the name of the main
4829 configuration file - useful now that there may be more than one file in a
4830 list (.included file names were always shown).
4832 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4833 for those rare installations that do not start the daemon as root or run it
4834 setuid root. I've cut out the call to initgroups() if the daemon is not
4837 60. The Exim user and group can now be bound into the binary as text strings
4838 that are looked up at the start of Exim's processing.
4840 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4842 62. Added $mailstore_basename variable.
4844 63. Installed patch to sieve.c from Michael Haardt.
4846 64. When Exim failed to open the panic log after failing to open the main log,
4847 the original message it was trying to log was written to stderr and debug
4848 output, but if they were not available (the usual case in production), it
4849 was lost. Now it is written to syslog before the two lines that record the
4850 failures to open the logs.
4852 65. Users' Exim filters run in subprocesses under the user's uid. It is
4853 possible for a "deliver" command or an alias in a "personal" command to
4854 provoke an address rewrite. If logging of address rewriting is configured,
4855 this fails because the process is not running as root or exim. There may be
4856 a better way of dealing with this, but for the moment (because 4.30 needs
4857 to be released), I have disabled address rewrite logging when running a
4858 filter in a non-root, non-exim process.
4864 1. The buildconfig auxiliary program wasn't quoting the value set for
4865 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4866 not defined. This bug was masked in 4.22 by the effect that was fixed in
4869 2. Some messages that were rejected after a message id was allocated were
4870 shown as "incomplete" by exigrep. It no longer does this for messages that
4871 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4873 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4874 have allow_domain_literals set, the ID did not get logged in the <= line.
4875 Domain literals are now always recognized in Message-ID: header lines.
4877 4. The first argument for a ${extract expansion item is the key name or field
4878 number. Leading and trailing spaces in this item were not being ignored,
4879 causing some misleading effects.
4881 5. When deliver_drop_privilege was set, single queue runner processes started
4882 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4883 same command in the process it spins off) were not dropping privilege.
4885 6. When the daemon running as "exim" started a queue runner, it always
4886 re-executed Exim in the spun-off process. This is a waste of effort when
4887 deliver_drop_privilege is set. The new process now just calls the
4888 queue-runner function directly.
4894 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4897 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4898 ignored. Though the use of -oP was forcing the writing of a pid file, it
4899 was always written to the default place.
4901 3. If the message "no IP address found for host xxxx" is generated during
4902 incoming verification, it is now followed by identification of the incoming
4903 connection (so you can more easily find what provoked it).
4905 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4907 5. Added some features to "harden" Exim a bit more against certain attacks:
4909 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4910 be put in Local/Makefile. This is like the never_users runtime option,
4911 but it cannot be overridden. The default setting is "root".
4913 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4914 prefix string with which any file named in a -C command line option
4917 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4918 is retained for -C and -D only if the caller of Exim is root. Without
4919 it, the exim user may also use -C and -D and retain privilege.
4921 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4922 command line option is disabled.
4924 6. Macro names set by the -D option must start with an upper case letter, just
4925 like macro names defined in the configuration file.
4927 7. Added "dereference=" facility to LDAP.
4929 8. Two instances of the typo "uknown" in the source files are fixed.
4931 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4932 the Configure-Makefile script screwed up while processing it.
4934 10. Incorporated PCRE 4.4.
4936 11. The SMTP synchronization check was not operating right at the start of an
4937 SMTP session. For example, it could not catch a HELO sent before the client
4938 waited for the greeting. There is now a check for outstanding input at the
4939 point when the greeting is written. Because of the duplex, asynchronous
4940 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4941 way, but not yet received, when the check is performed.
4943 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4944 on TCP/IP sockets, because this apparently causes some broken clients to
4947 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4948 unchanged) from the Cygwin maintainer.
4950 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4951 was supported instead of testing for mbx. Effectively a typo.
4953 15. The spa authenticator server code was not checking that the input it
4954 received was valid base64.
4956 16. The debug output line for the "set" modifier in ACLs was not showing the
4957 name of the variable that was being set.
4959 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4961 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4962 lookup on its own. Something else was needed to trigger the lookup. For
4963 example, a match in host_lookup or the need for a host name in a host list.
4964 Now, if $sender_host_name is referenced and the host name has not yet been
4965 looked up, a lookup is performed. If the lookup fails, the variable remains
4966 empty, and $host_lookup_failed is set to "1".
4968 19. Added "eqi" as a case-independent comparison operator.
4970 20. The saslauthd authentication condition could segfault if neither service
4971 nor realm was specified.
4973 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4974 error message that was logged was misleading, and incoming SMTP
4975 connections were dropped. The message is now more accurate, and temporary
4976 errors are given to SMTP connections.
4978 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4979 (except RSET) with a 421 error, until QUIT is received. However, it was
4980 failing to send a response to QUIT.
4982 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4983 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4984 ACL is now run after the helo_try_verify_hosts code.
4986 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4987 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4988 case-independent, so other case variants are also recognized). Apparently
4989 some systems use these upper case variants.
4991 25. If more than two messages were waiting for the same host, and a transport
4992 filter was specified for the transport, Exim sent two messages over the
4993 same TCP/IP connection, and then failed with "socket operation on non-
4994 socket" when it tried to send the third.
4996 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4998 27. The extern definition of crypt16() in expand.c was not being excluded when
4999 the OS had its own crypt16() function.
5001 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5002 as a preferred synonym for return_size_limit, both as an option and as an
5005 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5007 30. Changed the default configuration ACL to relax the local part checking rule
5008 for addresses that are not in any local domains. For these addresses,
5009 slashes and pipe symbols are allowed within local parts, but the sequence
5010 /../ is explicitly forbidden.
5012 31. SPA server authentication was not clearing the challenge buffer before
5015 32. log_message in a "warn" ACL statement was writing to the reject log as
5016 well as to the main log, which contradicts the documentation and doesn't
5017 seem right (because no rejection is happening). So I have stopped it.
5019 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5020 However, I am unable to do any testing of this.
5022 34. Fixed an infelicity in the appendfile transport. When checking directories
5023 for a mailbox, to see if any needed to be created, it was accidentally
5024 using path names with one or more superfluous leading slashes; tracing
5025 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5027 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5028 contents are added to the log line that is written for every discarded
5029 recipient. (Previously a log_message setting was ignored.)
5031 36. The ${quote: operator now quotes the string if it is empty.
5033 37. The install script runs exim in order to find its version number. If for
5034 some reason other than non-existence or emptiness, which it checks, it
5035 could not run './exim', it was installing it with an empty version number,
5036 i.e. as "exim-". This error state is now caught, and the installation is
5039 38. An argument was missing from the function that creates an error message
5040 when Exim fails to connect to the socket for saslauthd authentication.
5041 This could cause Exim to crash, or give a corrupted message.
5043 39. Added isip, isip4, and isip6 to ${if conditions.
5045 40. The ACL variables $acl_xx are now saved with the message, and can be
5046 accessed later in routers, transports, and filters.
5048 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5049 strings in the file are not string-expanded.
5051 42. If a MAIL command specified a SIZE value that was too large to fit into an
5052 int variable, the check against message_size_limit failed. Such values are
5053 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5054 day this will have to be increased, but I don't think I want to be around
5055 when emails are that large.
5062 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5063 iconv() is not standard in FreeBSD.
5065 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5066 IPv6 enabled. The observed symptom was a segmentation fault on return from
5067 the function os_common_find_running_interfaces() in src/os.c.
5069 3. In the check_special_case() function in daemon.c I had used "errno" as an
5070 argument name, which causes warnings on some systems. This was basically a
5071 typo, since it was named "eno" in the comments!
5073 4. The code that waits for the clock to tick (at a resolution of some fraction
5074 of a second) so as to ensure message-id uniqueness was always waiting for
5075 at least one whole tick, when it could have waited for less. [This is
5076 almost certainly not relevant at current processor speeds, where it is
5077 unlikely to ever wait at all. But we try to future-proof.]
5079 5. The function that sleeps for a time interval that includes fractions of a
5080 second contained a race. It did not block SIGALRM between setting the
5081 timer, and suspending (a couple of lines later). If the interval was short
5082 and the sigsuspend() was delayed until after it had expired, the suspension
5083 never ended. On busy systems this could lead to processes getting stuck for
5086 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5087 process, before it forks any delivery processes. The open lookup caching
5088 mechanism meant that the open file or database connection was passed into
5089 the delivery process. The problem was that delivery processes always tidy
5090 up cached lookup data. This could cause a problem for the next delivery
5091 process started by the queue runner, because the external queue runner
5092 process does not know about the closure. So the next delivery process
5093 still has data in the lookup cache. In the case of a file lookup, there was
5094 no problem because closing a file descriptor in a subprocess doesn't affect
5095 the parent. However, if the lookup was caching a connection to a database,
5096 the connection was closed, and the second delivery process was likely to
5097 see errors such as "PGSQL: query failed: server closed the connection
5098 unexpectedly". The problem has been fixed by closing all cached lookups
5099 in a queue runner before running a delivery process.
5101 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5102 seem to have the "const" qualifier which it has on other OS. I've
5105 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5106 *of the same type* (client or server) with the same public name that an
5107 error should be diagnosed.
5109 9. When Exim looked up a host name for an IP address, but failed to find the
5110 original IP address when looking up the host name (a safety check), it
5111 output the message "<ip address> does not match any IP for NULL", which was
5112 confusing, to say the least. The bug was that the host name should have
5113 appeared instead of "NULL".
5115 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5116 user that is built into the binary, and the -C or -D options is used, root
5117 privilege is dropped before the configuration file is read. In addition,
5118 logging is switched to stderr instead of the normal log files. If the
5119 configuration then re-defines the Exim user, the unprivileged environment
5120 is probably not what is expected, so Exim logs a panic warning message (but
5123 However, if deliver_drop_privilege is set, the unprivileged state may well
5124 be exactly what is intended, so the warning has been cut out in that case,
5125 and Exim is allowed to try to write to its normal log files.
5131 1. smtp_return_error_details was not giving details for temporary sender
5132 or receiver verification errors.
5134 2. Diagnose a configuration error if two authenticators have the same public
5137 3. Exim used not to create the message log file for a message until the first
5138 delivery attempt. This could be confusing when incoming messages were held
5139 for policy or load reasons. The message log file is now created at the time
5140 the message is received, and an initial "Received" line is written to it.
5142 4. The automatically generated man page for command line options had a minor
5143 bug that caused no ill effects; however, a more serious problem was that
5144 the procedure for building the man page automatically didn't always
5145 operate. Consequently, release 4.20 contains an out-of-date version. This
5146 shouldn't happen again.
5148 5. When building Exim with embedded Perl support, the script that builds the
5149 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5150 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5152 6. The freeze_tell option was not being used for messages that were frozen on
5153 arrival, either by an ACL or by local_scan().
5155 7. Added the smtp_incomplete_transaction log selector.
5157 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5158 was accepting AUTH without a new EHLO.
5160 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5161 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5164 10. Logging of TCP/IP connections (when configured) now happens in the main
5165 daemon process instead of the child process, so that the TCP/IP connection
5166 count is more accurate (but it can never be perfect).
5168 11. The use of "drop" in a nested ACL was not being handled correctly in the
5169 outer ACL. Now, if condition failure induced by the nested "drop" causes
5170 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5171 or "require"), the connection is dropped.
5173 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5174 that yield "discard" can now be used with an "accept" or a "discard" verb,
5175 but an error is generated for any others (because I can't see a useful way
5176 to define what should happen).
5178 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5179 are now processed in the same way as lines in the Exim configuration file.
5180 In particular, continuation lines are supported.
5182 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5184 15. Added -ti meaning -t -i.
5186 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5187 domains, and warn by logging if others are found.
5189 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5190 returned from GIFCONF when Exim is trying to find the list of interfaces on
5191 a host. The code in os.c has been modified to copy each ifreq to an aligned
5192 structure in all cases.
5194 Also, in some cases, the returned ifreq's were being copied to a 'struct
5195 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5196 means the last couple of bytes of an IPv6 address could be chopped if the
5197 ifreq contained only a normal sockaddr (14 bytes storage).
5199 18. Named domain lists were not supported in the hosts_treat_as_local option.
5200 An entry such as +xxxx was not recognized, and was treated as a literal
5203 19. Ensure that header lines added by a DATA ACL are included in the reject log
5204 if the ACL subsequently rejects the message.
5206 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5207 MD5 (which is deprecated).
5209 21. When testing a filter file using -bf, Exim was writing a message when it
5210 took the sender from a "From " line in the message, but it was not doing so
5211 when it took $return_path from a Return-Path: header line. It now does.
5213 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5214 with a valid header line field name (a series of printing characters
5215 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5217 23. Changed "disc" in the source to "disk" to conform to the documentation and
5218 the book and for uniformity.
5220 24. Ignore Sendmail's -Ooption=value command line item.
5222 25. When execve() failed while trying to run a command in a pipe transport,
5223 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5224 could be confused with a return value of 69 from the command itself. This
5225 has been changed to 127, the value the shell returns if it is asked to run
5226 a non-existent command. The wording for the related log line suggests a
5227 non-existent command as the problem.
5229 26. If received_header_text expands to an empty string, do not add a Received:
5230 header line to the message. (Well, it adds a token one on the spool, but
5231 marks it "old" so that it doesn't get used or transmitted.)
5233 27. Installed eximstats 1.28 (addition of -nt option).
5235 28. There was no check for failure on the call to getsockname() in the daemon
5236 code. This can fail if there is a shortage of resources on the system, with
5237 ENOMEM, for example. A temporary error is now given on failure.
5239 29. Contrary to the C standard, it seems that in some environments, the
5240 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5241 program. Exim now does this explicitly; it affects the formatting of
5242 timestamps using strftime().
5244 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5245 complaints. I've now initialized all the variables, to avoid this.
5247 32. Header lines added by a system filter were not being "seen" during
5248 transport-time rewrites.
5250 33. The info_callback() function passed to OpenSSL is set up with type void
5251 (*)(SSL *, int, int), as described somewhere. However, when calling the
5252 function (actually a macro) that sets it up, the type void(*)() is
5253 expected. I've put in a cast to prevent warnings from picky compilers.
5255 34. If a DNS black list lookup found a CNAME record, but there were no A
5256 records associated with the domain it pointed at, Exim crashed.
5258 35. If a DNS black list lookup returned more than one A record, Exim ignored
5259 all but the first. It now scans all returned addresses if a particular IP
5260 value is being sought. In this situation, the contents of the
5261 $dnslist_value variable are a list of all the addresses, separated by a
5264 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5265 to accept a host name and all its aliases if the forward lookup for any of
5266 them yielded the IP address of the incoming connection. Now it accepts only
5267 those names whose forward lookup yields the correct IP address. Any other
5268 names are discarded. This closes a loophole whereby a rogue DNS
5269 administrator could create reverse DNS records to break through a
5270 wildcarded host restriction in an ACL.
5272 37. If a user filter or a system filter that ran in a subprocess used any of
5273 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5274 the wrong values were passed to the pipe command ($thisaddress had the
5275 value of $0, $0 had the value of $1, etc). This bug was introduced by
5276 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5278 38. Improved the line breaking for long SMTP error messages from ACLs.
5279 Previously, if there was no break point between 40 and 75 characters, Exim
5280 left the rest of the message alone. Two changes have been made: (a) I've
5281 reduced the minimum length to 35 characters; (b) if it can't find a break
5282 point between 35 and 75 characters, it looks ahead and uses the first one
5283 that it finds. This may give the occasional overlong line, but at least the
5284 remaining text gets split now.
5286 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5287 file descriptors might be low, and that setting 1000 would always raise it.
5288 It turns out that in some environments, the limit is already over 1000 and
5289 that lowering it causes trouble. So now Exim takes care not to decrease it.
5291 40. When delivering a message, the value of $return_path is set to $sender_
5292 address at the start of routing (routers may change the value). By an
5293 oversight, this default was not being set up when an address was tested by
5294 -bt or -bv, which affected the outcome if any router or filter referred to
5297 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5298 log only when "message" or "log_message" are set. However, if one of the
5299 conditions was an address verification, or a call to a nested ACL, the
5300 messages generated by the underlying test were being passed through. This
5301 no longer happens. The underlying message is available in $acl_verify_
5302 message for both "message" and "log_message" expansions, so it can be
5303 passed through if needed.
5305 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5306 new expansion $bh_ to give the encoded byte string without charset
5307 translation. Translation happens only if iconv() is available; HAVE_ICONV
5308 indicates this at build time. HEADERS_CHARSET gives the charset to
5309 translate to; headers_charset can change it in the configuration, and
5310 "headers charset" can change it in an individual filter file.
5312 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5313 that creates RFC 2047 encoded "words" labels them as that charset instead
5314 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5315 expansion operator; (ii) when Exim creates a From: line for a local
5316 message; (iii) when a header line is rewritten to include a "phrase" part.
5318 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5319 buggy, causing it to skip the first lines of messages whose message ID
5320 ended in 'D'. This would not have bitten before Exim release 4.14, because
5321 message IDs were unlikely to end in 'D' before then. The effect was to have
5322 incorrect size information for certain domains.
5324 45. #include "config.h" was missing at the start of the crypt16.c module. This
5325 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5328 46. If there was a timeout during a "random" callout check, Exim treated it as
5329 a failure of the random address, and carried on sending RSET and the real
5330 address. If the delay was just some slowness somewhere, the response to the
5331 original RCPT would be taken as a response to RSET and so on, causing
5332 mayhem of various kinds.
5334 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5335 when I implemented it. It didn't allow for the fact that some option values
5336 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5337 the right test for positive values.
5339 48. Domain names in DNS records are case-independent. Exim always looks them up
5340 in lower case. Some resolvers return domain names in exactly the case they
5341 appear in the zone file, that is, they may contain uppercase letters. Not
5342 all resolvers do this - some return always lower case. Exim was treating a
5343 change of case by a resolver as a change of domain, similar to a widening
5344 of a domain abbreviation. This triggered its re-routing code and so it was
5345 trying to route what was effectively the same domain again. This normally
5346 caused routing to fail (because the router wouldn't handle the domain
5347 twice). Now Exim checks for this case specially, and just changes the
5348 casing of the domain that it ultimately uses when it transmits the message
5351 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5354 50. If a filter generated a file delivery with a non-absolute name (possible if
5355 no home directory exists for the router), the forbid_file option was not
5358 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5359 the existing equality matching.
5361 52. Exim was using ints instead of ino_t variables in some places where it was
5362 dealing with inode numbers.
5364 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5365 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5366 called TMPDIR, and if it finds it is different, it changes its value.
5368 54. The smtp_printf() function is now made available to local_scan() so
5369 additional output lines can be written before returning. There is also an
5370 smtp_fflush() function to enable the detection of a dropped connection.
5371 The variables smtp_input and smtp_batched_input are exported to
5374 55. Changed the default runtime configuration: the message "Unknown user"
5375 has been removed from the ACL, and instead placed on the localuser router,
5376 using the cannot_route_message feature. This means that any verification
5377 failures that generate their own messages won't get overridden. Similarly,
5378 the "Unrouteable address" message that was in the ACL for unverifiable
5379 relay addresses has also been removed.
5381 56. Added hosts_avoid_esmtp to the smtp transport.
5383 57. The exicyclog script was not checking for the esoteric option
5384 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5385 will work only if exicyclog is run under the appropriate euid.
5387 58. Following a discussion on the list, the rules by which Exim recognises line
5388 endings on incoming messages have been changed. The -dropcr and drop_cr
5389 options are now no-ops, retained only for backwards compatibility. The
5390 following line terminators are recognized: LF CRLF CR. However, special
5391 processing applies to CR:
5393 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5394 nor a local message in the state where . is a terminator.
5396 (ii) If a bare CR is encountered in a header line, an extra space is added
5397 after the line terminator so as not to end the header. The reasoning
5398 behind this is that bare CRs in header lines are most likely either
5399 to be mistakes, or people trying to play silly games.
5401 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5402 was being incorrectly given as 18 bytes larger than it should have been.
5403 This is a VOB (very old bug).
5405 60. This may never have affected anything current, but just in case it has:
5406 When the local host is found other than at the start of a list of hosts,
5407 the local host, those with the same MX, and any that follow, are discarded.
5408 When the list in question was part of a longer list of hosts, the following
5409 hosts (not currently being processed) were also being discarded. This no
5410 longer happens. I'm not sure if this situation could ever has previously
5413 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5416 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5417 header. This is something that is recommended in a new Internet Draft, and
5418 is something that is documented as being done by Sendmail. There are two
5419 possible values. For messages generated by the autoreply transport, Exim
5422 Auto-Submitted: auto-replied
5424 whereas for all other generated messages (e.g. bounces) it adds
5426 Auto-Submitted: auto-generated
5428 63. The "personal" condition in filters now includes a test for the
5429 Auto-Submitted: header. If it contains the string "auto-" the message it
5430 not considered personal.
5432 64. Added rcpt_include_affixes as a generic transport option.
5434 65. Added queue_only_override (default true).
5436 66. Added the syslog_duplication option.
5438 67. If what should have been the first header line of a message consisted of
5439 a space followed by a colon, Exim was mis-interpreting it as a header line.
5440 It isn't of course - it is syntactically invalid and should therefore be
5441 treated as the start of the message body. The misbehaviour could have
5442 caused a number of strange effects, including loss of data in subsequent
5443 header lines, and spool format errors.
5445 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5446 client host had authenticated. This control can now be exercised by an ACL
5447 for more flexibility.
5449 69. By default, callouts do not happen when testing with -bh. There is now a
5450 variant, -bhc, which does actually run the callout code, including
5451 consulting and updating the callout cache.
5453 70. Added support for saslauthd authentication, courtesy of Alexander
5456 71. If statvfs() failed on the spool or log directories while checking their
5457 size for availability, Exim confusingly gave the error "space shortage".
5458 Furthermore, in debugging mode it crashed with a floating point exception.
5459 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5460 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5461 serious problem, Exim now writes to the main and panic logs when this
5462 happens, with details of the failure. It then refuses to accept the
5463 incoming message, giving the message "spool directory problem" or "log
5464 directory problem" with a 421 code for SMTP messages.
5466 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5467 0, 1, and 2 exist, because some OS complain for execs without them (see
5468 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5469 descriptors. However, the code omitted to check that the open succeeded,
5470 causing mysterious errors if for some reason the permissions on /dev/null
5471 got screwed. Now Exim writes a message to the main and panic logs, and
5472 bombs out if it can't open /dev/null.
5474 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5475 interact so that it is all more flexible. It is supposed to remain
5476 backwards compatible. Also added extra_local_interfaces.
5478 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5479 to bomb out with an assertion failure - to the client this appears as a
5480 connection drop. This problem occurs in the part of the code that was taken
5481 from the Samba project. Fortunately, the assertion is in a very simple
5482 function, so I have fixed this by reproducing the function inline in the
5483 one place where it is called, and arranging for authentication to fail
5484 instead of killing the process with assert().
5486 75. The SPA client code was not working when the server requested OEM rather
5487 than Unicode encoding.
5489 76. Added code to make require_files with a specific uid setting more usable in
5490 the case where statting the file as root fails - usually a non-root-mounted
5491 NFS file system. When this happens and the failure is EACCES, Exim now
5492 forks a subprocess and does the per-uid checking as the relevant uid.
5494 77. Added process_log_path.
5496 78. If log_file_path was not explicitly set, a setting of check_log_space or
5497 check_log_inodes was ignored.
5499 79. If a space check for the spool or log partitions fails, the incident is now
5500 logged. Of course, in the latter case the data may get lost...
5502 80. Added the %p formatting code to string_format() so that it can be used to
5503 print addresses in debug_print(). Adjusted all the address printing in the
5504 debugging in store.c to use %p rather than %d.
5506 81. There was a concern that a line of code in smtp_in.c could overflow a
5507 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5508 initially expressed, the concern was not well-founded, because trailing
5509 spaces are removed early. However, if the trailing spaces were followed by
5510 a NULL, they did not get removed, so the overflow was possible. Two fixes
5513 (a) I re-wrote the offending code in a cleaner fashion.
5514 (b) If an incoming SMTP command contains a NULL character, it is rejected
5517 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5518 runs initgroups(), so that if the Exim user is in any additional groups,
5519 they will be used during message reception.
5525 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob