1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Solaris build fix for Oracle's LDAP libraries.
8 Bugzilla 1109, patch from Stephen Usher.
10 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
12 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
15 TF/02 Fix a couple more cases where we did not log the error message
16 when unlink() failed. See also change 4.74-TF/03.
18 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
19 lock up or crash if it happened to be inside a call to libc when it
20 got a SIGUSR1 from exiwhat.
22 The SIGUSR1 handler appends the current process status to the process
23 log which is later printed by exiwhat. It used to use the general
24 purpose logging code to do this, but several functions it calls are
27 The new output code in the SIGUSR1 handler is specific to the process
28 log, and simple enough that it's easy to inspect for signal safety.
29 Removing some special cases also simplifies the general logging code.
30 Removing the spurious timestamps from the process log simplifies
33 TF/04 Improved ratelimit ACL condition.
35 The /noupdate option has been deprecated in favour of /readonly which
36 has clearer semantics. The /leaky, /strict, and /readonly update modes
37 are mutually exclusive. The update mode is no longer included in the
38 database key; it just determines when the database is updated. (This
39 means that when you upgrde Exim will forget old rate measurements.)
41 Exim now checks that the per_* options are used with an update mode that
42 makes sense for the current ACL. For example, when Exim is processing a
43 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
44 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
45 must specify per_mail/readonly. If you omit the update mode it defaults to
46 /leaky where that makes sense (as before) or /readonly where required.
48 The /noupdate option is now undocumented but still supported for
49 backwards compatibility. It is equivalent to /readonly except that in
50 ACLs where /readonly is required you may specify /leaky/noupdate or
51 /strict/noupdate which are treated the same as /readonly.
53 A useful new feature is the /count= option. This is a generalization
54 of the per_byte option, so that you can measure the throughput of other
55 aggregate values. For example, the per_byte option is now equivalent
56 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
58 The per_rcpt option has been generalized using the /count= mechanism
59 (though it's more complicated than the per_byte equivalence). When it is
60 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
61 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
62 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
63 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
64 Note that using per_rcpt with a non-readonly update mode in more than
65 one ACL will cause the recipients to be double-counted. (The per_mail
66 and per_byte options don't have this problem.)
68 The handling of very low rates has changed slightly. If the computed rate
69 is less than the event's count (usually one) then this event is the first
70 after a long gap. In this case the rate is set to the same as this event's
71 count, so that the first message of a spam run is counted properly.
73 The major new feature is a mechanism for counting the rate of unique
74 events. The new per_addr option counts the number of different
75 recipients that someone has sent messages to in the last time period. It
76 behaves like per_rcpt if all the recipient addresses are different, but
77 duplicate recipient addresses do not increase the measured rate. Like
78 the /count= option this is a general mechanism, so the per_addr option
79 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
80 example, measure the rate that a client uses different sender addresses
81 with the options per_mail/unique=$sender_address. There are further
82 details in the main documentation.
84 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
86 TF/06 Removed a few PCRE remnants.
88 TF/07 Automatically extract Exim's version number from tags in the git
89 repository when doing development or release builds.
91 PP/02 Raise smtp_cmd_buffer_size to 16kB.
92 Bugzilla 879. Patch from Paul Fisher.
94 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
95 Heavily based on revision 40f9a89a from Simon Arlott's tree.
98 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
100 PP/05 Variable $av_failed, true if the AV scanner deferred.
101 Bugzilla 1078. Patch from John Horne.
103 PP/06 Stop make process more reliably on build failure.
104 Bugzilla 1087. Patch from Heiko Schlittermann.
106 PP/07 Make maildir_use_size_file an _expandable_ boolean.
107 Bugzilla 1089. Patch from Heiko Schlittermann.
109 PP/08 Handle ${run} returning more data than OS pipe buffer size.
110 Bugzilla 1131. Patch from Holger Weiß.
112 PP/09 Handle IPv6 addresses with SPF.
113 Bugzilla 860. Patch from Wolfgang Breyha.
119 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
121 PP/02 Harmonised TLS library version reporting; only show if debugging.
122 Layout now matches that introduced for other libraries in 4.74 PP/03.
124 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
126 PP/04 New "dns_use_edns0" global option.
128 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
131 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
132 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
134 TK/01 Updated PolarSSL code to 0.14.2.
135 Bugzilla 1097. Patch from Andreas Metzler.
137 PP/07 Catch divide-by-zero in ${eval:...}.
140 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
143 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
144 format-string attack -- SECURITY: remote arbitrary code execution.
146 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
147 time unintentionally subject to list matching rules, letting the header
148 cause arbitrary Exim lookups (of items which can occur in lists, *not*
149 arbitrary string expansion). This allowed for information disclosure.
151 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
152 INT_MIN/-1 -- value coerced to INT_MAX.
158 NM/01 Workround for PCRE version dependency in version reporting
161 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
162 This fixes portability to compilers other than gcc, notably
163 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
165 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
166 makefiles for portability to HP-UX and POSIX correctness.
168 PP/01 Permit LOOKUP_foo enabling on the make command-line.
169 Also via indented variable definition in the Makefile.
170 (Debugging by Oliver Heesakkers).
172 PP/02 Restore caching of spamd results with expanded spamd_address.
173 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
175 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
176 Improves build reliability. Fix from: Frank Elsner
178 NM/02 Fix wide character breakage in the rfc2047 coding
179 Fixes bug 1064. Patch from Andrey N. Oktyabrski
181 NM/03 Allow underscore in dnslist lookups
182 Fixes bug 1026. Patch from Graeme Fowler
184 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
185 Code patches from Adam Ciarcinski of NetBSD.
187 NM/04 Fixed exiqgrep to cope with mailq missing size issue
190 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
191 is logged, to avoid truncation. Patch from John Horne.
193 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
194 Patch from Jakob Hirsch.
196 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
197 SQL string expansion failure details.
198 Patch from Andrey Oktyabrski.
200 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
201 Patch from Simon Arlott.
203 PP/09 New lookups functionality failed to compile on old gcc which rejects
204 extern declarations in function scope.
205 Patch from Oliver Fleischmann
207 PP/10 Use sig_atomic_t for flags set from signal handlers.
208 Check getgroups() return and improve debugging.
209 Fixed developed for diagnosis in bug 927 (which turned out to be
212 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
213 Patch from Mark Zealey.
215 PP/12 Bugzilla 1056: Improved spamd server selection.
216 Patch from Mark Zealey.
218 PP/13 Bugzilla 1086: Deal with maildir quota file races.
219 Based on patch from Heiko Schlittermann.
221 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
222 Patch from Uwe Doering, sign-off by Michael Haardt.
224 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
225 variable declaration deep within a block. Bug and patch from
228 PP/15 lookups-Makefile IRIX compatibilty coercion.
230 PP/16 Make DISABLE_DKIM build knob functional.
232 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
233 Patch by Simon Arlott
235 TF/03 Fix valgrind.h portability to C89 compilers that do not support
236 variable argument macros. Our copy now differs from upstream.
242 TF/01 Failure to get a lock on a hints database can have serious
243 consequences so log it to the panic log.
245 TF/02 Log LMTP confirmation messages in the same way as SMTP,
246 controlled using the smtp_confirmation log selector.
248 TF/03 Include the error message when we fail to unlink a spool file.
250 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
251 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
252 for maintaining out-of-tree patches for some time.
254 PP/01 Bugzilla 139: Documentation and portability issues.
255 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
256 Handle per-OS dynamic-module compilation flags.
258 PP/02 Let /dev/null have normal permissions.
259 The 4.73 fixes were a little too stringent and complained about the
260 permissions on /dev/null. Exempt it from some checks.
261 Reported by Andreas M. Kirchwitz.
263 PP/03 Report version information for many libraries, including
264 Exim version information for dynamically loaded libraries. Created
265 version.h, now support a version extension string for distributors
266 who patch heavily. Dynamic module ABI change.
268 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
269 privilege escalation vulnerability whereby the Exim run-time user
270 can cause root to append content of the attacker's choosing to
273 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
276 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
277 If dropping privileges for untrusted macros, we disabled normal logging
278 on the basis that it would fail; for the Exim run-time user, this is not
279 the case, and it resulted in successful deliveries going unlogged.
280 Fixed. Reported by Andreas Metzler.
286 PP/01 Date: & Message-Id: revert to normally being appended to a message,
287 only prepend for the Resent-* case. Fixes regression introduced in
288 Exim 4.70 by NM/22 for Bugzilla 607.
290 PP/02 Include check_rfc2047_length in configure.default because we're seeing
291 increasing numbers of administrators be bitten by this.
293 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
295 PP/03 Bugzilla 994: added openssl_options main configuration option.
297 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
299 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
301 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
303 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
304 without a peer certificate, leading to a segfault because of an
305 assumption that peers always have certificates. Be a little more
306 paranoid. Problem reported by Martin Tscholak.
308 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
309 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
310 NB: ClamAV planning to remove STREAM in "middle of 2010".
311 CL also introduces -bmalware, various -d+acl logging additions and
312 more caution in buffer sizes.
314 PP/09 Implemented reverse_ip expansion operator.
316 PP/10 Bugzilla 937: provide a "debug" ACL control.
318 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
320 PP/12 Bugzilla 973: Implement --version.
322 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
324 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
326 PP/15 Bugzilla 816: support multiple condition rules on Routers.
328 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
329 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
330 ignore trailing whitespace.
332 JJ/02 prevent non-panic DKIM error from being sent to paniclog
334 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
337 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
338 Notification from Dr Andrew Aitchison.
340 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
341 ExtendedDetectionInfo response format.
342 Notification from John Horne.
344 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
347 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
348 XSL and documented dependency on system catalogs, with examples of how
351 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
354 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
355 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
356 configuration file which is writeable by the Exim user or group.
358 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
359 of configuration files to cover files specified with the -C option if
360 they are going to be used with root privileges, not just the default
363 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
364 option (effectively making it always true).
366 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
367 files to be used while preserving root privileges.
369 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
370 that rogue child processes cannot use them.
372 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
373 run-time user, instead of root.
375 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
376 Exim run-time user without dropping privileges.
378 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
379 result string, instead of calling string_vformat() twice with the same
382 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
383 for other users. Others should always drop root privileges if they use
384 -C on the command line, even for a whitelisted configure file.
386 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
388 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
394 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
395 $data_path, and $header_path variables; fixed documentation bugs and
398 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
399 exipick to access non-standard spools, including the "frozen" queue
402 NM/01 Bugzilla 965: Support mysql stored procedures.
403 Patch from Alain Williams
405 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
407 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
408 Patch from Andreas Metzler
410 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
411 Patch from Kirill Miazine
413 NM/05 Bugzilla 671: Added umask to procmail example.
415 JJ/03 installed exipick 20100323.0, fixing doc bug
417 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
418 directory. Notification and patch from Dan Rosenberg.
420 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
422 TK/02 Improve log output when DKIM signing operation fails.
424 MH/01 Treat the transport option dkim_domain as a colon separated
425 list, not as a single string, and sign the message with each element,
426 omitting multiple occurences of the same signer.
428 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
429 Bugzilla 985, 986. Patch by Simon Arlott
431 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
432 Patch by Simon Arlott
434 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
435 MBX locking. Notification from Dan Rosenberg.
441 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
443 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
445 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
447 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
449 NM/04 Bugzilla 915: Flag broken perl installation during build.
455 TK/01 Added patch by Johannes Berg that expands the main option
456 "spamd_address" if it starts with a dollar sign.
458 TK/02 Write list of recipients to X-Envelope-Sender header when building
459 the mbox-format spool file for content scanning (suggested by Jakob
462 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
463 (http://www.dcc-servers.net/) support via dccifd. Activated by
464 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
466 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
467 by Mark Daniel Reidel <mr@df.eu>.
469 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
470 When building exim an external PCRE library is now needed -
471 PCRE is a system library on the majority of modern systems.
472 See entry on PCRE_LIBS in EDITME file.
474 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
475 conversation. Added nologin parameter to request.
476 Patch contributed by Kirill Miazine.
478 TF/01 Do not log submission mode rewrites if they do not change the address.
480 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
482 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
483 log files in place. Contributed by Roberto Lima.
485 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
487 TF/03 Bugzilla 615: When checking the local_parts router precondition
488 after a local_part_suffix or local_part_prefix option, Exim now
489 does not use the address's named list lookup cache, since this
490 contains cached lookups for the whole local part.
492 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
493 Robert Millan. Documentation is in experimental-spec.txt.
495 TF/04 Bugzilla 668: Fix parallel build (make -j).
497 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
499 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
500 Patch provided by Jan Srzednicki.
502 TF/05 Leading white space used to be stripped from $spam_report which
503 wrecked the formatting. Now it is preserved.
505 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
506 that they are available at delivery time.
508 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
510 TF/08 TLS error reporting now respects the incoming_interface and
511 incoming_port log selectors.
513 TF/09 Produce a more useful error message if an SMTP transport's hosts
514 setting expands to an empty string.
516 NM/06 Bugzilla 744: EXPN did not work under TLS.
517 Patch provided by Phil Pennock.
519 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
520 Patch provided by Richard Godbee.
522 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
523 acl_smtp_notquit, added index entry.
525 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
526 Patch provided by Eugene Bujak.
528 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
529 accept(). Patch provided by Maxim Dounin.
531 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
532 Patch provided by Phil Pennock.
534 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
536 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
537 Patch provided by Brad "anomie" Jorsch.
539 NM/14 Bugzilla 622: Added timeout setting to transport filter.
540 Patch provided by Dean Brooks.
542 TK/05 Add native DKIM support (does not depend on external libraries).
544 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
545 Patch provided by Graeme Fowler.
547 NM/16 Bugzilla 851: Documentation example syntax fix.
549 NM/17 Changed NOTICE file to remove references to embedded PCRE.
551 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
554 NM/19 Bugzilla 745: TLS version reporting.
555 Patch provided by Phil Pennock.
557 NM/20 Bugzilla 167: bool: condition support.
558 Patch provided by Phil Pennock.
560 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
561 clients. Patch provided by Phil Pennock.
563 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
564 Patch provided by Brad "anomie" Jorsch.
566 NM/23 Bugzilla 687: Fix misparses in eximstats.
567 Patch provided by Heiko Schlittermann.
569 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
570 Patch provided by Heiko Schlittermann.
572 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
573 plus update to original patch.
575 NM/26 Bugzilla 799: Documentation correction for ratelimit.
577 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
578 Patch provided by David Brownlee.
580 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
582 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
584 NM/30 Bugzilla 888: TLS documentation bugfixes.
586 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
588 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
589 Unlike the original bugzilla I have changed all shell scripts in src tree.
591 NM/33 Bugzilla 898: Transport filter timeout fix.
592 Patch by Todd Rinaldo.
594 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
595 Patch by Serge Demonchaux.
597 NM/35 Bugzilla 39: Base64 decode bug fixes.
598 Patch by Jakob Hirsch.
600 NM/36 Bugzilla 909: Correct connect() call in dcc code.
602 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
604 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
606 NM/39 Bugzilla 911: Fixed MakeLinks build script.
612 TK/01 Add preliminary DKIM support. Currently requires a forked version of
613 ALT-N's libdkim that I have put here:
614 http://duncanthrax.net/exim-experimental/
616 Note to Michael Haardt: I had to rename some vars in sieve.c. They
617 were called 'true' and it seems that C99 defines that as a reserved
618 keyword to be used with 'bool' variable types. That means you could
619 not include C99-style headers which use bools without triggering
620 build errors in sieve.c.
622 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
623 as mailq or other aliases. Changed the --help handling significantly
624 to do whats expected. exim_usage() emits usage/help information.
626 SC/01 Added the -bylocaldomain option to eximstats.
628 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
630 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
632 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
638 PH/01 Another patch from the Sieve maintainer.
640 PH/02 When an IPv6 address is converted to a string for single-key lookup
641 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
642 dots are used instead of colons so that keys in lsearch files need not
643 contain colons. This was done some time before quoting was made available
644 in lsearch files. However, iplsearch files do require colons in IPv6 keys
645 (notated using the quote facility) so as to distinguish them from IPv4
646 keys. This meant that lookups for IP addresses in host lists did not work
647 for iplsearch lookups.
649 This has been fixed by arranging for IPv6 addresses to be expressed with
650 colons if the lookup type is iplsearch. This is not incompatible, because
651 previously such lookups could never work.
653 The situation is now rather anomolous, since one *can* have colons in
654 ordinary lsearch keys. However, making the change in all cases is
655 incompatible and would probably break a number of configurations.
657 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
660 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
661 conversion specification without a maximum field width, thereby enabling
662 a rogue spamd server to cause a buffer overflow. While nobody in their
663 right mind would setup Exim to query an untrusted spamd server, an
664 attacker that gains access to a server running spamd could potentially
665 exploit this vulnerability to run arbitrary code as the Exim user.
667 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
668 $primary_hostname instead of what libspf2 thinks the hosts name is.
670 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
671 a directory entry by the name of the lookup key. Previously, if a
672 symlink pointed to a non-existing file or a file in a directory that
673 Exim lacked permissions to read, a lookup for a key matching that
674 symlink would fail. Now it is enough that a matching directory entry
675 exists, symlink or not. (Bugzilla 503.)
677 PH/03 The body_linecount and body_zerocount variables are now exported in the
680 PH/04 Added the $dnslist_matched variable.
682 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
683 This means they are set thereafter only if the connection becomes
686 PH/06 Added the client_condition to authenticators so that some can be skipped
687 by clients under certain conditions.
689 PH/07 The error message for a badly-placed control=no_multiline_responses left
690 "_responses" off the end of the name.
692 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
694 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
695 (without spaces) instead of just copying the configuration text.
697 PH/10 Added the /noupdate option to the ratelimit ACL condition.
699 PH/11 Added $max_received_linelength.
701 PH/12 Added +ignore_defer and +include_defer to host lists.
703 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
704 way in which PCRE > 7.0 is built.
706 PH/14 Implemented queue_only_load_latch.
708 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
709 MAIL command. The effect was to mangle the value on 64-bit systems.
711 PH/16 Another patch from the Sieve maintainer.
713 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
715 PH/18 If a system quota error occurred while trying to create the file for
716 a maildir delivery, the message "Mailbox is full" was not appended to the
717 bounce if the delivery eventually timed out. Change 4.67/27 below applied
718 only to a quota excession during the actual writing of the file.
720 PH/19 It seems that peer DN values may contain newlines (and other non-printing
721 characters?) which causes problems in log lines. The DN values are now
722 passed through string_printing() before being added to log lines.
724 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
725 and InterBase are left for another time.)
727 PH/21 Added message_body_newlines option.
729 PH/22 Guard against possible overflow in moan_check_errorcopy().
731 PH/23 POSIX allows open() to be a macro; guard against that.
733 PH/24 If the recipient of an error message contained an @ in the local part
734 (suitably quoted, of course), incorrect values were put in $domain and
735 $local_part during the evaluation of errors_copy.
741 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
742 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
745 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
746 issue a MAIL command.
748 PH/02 In an ACL statement such as
750 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
752 if a client was not listed at all, or was listed with a value other than
753 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
754 the condition was not true (as it should be), so access was not denied.
755 The bug was that the ! inversion was incorrectly passed on to the second
756 item. This has been fixed.
758 PH/03 Added additional dnslists conditions == and =& which are different from
759 = and & when the dns lookup returns more than one IP address.
761 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
762 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
764 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
765 FSYNC, which compiles an option called disable_fsync that allows for
766 bypassing fsync(). The documentation is heavily laced with warnings.
768 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
770 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
771 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
772 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
773 including adding "make clean"; (3) Added -fPIC when compiling the test
774 dynamically loaded module, to get rid of a warning.
776 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
777 message fails, move_frozen_messages = true and ignore_bounce_errors_after
778 = 0s. The bug is otherwise harmless.
780 PH/07 There was a bug in the dovecot authenticator such that the value of
781 $auth1 could be overwritten, and so not correctly preserved, after a
782 successful authentication. This usually meant that the value preserved by
783 the server_setid option was incorrect.
785 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
787 PH/09 Installed PCRE release 7.0.
789 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
790 run for batched SMTP input. It is now run at the start of every message
791 in the batch. While fixing this I discovered that the process information
792 (output by running exiwhat) was not always getting set for -bs and -bS
793 input. This is fixed, and it now also says "batched" for BSMTP.
795 PH/11 Added control=no_pipelining.
797 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
798 patch, slightly modified), and move the expansion of helo_data till after
799 the connection is made in the smtp transport (so it can use these
802 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
804 PH/14 Added log_selector = +pid.
806 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
808 PH/16 Add ${if forany and ${if forall.
810 PH/17 Added dsn_from option to vary the From: line in DSNs.
812 PH/18 Flush SMTP output before performing a callout, unless control =
813 no_callout_flush is set.
815 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
816 was true (the default) a successful delivery failed to delete the retry
817 item, thus causing premature timeout of the address. The bug is now
820 PH/20 Added hosts_avoid_pipelining to the smtp transport.
822 PH/21 Long custom messages for fakedefer and fakereject are now split up
823 into multiline reponses in the same way that messages for "deny" and
824 other ACL rejections are.
826 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
827 with slight modification.
829 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
830 draft, changing the syntax and factoring some duplicate code".
832 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
833 for deliveries of the second and subsequent messages over the same SMTP
836 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
837 ${reduce, with only minor "tidies".
839 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
841 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
842 expansion side effects.
844 PH/27 When a message times out after an over-quota error from an Exim-imposed
845 quota, the bounce message says "mailbox is full". This message was not
846 being given when it was a system quota that was exceeded. It now should
849 MH/03 Made $recipients available in local_scan(). local_scan() already has
850 better access to the recipient list through recipients_list[], but
851 $recipients can be useful in postmaster-provided expansion strings.
853 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
854 in the case of a MAIL command with additional options following the
855 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
856 were accidentally chopped off.
858 PH/29 SMTP synchronization checks are implemented when a command is read -
859 there is a check that no more input is waiting when there shouldn't be
860 any. However, for some commands, a delay in an ACL can mean that it is
861 some time before the response is written. In this time, more input might
862 arrive, invalidly. So now there are extra checks after an ACL has run for
863 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
864 pipelining has not been advertised.
866 PH/30 MH's patch to allow iscntrl() characters to be list separators.
868 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
869 returned in the SMTP response when smtp_return_error_details was false.
872 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
873 instead of the C I/O that was originally supplied, because problems were
876 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
877 Exim which did not show up earlier: it was assuming that a call to
878 SSL_CTX_set_info_callback() might give an error value. In fact, there is
879 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
880 was a macro that became an assignment, so it seemed to work. This has
881 changed to a proper function call with a void return, hence the compile
882 error. Exim's code has been fixed.
884 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
887 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
889 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
891 PH/37 If a message is not accepted after it has had an id assigned (e.g.
892 because it turns out to be too big or there is a timeout) there is no
893 "Completed" line in the log. When some messages of this type were
894 selected by exigrep, they were listed as "not completed". Others were
895 picked up by some special patterns. I have improved the selection
896 criteria to be more general.
898 PH/38 The host_find_failed option in the manualroute router can now be set
899 to "ignore", to completely ignore a host whose IP address cannot be
900 found. If all hosts are ignored, the behaviour is controlled by the new
901 host_all_ignored option.
903 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
904 homing or because of multiple MX records with /mx) generated more than
905 one IP address, and the following item turned out to be the local host,
906 all the secondary addresses of the first item were incorrectly removed
907 from the list, along with the local host and any following hosts (which
908 is what is supposed to happen).
910 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
911 whoever called Exim into the -H file. In the case of the daemon it was
912 behaving confusingly. When first started, it used values for whoever
913 started the daemon, but after a SIGHUP it used the Exim user (because it
914 calls itself on a restart). I have changed the code so that it now always
917 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
918 message are rejected with the same error (e.g. no authentication or bad
919 sender address), and a DATA command is nevertheless sent (as can happen
920 with PIPELINING or a stupid MUA), the error message that was given to the
921 RCPT commands is included in the rejection of the DATA command. This is
922 intended to be helpful for MUAs that show only the final error to their
925 PH/42 Another patch from the Sieve maintainer.
927 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
928 Eximstats - Fixed some broken HTML links and added missing column headers
930 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
931 columns for Rejects, Temp Rejects, Ham, and Spam rows.
933 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
935 PH/43 Yet another patch from the Sieve maintainer.
937 PH/44 I found a way to check for a TCP/IP connection going away before sending
938 the response to the final '.' that terminates a message, but only in the
939 case where the client has not sent further data following the '.'
940 (unfortunately, this is allowed). However, in many cases there won't be
941 any further data because there won't be any more messages to send. A call
942 to select() can be used: if it shows that the input is "ready", there is
943 either input waiting, or the socket has been closed. An attempt to read
944 the next input character can distinguish the two cases. Previously, Exim
945 would have sent an OK response which the client would never have see.
946 This could lead to message repetition. This fix should cure that, at
947 least in a lot of common cases.
949 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
950 advertised in response to EHLO.
956 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
957 fixed by 4.65/MH/01 (is this a record?) are fixed:
959 (i) An empty string was always treated as zero by the numeric comparison
960 operators. This behaviour has been restored.
962 (ii) It is documented that the numeric comparison operators always treat
963 their arguments as decimal numbers. This was broken in that numbers
964 starting with 0 were being interpreted as octal.
966 While fixing these problems I realized that there was another issue that
967 hadn't been noticed. Values of message_size_limit (both the global option
968 and the transport option) were treated as octal if they started with 0.
969 The documentation was vague. These values are now always treated as
970 decimal, and I will make that clear in the documentation.
976 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
977 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
980 MH/01 Don't check that the operands of numeric comparison operators are
981 integers when their expansion is in "skipping" mode (fixes bug
982 introduced by 4.64-PH/07).
984 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
985 child addresses, Exim now panics and dies. Previously, because the count
986 is held in a short int, deliveries were likely to be lost. As such a
987 large number of recipients for a single message is ridiculous
988 (performance will be very, very poor), I have chosen to impose a limit
989 rather than extend the field.
995 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
996 leftover -K file (the existence of which was triggered by #402).
997 While we were at it, introduced process PID as part of the -K
998 filename. This should rule out race conditions when creating
1001 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1002 processing considerably. Previous code took too long for large mails,
1003 triggering a timeout which in turn triggers #401.
1005 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1006 in the DK code in transports.c. sendfile() is not really portable,
1007 hence the _LINUX specificness.
1009 TF/01 In the add_headers option to the mail command in an Exim filter,
1010 there was a bug that Exim would claim a syntax error in any
1011 header after the first one which had an odd number of characters
1014 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1015 callout verification, Exim cached a "reject" for the entire domain. This
1016 is correct for most verifications, but it is not correct for a recipient
1017 verification with use_sender or use_postmaster set, because in that case
1018 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1019 case of MAIL FROM:<> rejection from other early rejections (e.g.
1020 rejection of HELO). When verifying a recipient using a non-null MAIL
1021 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1022 Whatever the result of the callout, the value of the domain cache is
1023 left unchanged (for any other kind of callout, getting as far as trying
1024 RCPT means that the domain itself is ok).
1026 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1029 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1030 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1031 session, the daemon ignores ECONNECT errors and logs others; it now
1032 ignores EPIPE as well.
1034 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1035 (quoted-printable decoding).
1037 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1038 later the small subsequent patch to fix an introduced bug.
1040 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1042 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1044 PH/08 An error is now given if message_size_limit is specified negative.
1046 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1047 to be given (somewhat) arbitrary names.
1049 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1052 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1053 miscellaneous code fixes
1055 PH/10 Added the log_reject_target ACL modifier to specify where to log
1058 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1059 hostname. This is wrong, because it relates to the incoming message (and
1060 probably the interface on which it is arriving) and not to the outgoing
1061 callout (which could be using a different interface). This has been
1062 changed to use the value of the helo_data option from the smtp transport
1063 instead - this is what is used when a message is actually being sent. If
1064 there is no remote transport (possible with a router that sets up host
1065 addresses), $smtp_active_hostname is used.
1067 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1068 tweaks were necessary in order to get it to work (see also 21 below):
1069 (a) The code assumed that strncpy() returns a negative number on buffer
1070 overflow, which isn't the case. Replaced with Exim's string_format()
1072 (b) There were several signed/unsigned issues. I just did the minimum
1073 hacking in of casts. There is scope for a larger refactoring.
1074 (c) The code used strcasecmp() which is not a standard C function.
1075 Replaced with Exim's strcmpic() function.
1076 (d) The code set only $1; it now sets $auth1 as well.
1077 (e) A simple test gave the error "authentication client didn't specify
1078 service in request". It would seem that Dovecot has changed its
1079 interface. Fortunately there's a specification; I followed it and
1080 changed what the client sends and it appears to be working now.
1082 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1085 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1086 address A is aliased to B and C, where B exists and C does not. Without
1087 -v the output is "A verified" because verification stops after a
1088 successful redirection if more than one address is generated. However,
1089 with -v the child addresses are also verified. Exim was outputting "A
1090 failed to verify" and then showing the successful verification for C,
1091 with its parentage. It now outputs "B failed to verify", showing B's
1092 parentage before showing the successful verification of C.
1094 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1095 look up a TXT record in a specific list after matching in a combined
1098 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1099 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1100 they consult the DNS. I had assumed they would set it the way they
1101 wanted; and indeed my experiments on Linux seem to show that in some
1102 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1103 To be on the safe side, however, I have now made the interface to
1104 host_find_byname() similar to host_find_bydns(), with an argument
1105 containing the DNS resolver options. The host_find_byname() function now
1106 sets these options at its start, just as host_find_bydns() does. The smtp
1107 transport options dns_qualify_single and dns_search_parents are passed to
1108 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1109 of host_find_byname() use the default settings of RES_DEFNAMES
1110 (qualify_single) but not RES_DNSRCH (search_parents).
1112 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1113 spool_read_header() do less string testing, by means of a preliminary
1114 switch on the second character of optional "-foo" lines. (This is
1115 overdue, caused by the large number of possibilities that now exist.
1116 Originally there were few.) While I was there, I also converted the
1117 str(n)cmp tests so they don't re-test the leading "-" and the first
1118 character, in the hope this might squeeze out yet more improvement.
1120 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1121 flag allowing group syntax was set by the header_syntax check but not
1122 turned off, possible causing trouble later; (2) The flag was not being
1123 set at all for the header_verify test, causing "group"-style headers to
1124 be rejected. I have now set it in this case, and also caused header_
1125 verify to ignore an empty address taken from a group. While doing this, I
1126 came across some other cases where the code for allowing group syntax
1127 while scanning a header line wasn't quite right (mostly, not resetting
1128 the flag correctly in the right place). These bugs could have caused
1129 trouble for malformed header lines. I hope it is now all correct.
1131 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1132 with the "reply" argument non-NULL. The code, however (which originally
1133 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1134 but it didn't always do it. This confused somebody who was copying the
1135 code for some other use. I have removed all the tests.
1137 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1138 feature that was used to support insecure browsers during the U.S. crypto
1139 embargo. It requires special client support, and Exim is probably the
1140 only MTA that supported it -- and would never use it because real RSA is
1141 always available. This code has been removed, because it had the bad
1142 effect of slowing Exim down by computing (never used) parameters for the
1143 RSA_EXPORT functionality.
1145 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1146 authenticator to fail if there's a tab character in the incoming data
1147 (there should never be unless someone is messing about, as it's supposed
1148 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1149 if the connection is using TLS or if the remote IP is the same as the
1150 local IP, and the "valid-client-cert option" if a client certificate has
1153 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1154 authenticators. This can be used for authorization after authentication
1155 succeeds. (In the case of plaintext, it servers for both authentication
1158 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1159 if any retry times were supplied.
1161 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1162 connection, where there is no client IP address to check. In this
1163 situation, the verify now always succeeds.
1165 PH/25 Applied John Jetmore's -Mset patch.
1167 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1169 PH/27 In a string expansion for a processed (not raw) header when multiple
1170 headers of the same name were present, leading whitespace was being
1171 removed from all of them, but trailing whitespace was being removed only
1172 from the last one. Now trailing whitespace is removed from each header
1173 before concatenation. Completely empty headers in a concatenation (as
1174 before) are ignored.
1176 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1177 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1179 PH/29 [Removed. This was a change that I later backed out, and forgot to
1180 correct the ChangeLog entry (that I had efficiently created) before
1181 committing the later change.]
1183 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1184 address errors (4xx response to RCPT) over the same connection as other
1185 messages routed to the same hosts. Such deliveries are always "forced",
1186 so retry times are not inspected. This resulted in far too many retries
1187 for the affected addresses. The effect occurred only when there were more
1188 hosts than the hosts_max_try setting in the smtp transport when it had
1189 the 4xx errors. Those hosts that it had tried were not added to the list
1190 of hosts for which the message was waiting, so if all were tried, there
1191 was no problem. Two fixes have been applied:
1193 (i) If there are any address or message errors in an SMTP delivery, none
1194 of the hosts (tried or untried) are now added to the list of hosts
1195 for which the message is waiting, so the message should not be a
1196 candidate for sending over the same connection that was used for a
1197 successful delivery of some other message. This seems entirely
1198 reasonable: after all the message is NOT "waiting for some host".
1199 This is so "obvious" that I'm not sure why it wasn't done
1200 previously. Hope I haven't missed anything, but it can't do any
1201 harm, as the worst effect is to miss an optimization.
1203 (ii) If, despite (i), such a delivery is accidentally attempted, the
1204 routing retry time is respected, so at least it doesn't keep
1205 hammering the server.
1207 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1208 in ${readsocket because some servers need this prod.
1210 PH/32 Added some extra debug output when updating a wait-xxx database.
1212 PH/33 The hint "could be header name not terminated by colon", which has been
1213 given for certain expansion errors for a long time, was not being given
1214 for the ${if def:h_colon_omitted{... case.
1216 PH/34 The spec says: "With one important exception, whenever a domain list is
1217 being scanned, $domain contains the subject domain." There was at least
1218 one case where this was not true.
1220 PH/35 The error "getsockname() failed: connection reset by peer" was being
1221 written to the panic log as well as the main log, but it isn't really
1222 panic-worthy as it just means the connection died rather early on. I have
1223 removed the panic log writing for the ECONNRESET error when getsockname()
1226 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1227 runs only) independently of the message's sender address. This meant
1228 that, if the 4xx error was in fact related to the sender, a different
1229 message to the same recipient with a different sender could confuse
1230 things. In particualar, this can happen when sending to a greylisting
1231 server, but other circumstances could also provoke similar problems.
1232 I have changed the default so that the retry time for these errors is now
1233 based a combination of the sender and recipient addresses. This change
1234 can be overridden by setting address_retry_include_sender=false in the
1237 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1238 remote server are returned as part of bounce messages. This was not
1239 happening for LMTP over a pipe (the lmtp transport), but now it is the
1240 same for both kinds of LMTP.
1242 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1243 in header lines that were in fact CNAMEs. This is no longer the case.
1245 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1246 and queue runs started by the daemon processed all messages. This has
1247 been fixed so that -R and -S can now usefully be given with -q<time>.
1249 PH/40 Import PCRE release 6.7 (fixes some bugs).
1251 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1253 PH/42 Give an error if -q is specified more than once.
1255 PH/43 Renamed the variables $interface_address and $interface_port as
1256 $received_ip_address and $received_port, to make it clear that these
1257 values apply to message reception, and not to the outgoing interface when
1258 a message is delivered. (The old names remain recognized, of course.)
1260 PH/44 There was no timeout on the connect() call when using a Unix domain
1261 socket in the ${readsocket expansion. There now is.
1263 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1264 be meaningful with "accept".
1266 SC/01 Eximstats V1.43
1267 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1269 SC/02 Eximstats V1.44
1270 Use a glob alias rather than an array ref in the generated
1271 parser. This improves both readability and performance.
1273 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1274 Collect SpamAssassin and rejection statistics.
1275 Don't display local sender or destination tables unless
1276 there is data to show.
1277 Added average volumes into the top table text output.
1279 SC/04 Eximstats V1.46
1280 Collect data on the number of addresses (recipients)
1281 as well as the number of messages.
1283 SC/05 Eximstats V1.47
1284 Added 'Message too big' to the list of mail rejection
1285 reasons (thanks to Marco Gaiarin).
1287 SC/06 Eximstats V1.48
1288 Mainlog lines which have GMT offsets and are too short to
1289 have a flag are now skipped.
1291 SC/07 Eximstats V1.49 (Alain Williams)
1292 Added the -emptyok flag.
1294 SC/08 Eximstats V1.50
1295 Fixes for obtaining the IP address from reject messages.
1297 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1298 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1299 whitesspace changes from 4.64-PH/27
1301 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1304 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1305 are found, allow negative numbers in numeric criteria)
1307 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1309 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1310 to match changes made in 4.64-PH/43
1312 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1314 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1315 "domain" is actually an IP address, and force a failure. This locks out
1316 those revolvers/nameservers that support "A-for-A" lookups, in
1317 contravention of the specifications.
1319 PH/48 When a host name was looked up from an IP address, and the subsequent
1320 forward lookup of the name timed out, the host name was left in
1321 $sender_host_name, contrary to the specification.
1323 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1324 restricted to single-key lookups, Exim was not diagnosing an error if
1325 * or *@ was used with a query-style lookup.
1327 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1329 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1330 long ago, but noone interested enough thought of it. Let's just say that
1331 the "1.1" means that there are some new functions that weren't there at
1332 some point in the past.
1334 PH/51 Error processing for expansion failure of helo_data from an smtp
1335 transport during callout processing was broken.
1337 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1338 tested/used via the -bh/-bhc/-bs options.
1340 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1341 bug, fixed in subsequent PCRE releases).
1343 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1344 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1346 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1352 SC/01 Use a glob alias rather than an array ref in eximstats generated
1353 parser. This improves both readability and performance.
1355 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1356 Don't display local sender or destination tables in eximstats unless
1357 there is data to show.
1358 Added average volumes into the eximstats top table text output.
1360 SC/03 Collect data on the number of addresses (recipients) as well
1361 as the number of messages in eximstats.
1363 TF/01 Correct an error in the documentation for the redirect router. Exim
1364 does (usually) call initgroups() when daemonizing.
1366 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1367 with consistent privilege compared to when running as a daemon.
1369 TF/03 Note in the spec that $authenticated_id is not set for local
1370 submissions from trusted users.
1372 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1373 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1375 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1376 by adding some example configuration directives to the default
1377 configuration file. A little bit of work is required to uncomment the
1378 directives and define how usernames and passwords are checked, but
1379 there is now a framework to start from.
1381 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1382 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1383 without this. I don't know how relevant this is to other LDAP libraries.
1385 PH/02 Add the verb name to the "unknown ACL verb" error.
1387 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1389 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1391 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1392 directory not expanded when it should be if an expanded home directory
1393 was set for the address (which is overridden by the transport).
1395 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1398 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1399 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1400 because it is too late at that time, and has no effect.
1402 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1403 security issue with \' (bugzilla #107). I could not use the
1404 PQescapeStringConn() function, because it needs a PGconn value as one of
1407 PH/08 When testing addresses using -bt, indicate those final addresses that
1408 are duplicates that would not cause an additional delivery. At least one
1409 person was confused, thinking that -bt output corresponded to deliveries.
1410 (Suppressing duplicates isn't a good idea as you lose the information
1411 about possibly different redirections that led to the duplicates.)
1413 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1414 systems where poll() doesn't work, in particular OS X.
1416 PH/10 Added more information to debugging output for retry time not reached.
1418 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1419 operations in malware.c.
1421 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1424 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1425 syslog_duplication set false, log lines that would normally be written
1426 both the the main log and to the reject log were not written to syslog at
1429 PH/14 In the default configuration, change the use of "message" in ACL warn
1430 statements to "add_header".
1432 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1433 not followed by a command (e.g. "seen endif").
1435 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1436 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1439 PH/17 Added extra conditions to the default value of delay_warning_condition
1443 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1444 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1445 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1448 The Auto-Submitted: and various List- headers are standardised, whereas I
1449 don't think Precedence: ever was.
1451 PH/18 Refactored debugging code in route_finduser() to show more information,
1452 in particular, the error code if getpwnam() issues one.
1454 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1455 This is apparently needed in addition to the PH/07 change above to avoid
1456 any possible encoding problems.
1458 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1459 but not after initializing Perl.
1461 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1462 output them only if debugging. By default they are written stderr,
1463 apparently, which is not desirable.
1465 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1468 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1471 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1473 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1474 authenticated or an ident call has been made. Suppress the default
1475 values for $authenticated_id and $authenticated_sender (but permit -oMai
1476 and -oMas) when testing with -bh.
1478 PH/24 Re-jigged the order of the tests in the default configuration so that the
1479 tests for valid domains and recipients precede the DNS black list and CSA
1480 tests, on the grounds that those ones are more expensive.
1482 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1483 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1484 This bug exists in every version of Exim that I still have, right back to
1487 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1488 However, an attempt to turn on case-sensitivity in a regex key by
1489 including (?-i) didn't work because the subject string was already
1490 lowercased, and the effects were non-intuitive. It turns out that a
1491 one-line patch can be used to allow (?-i) to work as expected.
1497 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1498 other effects) broke the use of negated acl sub-conditions.
1500 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1503 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1504 "Deny" causes Exim to reject the incoming connection with a 554 error.
1505 Unfortunately, if there is a major crisis, such as a disk failure,
1506 tcp-wrappers gives "deny", whereas what one would like would be some
1507 kind of temporary error. A kludge has been added to help with this.
1508 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1509 554 error is used if errno is still zero or contains ENOENT (which occurs
1510 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1513 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1515 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1516 errors. Otherwise a message that provokes a temporary error (when other
1517 messages do not) can cause a whole host to time out.
1519 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1520 addresses were routed directly to files or pipes from a redirect router.
1521 File deliveries just didn't batch; pipe deliveries might have suffered
1524 PH/06 A failure to get a lock for a hints database would erroneously always say
1525 "Failed to get write lock", even when it was really a read lock.
1527 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1528 of 0600. This has been changed to use the value of the lockfile_mode
1529 option (which defaults to 0600).
1531 PH/08 Applied small patch from the Sieve maintainer.
1533 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1534 folder from quota calculations, a direct delivery into this folder messed
1535 up the contents of the maildirsize file. This was because the regex was
1536 used only to exclude .Trash (or whatever) when the size of the mailbox
1537 was calculated. There was no check that a delivery was happening into an
1538 excluded directory. This bug has been fixed by ignoring all quota
1539 processing for deliveries into excluded directories.
1541 PH/10 Added the maildirfolder_create_regex option to appendfile.
1547 PH/01 The code for finding all the local interface addresses on a FreeBSD
1548 system running IPv6 was broken. This may well have applied to all BSD
1549 systems, as well as to others that have similar system calls. The broken
1550 code found IPv4 interfaces correctly, but gave incorrect values for the
1551 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1552 that it would not match correctly against @[] and not recognize the IPv6
1555 PH/02 The ipliteral router was not recognizing addresses of the form user@
1556 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1558 PH/03 Added disable_ipv6.
1560 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1561 decoded form, because it is most often used to construct To: headers
1562 lines in autoreplies, and the decoded form may well be syntactically
1563 invalid. However, $reply_address has leading white space removed, and all
1564 newlines turned into spaces so that the autoreply transport does not
1567 PH/05 If group was specified without a user on a router, and no group or user
1568 was specified on a transport, the group from the router was ignored.
1570 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1571 for visible compile-time settings that can be used to change these
1572 numbers, for those that want even more. Backwards compatibility with old
1573 spool files has been maintained. However, going back to a previous Exim
1574 release will lost any variables that are in spool files.
1576 PH/07 Two small changes when running in the test harness: increase delay when
1577 passing a TCP/IP connection to a new process, in case the original
1578 process has to generate a bounce, and remove special handling of
1579 127.0.0.2 (sic), which is no longer necessary.
1581 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1582 be the same on different OS.
1584 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1587 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1588 whether --show-vars was specified or not
1590 JJ/02 exipick: Added support for new ACL variable spool format introduced
1593 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1594 syntactically invalid From: or Reply-to: line, and a filter used this to
1595 generate an autoreply, and therefore failed to obtain an address for the
1596 autoreply, Exim could try to deliver to a non-existent relative file
1597 name, causing unrelated and misleading errors. What now happens is that
1598 it logs this as a hard delivery error, but does not attempt to create a
1601 PH/11 The exinext utility has a -C option for testing purposes, but although
1602 the given file was scanned by exinext itself; it wasn't being passed on
1603 when Exim was called.
1605 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1606 an end-of-file indication when reading a command response.
1608 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1609 compiled. In many other places in Exim, IPv6 addresses are always
1610 recognized, so I have changed this. It also means that IPv4 domain
1611 literals of the form [IPV4:n.n.n.n] are now always recognized.
1613 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1614 used if the router is not running as root, for example, when verifying at
1615 ACL time, or when using -bh. The debugging output from this situation was
1616 non-existent - all you got was a failure to exec. I have made two
1619 (a) Failures to set uid/gid, the current directory, or a process leader
1620 in a subprocess such as that created by queryprogram now generate
1621 suitable debugging ouput when -d is set.
1623 (b) The queryprogram router detects when it is not running as root,
1624 outputs suitable debugging information if -d is set, and then runs
1625 the subprocess without attempting to change uid/gid.
1627 PH/15 Minor change to Makefile for building test_host (undocumented testing
1630 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1631 additional section of a DNS packet that returns MX or SRV records.
1632 Instead, it always explicitly searches for A/AAAA records. This avoids
1633 major problems that occur when a DNS server includes only records of one
1634 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1635 fixed another bug: if SRV records were looked up and the corresponding
1636 address records were *not* found in the additional section, the port
1637 values from the SRV records were lost.
1639 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1640 using the correct key (the original address) when searching the retry
1641 rules in order to find which one to use for generating the retry hint.
1643 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1644 adding the default one. Similarly, if it contains a Reply-To: header, the
1645 errors_reply_to option, if set, is not used.
1647 PH/19 When calculating a retry time, Exim used to measure the "time since
1648 failure" by looking at the "first failed" field in the retry record. Now
1649 it does not use this if it is later than than the arrival time of the
1650 message. Instead it uses the arrival time. This makes for better
1651 behaviour in cases where some deliveries succeed, thus re-setting the
1652 "first failed" field. An example is a quota failure for a huge message
1653 when small messages continue to be delivered. Without this change, the
1654 "time since failure" will always be short, possible causing more frequent
1655 delivery attempts for the huge message than are intended.
1656 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1658 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1659 $1, $2, $3) because the numerical variables can be reset during some
1660 expansion items (e.g. "match"), thereby losing the authentication data.
1662 PH/21 Make -bV show the size of off_t variables so that the test suite can
1663 decide whether to run tests for quotas > 2G.
1665 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1666 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1667 filecount value is greater than 2G or if a quota value is greater than 2G
1668 on a system where the size of off_t is not greater than 4, a panic error
1671 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1672 never match. The debug and -bh output now contains an explicit error
1673 message indicating a malformed IPv4 address or mask.
1675 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1676 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1677 PH/23 above applies.
1679 PH/25 Do not write to syslog when running in the test harness. The only
1680 occasion when this arises is a failure to open the main or panic logs
1681 (for which there is an explicit test).
1683 PH/26 Added the /no_tell option to "control=freeze".
1685 PH/27 If a host name lookup failed very early in a connection, for example, if
1686 the IP address matched host_lookup and the reverse lookup yielded a name
1687 that did not have a forward lookup, an error message of the form "no IP
1688 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1689 could be logged. Now it outputs the IP address instead of "NULL".
1691 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1692 allows the sender and the authenticated sender to be set when
1693 submitting a message from within Exim. Since child_open_exim() is
1694 documented for local_scan(), the new function should be too.
1696 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1697 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1698 results in an empty string is now treated as unset.
1700 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1702 PH/31 Added sender_verify_fail logging option.
1704 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1705 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1706 tidied the source and removed it altogether.
1708 PH/33 When a queue run was abandoned because the load average was too high, a
1709 log line was always written; now it is written only if the queue_run log
1710 selector is set. In addition, the log line for abandonment now contains
1711 information about the queue run such as the pid. This is always present
1712 in "start" and "stop" lines but was omitted from the "abandon" line.
1714 PH/34 Omit spaces between a header name and the colon in the error message that
1715 is given when verify = headers_syntax fails (if there are lots of them,
1716 the message gets confusing).
1718 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1719 names, as there are now some PTR records that contain slashes. This check
1720 is only to protect against broken name servers that fall over on strange
1721 characters, so the fact that it applies to all lookups doesn't matter.
1723 PH/36 Now that the new test suite is complete, we can remove some of the
1724 special code in Exim that was needed for the old test suite. For example,
1725 sorting DNS records because real resolvers return them in an arbitrary
1726 order. The new test suite's fake resolver always returns records in the
1729 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1730 bounces) except when queue_only is set, to avoid logging races between
1731 the different processes.
1733 PH/38 Panic-die if .include specifies a non-absolute path.
1735 PH/39 A tweak to the "H" retry rule from its user.
1737 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1738 a label. They prevented compilation on older perls.
1740 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1741 a warning to be raised on newish perls.
1743 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1744 on queue. Changes to match documented behaviour of showing count of
1745 messages matching specified criteria.
1747 PH/40 Changed the default ident timeout from 30s to 5s.
1749 PH/41 Added support for the use of login_cap features, on those BSD systems
1750 that have them, for controlling the resources used by pipe deliveries.
1752 PH/42 The content-scanning code uses fopen() to create files in which to put
1753 message data. Previously it was not paying any attention to the mode of
1754 the files. Exim runs with umask(0) because the rest of the code creates
1755 files with open(), and sets the required mode explicitly. Thus, these
1756 files were ending up world-writeable. This was not a big issue, because,
1757 being within the spool directory, they were not world-accessible. I have
1758 created a function called modefopen, which takes an additional mode
1759 argument. It sets umask(777), creates the file, chmods it to the required
1760 mode, then resets the umask. All the relevant calls to fopen() in the
1761 content scanning code have been changed to use this function.
1763 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1764 to 24 hours. This avoids potential overflow problems when processing G
1765 and H retry rules. I suspect nobody ever tinkers with this value.
1767 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1769 PH/45 When the plaintext authenticator is running as a client, the server's
1770 challenges are checked to ensure they are valid base64 strings. By
1771 default, the authentication attempt is cancelled if an invalid string is
1772 received. Setting client_ignore_invalid_base64 true ignores these errors.
1773 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1774 they are received. Thus, the responses can be made to depend on the
1775 challenges. If an invalid string is ignored, an empty string is placed in
1778 PH/46 Messages that are created by the autoreply transport now contains a
1779 References: header, in accordance with RFCs 2822 and 3834.
1781 PH/47 Added authenticated_sender_force to the smtp transport.
1783 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1785 PH/49 Installed latest patch from the Sieve maintainer.
1787 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1788 was also set, the appendfile transport was unnecessarily scanning a
1789 directory of message files (e.g. for maildir delivery) to find the count
1790 of files (along with the size), even though it did not need this
1791 information. It now does the scan only if it needs to find either the
1792 size of the count of files.
1794 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1796 PH/52 Two bugs concerned with error handling when the smtp transport is
1799 (i) Exim was not creating retry information for temporary errors given
1800 for individual recipients after the DATA command when the smtp transport
1801 was used in LMTP mode. This meant that they could be retried too
1802 frequently, and not timed out correctly.
1804 (ii) Exim was setting the flag that allows error details to be returned
1805 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1806 recipients that were returned after the DATA command.
1808 PH/53 This is related to PH/52, but is more general: for any failing address,
1809 when detailed error information was permitted to be returned to the
1810 sender, but the error was temporary, then after the final timeout, only
1811 "retry timeout exceeded" was returned. Now it returns the full error as
1812 well as "retry timeout exceeded".
1814 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1815 do this, and (what is worse) MTAs that accept it.
1817 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1818 will now be deprecated.
1820 PH/56 New os.c-cygwin from the Cygwin maintainer.
1822 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1823 formats (previously only available in exim formats via -bpr, -bpru,
1824 and -bpra. Now also available in native and exiqgrep formats)
1826 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1827 with very large, slow to parse queues
1829 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1831 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1833 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1834 responses to authentication challenges, though it was showing the
1835 challenges; (ii) I've removed the CR characters from the debug output for
1838 PH/58 Allow for the insertion of a newline as well as a space when a string
1839 is turned into more than one encoded-word during RFC 2047 encoding. The
1840 Sieve code now uses this.
1842 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1843 data_4xx, lost_connection, tls_required.
1845 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1846 message was being sent as an SMTP response.
1848 PH/61 Add -l and -k options to exicyclog.
1850 PH/62 When verifying, if an address was redirected to one new address, so that
1851 verification continued, and the new address failed or deferred after
1852 having set something in $address_data, the value of $address_data was not
1853 passed back to the ACL. This was different to the case when no
1854 redirection occurred. The value is now passed back in both cases.
1856 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1857 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1858 use login_cap.h, so on its own it isn't the distinguishing feature. The
1859 new name refers directly to the setclassresources() function.
1861 PH/65 Added configuration files for NetBSD3.
1863 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1865 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1866 is preferred over IPv4.
1868 PH/68 The bounce_return_message and bounce_return_body options were not being
1869 honoured for bounces generated during the reception of non-SMTP messages.
1870 In particular, this applied to messages rejected by the ACL. This bug has
1871 been fixed. However, if bounce_return_message is true and bounce_return_
1872 body is false, the headers that are returned for a non-SMTP message
1873 include only those that have been read before the error was detected.
1874 (In the case of an ACL rejection, they have all been read.)
1876 PH/69 The HTML version of the specification is now built in a directory called
1877 spec_html instead of spec.html, because the latter looks like a path with
1878 a MIME-type, and this confuses some software.
1880 PH/70 Catch two compiler warnings in sieve.c.
1882 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1883 function verify_get_ident() calls ip_connect() to connect a socket, but
1884 if the "connect()" function timed out, ip_connect() used to close the
1885 socket. However, verify_get_ident() also closes the socket later, and in
1886 between Exim writes to the log, which may get opened at this point. When
1887 the socket was closed in ip_connect(), the log could get the same file
1888 descriptor number as the socket. This naturally causes chaos. The fix is
1889 not to close the socket in ip_connect(); the socket should be closed by
1890 the function that creates it. There was only one place in the code where
1891 this was missing, in the iplookup router, which I don't think anybody now
1892 uses, but I've fixed it anyway.
1894 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1895 well as to direct DNS lookups. Otherwise the handling of names in host
1896 lists is inconsistent and therefore confusing.
1902 PH/01 Two changes to the default runtime configuration:
1904 (1) Move the checks for relay_from_hosts and authenticated clients from
1905 after to before the (commented out) DNS black list checks.
1907 (2) Add control=submission to the relay_from_hosts and authenticated
1908 clients checks, on the grounds that messages accepted by these
1909 statements are most likely to be submissions.
1911 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1913 (1) Generate an error if the third argument for the ${prvs expansion is
1916 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1919 (3) Reset the variables that are obtained from the first argument of
1920 ${prvscheck and used in the second argument before leaving the code,
1921 because their memory is reclaimed, so using them afterwards may do
1924 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1925 one (it's much easier than Tom thought :-).
1927 (5) Because of (4), we can now allow for the use of $prvscheck_result
1928 inside the third argument.
1930 PH/03 For some reason, the default setting of PATH when running a command from
1931 a pipe transport was just "/usr/bin". I have changed it to
1934 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1935 anything to be listed in the output from -bV.
1937 PH/05 When a filter generated an autoreply, the entire To: header line was
1938 quoted in the delivery log line, like this:
1940 => >A.N.Other <ano@some.domain> <original@ddress> ...
1942 This has been changed so that it extracts the operative address. There
1943 may be more than one such address. If so, they are comma-separated, like
1946 => >ano@some.domain,ona@other.domain <original@ddress> ...
1948 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1949 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1950 not being looked up in the rDNS to get a host name, Exim was showing the
1951 IP address twice in Received: lines, even though the IP addresses were
1952 identical. For example:
1954 Received: from [1.2.3.4] (helo=[1.2.3.4])
1956 However, if the real host name was known, it was omitting the HELO data
1957 if it matched the actual IP address. This has been tidied up so that it
1958 doesn't show the same IP address twice.
1960 PH/07 When both +timestamp and +memory debugging was on, the value given by
1961 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1962 called by the debug printing, thereby overwriting the timestamp buffer.
1963 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1965 PH/08 When the original message was included in an autoreply transport, it
1966 always said "this is a copy of the message, including all the headers",
1967 even if body_only or headers_only was set. It now gives an appropriate
1970 PH/09 Applied a patch from the Sieve maintainer which:
1972 o fixes some comments
1973 o adds the (disabled) notify extension core
1974 o adds some debug output for the result of if/elsif tests
1975 o points to the current vacation draft in the documentation
1976 and documents the missing references header update
1980 o fixes a bug in processing the envelope test (when testing
1981 multiple envelope elements, the last element determinted the
1984 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1985 Electronic Mail") by including:
1987 Auto-submitted: auto-generated
1989 in the messages that it generates (bounce messages and others, such as
1990 warnings). In the case of bounce messages for non-SMTP mesages, there was
1991 also a typo: it was using "Auto_submitted" (underscore instead of
1992 hyphen). Since every message generated by Exim is necessarily in response
1993 to another message, thes have all been changed to:
1995 Auto-Submitted: auto-replied
1997 in accordance with these statements in the RFC:
1999 The auto-replied keyword:
2001 - SHOULD be used on messages sent in direct response to another
2002 message by an automatic process,
2004 - MUST NOT be used on manually-generated messages,
2006 - MAY be used on Delivery Status Notifications (DSNs) and Message
2007 Disposition Notifications (MDNs),
2009 - MUST NOT be used on messages generated by automatic or periodic
2010 processes, except for messages which are automatic responses to
2013 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2014 to the default Received: header definition.
2016 PH/12 Added log selector acl_warn_skipped (default on).
2018 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2019 variables because (a) they are in the wrong storage pool and (b) even if
2020 they were copied, it wouldn't work properly because of the caching.
2022 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2023 checking when decoding. Apparently there are clients that generate
2024 overlong encoded strings. Why am I not surprised?
2026 PH/15 If the first argument of "${if match_address" was not empty, but did not
2027 contain an "@" character, Exim crashed. Now it writes a panic log message
2028 and treats the condition as false.
2030 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2032 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2033 extension "envelope-auth". The code is finished and in agreement with
2034 other implementations, but there is no documentation so far and in fact,
2035 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2036 not changing the active code.
2038 Print executed "if" and "elsif" statements when debugging is used. This
2039 helps a great deal to understand what a filter does.
2041 Document more things not specified clearly in RFC3028. I had all this
2042 sorted out, when out of a sudden new issues came to my mind. Oops."
2044 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2047 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2048 canonical form (as documented). However, after a host name lookup from
2049 the IP address, check_host() was doing a simple string comparison with
2050 addresses acquired from the DNS when checking that the found name did
2051 have the original IP as one of its addresses. Since any found IPv6
2052 addresses are likely to be in abbreviated form, the comparison could
2053 fail. Luckily, there already exists a function for doing the comparison
2054 by converting both addresses to binary, so now that is used instead of
2055 the text comparison.
2057 PH/20 There was another similar case to PH/19, when a complete host name was
2058 given in a host list; looking up its IP address could give an abbreviated
2059 form, whereas the current host's name might or might not be abbreviated.
2060 The same fix has been applied.
2066 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2067 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2070 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2071 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2073 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2075 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2076 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2077 Also, for "Auto-Submitted", treat anything other than "no" as
2078 non-personal, in accordance with RFC 3834. (Previously it treated
2079 anything starting "auto-" as non-personal.)
2081 TF/01 The control=submission/name=... option had a problem with syntax
2082 errors if the name included a slash character. The /name= option
2083 now slurps the rest of the string, so it can include any characters
2084 but it must come last in the list of options (after /sender_retain
2087 PH/05 Some modifications to the interface to the fake nameserver for the new
2095 TK/01 Added the "success_on_redirect" address verification option. See
2096 NewStuff for rationale and an example.
2098 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2100 PH/02 Patch to exigrep to allow it to work on syslog lines.
2102 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2103 fread() to skip over the body file's header line, because in Cygwin the
2104 header line is locked and is inaccessible.
2106 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2107 co-exist for some time) to make it clear that it is the Exim ID that is
2108 referenced, not the Message-ID: header line.
2110 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2111 string_format() function, because snprintf() does not exist on all
2114 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2115 ${stat: expansion item.
2117 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2118 protocol synchronization error", to keep the pedants happy.
2120 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2121 well as for IRIX systems, when gcc is being used. See the host.c source
2124 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2126 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2129 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2130 transport and to the smtp transport in LMTP mode.
2132 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2134 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2135 run a filter in a subprocess. This could lead to confusion in subsequent
2136 lookups in the parent process. There should also be a search_tidyup() at
2137 the end of the subprocess.
2139 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2140 only if the host matched helo_try_verify_hosts, which caused the
2141 verification to occur when the EHLO/HELO command was issued. The ACL just
2142 tested the remembered result. Now, if a previous verification attempt has
2143 not happened, "verify = helo" does it there and then.
2145 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2147 TK/03 Fix log output including CR from clamd.
2149 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2150 exist provoked a memory error which could cause a segfault.
2152 PH/15 Installed PCRE 6.2
2154 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2156 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2157 of the problem. Specifically, suggested +O2 rather than +O1 for the
2160 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2162 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2163 record, the port was not correctly logged when the outgoing_port log
2164 selector was set (it logged the transort's default port).
2166 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2167 fallback_hosts, and "hosts" in the smtp transport.
2169 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2170 host errors such as "Connection refused".
2172 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2173 authentication with radiusclient 0.4.9:
2175 - Error returned from rc_read_config was caught wrongly
2176 - Username/password not passed on to radius server due to wrong length.
2178 The presumption is that some radiusclient API changes for 4.51/PH/17
2179 were not taken care of correctly. The code is still untested by me (my
2180 Linux distribution still has 0.3.2 of radiusclient), but it was
2181 contributed by a Radius user.
2183 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2184 expanding the "port" option of the smtp transport.
2186 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2187 while reading a MIME header. Thanks to Tom Hughes for a patch.
2189 PH/24 Include config.h inside local_scan.h so that configuration settings are
2192 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2193 that in an ACL for RCPT (for example), you can examine exactly what was
2196 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2197 commands, but it was not correctly comparing the address with the actual
2198 client host address. Thus, it would show the EHLO address in Received:
2199 header lines when this was not necessary.
2201 PH/27 Added the % operator to ${eval:}.
2203 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2204 it should be ignoring failures (because with -C, for example, it has lost
2205 privilege). It wasn't ignoring creation failures other than "already
2208 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2211 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2212 because an input line was too long, either on its own, or by virtue of
2213 too many continuations, the temporary file was not being removed, and the
2214 return code was incorrect.
2216 PH/31 Missing "BOOL" in function definition in filtertest.c.
2218 PH/32 Applied Sieve patches from the maintainer.
2220 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2222 PH/33 Added "verify = not_blind".
2224 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2225 Local/Makefile (with some defaults set). These are used in built scripts
2226 such as exicyclog, but they have never been used in the exim_install
2227 script (though there are many overriding facilities there). I have
2228 arranged that the exim_install script now takes note of these two
2231 PH/35 Installed configuration files for Dragonfly.
2233 PH/36 When a locally submitted message by a trusted user did not contain a
2234 From: header, and the sender address was obtained from -f or from an SMTP
2235 MAIL command, and the trusted user did not use -F to supply a sender
2236 name, $originator_name was incorrectly used when constructing a From:
2237 header. Furthermore, $originator_name was used for submission mode
2238 messages from external hosts without From: headers in a similar way,
2239 which is clearly wrong.
2241 PH/37 Added control=suppress_local_fixups.
2243 PH/38 When log_selector = +received_sender was set, and the addition of the
2244 sender made the log line's construction buffer exactly full, or one byte
2245 less than full, an overflow happened when the terminating "\n" was
2248 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2249 when the result of a list match is failure because a DNS lookup failed.
2251 PH/40 RM_COMMAND is now used in the building process.
2253 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2254 the "build-* directories that it finds.
2256 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2257 address in a domain literal was a prefix of an interface address.
2259 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2260 when verifying a sender address, unless rewrite_headers is false.
2262 PH/44 Wrote a long comment about why errors_to addresses are verified as
2263 recipients, not senders.
2265 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2266 the ratelimit ACL was added.
2268 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2270 PH/46 Added extra information about PostgreSQL errors to the error string.
2272 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2273 avoiding the need to install special zones in a real server. This is
2274 backwards compatible; if it can't find the fake resolver, it drops back.
2275 Thus, both old and new test suites can be run.
2277 TF/02 Added util/ratelimit.pl
2279 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2280 clock is set back in time.
2282 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2283 Candler <B.Candler@pobox.com>.
2285 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2286 OK for addresses that are the result of redirections.
2288 PH/48 A number of further additions for the benefit of the new test suite,
2289 including a fake gethostbyname() that interfaces to the fake DNS resolver
2292 TF/06 The fix for widen_domains has also been applied to qualify_single and
2293 search_parents which are the other dnslookup options that can cause
2296 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2299 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2301 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2302 service" if it encountered an MX record with an empty target hostname.
2303 The message is now "an MX or SRV record indicated no SMTP service".
2305 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2306 if the DNS of the sending site is misconfigured. This is quite a
2307 common situation. This change restores the behaviour of treating a
2308 helo verification defer as a failure.
2310 PH/51 If self=fail was set on a router, the bounce message did not include the
2311 actual error message.
2317 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2319 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2320 command itself ended in error, the underlying message about the transport
2321 filter timeout was being overwritten with the pipe command error. Now the
2322 underlying error message should be appended to the second error message.
2324 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2326 PH/02 Reduce the amount of output that "make" produces by default. Full output
2327 can still be requested.
2329 PH/03 The warning log line about a condition test deferring for a "warn" verb
2330 was being output only once per connection, rather than after each
2331 occurrence (because it was using the same function as for successful
2332 "warn" verbs). This seems wrong, so I have changed it.
2334 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2335 it should not have, which might have caused a crash in the right
2336 circumstances, but probably never did.
2338 PH/04 Installed a modified version of Tony Finch's patch to make submission
2339 mode fix the return path as well as the Sender: header line, and to
2340 add a /name= option so that you can make the user's friendly name appear
2343 TF/03 Added the control = fakedefer ACL modifier.
2345 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2346 Mark Lowes for thorough testing.
2348 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2350 TK/03 Merged latest SRS patch from Miles Wilton.
2352 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2353 with the definition in sysexits.h (which is #included earlier).
2354 Fortunately, Exim does not actually use EX_OK. The code used to try to
2355 preserve the sysexits.h value, by assumimg that macro definitions were
2356 scanned for macro replacements. I have been disabused of this notion,
2357 so now the code just undefines EX_OK before #including unistd.h.
2359 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2360 in the smtp transport. When a block could not be written in a single
2361 write() function, the timeout was being re-applied to each part-write.
2362 This seems wrong - if the receiver was accepting one byte at a time it
2363 would take for ever. The timeout is now adjusted when this happens. It
2364 doesn't have to be particularly precise.
2366 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2367 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2369 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2370 without a domain if the check to <postmaster@domain> fails.
2372 SC/01 Eximstats: added -xls and the ability to specify output files
2373 (patch written by Frank Heydlauf).
2375 SC/02 Eximstats: use FileHandles for outputing results.
2377 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2379 SC/04 Eximstats: fixed display of large numbers with -nvr option
2381 SC/05 Eximstats: fixed merging of reports with empty tables.
2383 SC/06 Eximstats: added the -include_original_destination flag
2385 SC/07 Eximstats: removed tabs and trailing whitespace.
2387 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2389 TK/06 MBOX spool code: Add real "From " MBOX separator line
2390 so the .eml file is really in mbox format (even though
2391 most programs do not really care). Patch from Alex Miller.
2393 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2394 The latter is generated from $received_to and is only set if the
2395 message has one envelope recipient. SA can use these headers,
2396 obviously out-of-the-box. Patch from Alex Miller.
2398 PH/08 The ${def test on a variable was returning false if the variable's
2399 value was "0", contrary to what the specification has always said!
2400 The result should be true unless the variable is empty.
2402 PH/09 The syntax error of a character other than { following "${if
2403 def:variable_name" (after optional whitespace) was not being diagnosed.
2404 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2405 accidental colon was present, for example, could give incorrect results.
2407 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2408 result is used (not including appendfile, where other changes are about
2411 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2412 This involved changing a lot of size variables from int to off_t. It
2413 should work with maildirs and everything.
2415 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2416 spamd dying while we are connected to it.
2418 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2421 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2422 function for BDB 4.3.
2424 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2426 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2427 the sources for PCRE have been re-arranged and more files are now
2430 PH/15 The code I had for printing potentially long long variables in PH/11
2431 above was not the best (it lost precision). The length of off_t variables
2432 is now inspected at build time, and an appropriate printing format (%ld
2433 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2434 to be "long long int" or "long int". This is needed for the internal
2435 formatting function string_vformat().
2437 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2438 the configuration file to be ":syslog", then the script "guesses" where
2439 the logs files are, rather than using the compiled in default. In our
2440 case the guess is not the same as the compiled default, so the script
2441 suddenly stopped working when I started to use syslog. The patch checks
2442 to see if log_file_path is "". If so, it attempts to read it from exim
2443 with no configuration file to get the compiled in version, before it
2444 falls back to the previous guessing code."
2446 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2447 implementing BATV in an Exim configuration. See NewStuff for the gory
2450 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2451 Makefile that are specific to HP-UX.
2453 PH/18 If the "use_postmaster" option was set for a recipient callout together
2454 with the "random" option, the postmaster address was used as the MAIL
2455 FROM address for the random test, but not for the subsequent recipient
2456 test. It is now used for both.
2458 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2459 patch removes a few documentation additions to RFC 3028, because the
2460 latest draft now contains them. It adds the new en;ascii-case comparator
2461 and a new error check for 8bit text in MIME parts. Comparator and
2462 require names are now matched exactly. I enabled the subaddress
2463 extension, but it is not well tested yet (read: it works for me)."
2465 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2466 rework some of the code of TK/09 above to avoid the hardwired use of
2467 "%lld" and "long long". Replaced the call to snprintf() with a call to
2470 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2471 records point to non-existent hosts", "retry timeout exceeded", and
2472 "retry time not reached for any host after a long failure period".
2474 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2475 experimental DomainKeys support:
2477 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2478 (2) On an error such as an illegally used "control", the wrong name for
2479 the control was given.
2481 These problems did NOT occur unless DomainKeys support was compiled.
2483 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2485 PH/24 Added ${if match_ip condition.
2487 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2488 databases so that it will be absolutely obvious if a crash occurs in the
2489 DB library. This is a regular occurrence (often caused by mis-matched
2492 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2493 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2494 picked up on a user's system that detects such things. There doesn't seem
2495 to be a gcc warning option for this - only an attribute that has to be
2496 put on the function's prototype. It seems that in Fedora Core 4 they have
2497 set this on a number of new functions. No doubt there will be more in due
2500 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2501 specify a transport. However, if an address that was verified by such a
2502 router was the subject of a callout, Exim crashed because it tried to
2503 read the rcpt_include_affixes from the non-existent transport. Now it
2504 just assumes that the setting of that option is false. This bug was
2505 introduced by 4.51/PH/31.
2507 PH/28 Changed -d+all to exclude +memory, because that information is very
2508 rarely of interest, but it makes the output a lot bigger. People tend to
2509 do -d+all out of habit.
2511 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2512 code in os-type was giving problems when libc.so lives in lib64, like on
2515 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2516 aren't the modern standard, and it seems that some systems' include files
2517 don't always have them. Exim was already checking for some of the newer
2518 ones like T_AAAA, and defining it itself. I've added checks for all the
2519 record types that Exim uses.
2521 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2522 not automatically generating a new one, as it is supposed to. This
2523 prevented TLS from working. If the file did exist, but contained invalid
2524 data, a new version was generated, as expected. It was only the case of a
2525 non-existent file that was broken.
2527 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2528 with a change in libdomainkeys > 0.64.
2530 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2531 from DNS. If the selector record carries the flag, it now has
2532 precedence over the domain-wide flag.
2534 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2536 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2537 the use of a port name that isn't defined in /etc/services) occurred, the
2538 message was deferred as in a normal delivery, and thus remained on the
2539 spool, instead of being failed because of the mua_wrapper setting. This
2540 is now fixed, and I tidied up some of the mua_wrapper messages at the
2543 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2544 the messages in a hash of arrays rather than using individual hashes.
2545 This is a bit cleaner and results in dramatic memory savings, albeit
2546 at a slight CPU cost.
2548 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2549 as requested by Marc Sherman.
2551 SC/10 Eximstats: added histograms for user specified patterns as requested
2554 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2556 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2557 fopen() in the content-scanning modules that did not already have it.
2563 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2564 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2566 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2568 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2570 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2571 to test Sieve filters that use "vacation".
2573 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2574 that changes the way the GnuTLS parameters are stored in the cache file.
2575 The new format can be generated externally. For backward compatibility,
2576 if the data in the cache doesn't make sense, Exim assumes it has read an
2577 old-format file, and it generates new data and writes a new file. This
2578 means that you can't go back to an older release without removing the
2581 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2582 work if there are any delivery delays because "one_time" forces the
2583 parent to be marked "delivered", so its unseen clone is never tried
2584 again. For this reason, Exim now forbids the simultaneous setting of
2587 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2588 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2589 entry if you want to know the details. The fix, however, neglected to
2590 consider the case when local delivery batching is involved. The test for
2591 "previously delivered" was not happening when checking to see if an
2592 address could be batched with a previous (undelivered) one; under
2593 certain circumstances this could lead to multiple deliveries to the same
2596 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2597 in its include files, and this causes problems building Exim.
2599 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2600 header_syntax) but Exim was just ignoring anything given after a slash.
2601 In particular, this caused confusion with an attempt to use "verify =
2602 reverse_host_lookup/defer_ok". An error is now given when options are
2603 supplied for verify items that do not have them. (Maybe reverse_host_
2604 lookup should have a defer_ok option, but that's a different point.)
2606 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2607 defined by RFC 821) to 2048, because there were problems with some AUTH
2608 commands, and RFC 1869 says the size should be increased for extended
2609 SMTP commands that take arguments.
2611 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2614 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2615 "unknown" error; now it says that the functionality isn't in the binary.
2617 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2618 an address' error message when a string expansion fails (syntax or
2619 whatever). Otherwise the password may appear in the log. Following change
2620 PH/42 below, there is no longer a chance of it appearing in a bounce
2623 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2625 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2626 of its addresses was ever tried. (Bugzilla bug #2.)
2628 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2629 the result incorrectly in the debug output. (It correctly added a newline
2630 to what was transported.)
2632 TF/01 Added $received_time.
2634 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2635 commented out examples of how to interface to a virus scanner and to
2636 SpamAssassin. Also added commented examples of av_scanner and
2637 spamd_address settings.
2639 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2640 and controls are allowed in which ACLs. There were a couple of minor
2641 errors. Some of the entries in the conditions table (which is a table of
2642 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2643 negation of where the condition IS allowed.
2645 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2647 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2648 header file does not have a version number, so I've had to invent a new
2649 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2650 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2651 radiusclient), but it was contributed by a Radius user.
2653 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2654 files or directories, for OpenSSL.
2656 PH/19 When an Exim process that is running as root has to create an Exim log
2657 file, it does so in a subprocess that runs as exim:exim so as to get the
2658 ownership right at creation (otherwise, other Exim processes might see
2659 the file with the wrong ownership). There was no test for failure of this
2660 fork() call, which would lead to the process getting stuck as it waited
2661 for a non-existent subprocess. Forks do occasionally fail when resources
2662 run out. I reviewed all the other calls to fork(); they all seem to check
2665 PH/20 When checking for unexpected SMTP input at connect time (before writing
2666 the banner), Exim was not dealing correctly with a non-positive return
2667 from the read() function. If the client had disconnected by this time,
2668 the result was a log entry for a synchronization error with an empty
2669 string after "input=" when read() returned zero. If read() returned -1
2670 (an event I could not check), uninitialized data bytes were printed.
2671 There were reports of junk text (parts of files, etc) appearing after
2674 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2676 PH/22 Added support for macro redefinition, and (re)definition in between
2677 driver and ACL definitions.
2679 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2680 forgetting to use the resulting value; it was using the unexpanded value.
2682 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2683 hadn't been configured. The fix is from Juergen Kreileder, who
2684 understands it better than I do:
2686 "Here's what I see happening with three configured cyrus_sasl
2687 authenticators configured (plain, login, cram-md5):
2689 On startup auth_cyrus_sasl_init() gets called for each of these.
2690 This means three calls to sasl_listmech() without a specified mech_list.
2691 => SASL tests which mechs of all available mechs actually work
2692 => three warnings about OTP not working
2693 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2695 With the patch, sasl_listmech() also gets called three times. But now
2696 SASL's mech_list option is set to the server_mech specified in the the
2697 authenticator. Or in other words, the answer from sasl_listmech()
2698 gets limited to just the mech you're testing for (which is different
2700 => the return list contains just 'plain' or 'login', 'cram-md5' or
2701 nothing depending on the value of ob->server_mech.
2703 I've just tested the patch: Authentication still works fine,
2704 unavailable mechs specified in the exim configuration are still
2705 caught, and the auth.log warnings about OTP are gone."
2707 PH/25 When debugging is enabled, the contents of the command line are added
2708 to the debugging output, even when log_selector=+arguments is not
2711 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2712 answer is "GNU", and only if the return is "GNU/something" is the answer
2715 PH/27 $acl_verify_message is now set immediately after the failure of a
2716 verification in an ACL, and so is available in subsequent modifiers. In
2717 particular, the message can be preserved by coding like this:
2719 warn !verify = sender
2720 set acl_m0 = $acl_verify_message
2722 Previously, $acl_verify_message was set only while expanding "message"
2723 and "log_message" when a very denied access.
2725 PH/28 Modified OS/os.c-Linux with
2727 -#ifndef OS_LOAD_AVERAGE
2728 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2730 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2731 nomenclature these days.)
2733 PH/29 Installed patch from the Sieve maintainer that adds the options
2734 sieve_useraddress and sieve_subaddress to the redirect router.
2736 PH/30 In these circumstances:
2737 . Two addresses routed to the same list of hosts;
2738 . First host does not offer TLS;
2739 . First host accepts first address;
2740 . First host gives temporary error to second address;
2741 . Second host offers TLS and a TLS session is established;
2742 . Second host accepts second address.
2743 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2744 and peerdn, if requested) that were in fact used only for the second
2747 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2748 attention to any local part prefix or suffix that was matched by the
2749 router that accepted the address. It now behaves in the same way as it
2750 does for delivery: the affixes are removed from the local part unless
2751 rcpt_include_affixes is set on the transport.
2753 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2754 timeout during the DATA phase of an incoming message.
2756 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2757 applied a patch sanctioned by the Sieve maintainer.
2759 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2760 the uid or gid is negative. A case of a negative gid caused this to be
2761 noticed. The fix allows for either to be negative.
2763 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2764 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2767 PH/36 The change PH/12 above was broken. Fixed it.
2769 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2770 the grounds that routing the same address twice would always produce the
2771 same answer. This might have been true once, but it is certainly no
2772 longer true now. Routing a child address may depend on the previous
2773 routing that produced that child. Some complicated redirection strategies
2774 went wrong when messages had multiple recipients, and made Exim's
2775 behaviour dependent on the order in which the addresses were given.
2777 I have moved the duplicate checking until after the routing is complete.
2778 Exim scans the addresses that are assigned to local and remote
2779 transports, and removes any duplicates. This means that more work will be
2780 done, as duplicates will always all be routed, but duplicates are
2781 presumably rare, so I don't expect this is of any significance.
2783 For deliveries to pipes, files, and autoreplies, the duplicate checking
2784 still happens during the routing process, since they are not going to be
2787 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2788 It corrects a timeout issue with spamd. This is Ian's comment: "The
2789 background is that sometimes spamd either never reads data from a
2790 connection it has accepted, or it never writes response data. The exiscan
2791 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2792 blindly assumes that writes won't block so it may never time out."
2794 PH/39 Allow G after quota size as well as K and M.
2796 PH/40 The value set for $authenticated_id in an authenticator may not contain
2797 binary zeroes or newlines because the value is written to log lines and
2798 to spool files. There was no check on this. Now the value is run through
2799 the string_printing() function so that such characters are converted to
2800 printable escape sequences.
2802 PH/41 $message_linecount is a new variable that contains the total number of
2803 lines in the message. Compare $body_linecount, which is the count for the
2806 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2807 bounce and delay warning messages, except in certain special cases, which
2810 (a) An SMTP error message from a remote host;
2811 (b) A message specified in a :fail: redirection;
2812 (c) A message specified in a "fail" command in a system filter;
2813 (d) A message specified in a FAIL return from the queryprogram router;
2814 (e) A message specified by the cannot_route_message router option.
2816 In these cases only, Exim does include the error details in bounce and
2817 warning messages. There are also a few cases where bland messages such
2818 as "unrouteable address" or "local delivery error" are given.
2820 PH/43 $value is now also set for the "else" part of a ${run expansion.
2822 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2823 being worked on, but at least Exim now implements the latest version to
2826 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2827 process to complete was treated as a delivery failure, a timeout while
2828 writing the message to the pipe was logged, but erroneously treated as a
2829 successful delivery. Such timeouts include transport filter timeouts. For
2830 consistency with the overall process timeout, these timeouts are now
2831 treated as errors, giving rise to delivery failures by default. However,
2832 there is now a new Boolean option for the pipe transport called
2833 timeout_defer, which, if set TRUE, converts the failures into defers for
2834 both kinds of timeout. A transport filter timeout is now identified in
2837 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2838 systems where "make" and "gmake" are different, calling "gmake" at top
2839 level broke things. I've arranged for the value of $(MAKE) to be passed
2840 from the Makefile to this script so that it can call the same version of
2844 A note about Exim versions 4.44 and 4.50
2845 ----------------------------------------
2847 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2848 changes of various kinds. As a consequence, a big documentation update was
2849 needed. This delayed the release for rather longer than seemed good, especially
2850 in the light of a couple of (minor) security issues. Therefore, the changes
2851 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2852 release. So 4.44 and 4.50 are in effect two different branches that both start
2855 I have left the 4.50 change log unchanged; it contains all the changes since
2856 4.43. The change log for 4.44 is below; many of its items are identical to
2857 those for 4.50. This seems to be the most sensible way to preserve the
2858 historical information.
2864 1. Minor wording change to the doc/README.SIEVE file.
2866 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2867 computation of the current number of files was incorrect.
2869 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2870 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2873 4. Give more explanation in the error message when the command for a transport
2874 filter fails to execute.
2876 5. There are several places where Exim runs a non-Exim command in a
2877 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2878 was being done only for the command run by the queryprogram router. It is
2879 now done for all such subprocesses. The other cases are: ${run, transport
2880 filters, and the commands run by the lmtp and pipe transports.
2882 6. Added CONFIGURE_GROUP build-time option.
2884 7. Some older OS have a limit of 256 on the maximum number of file
2885 descriptors. Exim was using setrlimit() to set 1000 as a large value
2886 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2887 systems. I've change it so that if it can't get 1000, it tries for 256.
2889 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2890 was an oversight, and furthermore, ever since the addition of extra
2891 controls (e.g. 4.43/32), the checks on when to allow different forms of
2892 "control" were broken. There should now be diagnostics for all cases when a
2893 control that does not make sense is encountered.
2895 9. Added the /retain_sender option to "control=submission".
2897 10. $recipients is now available in the predata ACL (oversight).
2899 11. Tidy the search cache before the fork to do a delivery from a message
2900 received from the command line. Otherwise the child will trigger a lookup
2901 failure and thereby defer the delivery if it tries to use (for example) a
2902 cached ldap connection that the parent has called unbind on.
2904 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2905 of $address_data from the recipient verification was clobbered by the
2906 sender verification.
2908 13. The value of address_data from a sender verification is now available in
2909 $sender_address_data in subsequent conditions in the ACL statement.
2911 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2913 15. Added a new option "connect=<time>" to callout options, to set a different
2916 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2917 was its contents. (It was OK if the option was not defined at all.)
2919 17. A "Completed" log line is now written for messages that are removed from
2920 the spool by the -Mrm option.
2922 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2923 information about exactly what failed.
2925 19. Added -dd to debug only the daemon process.
2927 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2928 handles timeouts, both on the server side and network timeouts. Renamed the
2929 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2931 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2932 It is now set to "smtps".
2934 22. $host_address is now set to the target address during the checking of
2935 ignore_target_hosts.
2937 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2938 being passed; this would have caused $sender_host_name to have been used if
2939 matching the list had actually called for a host name (not very likely,
2940 since this list is usually IP addresses). A host name is now passed as
2943 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2944 code that specifies a non-restarting handler (typically sigaction() in
2945 modern systems) in an attempt to fix a rare and obscure crash bug.
2947 25. Narrowed the window for a race in the daemon that could cause it to ignore
2948 SIGCHLD signals. This is not a major problem, because they are used only to
2949 wake it up if nothing else does.
2951 26. A malformed maildirsize file could cause Exim to calculate negative values
2952 for the mailbox size or file count. Odd effects could occur as a result.
2953 The maildirsize information is now recalculated if the size or filecount
2956 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2957 support for a long time. Removed HAVE_SYS_VFS_H.
2959 28. Installed the latest version of exipick from John Jetmore.
2961 29. In an address list, if the pattern was not a regular expression, an empty
2962 subject address (from a bounce message) matched only if the pattern was an
2963 empty string. Non-empty patterns were not even tested. This was the wrong
2964 because it is perfectly reasonable to use an empty address as part of a
2965 database query. An empty address is now tested by patterns that are
2966 lookups. However, all the other forms of pattern expect the subject to
2967 contain a local part and a domain, and therefore, for them, an empty
2968 address still always fails if the pattern is not itself empty.
2970 30. Exim went into a mad DNS loop when attempting to do a callout where the
2971 host was specified on an smtp transport, and looking it up yielded more
2972 than one IP address.
2974 31. Re-factored the code for checking spool and log partition space into a
2975 function that finds that data and another that does the check. The former
2976 is then used to implement four new variables: $spool_space, $log_space,
2977 $spool_inodes, and $log_inodes.
2979 32. The RFC2047 encoding function was originally intended for short strings
2980 such as real names; it was not keeping to the 75-character limit for
2981 encoded words that the RFC imposes. It now respects the limit, and
2982 generates multiple encoded words if necessary. To be on the safe side, I
2983 have increased the buffer size for the ${rfc2047: expansion operator from
2986 33. It is now permitted to omit both strings after an "if" condition; if the
2987 condition is true, the result is "true". As before, when the second string
2988 is omitted, a false condition yields an empty string. This makes it less
2989 cumbersome to write custom ACL and router conditions.
2991 34. Failure to deliver a bounce message always caused it to be frozen, even if
2992 there was an errors_to setting on the router. The errors_to setting is now
2995 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2996 canonical form (fully expanded) before being placed in
2997 $sender_host_address.
2999 36. The table in the code that translates DNS record types into text (T_A to
3000 "A" for instance) was missing entries for NS and CNAME. It is just possible
3001 that this could have caused confusion if both these types were looked up
3002 for the same domain, because the text type is used as part of Exim's
3003 per-process caching. But the chance of anyone hitting this buglet seems
3006 37. The dnsdb lookup has been extended in a number of ways.
3008 (1) There is a new type, "zns", which walks up the domain tree until it
3009 finds some nameserver records. It should be used with care.
3011 (2) There is a new type, "mxh", which is like "mx" except that it returns
3012 just the host names, not the priorities.
3014 (3) It is now possible to give a list of domains (or IP addresses) to be
3015 looked up. The behaviour when one of the lookups defers can be
3016 controlled by a keyword.
3018 (4) It is now possible to specify the separator character for use when
3019 multiple records are returned.
3021 38. The dnslists ACL condition has been extended: it is now possible to supply
3022 a list of IP addresses and/or domains to be looked up in a particular DNS
3025 39. Added log_selector=+queue_time_overall.
3027 40. When running the queue in the test harness, wait just a tad after forking a
3028 delivery process, to get repeatability of debugging output.
3030 41. Include certificate and key file names in error message when GnuTLS fails
3031 to set them up, because the GnuTLS error message doesn't include the name
3032 of the failing file when there is a problem reading it.
3034 42. Allow both -bf and -bF in the same test run.
3036 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3038 44. The "Exiscan patch" is now merged into the mainline Exim source.
3040 45. Sometimes the final signoff response after QUIT could fail to get
3041 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3042 < 0 before doing a fflush(). This bug looks as though it goes back to the
3043 introduction of TLS in release 3.20, but "sometimes" must have been rare
3044 because the tests only now provoked it.
3046 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3047 (this can affect the format of dates).
3049 47. exim_tidydb, when checking for the continued existence of a message for
3050 which it has found a message-specific retry record, was not finding
3051 messages that were in split spool directories. Consequently, it was
3052 deleting retry records that should have stayed in existence.
3054 48. Steve fixed some bugs in eximstats.
3056 49. The SPA authentication driver was not abandoning authentication and moving
3057 on to the next authenticator when an expansion was forced to fail,
3058 contradicting the general specification for all authenticators. Instead it
3059 was generating a temporary error. It now behaves as specified.
3061 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3062 (the order specifies the preference for clients). The order is now AES256,
3063 AES128, 3DES, ARCFOUR128.
3065 51. Small patch to Sieve code - explicitly set From: when generating an
3068 52. Exim crashed if a remote delivery caused a very long error message to be
3069 recorded - for instance if somebody sent an entire SpamAssassin report back
3070 as a large number of 550 error lines. This bug was coincidentally fixed by
3071 increasing the size of one of Exim's internal buffers (big_buffer) that
3072 happened as part of the Exiscan merge. However, to be on the safe side, I
3073 have made the code more robust (and fixed the comments that describe what
3076 53. Now that there can be additional text after "Completed" in log lines (if
3077 the queue_time_overall log selector is set), a one-byte patch to exigrep
3078 was needed to allow it to recognize "Completed" as not the last thing in
3081 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3082 patch that reportedly fixes this has been added. I am not expert enough to
3083 create a test for it. This is what the patch creator wrote:
3085 "I found a little strange behaviour of ldap code when working with
3086 Windows 2003 AD Domain, where users was placed in more than one
3087 Organization Units. When I tried to give exim partial DN, the exit code
3088 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3089 But simultaneously result of request was absolutely normal ldap result,
3090 so I produce this patch..."
3092 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3093 REFERENCE, so I have modified the code to exclude the patch when that macro
3096 55. Some experimental protocols are using DNS PTR records for new purposes. The
3097 keys for these records are domain names, not reversed IP addresses. The
3098 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3099 leaves it alone. Component reversal etc. now happens only for IP addresses.
3102 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3104 57. Double the size of the debug message buffer (to 2048) so that more of very
3105 long debug lines gets shown.
3107 58. The exicyclog utility now does better if the number of log files to keep
3108 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3110 59. Two changes related to the smtp_active_hostname option:
3112 (1) $smtp_active_hostname is now available as a variable.
3113 (2) The default for smtp_banner uses $smtp_active_hostname instead
3114 of $primary_hostname.
3116 60. The host_aton() function is supposed to be passed a string that is known
3117 to be a valid IP address. However, in the case of IPv6 addresses, it was
3118 not checking this. This is a hostage to fortune. Exim now panics and dies
3119 if the condition is not met. A case was found where this could be provoked
3120 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3121 components; fortuitously, this particular loophole had already been fixed
3122 by change 4.50/55 above.
3124 If there are any other similar loopholes, the new check in host_aton()
3125 itself should stop them being exploited. The report I received stated that
3126 data on the command line could provoke the exploit when Exim was running as
3127 exim, but did not say which command line option was involved. All I could
3128 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3129 running as the user.
3132 61. There was a buffer overflow vulnerability in the SPA authentication code
3133 (which came originally from the Samba project). I have added a test to the
3134 spa_base64_to_bits() function which I hope fixes it.
3137 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3138 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3140 63. The daemon start-up calls getloadavg() while still root for those OS that
3141 need the first call to be done as root, but it missed one case: when
3142 deliver_queue_load_max is set with deliver_drop_privilege. This is
3143 necessary for the benefit of the queue runner, because there is no re-exec
3144 when deliver_drop_privilege is set.
3146 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3147 This has been fixed.
3149 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3150 was in use, was not putting the data itself into the right store pool;
3151 consequently, it could be overwritten for a subsequent message in the same
3152 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3155 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3157 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3158 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3159 the function were treating the return as a boolean value, which happened to
3160 work because 0=false and not-0=true, but is not correct code.
3162 68. The host_aton() function was not handling scoped IPv6 addresses (those
3163 with, for example, "%eth0" on the end) correctly.
3165 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3166 negated items (that is, ~something) in unsigned ints. Some compilers
3167 apparently mutter when there is no cast.
3169 70. If an address verification called from an ACL failed, and did not produce a
3170 user-specific message (i.e. there was only a "system" message), nothing was
3171 put in $acl_verify_message. In this situation, it now puts the system
3174 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3175 session; change 4.31/43 added the unwanted input to the log line - except
3176 that it did not do this in the start of session case. It now does.
3178 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3179 This is wrong and can cause the other end to generate a synchronization
3180 error if it is another Exim or anything else that does the synchronization
3181 check. A QUIT command is no longer sent after a timeout.
3183 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3184 during host lookups.
3186 74. The defer_ok option of callout verification was not working if it was used
3187 when verifying addresses in header lines, that is, for this case:
3189 verify = header_sender/callout=defer_ok
3191 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3192 those file descriptors could be used for SMTP connections. If anything
3193 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3194 could be sent to the SMTP client, causing chaos. The daemon now opens
3195 stdin, stdout, and stderr to /dev/null when it puts itself into the
3198 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3199 log by default. The user can override this with suitable Perl magic.
3201 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3202 the log message when discard triggers, was not working for the DATA ACL or
3203 for the non-SMTP ACL.
3205 78. Error message wording change in sieve.c.
3207 79. If smtp_accept_max_per_host was set, the number of connections could be
3208 restricted to fewer than expected, because the daemon was trying to set up
3209 a new connection before checking whether the processes handling previous
3210 connections had finished. The check for completed processes is now done
3211 earlier. On busy systems, this bug wouldn't be noticed because something
3212 else would have woken the daemon, and it would have reaped the completed
3215 80. If a message was submitted locally by a user whose login name contained one
3216 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3217 It caused a spool format error. I have fixed the spool reading code. A
3218 related problem was that the "from" clause in the Received: line became
3219 illegal because of the space(s). It is now covered by ${quote_local_part.
3221 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3224 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3226 83. Patch from Sieve maintainer to fix unterminated string problem in
3227 "vacation" handling.
3229 84. Some minor changes to the Linux configuration files to help with other
3230 OS variants using glibc.
3232 85. One more patch for Sieve to update vacation handling to latest spec.
3235 ----------------------------------------------------
3236 See the note above about the 4.44 and 4.50 releases.
3237 ----------------------------------------------------
3243 1. Change 4.43/35 introduced a bug that caused file counts to be
3244 incorrectly computed when quota_filecount was set in an appendfile
3247 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3248 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3251 3. Give more explanation in the error message when the command for a transport
3252 filter fails to execute.
3254 4. There are several places where Exim runs a non-Exim command in a
3255 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3256 was being done only for the command run by the queryprogram router. It is
3257 now done for all such subprocesses. The other cases are: ${run, transport
3258 filters, and the commands run by the lmtp and pipe transports.
3260 5. Some older OS have a limit of 256 on the maximum number of file
3261 descriptors. Exim was using setrlimit() to set 1000 as a large value
3262 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3263 systems. I've change it so that if it can't get 1000, it tries for 256.
3265 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3266 was an oversight, and furthermore, ever since the addition of extra
3267 controls (e.g. 4.43/32), the checks on when to allow different forms of
3268 "control" were broken. There should now be diagnostics for all cases when a
3269 control that does not make sense is encountered.
3271 7. $recipients is now available in the predata ACL (oversight).
3273 8. Tidy the search cache before the fork to do a delivery from a message
3274 received from the command line. Otherwise the child will trigger a lookup
3275 failure and thereby defer the delivery if it tries to use (for example) a
3276 cached ldap connection that the parent has called unbind on.
3278 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3279 of $address_data from the recipient verification was clobbered by the
3280 sender verification.
3282 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3283 was its contents. (It was OK if the option was not defined at all.)
3285 11. A "Completed" log line is now written for messages that are removed from
3286 the spool by the -Mrm option.
3288 12. $host_address is now set to the target address during the checking of
3289 ignore_target_hosts.
3291 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3292 being passed; this would have caused $sender_host_name to have been used if
3293 matching the list had actually called for a host name (not very likely,
3294 since this list is usually IP addresses). A host name is now passed as
3297 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3298 code that specifies a non-restarting handler (typically sigaction() in
3299 modern systems) in an attempt to fix a rare and obscure crash bug.
3301 15. Narrowed the window for a race in the daemon that could cause it to ignore
3302 SIGCHLD signals. This is not a major problem, because they are used only to
3303 wake it up if nothing else does.
3305 16. A malformed maildirsize file could cause Exim to calculate negative values
3306 for the mailbox size or file count. Odd effects could occur as a result.
3307 The maildirsize information is now recalculated if the size or filecount
3310 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3311 support for a long time. Removed HAVE_SYS_VFS_H.
3313 18. Updated exipick to current release from John Jetmore.
3315 19. Allow an empty sender to be matched against a lookup in an address list.
3316 Previously the only cases considered were a regular expression, or an
3319 20. Exim went into a mad DNS lookup loop when doing a callout where the
3320 host was specified on the transport, if the DNS lookup yielded more than
3323 21. The RFC2047 encoding function was originally intended for short strings
3324 such as real names; it was not keeping to the 75-character limit for
3325 encoded words that the RFC imposes. It now respects the limit, and
3326 generates multiple encoded words if necessary. To be on the safe side, I
3327 have increased the buffer size for the ${rfc2047: expansion operator from
3330 22. Failure to deliver a bounce message always caused it to be frozen, even if
3331 there was an errors_to setting on the router. The errors_to setting is now
3334 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3335 canonical form (fully expanded) before being placed in
3336 $sender_host_address.
3338 24. Updated eximstats to version 1.33
3340 25. Include certificate and key file names in error message when GnuTLS fails
3341 to set them up, because the GnuTLS error message doesn't include the name
3342 of the failing file when there is a problem reading it.
3344 26. Expand error message when OpenSSL has problems setting up cert/key files.
3347 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3348 (this can affect the format of dates).
3350 28. exim_tidydb, when checking for the continued existence of a message for
3351 which it has found a message-specific retry record, was not finding
3352 messages that were in split spool directories. Consequently, it was
3353 deleting retry records that should have stayed in existence.
3355 29. eximstats updated to version 1.35
3356 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3357 1.35 - bugfix such that pie charts by volume are generated correctly
3359 30. The SPA authentication driver was not abandoning authentication and moving
3360 on to the next authenticator when an expansion was forced to fail,
3361 contradicting the general specification for all authenticators. Instead it
3362 was generating a temporary error. It now behaves as specified.
3364 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3365 (the order specifies the preference for clients). The order is now AES256,
3366 AES128, 3DES, ARCFOUR128.
3368 31. Small patch to Sieve code - explicitly set From: when generating an
3371 32. Exim crashed if a remote delivery caused a very long error message to be
3372 recorded - for instance if somebody sent an entire SpamAssassin report back
3373 as a large number of 550 error lines. This bug was coincidentally fixed by
3374 increasing the size of one of Exim's internal buffers (big_buffer) that
3375 happened as part of the Exiscan merge. However, to be on the safe side, I
3376 have made the code more robust (and fixed the comments that describe what
3379 33. Some experimental protocols are using DNS PTR records for new purposes. The
3380 keys for these records are domain names, not reversed IP addresses. The
3381 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3382 leaves it alone. Component reversal etc. now happens only for IP addresses.
3385 34. The host_aton() function is supposed to be passed a string that is known
3386 to be a valid IP address. However, in the case of IPv6 addresses, it was
3387 not checking this. This is a hostage to fortune. Exim now panics and dies
3388 if the condition is not met. A case was found where this could be provoked
3389 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3390 components; fortuitously, this particular loophole had already been fixed
3391 by change 4.50/55 or 4.44/33 above.
3393 If there are any other similar loopholes, the new check in host_aton()
3394 itself should stop them being exploited. The report I received stated that
3395 data on the command line could provoke the exploit when Exim was running as
3396 exim, but did not say which command line option was involved. All I could
3397 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3398 running as the user.
3401 35. There was a buffer overflow vulnerability in the SPA authentication code
3402 (which came originally from the Samba project). I have added a test to the
3403 spa_base64_to_bits() function which I hope fixes it.
3406 36. The daemon start-up calls getloadavg() while still root for those OS that
3407 need the first call to be done as root, but it missed one case: when
3408 deliver_queue_load_max is set with deliver_drop_privilege. This is
3409 necessary for the benefit of the queue runner, because there is no re-exec
3410 when deliver_drop_privilege is set.
3412 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3413 was in use, was not putting the data itself into the right store pool;
3414 consequently, it could be overwritten for a subsequent message in the same
3415 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3418 38. Sometimes the final signoff response after QUIT could fail to get
3419 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3420 < 0 before doing a fflush(). This bug looks as though it goes back to the
3421 introduction of TLS in release 3.20, but "sometimes" must have been rare
3422 because the tests only now provoked it.
3428 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3429 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3430 Now it returns an integer. A number of calls were still expecting a T/F
3431 return. Fortuitously, in all cases, the tests worked in OK situations,
3432 which is the norm. However, things would have gone wrong on any write
3433 failures on the smtp file descriptor. This function is used when sending
3434 messages over SMTP and also when doing verify callouts.
3436 2. When Exim is called to do synchronous delivery of a locally submitted
3437 message (the -odf or -odi options), it no longer closes stderr before doing
3440 3. Implemented the mua_wrapper option.
3442 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3444 5. Implemented the functions header_remove(), header_testname(),
3445 header_add_at_position(), and receive_remove_recipient(), and exported them
3448 6. If an ACL "warn" statement specified the addition of headers, Exim already
3449 inserted X-ACL-Warn: at the start if there was no header name. However, it
3450 was not making this test for the second and subsequent header lines if
3451 there were newlines in the string. This meant that an invalid header could
3452 be inserted if Exim was badly configured.
3454 7. Allow an ACL "warn" statement to add header lines at the start or after all
3455 the Received: headers, as well as at the end.
3457 8. Added the rcpt_4xx retry error code.
3459 9. Added postmaster_mailfrom=xxx to callout verification option.
3461 10. Added mailfrom=xxxx to the callout verification option, for verify=
3464 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3465 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3467 12. Inserted some casts to stop certain compilers warning when using pointer
3468 differences as field lengths or precisions in printf-type calls (mostly
3469 affecting debugging statements).
3471 13. Added optional readline() support for -be (dynamically loaded).
3473 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3474 same clock tick as a message's arrival, so that its received time was the
3475 same as the "first fail" time on the retry record, and that message
3476 remained on the queue past the ultimate address timeout, every queue runner
3477 would try a delivery (because it was past the ultimate address timeout) but
3478 after another failure, the ultimate address timeout, which should have then
3479 bounced the address, did not kick in. This was a "< instead of <=" error;
3480 in most cases the first failure would have been in the next clock tick
3481 after the received time, and all would be well.
3483 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3484 being recognized when the domain list was tested by the match_domain
3485 condition in an expansion string.
3487 16. Added the ${str2b64: operator.
3489 17. Exim was always calling setrlimit() to set a large limit for the number of
3490 processes, without checking whether the existing limit was already
3491 adequate. (It did check for the limit on file descriptors.) Furthermore,
3492 errors from getrlimit() and setrlimit() were being ignored. Now they are
3493 logged to the main and panic logs, but Exim does carry on, to try to do its
3494 job under whatever limits there are.
3496 18. Imported PCRE 5.0.
3498 19. Trivial typo in log message " temporarily refused connection" (the leading
3501 20. If the log selector return_path_on_delivery was set and an address was
3502 redirected to /dev/null, the delivery process crashed because it assumed
3503 that a return path would always be set for a "successful" delivery. In this
3504 case, the whole delivery is bypassed as an optimization, and therefore no
3507 21. Internal re-arrangement: the function for sending a challenge and reading
3508 a response while authentication was assuming a zero-terminated challenge
3509 string. It's now changed to take a pointer and a length, to allow for
3510 binary data in such strings.
3512 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3514 23. Exim was not respecting finduser_retries when seeking the login of the
3515 uid under which it was called; it was always trying 10 times. (The default
3516 setting of finduser_retries is zero.) Also, it was sleeping after the final
3517 failure, which is pointless.
3519 24. Implemented tls_on_connect_ports.
3521 25. Implemented acl_smtp_predata.
3523 26. If the domain in control=submission is set empty, Exim assumes that the
3524 authenticated id is a complete email address when it generates From: or
3525 Sender: header lines.
3527 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3528 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3529 chown and chgrp in /bin and hostname in /usr/bin.
3531 28. Exim was keeping the "process log" file open after each use, just as it
3532 does for the main log. This opens the possibility of it remaining open for
3533 long periods when the USR1 signal hits a daemon. Occasional processlog
3534 errors were reported, that could have been caused by this. Anyway, it seems
3535 much more sensible not to leave this file open at all, so that is what now
3538 29. The long-running daemon process does not normally write to the log once it
3539 has entered its main loop, and it closes the log before doing so. This is
3540 so that log files can straightforwardly be renamed and moved. However,
3541 there are a couple of unusual error situations where the daemon does write
3542 log entries, and I had neglected to close the log afterwards.
3544 30. The text of an SMTP error response that was received during a remote
3545 delivery was being truncated at 512 bytes. This is too short for some of
3546 the long messages that one sometimes sees. I've increased the limit to
3549 31. It is now possible to make retry rules that apply only when a message has a
3550 specific sender, in particular, an empty sender.
3552 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3553 it possible to be selective about when SMTP synchronization is enforced.
3555 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3557 32. Implemented hosts_connection_nolog.
3559 33. Added an ACL for QUIT.
3561 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3564 35. Added mailbox_size and mailbox_filecount to appendfile.
3566 36. Added control = no_multiline_responses to ACLs.
3568 37. There was a bug in the logic of the code that waits for the clock to tick
3569 in the case where the clock went backwards by a substantial amount such
3570 that the microsecond fraction of "now" was more than the microsecond
3571 fraction of "then" (but the whole seconds number was less).
3573 38. Added support for the libradius Radius client library this is found on
3574 FreeBSD (previously only the radiusclient library was supported).
3580 1. When certain lookups returned multiple values in the form name=value, the
3581 quoting of the values was not always being done properly. Specifically:
3582 (a) If the value started with a double quote, but contained no whitespace,
3584 (b) If the value contained whitespace other than a space character (i.e.
3585 tabs or newlines or carriage returns) it was not quoted.
3586 This fix has been applied to the mysql and pgsql lookups by writing a
3587 separate quoting function and calling it from the lookup code. The fix
3588 should probably also be applied to nisplus, ibase and oracle lookups, but
3589 since I cannot test any of those, I have not disturbed their existing code.
3591 2. A hit in the callout cache for a specific address caused a log line with no
3592 reason for rejecting RCPT. Now it says "Previous (cached) callout
3593 verification failure".
3595 3. There was an off-by-one bug in the queryprogram router. An over-long
3596 return line was truncated at 256 instead of 255 characters, thereby
3597 overflowing its buffer with the terminating zero. As well as fixing this, I
3598 have increased the buffer size to 1024 (and made a note to document this).
3600 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3601 when Exim is waiting for an SMTP response from a remote server, Exim
3602 restarts its select() call on the socket, thereby resetting its timeout.
3603 This is not a problem when such interrupts are rare. Somebody set up a cron
3604 job to run exiwhat every 2 minutes, which is less than the normal select()
3605 timeout (5 or 10 minutes). This meant that the select() timeout never
3606 kicked in because it was always reset. I have fixed this by comparing the
3607 time when an interrupt arrives with the time at the start of the first call
3608 to select(). If more time than the timeout has elapsed, the interrupt is
3609 treated as a timeout.
3611 5. Some internal re-factoring in preparation for the addition of Sieve
3612 extensions (by MH). In particular, the "personal" test is moved to a
3613 separate function, and given an option for scanning Cc: and Bcc: (which is
3614 not set for Exim filters).
3616 6. When Exim created an email address using the login of the caller as the
3617 local part (e.g. when creating a From: or Sender: header line), it was not
3618 quoting the local part when it contained special characters such as @.
3620 7. Installed new OpenBSD configuration files.
3622 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3623 try to make them clearer.
3625 9. Callout options, other than the timeout value, were being ignored when
3626 verifying sender addresses in header lines. For example, when using
3628 verify = header_sender/callout=no_cache
3630 the cache was (incorrectly) being used.
3632 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3633 only the Cygwin environment.
3635 11. When return_path_on_delivery was set as a log selector, if different remote
3636 addresses in the same message used different return paths and parallel
3637 remote delivery occurred, the wrong values would sometimes be logged.
3638 (Whenever a remote delivery process finished, the return path value from
3639 the most recently started remote delivery process was logged.)
3641 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3642 lines when AUTH and/or TLS are in use. This is the "received protocol"
3643 field. Exim used to use "asmtp" for authenticated SMTP, without any
3644 indication (in the protocol name) for TLS use. Now it follows the RFC and
3645 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3646 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3647 names appear in log lines as well as in Received: header lines.
3649 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3650 extensions, and comparison tests, and to fix some bugs.
3652 14. Changes to the "personal" filter test:
3654 (1) The test was buggy in that it was just doing the equivalent of
3655 "contains" tests on header lines. For example, if a user's address was
3656 anne@some.where, the "personal" test would incorrectly be true for
3658 To: susanne@some.where
3660 This test is now done by extracting each address from the header in turn,
3661 and checking the entire address. Other tests that are part of "personal"
3662 are now done using regular expressions (for example, to check local parts
3663 of addresses in From: header lines).
3665 (2) The list of non-personal local parts in From: addresses has been
3666 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3667 taken from the Sieve specification recommendations.
3669 (3) If the message contains any header line starting with "List-" it is
3670 treated as non-personal.
3672 (4) The test for "circular" in the Subject: header line has been removed
3673 because it now seems ill-conceived.
3675 15. Minor typos in src/EDITME comments corrected.
3677 16. Installed latest exipick from John Jetmore.
3679 17. If headers_add on a router specified a text string that was too long for
3680 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3681 of string_sprintf() is now avoided.
3683 18. $message_body_size was not set (it was always zero) when running the DATA
3684 ACL and the local_scan() function.
3686 19. For the "mail" command in an Exim filter, no default was being set for
3687 the once_repeat time, causing a random time value to be used if "once" was
3688 specified. (If the value happened to be <= 0, no repeat happened.) The
3689 default is now 0s, meaning "never repeat". The "vacation" command was OK
3690 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3691 (I found it when inspecting the code).
3693 20. There is now an overall timeout for performing a callout verification. It
3694 defaults to 4 times the callout timeout, which applies to individual SMTP
3695 commands during the callout. The overall timeout applies when there is more
3696 than one host that can be tried. The timeout is checked before trying the
3697 next host. This prevents very long delays if there are a large number of
3698 hosts and all are timing out (e.g. when the network connections are timing
3699 out). The value of the overall timeout can be changed by specifying an
3700 additional sub-option for "callout", called "maxwait". For example:
3702 verify = sender/callout=5s,maxwait=20s
3704 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3705 to the end before writing, but this should make it even safer).
3707 22. Exim was forgetting that it had advertised PIPELINING for the second and
3708 subsequent messages on an SMTP connection. It was also not resetting its
3709 memory on STARTTLS and an internal HELO.
3711 23. When Exim logs an SMTP synchronization error within a session, it now
3712 records whether PIPELINING has been advertised or not.
3714 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3715 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3716 rather than long int.
3718 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3720 26. Added the never_mail option to autoreply.
3726 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3727 crash if the getsockname() call failed; this can happen if a connection is
3728 closed very soon after it is established. The problem was simply in the
3729 order in which certain operations were done, causing Exim to try to write
3730 to the SMTP stream before it had set up the file descriptor. The bug has
3731 been fixed by making things happen in the correct order.
3737 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3738 before the connection was closed, thus losing the rejection response.
3740 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3741 some early Solaris releases, but causes trouble in current releases where
3742 socklen_t is defined.
3744 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3747 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3750 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3751 that releasing the top of it at the end releases what was used for sub-
3752 expansions (unless the block got too big). However, discard this block if
3753 the first thing is a variable or header, so that we can use its block when
3754 it is dynamic (useful for very large $message_headers, for example).
3756 6. Lookups now cache *every* query, not just the most recent. A new, separate
3757 store pool is used for this. It can be recovered when all lookup caches are
3758 flushed. Lookups now release memory at the end of their result strings.
3759 This has involved some general refactoring of the lookup sources.
3761 7. Some code has been added to the store_xxx() functions to reduce the amount
3762 of flapping under certain conditions.
3764 8. log_incoming_interface used to affect only the <= reception log lines. Now
3765 it causes the local interface and port to be added to several more SMTP log
3766 lines, for example "SMTP connection from", and rejection lines.
3768 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3770 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3772 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3773 could be overwritten at the end of the current message (or the start of a
3774 new message if it was set in a HELO ACL). The value is now preserved for
3775 the duration of the SMTP connection.
3777 12. If a transport had a headers_rewrite setting, and a matching header line
3778 contained an unqualified address, that address was qualified, even if it
3779 did not match any rewriting rules. The underlying bug was that the values
3780 of the flags that permit the existence of unqualified sender and recipient
3781 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3782 non-local messages, and by -bnq for local messages) were not being
3783 preserved with the message after it was received.
3785 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3786 "next input=" as part of the text comprising the host identity instead of
3787 the correct text. The code was using the same buffer for two different
3788 strings. However, depending on which order the printing function evaluated
3789 its arguments, the bug did not always show up. Under Linux, for example, my
3790 test suite worked just fine.
3792 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3793 doesn't work with some older versions of Perl. It has been changed to "my",
3794 which in any case is probably the better facility to use.
3796 15. A really picky compiler found some instances of statements for creating
3797 error messages that either had too many or two few arguments for the format
3800 16. The size of the buffer for calls to the DNS resolver has been increased
3801 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3802 for addresses that have a lot of PTR records. This alleviates a problem; it
3803 does not fully solve it.
3805 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3806 buffer now truncates the list and logs the incident, which is the same
3807 action as happens when Exim is looking up a host name and its aliases.
3808 Previously in this situation something unpredictable would happen;
3809 sometimes it was "internal error: store_reset failed".
3811 18. If a server dropped the connection unexpectedly when an Exim client was
3812 using GnuTLS and trying to read a response, the client delivery process
3813 crashed while trying to generate an error log message.
3815 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3816 string, for example:
3818 warn message = H1: something\nH2: something
3820 the text was added as a single header line from Exim's point of view
3821 though it ended up OK in the delivered message. However, searching for the
3822 second and subsequent header lines using $h_h2: did not work. This has been
3823 fixed. Similarly, if a system filter added multiple headers in this way,
3824 the routers could not see them.
3826 20. Expanded the error message when iplsearch is called with an invalid key to
3827 suggest using net-iplsearch in a host list.
3829 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3830 ACLs are no longer actually imposed (and a message to that effect is
3833 22. If a "gecos" field in a passwd entry contained escaped characters, in
3834 particular, if it contained a \" sequence, Exim got it wrong when building
3835 a From: or a Sender: header from that name. A second bug also caused
3836 incorrect handling when an unquoted " was present following a character
3837 that needed quoting.
3839 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3840 was not being matched caselessly.
3842 24. Arranged for all hyphens in the exim.8 source to be escaped with
3845 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3846 properly. Recipient callout cache records were still being keyed to include
3847 the sender, even when use_sender was set false. This led to far more
3848 callouts that were necessary. The sender is no longer included in the key
3849 when use_sender is false.
3851 26. Added "control = submission" modifier to ACLs.
3853 27. Added the ${base62d: operator to decode base 62 numbers.
3855 28. dnsdb lookups can now access SRV records.
3857 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3858 the configuration file.
3860 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3861 (-v) mode. This makes the output for a verbose queue run more intelligible.
3863 31. Added a use_postmaster feature to recipient callouts.
3865 32. Added the $body_zerocount variable, containing the number of binary zero
3866 bytes in the message body.
3868 33. The time of last modification of the "new" subdirectory is now used as the
3869 "mailbox time last read" when there is a quota error for a maildir
3872 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3874 35. Added +ignore_unknown as a special item in host lists.
3876 36. Code for decoding IPv6 addresses in host lists is now included, even if
3877 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3878 address was recognized as an IP address, but was then not correctly decoded
3879 into binary, causing unexpected and incorrect effects when compared with
3886 1. Very minor rewording of debugging text in manualroute to say "list of
3887 hosts" instead of "hostlist".
3889 2. If verify=header_syntax was set, and a header line with an unqualified
3890 address (no domain) and a large number of spaces between the end of the
3891 name and the colon was received, the reception process suffered a buffer
3892 overflow, and (when I tested it) crashed. This was caused by some obsolete
3893 code that should have been removed. The fix is to remove it!
3895 3. When running in the test harness, delay a bit after writing a bounce
3896 message to get a bit more predictability in the log output.
3898 4. Added a call to search_tidyup() just before forking a reception process. In
3899 theory, someone could use a lookup in the expansion of smtp_accept_max_
3900 per_host which, without the tidyup, could leave open a database connection.
3902 5. Added the variables $recipient_data and $sender_data which get set from a
3903 lookup success in an ACL "recipients" or "senders" condition, or a router
3904 "senders" option, similar to $domain_data and $local_part_data.
3906 6. Moved the writing of debug_print from before to after the "senders" test
3909 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3910 problems for message scanning, either using a data ACL, or using
3911 local_scan() because the Received: header was not generated till after they
3912 were called (in order to set the time as the time of reception completion).
3913 I have revised the way this works. The header is now generated after the
3914 body is received, but before the ACL or local_scan() are called. After they
3915 are run, the timestamp in the header is updated.
3921 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3922 before starting a queue runner without re-exec. This happened only when
3923 deliver_drop_privilege was set or when the Exim user was set to root. The
3924 effect of the bug was that timeouts during subsequent deliveries caused
3925 crashes instead of being properly handled. The handler is now left at its
3926 default (and expected) setting.
3928 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3929 message, again when deliver_drop_privilege is set or Exim is run as root.
3930 The bug described in (1) was not present in this case, but the tidying up
3931 of the other signals was missing. I have made the two cases consistent.
3933 3. The ignore_target_hosts setting on a manualroute router was being ignored
3934 for hosts that were looked up using the /MX notation.
3936 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3939 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3940 operated on the sender address. After changing the $sender_address to <>
3941 for the sender address verify, Exim was re-instated it as the original
3942 (before rewriting) address, but remembering that it had rewritten it, so it
3943 wasn't rewriting it again. This bug also had the effect of breaking the
3944 sender address verification caching when the sender address was rewritten.
3946 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3947 This has been changed so that if the ip literal address matches
3948 ignore_target_hosts, the router declines.
3950 7. Added expansion conditions match_domain, match_address, and match_local_
3951 part (NOT match_host).
3953 8. The placeholder for the Received: header didn't have a length field set.
3955 9. Added code to Exim itself and to exim_lock to test for a specific race
3956 condition that could lead to file corruption when using MBX delivery. The
3957 issue is with the lockfile that is created in /tmp. If this file is removed
3958 after a process has opened it but before that process has acquired a lock,
3959 there is the potential for a second process to recreate the file and also
3960 acquire a lock. This could lead to two Exim processes writing to the file
3961 at the same time. The added code performs the same test as UW imapd; it
3962 checks after acquiring the lock that its file descriptor still refers to
3963 the same named file.
3965 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3966 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3969 11. Added the smtp_active_hostname option. If used, this will typically be made
3970 to depend on the incoming interface address. Because $interface_address is
3971 not set up until the daemon has forked a reception process, error responses
3972 that can happen earlier (such as "too many connections") no longer contain
3975 12. If an expansion in a condition on a "warn" statement fails because a lookup
3976 defers, the "warn" statement is abandoned, and the next ACL statement is
3977 processed. Previously this caused the whole ACL to be aborted.
3979 13. Added the iplsearch lookup type.
3981 14. Added ident_timeout as a log selector.
3983 15. Added tls_certificate_verified as a log selector.
3985 16. Added a global option tls_require_ciphers (compare the smtp transport
3986 option of the same name). This controls incoming TLS connections.
3988 17. I finally figured out how to make tls_require_ciphers do a similar thing
3989 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3990 before starting the TLS session.
3992 18. Tabs are now shown as \t in -bP output.
3994 19. If the log selector return_path_on_delivery was set, Exim crashed when
3995 bouncing a message because it had too many Received: header lines.
3997 20. If two routers both had headers_remove settings, and the first one included
3998 a superfluous trailing colon, the final name in the first list and the
3999 first name in the second list were incorrectly joined into one item (with a
4000 colon in the middle).
4006 1. Added -C and -D options to the exinext utility, mainly to make it easier
4007 to include in the automated testing, but these could be helpful when
4008 multiple configurations are in use.
4010 2. The exinext utility was not formatting the output nicely when there was
4011 an alternate port involved in the retry record key, nor when there was a
4012 message id as well (for retries that were specific to a specific message
4013 and a specific host). It was also confused by IPv6 addresses, because of
4014 the additional colons they contain. I have fixed the IPv4 problem, and
4015 patched it up to do a reasonable job for IPv6.
4017 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4018 delivery, the log line now contains "pipelined" if PIPELINING was used.
4020 4. An SMTP transport process used to panic and die if the bind() call to set
4021 an explicit outgoing interface failed. This has been changed; it is now
4022 treated in the same way as a connect() failure.
4024 5. A reference to $sender_host_name in the part of a conditional expansion
4025 that was being skipped was still causing a DNS lookup. This no longer
4028 6. The def: expansion condition was not recognizing references to header lines
4029 that used bh_ and bheader_.
4031 7. Added the _cache feature to named lists.
4033 8. The code for checking quota_filecount in the appendfile transport was
4034 allowing one more file than it should have been.
4036 9. For compatibility with Sendmail, the command line option
4044 and sets the incoming protocol and host name (for trusted callers). The
4045 host name and its colon can be omitted when only the protocol is to be set.
4046 Note the Exim already has two private options, -pd and -ps, that refer to
4047 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4048 "s", but I don't think that's a major issue.
4050 10. A number of refactoring changes to the code, none of which should affect
4053 (a) The number of logging options was getting close to filling up the
4054 32-bit word that was used as a bit map. I have split them into two classes:
4055 those that are passed in the argument to log_write(), and those that are
4056 only ever tested independently outside of that function. These are now in
4057 separate 32-bit words, so there is plenty of room for expansion again.
4058 There is no change in the user interface or the logging behaviour.
4060 (b) When building, for example, log lines, the code previously used a
4061 macro that called string_cat() twice, in order to add two strings. This is
4062 not really sufficiently general. Furthermore, there was one instance where
4063 it was actually wrong because one of the argument was used twice, and in
4064 one call a function was used. (As it happened, calling the function twice
4065 did not affect the overall behaviour.) The macro has been replaced by a
4066 function that can join an arbitrary number of extra strings onto a growing
4069 (c) The code for expansion conditions now uses a table and a binary chop
4070 instead of a serial search (which was left over from when there were very
4071 few conditions). Also, it now recognizes conditions like "pam" even when
4072 the relevant support is not compiled in: a suitably worded error message is
4073 given if an attempt is made to use such a condition.
4075 11. Added ${time_interval:xxxxx}.
4077 12. A bug was causing one of the ddress fields not to be passed back correctly
4078 from remote delivery subprocesses. The field in question was not being
4079 subsequently used, so this caused to problems in practice.
4081 13. Added new log selectors queue_time and deliver_time.
4083 14. Might have fixed a bug in maildirsizefile handling that threw up
4084 "unexpected character" debug warnings, and recalculated the data
4085 unnecessarily. In any case, I expanded the warning message to give more
4088 15. Added the message "Restricted characters in address" to the statements in
4089 the default ACL that block characters like @ and % in local parts.
4091 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4092 Three changes have been made:
4094 (a) There was a serious bug; a negative response to MAIL caused the whole
4095 recipient domain to be cached as invalid, thereby blocking all messages
4096 to all local parts at the same domain, from all senders. This bug has
4097 been fixed. The domain is no longer cached after a negative response to
4098 MAIL if the sender used is not empty.
4100 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4103 (c) A new callout option, "use_sender" has been added for people who want
4104 the modified behaviour.
4110 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4113 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4114 indeed breaks things for older releases.
4116 3. Added additional logging to the case where there is a problem reading data
4117 from a filter that is running in a subprocess using a pipe, in order to
4118 try to track down a specific problem.
4120 4. Testing facility fudge: when running in the test harness and attempting
4121 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4122 getting "No route to host". Convert this to a timeout.
4124 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4127 6. Some OS don't have socklen_t but use size_t instead. This affects the
4128 fifth argument of getsockopt() amongst other things. This is now
4129 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4130 can be set for individual OS. I have set it for SunOS5, OSF1, and
4131 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4132 some earlier ones do not.
4134 7. Change 4.30/15 was not doing the test caselessly.
4136 8. The standard form for an IPv6 address literal was being rejected by address
4137 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4138 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4139 this, as well as the form without the "IPv6" on the front (but only when
4140 address literals are enabled, of course).
4142 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4144 10. Exim crashed if a message with an empty sender address specified by -f
4145 encountered a router with an errors_to setting. This could be provoked only
4146 by a command such as
4150 where an empty string was supplied; "<>" did not hit this bug.
4152 11. Installed PCRE release 4.5.
4154 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4155 remained set. It is now erased.
4157 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4158 times from message ids (which are base 36 rather than the normal 62).
4160 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4161 were being counted as actual protocol errors, and logged if the log
4162 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4163 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4164 and DATA following a set of rejected RCPTs do not count as protocol errors.
4165 In other words, Exim assumes they were pipelined, though this may not
4166 actually be the case. Of course, in all cases the client gets an
4167 appropriate error code.
4169 15. If a lookup fails in an ACL condition, a message about the failure may
4170 be available; it is used if testing the ACL cannot continue, because most
4171 such messages specify what the cause of the deferral is. However, some
4172 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4173 that caused an old message to be retained and used if a later statement
4174 caused a defer, replacing the real cause of the deferral.
4176 16. If an IP address had so many PTR records that the DNS lookup buffer
4177 was not large enough to hold them, Exim could crash while trying to process
4178 the truncated data. It now detects and logs this case.
4180 17. Further to 4.21/58, another change has been made: if (and only if) the
4181 first line of a message (the first header line) ends with CRLF, a bare LF
4182 in a subsequent header line has a space inserted after it, so as not to
4183 terminate the header.
4185 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4186 unnecessarily, used atoi() instead of strtol(), and didn't check the
4187 termination when getting file sizes from file names by regex.
4189 19. Completely re-implemented the support for maildirsize files, in the light
4190 of a number of problems with the previous contributed implementation
4191 (4.30/29). In particular:
4193 . If the quota is zero, the maildirsize file is maintained, but no quota is
4196 . If the maildir directory does not exist, it is created before any attempt
4197 to write a maildirsize file.
4199 . The quota value in the file is just a cache; if the quota is changed in
4200 the transport, the new value overrides.
4202 . A regular expression is available for excluding directories from the
4205 20. The autoreply transport checks the characters in options that define the
4206 message's headers; it allows continued headers, but it was checking with
4207 isspace() after an embedded newline instead of explicitly looking for a
4210 21. If all the "regular" hosts to which an address was routed had passed their
4211 expiry times, and had not reached their retry times, the address was
4212 bounced, even if fallback hosts were defined. Now Exim should go on to try
4215 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4216 equivalent code in the SMTP transport. Some hosts send humungous responses
4217 to HELO/EHLO, more than 1024 it seems.
4219 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4220 gives compiler warnings in some environments. I've now done it "properly",
4223 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4224 (because of problems with the built-in one) was declared to return uschar *
4225 instead of char *, causing compiler failure.
4227 25. Fixed a file descriptor leak when processing alias/forward files.
4229 26. Fixed a minor format string issue in dbfn.c.
4231 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4233 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4234 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4237 29. When a sender address is verified, it is cached, to save repeating the test
4238 when there is more than one recipient in a message. However, when the
4239 verification involves a callout, it is possible for different callout
4240 options to be set for different recipients. It is too complicated to keep
4241 track of this in the cache, so now Exim always runs a verification when a
4242 callout is required, relying on the callout cache for the optimization.
4243 The overhead is duplication of the address routing, but this should not be
4246 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4247 to be verified with callout=postmaster, and the main callout worked but the
4248 postmaster check failed, the verification correctly failed. However, if a
4249 subsequent RCPT command asked for sender verification *without* the
4250 postmaster check, incorrect caching caused this verification also to fail,
4253 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4254 it was not caching the DNS options (qualify_single, search_parents) that
4255 were used when the lookup failed. A subsequent lookup with different
4256 options therefore always gave the same answer, though there were cases
4257 where it should not have. (Example: a "domains = !$mx_any" option on a
4258 dnslookup router: the "domains" option is always processed without any
4259 widening, but the router might have qualify_single set.) Now Exim uses the
4260 cached value only when the same options are set.
4262 32. Added John Jetmore's "exipick" utility to the distribution.
4264 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4265 than a timeout (e.g. a certificate is required, and is not provided), an
4266 Exim server now closes the connection immediately. Previously it waited for
4267 the client to close - but if the client is SSL, it seems that they each
4268 wait for each other, leading to a delay before one of them times out.
4270 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4271 maintained 0.8.x compatibility because I don't think many are using it, and
4272 it is clearly obsolete.
4274 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4277 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4278 client certificate was expired. A simple patch fixes this, though I don't
4279 understand the full logic of why the verify callback is called multiple
4282 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4283 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4284 which causes problems with some clients (such as the Certicom SSL Plus
4285 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4286 disables the coutermeasure allowing Eudora to connect."
4288 38. Exim was not checking that a write() to a log file succeeded. This could
4289 lead to Bad Things if a log got too big, in particular if it hit a file
4290 size limit. Exim now panics and dies if it cannot write to a log file, just
4291 as it does if it cannot open a log file.
4293 39. Modified OS/Makefile-Linux so that it now contains
4295 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4297 The two -D definitions ensure that Exim is compiled with large file
4298 support, which makes it possible to handle log files that are bigger than
4301 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4302 instance) a domain was checked against a named list that involved a lookup,
4303 causing $domain_data to be set, then another domain was checked against the
4304 same list, then the first domain was re-checked, the value of $domain_data
4305 after the final check could be wrong. In particular, if the second check
4306 failed, it could be set empty. This bug probably also applied to
4309 41. The strip_trailing_dot option was not being applied to the address given
4310 with the -f command-line option.
4312 42. The code for reading a message's header from the spool was incrementing
4313 $received_count, but never initializing it. This meant that the value was
4314 incorrect (doubled) while delivering a message in the same process in which
4315 it was received. In the most common configuration of Exim, this never
4316 happens - a fresh exec is done - but it can happen when
4317 deliver_drop_privilege is set.
4319 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4320 it now includes up to 150 characters of the unexpected data in the log
4323 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4324 and building data strings. The size of both of these buffers was 10 000
4325 bytes - far larger than anybody would *ever* want, thought I. Needless to
4326 say, somebody hit the limit. I have increased the maximum line length to
4327 20 000 and the maximum data length of concatenated lines to 100 000. I have
4328 also fixed two bugs, because there was no checking on these buffers. Tsk,
4329 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4330 buffer is too small.
4332 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4333 lsearch lookups. Now it does.
4335 46. When parsing a route_list item in a manualroute router, a fixed-length
4336 buffer was used for the list of hosts. I made this 1024 bytes long,
4337 thinking that nobody would ever have a list of hosts that long. Wrong.
4338 Somebody had a whole pile of complicated expansion conditions, and the
4339 string was silently truncated, leading to an expansion error. It turns out
4340 that it is easier to change to an unlimited length (owing to other changes
4341 that have happened since this code was originally written) than to build
4342 structure for giving a limitation error. The length of the item that
4343 expands into the list of hosts is now unlimited.
4345 47. The lsearch lookup could not handle data where the length of text line was
4346 more than 4095 characters. Such lines were truncated, leading to shortened
4347 data being returned. It should now handle lines of any length.
4349 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4350 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4353 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4354 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4355 no longer happens in comment lines. A list of replacements is now placed
4356 at the head of all of the source files, except those whose only change is
4357 to replace PERL_COMMAND in the very first #! line.
4359 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4360 messages on the queue, with a bottom-up merge sort, using code contributed
4361 by Michael Haardt. This should make operations like -bp somewhat faster on
4362 large queues. It won't affect queue runners, except when queue_run_in_order
4365 51. Installed eximstats 1.31 in the distribution.
4367 52. Added support for SRV lookups to the dnslookup router.
4369 53. If an ACL referred to $message_body or $message_body_end, the value was not
4370 reset for any messages that followed in the same SMTP session.
4372 54. The store-handling optimization for building very long strings was not
4373 differentiating between the different store pools. I don't think this
4374 actually made any difference in practice, but I've tidied it.
4376 55. While running the routers to verify a sender address, $sender_address
4377 was still set to the sender address. This is wrong, because when routing to
4378 send a bounce to the sender, it would be empty. Therefore, I have changed
4379 it so that, while verifying a sender address, $sender_address is set to <>.
4380 (There is no change to what happens when verifying a recipient address.)
4382 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4383 target A or AAAA records (if not already returned) without resetting the
4384 qualify_single or search_parents options of the DNS resolver. These are
4385 inappropriate in this case because the targets of MX and SRV records must
4386 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4387 target that, when qualified, matched something in the local domain. These
4388 two options are now turned off when doing these lookups.
4390 57. It seems that at least some releases of Reiserfs (which does not have the
4391 concept of a fixed number of inodes) returns zero and not -1 for the
4392 number of available inodes. This interacted badly with check_spool_inodes,
4393 which assumed that -1 was the "no such thing" setting. What I have done is
4394 to check that the total number of inodes is greater than zero before doing
4395 the test of how many are available.
4397 58. When a "warn" ACL statement has a log_message modifier, the message is
4398 remembered, and not repeated. This is to avoid a lot of repetition when a
4399 message has many recipients that cause the same warning to be written.
4400 Howewer, Exim was preserving the list of already written lines for an
4401 entire SMTP session, which doesn't seem right. The memory is now reset if a
4402 new message is started.
4404 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4405 debugging output unless log_rewrite was also set.
4407 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4408 of (int)(handle) when we know that handle contains (void *)(-1).
4410 61. The Exim daemon panic-logs an error return when it closes the incoming
4411 connection. However "connection reset by peer" seems to be common, and
4412 isn't really an error worthy of noting specially, so that particular error
4415 62. When Exim is trying to find all the local interfaces, it used to panic and
4416 die if the ioctl to get the interface flags failed. However, it seems that
4417 on at least one OS (Solaris 9) it is possible to have an interface that is
4418 included in the list of interfaces, but for which you get a failure error
4419 for this call. This happens when the interface is not "plumbed" into a
4420 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4421 failure of the "get flags" call assumes that the interface is down.
4423 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4424 makes life easier for people who are doing arithmetic on fields extracted
4425 from dates, where you often get leading zeros that should not be
4426 interpreted as octal.
4428 64. Added qualify_domain to the redirect router, to override the global
4431 65. If a pathologically long header line contained very many addresses (the
4432 report of this problem mentioned 10 000) and each of them was rewritten,
4433 Exim could use up a very large amount of memory. (It kept on making new
4434 copies of the header line as it rewrote, and never released the old ones.)
4435 At the expense of a bit more processing, the header rewriting function has
4436 been changed so that it no longer eats memory in this way.
4438 66. The generation of the Received: header has been moved from the time that a
4439 message starts to be received, to the time that it finishes. The timestamp
4440 in the Received: header should now be very close to that of the <= log
4441 line. There are two side-effects of this change:
4443 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4444 logged header lines no longer include the local Received: line, because
4445 it has not yet been created. The same applies to a copy of the message
4446 that is returned to a non-SMTP sender when a message is rejected.
4448 (b) When a filter file is tested using -bf, no additional Received: header
4449 is added to the test message. After some thought, I decided that this
4452 This change does not affect the value of $received_for. It is still set
4453 after address rewriting, but before local_scan() is called.
4455 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4457 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4458 gave an unhelpful panic error message, and a defer error. I have managed to
4459 change this behaviour so that it now rejects any supplied certificate,
4460 which seems right, as the list of acceptable certificates is empty.
4462 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4463 gave an unhelpful defer error. I have not managed to make this reject any
4464 supplied certificates, but the error message it gives is "no certificate
4465 supplied", which is not helpful.
4467 70. exigrep's output now also includes lines that are not associated with any
4468 message, but which match the given pattern. Implemented by a patch from
4469 Martin Sluka, which also tidied up the Perl a bit.
4471 71. Recipient callout verification, like sender verification, was using <> in
4472 the MAIL FROM command. This isn't really the right thing, since the actual
4473 sender may affect whether the remote host accepts the recipient or not. I
4474 have changed it to use the actual sender in the callout; this means that
4475 the cache record is now keyed on a recipient/sender pair, not just the
4476 recipient address. There doesn't seem to be a real danger of callout loops,
4477 since a callout by the remote host to check the sender would use <>.
4478 [SEE ABOVE: changed after hitting problems.]
4480 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4481 temporary errors. However, in the case of such a code being given after
4482 the end of a data transmission (i.e. after ".") Exim was failing to write
4483 a retry record for the message. (Yes, there was some broken host that was
4484 actually sending 8xx at this point.)
4486 73. An unknown lookup type in a host list could cause Exim to panic-die when
4487 the list was checked. (An example that provoked this was putting <; in the
4488 middle of a list instead of at the start.) If this happened during a DATA
4489 ACL check, a -D file could be left lying around. This kind of configuration
4490 error no longer causes Exim to die; instead it causes a defer errror. The
4491 incident is still logged to the main and panic logs.
4493 74. Buglet left over from Exim 3 conversion. The message "too many messages
4494 in one connection" was written to the rejectlog but not the mainlog, except
4495 when address rewriting (yes!) was being logged.
4497 75. Added write_rejectlog option.
4499 76. When a system filter was run not as root (that is, when system_filter_user
4500 was set), the values of the $n variables were not being returned to the
4501 main process; thus, they were not subsequently available in the $sn
4504 77. Added +return_path_on_delivery log selector.
4506 78. A connection timeout was being treated differently from recipients deferred
4507 when testing hosts_max_try with a message that was older than the host's
4508 retry timeout. (The host should not be counted, thus allowing all hosts to
4509 be tried at least once before bouncing.) This may have been the cause of an
4510 occasionally reported bug whereby a message would remain on the queue
4511 longer than the retry timeout, but would be bounced if a delivery was
4512 forced. I say "may" because I never totally pinned down the problem;
4513 setting up timeout/retry tests is difficult. See also the next item.
4515 79. The ultimate address timeout was not being applied to errors that involved
4516 a combination of host plus message (for example, a timeout on a MAIL
4517 command). When an address resolved to a number of possible hosts, and they
4518 were not all tried for each delivery (e.g. because of hosts_max_try), a
4519 message could remain on the queue longer than the retry timeout.
4521 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4524 81. Fixed an obscure SMTP outgoing bug which required at least the following
4525 conditions: (a) there was another message waiting for the same server;
4526 (b) the server returned 5xx to all RCPT commands in the first message so
4527 that the message was not completed; (c) the server dropped the connection
4528 or gave a negative response to the RSET that Exim sends to abort the
4529 transaction. The observed case was a dropped connection after DATA that had
4530 been sent in pipelining mode. That is, the server had advertised PIPELINING
4531 but was not implementing it correctly. The effect of the bug was incorrect
4532 behaviour, such as trying another host, and this could lead to a crash.
4538 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4539 and daemon.c were passed as pointers to ints; they should have been
4540 pointers to socklen_t variables (which are typically unsigned ints).
4542 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4545 3. Fixed a really odd bug that affected only the testing scheme; patching a
4546 certain fixed string in the binary changed the value of another string that
4547 happened to be identical to the end of the original first string.
4549 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4550 name", it returns that address as the IP address. On some operating
4551 systems (e.g. Solaris), it also passes back the IP address string as the
4552 "host name". However, on others (e.g. Linux), it passes back an empty
4553 string. Exim wasn't checking for this, and was changing the host name to an
4554 empty string, assuming it had been canonicized.
4556 5. Although rare, it is permitted to have more than one PTR record for a given
4557 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4558 all the names associated with an address, because they do in Solaris.
4559 However, it seems that they do not in Linux for data that comes from the
4560 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4561 I found this out when I moved to a new Linux workstation and tried to run
4562 the Exim test suite.
4564 To get round this problem I have changed the code so that it now does its
4565 own call to the DNS to look up PTR records when searching for a host name.
4566 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4567 addresses that are only in /etc/hosts are still found.
4569 This behaviour is, however, controlled by an option called host_lookup_
4570 order, which defaults to "bydns:byaddr". If people want to use the other
4571 order, or indeed, just use one or the other means of lookup, they can
4572 specify it in this variable.
4574 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4575 some operating systems, this comes back from gethostbyaddr() as an empty
4576 string, and this is what Exim used to test for. However, it seems that in
4577 other systems, "." is yielded. Exim now tests for this case too.
4579 7. The values of check_spool_space and check_log_space are now held internally
4580 as a number of kilobytes instead of an absolute number of bytes. If a
4581 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4582 kilobyte. This means that much larger values can be stored.
4584 8. Exim monitor: an attempt to get the action menu when not actually pointing
4585 at a message produces an empty menu entitled "No message selected". This
4586 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4587 no entries in it ("Shell widget menu has zero width and/or height"). So I
4588 have added a single, blank menu entry in this case.
4590 9. Added ${quote_local_part.
4592 10. MIME decoding is now applied to the contents of Subject: header lines when
4595 11. Now that a reference to $sender_host_address automatically causes a reverse
4596 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4597 host lookup before query-style lookups in lists that might use this
4598 variable. This has therefore been abolished, and the "net-" prefix is no
4599 longer necessary for query-style lookups.
4601 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4602 have been a typo for LFLAGS, so it has been changed.
4604 13. The install script calls Exim with "-C /dev/null" in order to find the
4605 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4606 to be output. Howeve, since Exim outputs its version number before the
4607 error, it didn't break the script. It just looked ugly. I fixed this by
4608 always allowing "-C /dev/null" if the caller is root.
4610 14. Ignore overlarge ACL variable number when reading spool file - insurance
4611 against a later release with more variables having written the file.
4613 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4614 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4615 this, as well as the form without the "IPv6" on the front.
4617 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4618 OS/Makefile-Darwin file.
4620 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4621 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4623 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4624 variable. However, some OS use ints for inodes. I've added cast to long int
4625 to get rid of the compiler warning.
4627 19. I had forgotten to lock out "/../" in configuration file names when
4628 ALT_CONFIG_PREFIX was set.
4630 20. Routers used for verification do not need to specify transports. However,
4631 if such a router generated a host list, and callout was configured, Exim
4632 crashed, because it could not find a port number from the (non-existent)
4633 transport. It now assumes port 25 in this circumstance.
4635 21. Added the -t option to exigrep.
4637 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4638 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4639 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4640 them is set, LOOKUP_LSEARCH is forced.
4642 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4644 24. Added sender and host information to the "rejected by local_scan()" log
4645 line; previously there was no indication of these.
4647 25. Added .include_if_exists.
4649 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4650 Linux. It turns out that not all file systems support this. Apparently some
4651 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4652 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4653 not supported on the file descriptor, is now ignored when Exim is trying to
4654 sync a directory. This applies only to Linux.
4656 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4658 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4659 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4662 29. Added support for maildirsize files from supplied patch (modified a bit).
4664 30. The use of :fail: followed by an empty string could lead Exim to respond to
4665 sender verification failures with (e.g.):
4667 550 Verification failed for <xxx>
4668 550 Sender verify failed
4670 where the first response line was missing the '-' that indicates it is not
4671 the final line of the response.
4673 31. The loop for finding the name of the user that called Exim had a hardwired
4674 limit of 10; it now uses the value of finduser_retries, which is used for
4675 all other user lookups.
4677 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4680 33. Exim was neglecting to zero errno before one call of strtol() when
4681 expanding a string and expecting an integer value. On some systems this
4682 resulted in spurious "integer overflow" errors. Also, it was casting the
4683 result into an int without checking.
4685 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4686 did not work. The code looks as if it has *never* worked, though it appears
4687 to have been documented since at least releast 1.62. I have made it work.
4689 35. The "timeout_DNS" error in retry rules, also documented since at least
4690 1.62, also never worked. As it isn't clear exactly what this means, and
4691 clearly it isn't a major issue, I have abolished the feature by treating it
4692 as "timeout", and writing a warning to the main and panic logs.
4694 36. The display of retry rules for -brt wasn't always showing the error code
4697 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4698 timeout_connect_A, timeout_connect_MX.
4700 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4701 to the empty sender.
4703 39. The daemon was not analysing the content of -oX till after it had closed
4704 stderr and disconnected from the controlling terminal. This meant that any
4705 syntax errors were only noted on the panic log, and the return code from
4706 the command was 0. By re-arranging the code a little, I've made the
4707 decoding happen first, so such errors now appear on stderr, and the return
4708 code is 1. However, the actual setting up of the sockets still happens in
4709 the disconnected process, so errors there are still only recorded on the
4712 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4713 connections (as happens on some IP stacks) was logged at start up time as
4714 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4715 it from "IPv6 and IPv4", which means that two separate sockets are being
4718 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4719 says whether AF_INET or AF_INET6 was passed as an argument.
4721 42. Exiwhat output was messed up when time zones were included in log
4724 43. Exiwhat now gives more information about the daemon's listening ports,
4725 and whether -tls-on-connect was used.
4727 44. The "port" option of the smtp transport is now expanded.
4729 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4730 silently ignored. Now an error message is written to the main and panic
4733 46. There's a new ACL modifier called "logwrite" which writes to a log file
4734 as soon as it is encountered.
4736 47. Added $local_user_uid and $local_user_gid at routing time.
4738 48. Exim crashed when trying to verify a sender address that was being
4741 49. Exim was recognizing only a space character after ".include". It now also
4742 recognizes a tab character.
4744 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4745 extracting the relevant information from the specification. The man page no
4746 longer contains scrambled data for the -d option, and I've added a section
4747 at the front about calling Exim under different names.
4749 51. Added "extra_headers" argument to the "mail" command in filter files.
4751 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4754 53. Installed eximstats 1.29.
4756 54. Added transport_filter_timeout as a generic transport option.
4758 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4759 Cc: header lines. This was required by RFC 822, but it not required by RFC
4762 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4763 incoming messages that did not have them. Now it does so only if the
4764 message originates locally, that is, if there is no associated remote host
4765 address. When Resent- header lines are present, this applies to the Resent-
4766 lines rather than the non-Resent- lines.
4768 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4769 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4771 58. Messages for configuration errors now include the name of the main
4772 configuration file - useful now that there may be more than one file in a
4773 list (.included file names were always shown).
4775 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4776 for those rare installations that do not start the daemon as root or run it
4777 setuid root. I've cut out the call to initgroups() if the daemon is not
4780 60. The Exim user and group can now be bound into the binary as text strings
4781 that are looked up at the start of Exim's processing.
4783 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4785 62. Added $mailstore_basename variable.
4787 63. Installed patch to sieve.c from Michael Haardt.
4789 64. When Exim failed to open the panic log after failing to open the main log,
4790 the original message it was trying to log was written to stderr and debug
4791 output, but if they were not available (the usual case in production), it
4792 was lost. Now it is written to syslog before the two lines that record the
4793 failures to open the logs.
4795 65. Users' Exim filters run in subprocesses under the user's uid. It is
4796 possible for a "deliver" command or an alias in a "personal" command to
4797 provoke an address rewrite. If logging of address rewriting is configured,
4798 this fails because the process is not running as root or exim. There may be
4799 a better way of dealing with this, but for the moment (because 4.30 needs
4800 to be released), I have disabled address rewrite logging when running a
4801 filter in a non-root, non-exim process.
4807 1. The buildconfig auxiliary program wasn't quoting the value set for
4808 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4809 not defined. This bug was masked in 4.22 by the effect that was fixed in
4812 2. Some messages that were rejected after a message id was allocated were
4813 shown as "incomplete" by exigrep. It no longer does this for messages that
4814 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4816 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4817 have allow_domain_literals set, the ID did not get logged in the <= line.
4818 Domain literals are now always recognized in Message-ID: header lines.
4820 4. The first argument for a ${extract expansion item is the key name or field
4821 number. Leading and trailing spaces in this item were not being ignored,
4822 causing some misleading effects.
4824 5. When deliver_drop_privilege was set, single queue runner processes started
4825 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4826 same command in the process it spins off) were not dropping privilege.
4828 6. When the daemon running as "exim" started a queue runner, it always
4829 re-executed Exim in the spun-off process. This is a waste of effort when
4830 deliver_drop_privilege is set. The new process now just calls the
4831 queue-runner function directly.
4837 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4840 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4841 ignored. Though the use of -oP was forcing the writing of a pid file, it
4842 was always written to the default place.
4844 3. If the message "no IP address found for host xxxx" is generated during
4845 incoming verification, it is now followed by identification of the incoming
4846 connection (so you can more easily find what provoked it).
4848 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4850 5. Added some features to "harden" Exim a bit more against certain attacks:
4852 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4853 be put in Local/Makefile. This is like the never_users runtime option,
4854 but it cannot be overridden. The default setting is "root".
4856 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4857 prefix string with which any file named in a -C command line option
4860 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4861 is retained for -C and -D only if the caller of Exim is root. Without
4862 it, the exim user may also use -C and -D and retain privilege.
4864 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4865 command line option is disabled.
4867 6. Macro names set by the -D option must start with an upper case letter, just
4868 like macro names defined in the configuration file.
4870 7. Added "dereference=" facility to LDAP.
4872 8. Two instances of the typo "uknown" in the source files are fixed.
4874 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4875 the Configure-Makefile script screwed up while processing it.
4877 10. Incorporated PCRE 4.4.
4879 11. The SMTP synchronization check was not operating right at the start of an
4880 SMTP session. For example, it could not catch a HELO sent before the client
4881 waited for the greeting. There is now a check for outstanding input at the
4882 point when the greeting is written. Because of the duplex, asynchronous
4883 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4884 way, but not yet received, when the check is performed.
4886 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4887 on TCP/IP sockets, because this apparently causes some broken clients to
4890 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4891 unchanged) from the Cygwin maintainer.
4893 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4894 was supported instead of testing for mbx. Effectively a typo.
4896 15. The spa authenticator server code was not checking that the input it
4897 received was valid base64.
4899 16. The debug output line for the "set" modifier in ACLs was not showing the
4900 name of the variable that was being set.
4902 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4904 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4905 lookup on its own. Something else was needed to trigger the lookup. For
4906 example, a match in host_lookup or the need for a host name in a host list.
4907 Now, if $sender_host_name is referenced and the host name has not yet been
4908 looked up, a lookup is performed. If the lookup fails, the variable remains
4909 empty, and $host_lookup_failed is set to "1".
4911 19. Added "eqi" as a case-independent comparison operator.
4913 20. The saslauthd authentication condition could segfault if neither service
4914 nor realm was specified.
4916 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4917 error message that was logged was misleading, and incoming SMTP
4918 connections were dropped. The message is now more accurate, and temporary
4919 errors are given to SMTP connections.
4921 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4922 (except RSET) with a 421 error, until QUIT is received. However, it was
4923 failing to send a response to QUIT.
4925 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4926 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4927 ACL is now run after the helo_try_verify_hosts code.
4929 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4930 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4931 case-independent, so other case variants are also recognized). Apparently
4932 some systems use these upper case variants.
4934 25. If more than two messages were waiting for the same host, and a transport
4935 filter was specified for the transport, Exim sent two messages over the
4936 same TCP/IP connection, and then failed with "socket operation on non-
4937 socket" when it tried to send the third.
4939 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4941 27. The extern definition of crypt16() in expand.c was not being excluded when
4942 the OS had its own crypt16() function.
4944 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4945 as a preferred synonym for return_size_limit, both as an option and as an
4948 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4950 30. Changed the default configuration ACL to relax the local part checking rule
4951 for addresses that are not in any local domains. For these addresses,
4952 slashes and pipe symbols are allowed within local parts, but the sequence
4953 /../ is explicitly forbidden.
4955 31. SPA server authentication was not clearing the challenge buffer before
4958 32. log_message in a "warn" ACL statement was writing to the reject log as
4959 well as to the main log, which contradicts the documentation and doesn't
4960 seem right (because no rejection is happening). So I have stopped it.
4962 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4963 However, I am unable to do any testing of this.
4965 34. Fixed an infelicity in the appendfile transport. When checking directories
4966 for a mailbox, to see if any needed to be created, it was accidentally
4967 using path names with one or more superfluous leading slashes; tracing
4968 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4970 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4971 contents are added to the log line that is written for every discarded
4972 recipient. (Previously a log_message setting was ignored.)
4974 36. The ${quote: operator now quotes the string if it is empty.
4976 37. The install script runs exim in order to find its version number. If for
4977 some reason other than non-existence or emptiness, which it checks, it
4978 could not run './exim', it was installing it with an empty version number,
4979 i.e. as "exim-". This error state is now caught, and the installation is
4982 38. An argument was missing from the function that creates an error message
4983 when Exim fails to connect to the socket for saslauthd authentication.
4984 This could cause Exim to crash, or give a corrupted message.
4986 39. Added isip, isip4, and isip6 to ${if conditions.
4988 40. The ACL variables $acl_xx are now saved with the message, and can be
4989 accessed later in routers, transports, and filters.
4991 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4992 strings in the file are not string-expanded.
4994 42. If a MAIL command specified a SIZE value that was too large to fit into an
4995 int variable, the check against message_size_limit failed. Such values are
4996 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4997 day this will have to be increased, but I don't think I want to be around
4998 when emails are that large.
5005 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5006 iconv() is not standard in FreeBSD.
5008 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5009 IPv6 enabled. The observed symptom was a segmentation fault on return from
5010 the function os_common_find_running_interfaces() in src/os.c.
5012 3. In the check_special_case() function in daemon.c I had used "errno" as an
5013 argument name, which causes warnings on some systems. This was basically a
5014 typo, since it was named "eno" in the comments!
5016 4. The code that waits for the clock to tick (at a resolution of some fraction
5017 of a second) so as to ensure message-id uniqueness was always waiting for
5018 at least one whole tick, when it could have waited for less. [This is
5019 almost certainly not relevant at current processor speeds, where it is
5020 unlikely to ever wait at all. But we try to future-proof.]
5022 5. The function that sleeps for a time interval that includes fractions of a
5023 second contained a race. It did not block SIGALRM between setting the
5024 timer, and suspending (a couple of lines later). If the interval was short
5025 and the sigsuspend() was delayed until after it had expired, the suspension
5026 never ended. On busy systems this could lead to processes getting stuck for
5029 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5030 process, before it forks any delivery processes. The open lookup caching
5031 mechanism meant that the open file or database connection was passed into
5032 the delivery process. The problem was that delivery processes always tidy
5033 up cached lookup data. This could cause a problem for the next delivery
5034 process started by the queue runner, because the external queue runner
5035 process does not know about the closure. So the next delivery process
5036 still has data in the lookup cache. In the case of a file lookup, there was
5037 no problem because closing a file descriptor in a subprocess doesn't affect
5038 the parent. However, if the lookup was caching a connection to a database,
5039 the connection was closed, and the second delivery process was likely to
5040 see errors such as "PGSQL: query failed: server closed the connection
5041 unexpectedly". The problem has been fixed by closing all cached lookups
5042 in a queue runner before running a delivery process.
5044 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5045 seem to have the "const" qualifier which it has on other OS. I've
5048 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5049 *of the same type* (client or server) with the same public name that an
5050 error should be diagnosed.
5052 9. When Exim looked up a host name for an IP address, but failed to find the
5053 original IP address when looking up the host name (a safety check), it
5054 output the message "<ip address> does not match any IP for NULL", which was
5055 confusing, to say the least. The bug was that the host name should have
5056 appeared instead of "NULL".
5058 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5059 user that is built into the binary, and the -C or -D options is used, root
5060 privilege is dropped before the configuration file is read. In addition,
5061 logging is switched to stderr instead of the normal log files. If the
5062 configuration then re-defines the Exim user, the unprivileged environment
5063 is probably not what is expected, so Exim logs a panic warning message (but
5066 However, if deliver_drop_privilege is set, the unprivileged state may well
5067 be exactly what is intended, so the warning has been cut out in that case,
5068 and Exim is allowed to try to write to its normal log files.
5074 1. smtp_return_error_details was not giving details for temporary sender
5075 or receiver verification errors.
5077 2. Diagnose a configuration error if two authenticators have the same public
5080 3. Exim used not to create the message log file for a message until the first
5081 delivery attempt. This could be confusing when incoming messages were held
5082 for policy or load reasons. The message log file is now created at the time
5083 the message is received, and an initial "Received" line is written to it.
5085 4. The automatically generated man page for command line options had a minor
5086 bug that caused no ill effects; however, a more serious problem was that
5087 the procedure for building the man page automatically didn't always
5088 operate. Consequently, release 4.20 contains an out-of-date version. This
5089 shouldn't happen again.
5091 5. When building Exim with embedded Perl support, the script that builds the
5092 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5093 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5095 6. The freeze_tell option was not being used for messages that were frozen on
5096 arrival, either by an ACL or by local_scan().
5098 7. Added the smtp_incomplete_transaction log selector.
5100 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5101 was accepting AUTH without a new EHLO.
5103 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5104 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5107 10. Logging of TCP/IP connections (when configured) now happens in the main
5108 daemon process instead of the child process, so that the TCP/IP connection
5109 count is more accurate (but it can never be perfect).
5111 11. The use of "drop" in a nested ACL was not being handled correctly in the
5112 outer ACL. Now, if condition failure induced by the nested "drop" causes
5113 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5114 or "require"), the connection is dropped.
5116 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5117 that yield "discard" can now be used with an "accept" or a "discard" verb,
5118 but an error is generated for any others (because I can't see a useful way
5119 to define what should happen).
5121 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5122 are now processed in the same way as lines in the Exim configuration file.
5123 In particular, continuation lines are supported.
5125 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5127 15. Added -ti meaning -t -i.
5129 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5130 domains, and warn by logging if others are found.
5132 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5133 returned from GIFCONF when Exim is trying to find the list of interfaces on
5134 a host. The code in os.c has been modified to copy each ifreq to an aligned
5135 structure in all cases.
5137 Also, in some cases, the returned ifreq's were being copied to a 'struct
5138 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5139 means the last couple of bytes of an IPv6 address could be chopped if the
5140 ifreq contained only a normal sockaddr (14 bytes storage).
5142 18. Named domain lists were not supported in the hosts_treat_as_local option.
5143 An entry such as +xxxx was not recognized, and was treated as a literal
5146 19. Ensure that header lines added by a DATA ACL are included in the reject log
5147 if the ACL subsequently rejects the message.
5149 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5150 MD5 (which is deprecated).
5152 21. When testing a filter file using -bf, Exim was writing a message when it
5153 took the sender from a "From " line in the message, but it was not doing so
5154 when it took $return_path from a Return-Path: header line. It now does.
5156 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5157 with a valid header line field name (a series of printing characters
5158 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5160 23. Changed "disc" in the source to "disk" to conform to the documentation and
5161 the book and for uniformity.
5163 24. Ignore Sendmail's -Ooption=value command line item.
5165 25. When execve() failed while trying to run a command in a pipe transport,
5166 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5167 could be confused with a return value of 69 from the command itself. This
5168 has been changed to 127, the value the shell returns if it is asked to run
5169 a non-existent command. The wording for the related log line suggests a
5170 non-existent command as the problem.
5172 26. If received_header_text expands to an empty string, do not add a Received:
5173 header line to the message. (Well, it adds a token one on the spool, but
5174 marks it "old" so that it doesn't get used or transmitted.)
5176 27. Installed eximstats 1.28 (addition of -nt option).
5178 28. There was no check for failure on the call to getsockname() in the daemon
5179 code. This can fail if there is a shortage of resources on the system, with
5180 ENOMEM, for example. A temporary error is now given on failure.
5182 29. Contrary to the C standard, it seems that in some environments, the
5183 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5184 program. Exim now does this explicitly; it affects the formatting of
5185 timestamps using strftime().
5187 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5188 complaints. I've now initialized all the variables, to avoid this.
5190 32. Header lines added by a system filter were not being "seen" during
5191 transport-time rewrites.
5193 33. The info_callback() function passed to OpenSSL is set up with type void
5194 (*)(SSL *, int, int), as described somewhere. However, when calling the
5195 function (actually a macro) that sets it up, the type void(*)() is
5196 expected. I've put in a cast to prevent warnings from picky compilers.
5198 34. If a DNS black list lookup found a CNAME record, but there were no A
5199 records associated with the domain it pointed at, Exim crashed.
5201 35. If a DNS black list lookup returned more than one A record, Exim ignored
5202 all but the first. It now scans all returned addresses if a particular IP
5203 value is being sought. In this situation, the contents of the
5204 $dnslist_value variable are a list of all the addresses, separated by a
5207 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5208 to accept a host name and all its aliases if the forward lookup for any of
5209 them yielded the IP address of the incoming connection. Now it accepts only
5210 those names whose forward lookup yields the correct IP address. Any other
5211 names are discarded. This closes a loophole whereby a rogue DNS
5212 administrator could create reverse DNS records to break through a
5213 wildcarded host restriction in an ACL.
5215 37. If a user filter or a system filter that ran in a subprocess used any of
5216 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5217 the wrong values were passed to the pipe command ($thisaddress had the
5218 value of $0, $0 had the value of $1, etc). This bug was introduced by
5219 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5221 38. Improved the line breaking for long SMTP error messages from ACLs.
5222 Previously, if there was no break point between 40 and 75 characters, Exim
5223 left the rest of the message alone. Two changes have been made: (a) I've
5224 reduced the minimum length to 35 characters; (b) if it can't find a break
5225 point between 35 and 75 characters, it looks ahead and uses the first one
5226 that it finds. This may give the occasional overlong line, but at least the
5227 remaining text gets split now.
5229 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5230 file descriptors might be low, and that setting 1000 would always raise it.
5231 It turns out that in some environments, the limit is already over 1000 and
5232 that lowering it causes trouble. So now Exim takes care not to decrease it.
5234 40. When delivering a message, the value of $return_path is set to $sender_
5235 address at the start of routing (routers may change the value). By an
5236 oversight, this default was not being set up when an address was tested by
5237 -bt or -bv, which affected the outcome if any router or filter referred to
5240 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5241 log only when "message" or "log_message" are set. However, if one of the
5242 conditions was an address verification, or a call to a nested ACL, the
5243 messages generated by the underlying test were being passed through. This
5244 no longer happens. The underlying message is available in $acl_verify_
5245 message for both "message" and "log_message" expansions, so it can be
5246 passed through if needed.
5248 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5249 new expansion $bh_ to give the encoded byte string without charset
5250 translation. Translation happens only if iconv() is available; HAVE_ICONV
5251 indicates this at build time. HEADERS_CHARSET gives the charset to
5252 translate to; headers_charset can change it in the configuration, and
5253 "headers charset" can change it in an individual filter file.
5255 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5256 that creates RFC 2047 encoded "words" labels them as that charset instead
5257 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5258 expansion operator; (ii) when Exim creates a From: line for a local
5259 message; (iii) when a header line is rewritten to include a "phrase" part.
5261 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5262 buggy, causing it to skip the first lines of messages whose message ID
5263 ended in 'D'. This would not have bitten before Exim release 4.14, because
5264 message IDs were unlikely to end in 'D' before then. The effect was to have
5265 incorrect size information for certain domains.
5267 45. #include "config.h" was missing at the start of the crypt16.c module. This
5268 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5271 46. If there was a timeout during a "random" callout check, Exim treated it as
5272 a failure of the random address, and carried on sending RSET and the real
5273 address. If the delay was just some slowness somewhere, the response to the
5274 original RCPT would be taken as a response to RSET and so on, causing
5275 mayhem of various kinds.
5277 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5278 when I implemented it. It didn't allow for the fact that some option values
5279 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5280 the right test for positive values.
5282 48. Domain names in DNS records are case-independent. Exim always looks them up
5283 in lower case. Some resolvers return domain names in exactly the case they
5284 appear in the zone file, that is, they may contain uppercase letters. Not
5285 all resolvers do this - some return always lower case. Exim was treating a
5286 change of case by a resolver as a change of domain, similar to a widening
5287 of a domain abbreviation. This triggered its re-routing code and so it was
5288 trying to route what was effectively the same domain again. This normally
5289 caused routing to fail (because the router wouldn't handle the domain
5290 twice). Now Exim checks for this case specially, and just changes the
5291 casing of the domain that it ultimately uses when it transmits the message
5294 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5297 50. If a filter generated a file delivery with a non-absolute name (possible if
5298 no home directory exists for the router), the forbid_file option was not
5301 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5302 the existing equality matching.
5304 52. Exim was using ints instead of ino_t variables in some places where it was
5305 dealing with inode numbers.
5307 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5308 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5309 called TMPDIR, and if it finds it is different, it changes its value.
5311 54. The smtp_printf() function is now made available to local_scan() so
5312 additional output lines can be written before returning. There is also an
5313 smtp_fflush() function to enable the detection of a dropped connection.
5314 The variables smtp_input and smtp_batched_input are exported to
5317 55. Changed the default runtime configuration: the message "Unknown user"
5318 has been removed from the ACL, and instead placed on the localuser router,
5319 using the cannot_route_message feature. This means that any verification
5320 failures that generate their own messages won't get overridden. Similarly,
5321 the "Unrouteable address" message that was in the ACL for unverifiable
5322 relay addresses has also been removed.
5324 56. Added hosts_avoid_esmtp to the smtp transport.
5326 57. The exicyclog script was not checking for the esoteric option
5327 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5328 will work only if exicyclog is run under the appropriate euid.
5330 58. Following a discussion on the list, the rules by which Exim recognises line
5331 endings on incoming messages have been changed. The -dropcr and drop_cr
5332 options are now no-ops, retained only for backwards compatibility. The
5333 following line terminators are recognized: LF CRLF CR. However, special
5334 processing applies to CR:
5336 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5337 nor a local message in the state where . is a terminator.
5339 (ii) If a bare CR is encountered in a header line, an extra space is added
5340 after the line terminator so as not to end the header. The reasoning
5341 behind this is that bare CRs in header lines are most likely either
5342 to be mistakes, or people trying to play silly games.
5344 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5345 was being incorrectly given as 18 bytes larger than it should have been.
5346 This is a VOB (very old bug).
5348 60. This may never have affected anything current, but just in case it has:
5349 When the local host is found other than at the start of a list of hosts,
5350 the local host, those with the same MX, and any that follow, are discarded.
5351 When the list in question was part of a longer list of hosts, the following
5352 hosts (not currently being processed) were also being discarded. This no
5353 longer happens. I'm not sure if this situation could ever has previously
5356 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5359 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5360 header. This is something that is recommended in a new Internet Draft, and
5361 is something that is documented as being done by Sendmail. There are two
5362 possible values. For messages generated by the autoreply transport, Exim
5365 Auto-Submitted: auto-replied
5367 whereas for all other generated messages (e.g. bounces) it adds
5369 Auto-Submitted: auto-generated
5371 63. The "personal" condition in filters now includes a test for the
5372 Auto-Submitted: header. If it contains the string "auto-" the message it
5373 not considered personal.
5375 64. Added rcpt_include_affixes as a generic transport option.
5377 65. Added queue_only_override (default true).
5379 66. Added the syslog_duplication option.
5381 67. If what should have been the first header line of a message consisted of
5382 a space followed by a colon, Exim was mis-interpreting it as a header line.
5383 It isn't of course - it is syntactically invalid and should therefore be
5384 treated as the start of the message body. The misbehaviour could have
5385 caused a number of strange effects, including loss of data in subsequent
5386 header lines, and spool format errors.
5388 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5389 client host had authenticated. This control can now be exercised by an ACL
5390 for more flexibility.
5392 69. By default, callouts do not happen when testing with -bh. There is now a
5393 variant, -bhc, which does actually run the callout code, including
5394 consulting and updating the callout cache.
5396 70. Added support for saslauthd authentication, courtesy of Alexander
5399 71. If statvfs() failed on the spool or log directories while checking their
5400 size for availability, Exim confusingly gave the error "space shortage".
5401 Furthermore, in debugging mode it crashed with a floating point exception.
5402 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5403 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5404 serious problem, Exim now writes to the main and panic logs when this
5405 happens, with details of the failure. It then refuses to accept the
5406 incoming message, giving the message "spool directory problem" or "log
5407 directory problem" with a 421 code for SMTP messages.
5409 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5410 0, 1, and 2 exist, because some OS complain for execs without them (see
5411 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5412 descriptors. However, the code omitted to check that the open succeeded,
5413 causing mysterious errors if for some reason the permissions on /dev/null
5414 got screwed. Now Exim writes a message to the main and panic logs, and
5415 bombs out if it can't open /dev/null.
5417 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5418 interact so that it is all more flexible. It is supposed to remain
5419 backwards compatible. Also added extra_local_interfaces.
5421 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5422 to bomb out with an assertion failure - to the client this appears as a
5423 connection drop. This problem occurs in the part of the code that was taken
5424 from the Samba project. Fortunately, the assertion is in a very simple
5425 function, so I have fixed this by reproducing the function inline in the
5426 one place where it is called, and arranging for authentication to fail
5427 instead of killing the process with assert().
5429 75. The SPA client code was not working when the server requested OEM rather
5430 than Unicode encoding.
5432 76. Added code to make require_files with a specific uid setting more usable in
5433 the case where statting the file as root fails - usually a non-root-mounted
5434 NFS file system. When this happens and the failure is EACCES, Exim now
5435 forks a subprocess and does the per-uid checking as the relevant uid.
5437 77. Added process_log_path.
5439 78. If log_file_path was not explicitly set, a setting of check_log_space or
5440 check_log_inodes was ignored.
5442 79. If a space check for the spool or log partitions fails, the incident is now
5443 logged. Of course, in the latter case the data may get lost...
5445 80. Added the %p formatting code to string_format() so that it can be used to
5446 print addresses in debug_print(). Adjusted all the address printing in the
5447 debugging in store.c to use %p rather than %d.
5449 81. There was a concern that a line of code in smtp_in.c could overflow a
5450 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5451 initially expressed, the concern was not well-founded, because trailing
5452 spaces are removed early. However, if the trailing spaces were followed by
5453 a NULL, they did not get removed, so the overflow was possible. Two fixes
5456 (a) I re-wrote the offending code in a cleaner fashion.
5457 (b) If an incoming SMTP command contains a NULL character, it is rejected
5460 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5461 runs initgroups(), so that if the Exim user is in any additional groups,
5462 they will be used during message reception.
5468 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob