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.
110 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
112 PP/02 Harmonised TLS library version reporting; only show if debugging.
113 Layout now matches that introduced for other libraries in 4.74 PP/03.
115 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
117 PP/04 New "dns_use_edns0" global option.
119 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
122 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
123 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
125 TK/01 Updated PolarSSL code to 0.14.2.
126 Bugzilla 1097. Patch from Andreas Metzler.
128 PP/07 Catch divide-by-zero in ${eval:...}.
131 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
134 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
135 format-string attack -- SECURITY: remote arbitrary code execution.
137 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
138 time unintentionally subject to list matching rules, letting the header
139 cause arbitrary Exim lookups (of items which can occur in lists, *not*
140 arbitrary string expansion). This allowed for information disclosure.
142 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
143 INT_MIN/-1 -- value coerced to INT_MAX.
149 NM/01 Workround for PCRE version dependency in version reporting
152 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
153 This fixes portability to compilers other than gcc, notably
154 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
156 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
157 makefiles for portability to HP-UX and POSIX correctness.
159 PP/01 Permit LOOKUP_foo enabling on the make command-line.
160 Also via indented variable definition in the Makefile.
161 (Debugging by Oliver Heesakkers).
163 PP/02 Restore caching of spamd results with expanded spamd_address.
164 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
166 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
167 Improves build reliability. Fix from: Frank Elsner
169 NM/02 Fix wide character breakage in the rfc2047 coding
170 Fixes bug 1064. Patch from Andrey N. Oktyabrski
172 NM/03 Allow underscore in dnslist lookups
173 Fixes bug 1026. Patch from Graeme Fowler
175 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
176 Code patches from Adam Ciarcinski of NetBSD.
178 NM/04 Fixed exiqgrep to cope with mailq missing size issue
181 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
182 is logged, to avoid truncation. Patch from John Horne.
184 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
185 Patch from Jakob Hirsch.
187 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
188 SQL string expansion failure details.
189 Patch from Andrey Oktyabrski.
191 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
192 Patch from Simon Arlott.
194 PP/09 New lookups functionality failed to compile on old gcc which rejects
195 extern declarations in function scope.
196 Patch from Oliver Fleischmann
198 PP/10 Use sig_atomic_t for flags set from signal handlers.
199 Check getgroups() return and improve debugging.
200 Fixed developed for diagnosis in bug 927 (which turned out to be
203 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
204 Patch from Mark Zealey.
206 PP/12 Bugzilla 1056: Improved spamd server selection.
207 Patch from Mark Zealey.
209 PP/13 Bugzilla 1086: Deal with maildir quota file races.
210 Based on patch from Heiko Schlittermann.
212 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
213 Patch from Uwe Doering, sign-off by Michael Haardt.
215 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
216 variable declaration deep within a block. Bug and patch from
219 PP/15 lookups-Makefile IRIX compatibilty coercion.
221 PP/16 Make DISABLE_DKIM build knob functional.
223 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
224 Patch by Simon Arlott
226 TF/03 Fix valgrind.h portability to C89 compilers that do not support
227 variable argument macros. Our copy now differs from upstream.
233 TF/01 Failure to get a lock on a hints database can have serious
234 consequences so log it to the panic log.
236 TF/02 Log LMTP confirmation messages in the same way as SMTP,
237 controlled using the smtp_confirmation log selector.
239 TF/03 Include the error message when we fail to unlink a spool file.
241 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
242 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
243 for maintaining out-of-tree patches for some time.
245 PP/01 Bugzilla 139: Documentation and portability issues.
246 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
247 Handle per-OS dynamic-module compilation flags.
249 PP/02 Let /dev/null have normal permissions.
250 The 4.73 fixes were a little too stringent and complained about the
251 permissions on /dev/null. Exempt it from some checks.
252 Reported by Andreas M. Kirchwitz.
254 PP/03 Report version information for many libraries, including
255 Exim version information for dynamically loaded libraries. Created
256 version.h, now support a version extension string for distributors
257 who patch heavily. Dynamic module ABI change.
259 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
260 privilege escalation vulnerability whereby the Exim run-time user
261 can cause root to append content of the attacker's choosing to
264 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
267 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
268 If dropping privileges for untrusted macros, we disabled normal logging
269 on the basis that it would fail; for the Exim run-time user, this is not
270 the case, and it resulted in successful deliveries going unlogged.
271 Fixed. Reported by Andreas Metzler.
277 PP/01 Date: & Message-Id: revert to normally being appended to a message,
278 only prepend for the Resent-* case. Fixes regression introduced in
279 Exim 4.70 by NM/22 for Bugzilla 607.
281 PP/02 Include check_rfc2047_length in configure.default because we're seeing
282 increasing numbers of administrators be bitten by this.
284 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
286 PP/03 Bugzilla 994: added openssl_options main configuration option.
288 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
290 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
292 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
294 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
295 without a peer certificate, leading to a segfault because of an
296 assumption that peers always have certificates. Be a little more
297 paranoid. Problem reported by Martin Tscholak.
299 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
300 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
301 NB: ClamAV planning to remove STREAM in "middle of 2010".
302 CL also introduces -bmalware, various -d+acl logging additions and
303 more caution in buffer sizes.
305 PP/09 Implemented reverse_ip expansion operator.
307 PP/10 Bugzilla 937: provide a "debug" ACL control.
309 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
311 PP/12 Bugzilla 973: Implement --version.
313 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
315 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
317 PP/15 Bugzilla 816: support multiple condition rules on Routers.
319 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
320 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
321 ignore trailing whitespace.
323 JJ/02 prevent non-panic DKIM error from being sent to paniclog
325 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
328 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
329 Notification from Dr Andrew Aitchison.
331 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
332 ExtendedDetectionInfo response format.
333 Notification from John Horne.
335 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
338 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
339 XSL and documented dependency on system catalogs, with examples of how
342 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
345 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
346 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
347 configuration file which is writeable by the Exim user or group.
349 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
350 of configuration files to cover files specified with the -C option if
351 they are going to be used with root privileges, not just the default
354 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
355 option (effectively making it always true).
357 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
358 files to be used while preserving root privileges.
360 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
361 that rogue child processes cannot use them.
363 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
364 run-time user, instead of root.
366 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
367 Exim run-time user without dropping privileges.
369 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
370 result string, instead of calling string_vformat() twice with the same
373 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
374 for other users. Others should always drop root privileges if they use
375 -C on the command line, even for a whitelisted configure file.
377 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
379 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
385 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
386 $data_path, and $header_path variables; fixed documentation bugs and
389 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
390 exipick to access non-standard spools, including the "frozen" queue
393 NM/01 Bugzilla 965: Support mysql stored procedures.
394 Patch from Alain Williams
396 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
398 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
399 Patch from Andreas Metzler
401 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
402 Patch from Kirill Miazine
404 NM/05 Bugzilla 671: Added umask to procmail example.
406 JJ/03 installed exipick 20100323.0, fixing doc bug
408 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
409 directory. Notification and patch from Dan Rosenberg.
411 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
413 TK/02 Improve log output when DKIM signing operation fails.
415 MH/01 Treat the transport option dkim_domain as a colon separated
416 list, not as a single string, and sign the message with each element,
417 omitting multiple occurences of the same signer.
419 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
420 Bugzilla 985, 986. Patch by Simon Arlott
422 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
423 Patch by Simon Arlott
425 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
426 MBX locking. Notification from Dan Rosenberg.
432 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
434 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
436 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
438 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
440 NM/04 Bugzilla 915: Flag broken perl installation during build.
446 TK/01 Added patch by Johannes Berg that expands the main option
447 "spamd_address" if it starts with a dollar sign.
449 TK/02 Write list of recipients to X-Envelope-Sender header when building
450 the mbox-format spool file for content scanning (suggested by Jakob
453 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
454 (http://www.dcc-servers.net/) support via dccifd. Activated by
455 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
457 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
458 by Mark Daniel Reidel <mr@df.eu>.
460 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
461 When building exim an external PCRE library is now needed -
462 PCRE is a system library on the majority of modern systems.
463 See entry on PCRE_LIBS in EDITME file.
465 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
466 conversation. Added nologin parameter to request.
467 Patch contributed by Kirill Miazine.
469 TF/01 Do not log submission mode rewrites if they do not change the address.
471 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
473 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
474 log files in place. Contributed by Roberto Lima.
476 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
478 TF/03 Bugzilla 615: When checking the local_parts router precondition
479 after a local_part_suffix or local_part_prefix option, Exim now
480 does not use the address's named list lookup cache, since this
481 contains cached lookups for the whole local part.
483 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
484 Robert Millan. Documentation is in experimental-spec.txt.
486 TF/04 Bugzilla 668: Fix parallel build (make -j).
488 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
490 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
491 Patch provided by Jan Srzednicki.
493 TF/05 Leading white space used to be stripped from $spam_report which
494 wrecked the formatting. Now it is preserved.
496 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
497 that they are available at delivery time.
499 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
501 TF/08 TLS error reporting now respects the incoming_interface and
502 incoming_port log selectors.
504 TF/09 Produce a more useful error message if an SMTP transport's hosts
505 setting expands to an empty string.
507 NM/06 Bugzilla 744: EXPN did not work under TLS.
508 Patch provided by Phil Pennock.
510 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
511 Patch provided by Richard Godbee.
513 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
514 acl_smtp_notquit, added index entry.
516 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
517 Patch provided by Eugene Bujak.
519 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
520 accept(). Patch provided by Maxim Dounin.
522 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
523 Patch provided by Phil Pennock.
525 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
527 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
528 Patch provided by Brad "anomie" Jorsch.
530 NM/14 Bugzilla 622: Added timeout setting to transport filter.
531 Patch provided by Dean Brooks.
533 TK/05 Add native DKIM support (does not depend on external libraries).
535 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
536 Patch provided by Graeme Fowler.
538 NM/16 Bugzilla 851: Documentation example syntax fix.
540 NM/17 Changed NOTICE file to remove references to embedded PCRE.
542 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
545 NM/19 Bugzilla 745: TLS version reporting.
546 Patch provided by Phil Pennock.
548 NM/20 Bugzilla 167: bool: condition support.
549 Patch provided by Phil Pennock.
551 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
552 clients. Patch provided by Phil Pennock.
554 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
555 Patch provided by Brad "anomie" Jorsch.
557 NM/23 Bugzilla 687: Fix misparses in eximstats.
558 Patch provided by Heiko Schlittermann.
560 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
561 Patch provided by Heiko Schlittermann.
563 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
564 plus update to original patch.
566 NM/26 Bugzilla 799: Documentation correction for ratelimit.
568 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
569 Patch provided by David Brownlee.
571 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
573 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
575 NM/30 Bugzilla 888: TLS documentation bugfixes.
577 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
579 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
580 Unlike the original bugzilla I have changed all shell scripts in src tree.
582 NM/33 Bugzilla 898: Transport filter timeout fix.
583 Patch by Todd Rinaldo.
585 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
586 Patch by Serge Demonchaux.
588 NM/35 Bugzilla 39: Base64 decode bug fixes.
589 Patch by Jakob Hirsch.
591 NM/36 Bugzilla 909: Correct connect() call in dcc code.
593 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
595 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
597 NM/39 Bugzilla 911: Fixed MakeLinks build script.
603 TK/01 Add preliminary DKIM support. Currently requires a forked version of
604 ALT-N's libdkim that I have put here:
605 http://duncanthrax.net/exim-experimental/
607 Note to Michael Haardt: I had to rename some vars in sieve.c. They
608 were called 'true' and it seems that C99 defines that as a reserved
609 keyword to be used with 'bool' variable types. That means you could
610 not include C99-style headers which use bools without triggering
611 build errors in sieve.c.
613 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
614 as mailq or other aliases. Changed the --help handling significantly
615 to do whats expected. exim_usage() emits usage/help information.
617 SC/01 Added the -bylocaldomain option to eximstats.
619 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
621 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
623 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
629 PH/01 Another patch from the Sieve maintainer.
631 PH/02 When an IPv6 address is converted to a string for single-key lookup
632 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
633 dots are used instead of colons so that keys in lsearch files need not
634 contain colons. This was done some time before quoting was made available
635 in lsearch files. However, iplsearch files do require colons in IPv6 keys
636 (notated using the quote facility) so as to distinguish them from IPv4
637 keys. This meant that lookups for IP addresses in host lists did not work
638 for iplsearch lookups.
640 This has been fixed by arranging for IPv6 addresses to be expressed with
641 colons if the lookup type is iplsearch. This is not incompatible, because
642 previously such lookups could never work.
644 The situation is now rather anomolous, since one *can* have colons in
645 ordinary lsearch keys. However, making the change in all cases is
646 incompatible and would probably break a number of configurations.
648 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
651 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
652 conversion specification without a maximum field width, thereby enabling
653 a rogue spamd server to cause a buffer overflow. While nobody in their
654 right mind would setup Exim to query an untrusted spamd server, an
655 attacker that gains access to a server running spamd could potentially
656 exploit this vulnerability to run arbitrary code as the Exim user.
658 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
659 $primary_hostname instead of what libspf2 thinks the hosts name is.
661 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
662 a directory entry by the name of the lookup key. Previously, if a
663 symlink pointed to a non-existing file or a file in a directory that
664 Exim lacked permissions to read, a lookup for a key matching that
665 symlink would fail. Now it is enough that a matching directory entry
666 exists, symlink or not. (Bugzilla 503.)
668 PH/03 The body_linecount and body_zerocount variables are now exported in the
671 PH/04 Added the $dnslist_matched variable.
673 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
674 This means they are set thereafter only if the connection becomes
677 PH/06 Added the client_condition to authenticators so that some can be skipped
678 by clients under certain conditions.
680 PH/07 The error message for a badly-placed control=no_multiline_responses left
681 "_responses" off the end of the name.
683 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
685 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
686 (without spaces) instead of just copying the configuration text.
688 PH/10 Added the /noupdate option to the ratelimit ACL condition.
690 PH/11 Added $max_received_linelength.
692 PH/12 Added +ignore_defer and +include_defer to host lists.
694 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
695 way in which PCRE > 7.0 is built.
697 PH/14 Implemented queue_only_load_latch.
699 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
700 MAIL command. The effect was to mangle the value on 64-bit systems.
702 PH/16 Another patch from the Sieve maintainer.
704 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
706 PH/18 If a system quota error occurred while trying to create the file for
707 a maildir delivery, the message "Mailbox is full" was not appended to the
708 bounce if the delivery eventually timed out. Change 4.67/27 below applied
709 only to a quota excession during the actual writing of the file.
711 PH/19 It seems that peer DN values may contain newlines (and other non-printing
712 characters?) which causes problems in log lines. The DN values are now
713 passed through string_printing() before being added to log lines.
715 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
716 and InterBase are left for another time.)
718 PH/21 Added message_body_newlines option.
720 PH/22 Guard against possible overflow in moan_check_errorcopy().
722 PH/23 POSIX allows open() to be a macro; guard against that.
724 PH/24 If the recipient of an error message contained an @ in the local part
725 (suitably quoted, of course), incorrect values were put in $domain and
726 $local_part during the evaluation of errors_copy.
732 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
733 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
736 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
737 issue a MAIL command.
739 PH/02 In an ACL statement such as
741 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
743 if a client was not listed at all, or was listed with a value other than
744 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
745 the condition was not true (as it should be), so access was not denied.
746 The bug was that the ! inversion was incorrectly passed on to the second
747 item. This has been fixed.
749 PH/03 Added additional dnslists conditions == and =& which are different from
750 = and & when the dns lookup returns more than one IP address.
752 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
753 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
755 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
756 FSYNC, which compiles an option called disable_fsync that allows for
757 bypassing fsync(). The documentation is heavily laced with warnings.
759 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
761 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
762 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
763 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
764 including adding "make clean"; (3) Added -fPIC when compiling the test
765 dynamically loaded module, to get rid of a warning.
767 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
768 message fails, move_frozen_messages = true and ignore_bounce_errors_after
769 = 0s. The bug is otherwise harmless.
771 PH/07 There was a bug in the dovecot authenticator such that the value of
772 $auth1 could be overwritten, and so not correctly preserved, after a
773 successful authentication. This usually meant that the value preserved by
774 the server_setid option was incorrect.
776 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
778 PH/09 Installed PCRE release 7.0.
780 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
781 run for batched SMTP input. It is now run at the start of every message
782 in the batch. While fixing this I discovered that the process information
783 (output by running exiwhat) was not always getting set for -bs and -bS
784 input. This is fixed, and it now also says "batched" for BSMTP.
786 PH/11 Added control=no_pipelining.
788 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
789 patch, slightly modified), and move the expansion of helo_data till after
790 the connection is made in the smtp transport (so it can use these
793 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
795 PH/14 Added log_selector = +pid.
797 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
799 PH/16 Add ${if forany and ${if forall.
801 PH/17 Added dsn_from option to vary the From: line in DSNs.
803 PH/18 Flush SMTP output before performing a callout, unless control =
804 no_callout_flush is set.
806 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
807 was true (the default) a successful delivery failed to delete the retry
808 item, thus causing premature timeout of the address. The bug is now
811 PH/20 Added hosts_avoid_pipelining to the smtp transport.
813 PH/21 Long custom messages for fakedefer and fakereject are now split up
814 into multiline reponses in the same way that messages for "deny" and
815 other ACL rejections are.
817 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
818 with slight modification.
820 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
821 draft, changing the syntax and factoring some duplicate code".
823 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
824 for deliveries of the second and subsequent messages over the same SMTP
827 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
828 ${reduce, with only minor "tidies".
830 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
832 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
833 expansion side effects.
835 PH/27 When a message times out after an over-quota error from an Exim-imposed
836 quota, the bounce message says "mailbox is full". This message was not
837 being given when it was a system quota that was exceeded. It now should
840 MH/03 Made $recipients available in local_scan(). local_scan() already has
841 better access to the recipient list through recipients_list[], but
842 $recipients can be useful in postmaster-provided expansion strings.
844 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
845 in the case of a MAIL command with additional options following the
846 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
847 were accidentally chopped off.
849 PH/29 SMTP synchronization checks are implemented when a command is read -
850 there is a check that no more input is waiting when there shouldn't be
851 any. However, for some commands, a delay in an ACL can mean that it is
852 some time before the response is written. In this time, more input might
853 arrive, invalidly. So now there are extra checks after an ACL has run for
854 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
855 pipelining has not been advertised.
857 PH/30 MH's patch to allow iscntrl() characters to be list separators.
859 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
860 returned in the SMTP response when smtp_return_error_details was false.
863 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
864 instead of the C I/O that was originally supplied, because problems were
867 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
868 Exim which did not show up earlier: it was assuming that a call to
869 SSL_CTX_set_info_callback() might give an error value. In fact, there is
870 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
871 was a macro that became an assignment, so it seemed to work. This has
872 changed to a proper function call with a void return, hence the compile
873 error. Exim's code has been fixed.
875 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
878 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
880 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
882 PH/37 If a message is not accepted after it has had an id assigned (e.g.
883 because it turns out to be too big or there is a timeout) there is no
884 "Completed" line in the log. When some messages of this type were
885 selected by exigrep, they were listed as "not completed". Others were
886 picked up by some special patterns. I have improved the selection
887 criteria to be more general.
889 PH/38 The host_find_failed option in the manualroute router can now be set
890 to "ignore", to completely ignore a host whose IP address cannot be
891 found. If all hosts are ignored, the behaviour is controlled by the new
892 host_all_ignored option.
894 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
895 homing or because of multiple MX records with /mx) generated more than
896 one IP address, and the following item turned out to be the local host,
897 all the secondary addresses of the first item were incorrectly removed
898 from the list, along with the local host and any following hosts (which
899 is what is supposed to happen).
901 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
902 whoever called Exim into the -H file. In the case of the daemon it was
903 behaving confusingly. When first started, it used values for whoever
904 started the daemon, but after a SIGHUP it used the Exim user (because it
905 calls itself on a restart). I have changed the code so that it now always
908 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
909 message are rejected with the same error (e.g. no authentication or bad
910 sender address), and a DATA command is nevertheless sent (as can happen
911 with PIPELINING or a stupid MUA), the error message that was given to the
912 RCPT commands is included in the rejection of the DATA command. This is
913 intended to be helpful for MUAs that show only the final error to their
916 PH/42 Another patch from the Sieve maintainer.
918 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
919 Eximstats - Fixed some broken HTML links and added missing column headers
921 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
922 columns for Rejects, Temp Rejects, Ham, and Spam rows.
924 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
926 PH/43 Yet another patch from the Sieve maintainer.
928 PH/44 I found a way to check for a TCP/IP connection going away before sending
929 the response to the final '.' that terminates a message, but only in the
930 case where the client has not sent further data following the '.'
931 (unfortunately, this is allowed). However, in many cases there won't be
932 any further data because there won't be any more messages to send. A call
933 to select() can be used: if it shows that the input is "ready", there is
934 either input waiting, or the socket has been closed. An attempt to read
935 the next input character can distinguish the two cases. Previously, Exim
936 would have sent an OK response which the client would never have see.
937 This could lead to message repetition. This fix should cure that, at
938 least in a lot of common cases.
940 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
941 advertised in response to EHLO.
947 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
948 fixed by 4.65/MH/01 (is this a record?) are fixed:
950 (i) An empty string was always treated as zero by the numeric comparison
951 operators. This behaviour has been restored.
953 (ii) It is documented that the numeric comparison operators always treat
954 their arguments as decimal numbers. This was broken in that numbers
955 starting with 0 were being interpreted as octal.
957 While fixing these problems I realized that there was another issue that
958 hadn't been noticed. Values of message_size_limit (both the global option
959 and the transport option) were treated as octal if they started with 0.
960 The documentation was vague. These values are now always treated as
961 decimal, and I will make that clear in the documentation.
967 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
968 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
971 MH/01 Don't check that the operands of numeric comparison operators are
972 integers when their expansion is in "skipping" mode (fixes bug
973 introduced by 4.64-PH/07).
975 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
976 child addresses, Exim now panics and dies. Previously, because the count
977 is held in a short int, deliveries were likely to be lost. As such a
978 large number of recipients for a single message is ridiculous
979 (performance will be very, very poor), I have chosen to impose a limit
980 rather than extend the field.
986 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
987 leftover -K file (the existence of which was triggered by #402).
988 While we were at it, introduced process PID as part of the -K
989 filename. This should rule out race conditions when creating
992 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
993 processing considerably. Previous code took too long for large mails,
994 triggering a timeout which in turn triggers #401.
996 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
997 in the DK code in transports.c. sendfile() is not really portable,
998 hence the _LINUX specificness.
1000 TF/01 In the add_headers option to the mail command in an Exim filter,
1001 there was a bug that Exim would claim a syntax error in any
1002 header after the first one which had an odd number of characters
1005 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1006 callout verification, Exim cached a "reject" for the entire domain. This
1007 is correct for most verifications, but it is not correct for a recipient
1008 verification with use_sender or use_postmaster set, because in that case
1009 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1010 case of MAIL FROM:<> rejection from other early rejections (e.g.
1011 rejection of HELO). When verifying a recipient using a non-null MAIL
1012 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1013 Whatever the result of the callout, the value of the domain cache is
1014 left unchanged (for any other kind of callout, getting as far as trying
1015 RCPT means that the domain itself is ok).
1017 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1020 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1021 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1022 session, the daemon ignores ECONNECT errors and logs others; it now
1023 ignores EPIPE as well.
1025 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1026 (quoted-printable decoding).
1028 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1029 later the small subsequent patch to fix an introduced bug.
1031 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1033 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1035 PH/08 An error is now given if message_size_limit is specified negative.
1037 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1038 to be given (somewhat) arbitrary names.
1040 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1043 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1044 miscellaneous code fixes
1046 PH/10 Added the log_reject_target ACL modifier to specify where to log
1049 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1050 hostname. This is wrong, because it relates to the incoming message (and
1051 probably the interface on which it is arriving) and not to the outgoing
1052 callout (which could be using a different interface). This has been
1053 changed to use the value of the helo_data option from the smtp transport
1054 instead - this is what is used when a message is actually being sent. If
1055 there is no remote transport (possible with a router that sets up host
1056 addresses), $smtp_active_hostname is used.
1058 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1059 tweaks were necessary in order to get it to work (see also 21 below):
1060 (a) The code assumed that strncpy() returns a negative number on buffer
1061 overflow, which isn't the case. Replaced with Exim's string_format()
1063 (b) There were several signed/unsigned issues. I just did the minimum
1064 hacking in of casts. There is scope for a larger refactoring.
1065 (c) The code used strcasecmp() which is not a standard C function.
1066 Replaced with Exim's strcmpic() function.
1067 (d) The code set only $1; it now sets $auth1 as well.
1068 (e) A simple test gave the error "authentication client didn't specify
1069 service in request". It would seem that Dovecot has changed its
1070 interface. Fortunately there's a specification; I followed it and
1071 changed what the client sends and it appears to be working now.
1073 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1076 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1077 address A is aliased to B and C, where B exists and C does not. Without
1078 -v the output is "A verified" because verification stops after a
1079 successful redirection if more than one address is generated. However,
1080 with -v the child addresses are also verified. Exim was outputting "A
1081 failed to verify" and then showing the successful verification for C,
1082 with its parentage. It now outputs "B failed to verify", showing B's
1083 parentage before showing the successful verification of C.
1085 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1086 look up a TXT record in a specific list after matching in a combined
1089 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1090 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1091 they consult the DNS. I had assumed they would set it the way they
1092 wanted; and indeed my experiments on Linux seem to show that in some
1093 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1094 To be on the safe side, however, I have now made the interface to
1095 host_find_byname() similar to host_find_bydns(), with an argument
1096 containing the DNS resolver options. The host_find_byname() function now
1097 sets these options at its start, just as host_find_bydns() does. The smtp
1098 transport options dns_qualify_single and dns_search_parents are passed to
1099 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1100 of host_find_byname() use the default settings of RES_DEFNAMES
1101 (qualify_single) but not RES_DNSRCH (search_parents).
1103 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1104 spool_read_header() do less string testing, by means of a preliminary
1105 switch on the second character of optional "-foo" lines. (This is
1106 overdue, caused by the large number of possibilities that now exist.
1107 Originally there were few.) While I was there, I also converted the
1108 str(n)cmp tests so they don't re-test the leading "-" and the first
1109 character, in the hope this might squeeze out yet more improvement.
1111 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1112 flag allowing group syntax was set by the header_syntax check but not
1113 turned off, possible causing trouble later; (2) The flag was not being
1114 set at all for the header_verify test, causing "group"-style headers to
1115 be rejected. I have now set it in this case, and also caused header_
1116 verify to ignore an empty address taken from a group. While doing this, I
1117 came across some other cases where the code for allowing group syntax
1118 while scanning a header line wasn't quite right (mostly, not resetting
1119 the flag correctly in the right place). These bugs could have caused
1120 trouble for malformed header lines. I hope it is now all correct.
1122 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1123 with the "reply" argument non-NULL. The code, however (which originally
1124 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1125 but it didn't always do it. This confused somebody who was copying the
1126 code for some other use. I have removed all the tests.
1128 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1129 feature that was used to support insecure browsers during the U.S. crypto
1130 embargo. It requires special client support, and Exim is probably the
1131 only MTA that supported it -- and would never use it because real RSA is
1132 always available. This code has been removed, because it had the bad
1133 effect of slowing Exim down by computing (never used) parameters for the
1134 RSA_EXPORT functionality.
1136 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1137 authenticator to fail if there's a tab character in the incoming data
1138 (there should never be unless someone is messing about, as it's supposed
1139 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1140 if the connection is using TLS or if the remote IP is the same as the
1141 local IP, and the "valid-client-cert option" if a client certificate has
1144 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1145 authenticators. This can be used for authorization after authentication
1146 succeeds. (In the case of plaintext, it servers for both authentication
1149 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1150 if any retry times were supplied.
1152 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1153 connection, where there is no client IP address to check. In this
1154 situation, the verify now always succeeds.
1156 PH/25 Applied John Jetmore's -Mset patch.
1158 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1160 PH/27 In a string expansion for a processed (not raw) header when multiple
1161 headers of the same name were present, leading whitespace was being
1162 removed from all of them, but trailing whitespace was being removed only
1163 from the last one. Now trailing whitespace is removed from each header
1164 before concatenation. Completely empty headers in a concatenation (as
1165 before) are ignored.
1167 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1168 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1170 PH/29 [Removed. This was a change that I later backed out, and forgot to
1171 correct the ChangeLog entry (that I had efficiently created) before
1172 committing the later change.]
1174 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1175 address errors (4xx response to RCPT) over the same connection as other
1176 messages routed to the same hosts. Such deliveries are always "forced",
1177 so retry times are not inspected. This resulted in far too many retries
1178 for the affected addresses. The effect occurred only when there were more
1179 hosts than the hosts_max_try setting in the smtp transport when it had
1180 the 4xx errors. Those hosts that it had tried were not added to the list
1181 of hosts for which the message was waiting, so if all were tried, there
1182 was no problem. Two fixes have been applied:
1184 (i) If there are any address or message errors in an SMTP delivery, none
1185 of the hosts (tried or untried) are now added to the list of hosts
1186 for which the message is waiting, so the message should not be a
1187 candidate for sending over the same connection that was used for a
1188 successful delivery of some other message. This seems entirely
1189 reasonable: after all the message is NOT "waiting for some host".
1190 This is so "obvious" that I'm not sure why it wasn't done
1191 previously. Hope I haven't missed anything, but it can't do any
1192 harm, as the worst effect is to miss an optimization.
1194 (ii) If, despite (i), such a delivery is accidentally attempted, the
1195 routing retry time is respected, so at least it doesn't keep
1196 hammering the server.
1198 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1199 in ${readsocket because some servers need this prod.
1201 PH/32 Added some extra debug output when updating a wait-xxx database.
1203 PH/33 The hint "could be header name not terminated by colon", which has been
1204 given for certain expansion errors for a long time, was not being given
1205 for the ${if def:h_colon_omitted{... case.
1207 PH/34 The spec says: "With one important exception, whenever a domain list is
1208 being scanned, $domain contains the subject domain." There was at least
1209 one case where this was not true.
1211 PH/35 The error "getsockname() failed: connection reset by peer" was being
1212 written to the panic log as well as the main log, but it isn't really
1213 panic-worthy as it just means the connection died rather early on. I have
1214 removed the panic log writing for the ECONNRESET error when getsockname()
1217 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1218 runs only) independently of the message's sender address. This meant
1219 that, if the 4xx error was in fact related to the sender, a different
1220 message to the same recipient with a different sender could confuse
1221 things. In particualar, this can happen when sending to a greylisting
1222 server, but other circumstances could also provoke similar problems.
1223 I have changed the default so that the retry time for these errors is now
1224 based a combination of the sender and recipient addresses. This change
1225 can be overridden by setting address_retry_include_sender=false in the
1228 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1229 remote server are returned as part of bounce messages. This was not
1230 happening for LMTP over a pipe (the lmtp transport), but now it is the
1231 same for both kinds of LMTP.
1233 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1234 in header lines that were in fact CNAMEs. This is no longer the case.
1236 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1237 and queue runs started by the daemon processed all messages. This has
1238 been fixed so that -R and -S can now usefully be given with -q<time>.
1240 PH/40 Import PCRE release 6.7 (fixes some bugs).
1242 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1244 PH/42 Give an error if -q is specified more than once.
1246 PH/43 Renamed the variables $interface_address and $interface_port as
1247 $received_ip_address and $received_port, to make it clear that these
1248 values apply to message reception, and not to the outgoing interface when
1249 a message is delivered. (The old names remain recognized, of course.)
1251 PH/44 There was no timeout on the connect() call when using a Unix domain
1252 socket in the ${readsocket expansion. There now is.
1254 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1255 be meaningful with "accept".
1257 SC/01 Eximstats V1.43
1258 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1260 SC/02 Eximstats V1.44
1261 Use a glob alias rather than an array ref in the generated
1262 parser. This improves both readability and performance.
1264 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1265 Collect SpamAssassin and rejection statistics.
1266 Don't display local sender or destination tables unless
1267 there is data to show.
1268 Added average volumes into the top table text output.
1270 SC/04 Eximstats V1.46
1271 Collect data on the number of addresses (recipients)
1272 as well as the number of messages.
1274 SC/05 Eximstats V1.47
1275 Added 'Message too big' to the list of mail rejection
1276 reasons (thanks to Marco Gaiarin).
1278 SC/06 Eximstats V1.48
1279 Mainlog lines which have GMT offsets and are too short to
1280 have a flag are now skipped.
1282 SC/07 Eximstats V1.49 (Alain Williams)
1283 Added the -emptyok flag.
1285 SC/08 Eximstats V1.50
1286 Fixes for obtaining the IP address from reject messages.
1288 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1289 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1290 whitesspace changes from 4.64-PH/27
1292 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1295 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1296 are found, allow negative numbers in numeric criteria)
1298 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1300 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1301 to match changes made in 4.64-PH/43
1303 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1305 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1306 "domain" is actually an IP address, and force a failure. This locks out
1307 those revolvers/nameservers that support "A-for-A" lookups, in
1308 contravention of the specifications.
1310 PH/48 When a host name was looked up from an IP address, and the subsequent
1311 forward lookup of the name timed out, the host name was left in
1312 $sender_host_name, contrary to the specification.
1314 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1315 restricted to single-key lookups, Exim was not diagnosing an error if
1316 * or *@ was used with a query-style lookup.
1318 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1320 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1321 long ago, but noone interested enough thought of it. Let's just say that
1322 the "1.1" means that there are some new functions that weren't there at
1323 some point in the past.
1325 PH/51 Error processing for expansion failure of helo_data from an smtp
1326 transport during callout processing was broken.
1328 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1329 tested/used via the -bh/-bhc/-bs options.
1331 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1332 bug, fixed in subsequent PCRE releases).
1334 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1335 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1337 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1343 SC/01 Use a glob alias rather than an array ref in eximstats generated
1344 parser. This improves both readability and performance.
1346 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1347 Don't display local sender or destination tables in eximstats unless
1348 there is data to show.
1349 Added average volumes into the eximstats top table text output.
1351 SC/03 Collect data on the number of addresses (recipients) as well
1352 as the number of messages in eximstats.
1354 TF/01 Correct an error in the documentation for the redirect router. Exim
1355 does (usually) call initgroups() when daemonizing.
1357 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1358 with consistent privilege compared to when running as a daemon.
1360 TF/03 Note in the spec that $authenticated_id is not set for local
1361 submissions from trusted users.
1363 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1364 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1366 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1367 by adding some example configuration directives to the default
1368 configuration file. A little bit of work is required to uncomment the
1369 directives and define how usernames and passwords are checked, but
1370 there is now a framework to start from.
1372 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1373 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1374 without this. I don't know how relevant this is to other LDAP libraries.
1376 PH/02 Add the verb name to the "unknown ACL verb" error.
1378 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1380 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1382 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1383 directory not expanded when it should be if an expanded home directory
1384 was set for the address (which is overridden by the transport).
1386 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1389 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1390 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1391 because it is too late at that time, and has no effect.
1393 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1394 security issue with \' (bugzilla #107). I could not use the
1395 PQescapeStringConn() function, because it needs a PGconn value as one of
1398 PH/08 When testing addresses using -bt, indicate those final addresses that
1399 are duplicates that would not cause an additional delivery. At least one
1400 person was confused, thinking that -bt output corresponded to deliveries.
1401 (Suppressing duplicates isn't a good idea as you lose the information
1402 about possibly different redirections that led to the duplicates.)
1404 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1405 systems where poll() doesn't work, in particular OS X.
1407 PH/10 Added more information to debugging output for retry time not reached.
1409 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1410 operations in malware.c.
1412 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1415 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1416 syslog_duplication set false, log lines that would normally be written
1417 both the the main log and to the reject log were not written to syslog at
1420 PH/14 In the default configuration, change the use of "message" in ACL warn
1421 statements to "add_header".
1423 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1424 not followed by a command (e.g. "seen endif").
1426 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1427 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1430 PH/17 Added extra conditions to the default value of delay_warning_condition
1434 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1435 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1436 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1439 The Auto-Submitted: and various List- headers are standardised, whereas I
1440 don't think Precedence: ever was.
1442 PH/18 Refactored debugging code in route_finduser() to show more information,
1443 in particular, the error code if getpwnam() issues one.
1445 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1446 This is apparently needed in addition to the PH/07 change above to avoid
1447 any possible encoding problems.
1449 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1450 but not after initializing Perl.
1452 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1453 output them only if debugging. By default they are written stderr,
1454 apparently, which is not desirable.
1456 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1459 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1462 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1464 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1465 authenticated or an ident call has been made. Suppress the default
1466 values for $authenticated_id and $authenticated_sender (but permit -oMai
1467 and -oMas) when testing with -bh.
1469 PH/24 Re-jigged the order of the tests in the default configuration so that the
1470 tests for valid domains and recipients precede the DNS black list and CSA
1471 tests, on the grounds that those ones are more expensive.
1473 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1474 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1475 This bug exists in every version of Exim that I still have, right back to
1478 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1479 However, an attempt to turn on case-sensitivity in a regex key by
1480 including (?-i) didn't work because the subject string was already
1481 lowercased, and the effects were non-intuitive. It turns out that a
1482 one-line patch can be used to allow (?-i) to work as expected.
1488 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1489 other effects) broke the use of negated acl sub-conditions.
1491 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1494 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1495 "Deny" causes Exim to reject the incoming connection with a 554 error.
1496 Unfortunately, if there is a major crisis, such as a disk failure,
1497 tcp-wrappers gives "deny", whereas what one would like would be some
1498 kind of temporary error. A kludge has been added to help with this.
1499 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1500 554 error is used if errno is still zero or contains ENOENT (which occurs
1501 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1504 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1506 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1507 errors. Otherwise a message that provokes a temporary error (when other
1508 messages do not) can cause a whole host to time out.
1510 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1511 addresses were routed directly to files or pipes from a redirect router.
1512 File deliveries just didn't batch; pipe deliveries might have suffered
1515 PH/06 A failure to get a lock for a hints database would erroneously always say
1516 "Failed to get write lock", even when it was really a read lock.
1518 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1519 of 0600. This has been changed to use the value of the lockfile_mode
1520 option (which defaults to 0600).
1522 PH/08 Applied small patch from the Sieve maintainer.
1524 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1525 folder from quota calculations, a direct delivery into this folder messed
1526 up the contents of the maildirsize file. This was because the regex was
1527 used only to exclude .Trash (or whatever) when the size of the mailbox
1528 was calculated. There was no check that a delivery was happening into an
1529 excluded directory. This bug has been fixed by ignoring all quota
1530 processing for deliveries into excluded directories.
1532 PH/10 Added the maildirfolder_create_regex option to appendfile.
1538 PH/01 The code for finding all the local interface addresses on a FreeBSD
1539 system running IPv6 was broken. This may well have applied to all BSD
1540 systems, as well as to others that have similar system calls. The broken
1541 code found IPv4 interfaces correctly, but gave incorrect values for the
1542 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1543 that it would not match correctly against @[] and not recognize the IPv6
1546 PH/02 The ipliteral router was not recognizing addresses of the form user@
1547 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1549 PH/03 Added disable_ipv6.
1551 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1552 decoded form, because it is most often used to construct To: headers
1553 lines in autoreplies, and the decoded form may well be syntactically
1554 invalid. However, $reply_address has leading white space removed, and all
1555 newlines turned into spaces so that the autoreply transport does not
1558 PH/05 If group was specified without a user on a router, and no group or user
1559 was specified on a transport, the group from the router was ignored.
1561 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1562 for visible compile-time settings that can be used to change these
1563 numbers, for those that want even more. Backwards compatibility with old
1564 spool files has been maintained. However, going back to a previous Exim
1565 release will lost any variables that are in spool files.
1567 PH/07 Two small changes when running in the test harness: increase delay when
1568 passing a TCP/IP connection to a new process, in case the original
1569 process has to generate a bounce, and remove special handling of
1570 127.0.0.2 (sic), which is no longer necessary.
1572 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1573 be the same on different OS.
1575 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1578 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1579 whether --show-vars was specified or not
1581 JJ/02 exipick: Added support for new ACL variable spool format introduced
1584 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1585 syntactically invalid From: or Reply-to: line, and a filter used this to
1586 generate an autoreply, and therefore failed to obtain an address for the
1587 autoreply, Exim could try to deliver to a non-existent relative file
1588 name, causing unrelated and misleading errors. What now happens is that
1589 it logs this as a hard delivery error, but does not attempt to create a
1592 PH/11 The exinext utility has a -C option for testing purposes, but although
1593 the given file was scanned by exinext itself; it wasn't being passed on
1594 when Exim was called.
1596 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1597 an end-of-file indication when reading a command response.
1599 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1600 compiled. In many other places in Exim, IPv6 addresses are always
1601 recognized, so I have changed this. It also means that IPv4 domain
1602 literals of the form [IPV4:n.n.n.n] are now always recognized.
1604 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1605 used if the router is not running as root, for example, when verifying at
1606 ACL time, or when using -bh. The debugging output from this situation was
1607 non-existent - all you got was a failure to exec. I have made two
1610 (a) Failures to set uid/gid, the current directory, or a process leader
1611 in a subprocess such as that created by queryprogram now generate
1612 suitable debugging ouput when -d is set.
1614 (b) The queryprogram router detects when it is not running as root,
1615 outputs suitable debugging information if -d is set, and then runs
1616 the subprocess without attempting to change uid/gid.
1618 PH/15 Minor change to Makefile for building test_host (undocumented testing
1621 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1622 additional section of a DNS packet that returns MX or SRV records.
1623 Instead, it always explicitly searches for A/AAAA records. This avoids
1624 major problems that occur when a DNS server includes only records of one
1625 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1626 fixed another bug: if SRV records were looked up and the corresponding
1627 address records were *not* found in the additional section, the port
1628 values from the SRV records were lost.
1630 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1631 using the correct key (the original address) when searching the retry
1632 rules in order to find which one to use for generating the retry hint.
1634 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1635 adding the default one. Similarly, if it contains a Reply-To: header, the
1636 errors_reply_to option, if set, is not used.
1638 PH/19 When calculating a retry time, Exim used to measure the "time since
1639 failure" by looking at the "first failed" field in the retry record. Now
1640 it does not use this if it is later than than the arrival time of the
1641 message. Instead it uses the arrival time. This makes for better
1642 behaviour in cases where some deliveries succeed, thus re-setting the
1643 "first failed" field. An example is a quota failure for a huge message
1644 when small messages continue to be delivered. Without this change, the
1645 "time since failure" will always be short, possible causing more frequent
1646 delivery attempts for the huge message than are intended.
1647 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1649 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1650 $1, $2, $3) because the numerical variables can be reset during some
1651 expansion items (e.g. "match"), thereby losing the authentication data.
1653 PH/21 Make -bV show the size of off_t variables so that the test suite can
1654 decide whether to run tests for quotas > 2G.
1656 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1657 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1658 filecount value is greater than 2G or if a quota value is greater than 2G
1659 on a system where the size of off_t is not greater than 4, a panic error
1662 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1663 never match. The debug and -bh output now contains an explicit error
1664 message indicating a malformed IPv4 address or mask.
1666 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1667 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1668 PH/23 above applies.
1670 PH/25 Do not write to syslog when running in the test harness. The only
1671 occasion when this arises is a failure to open the main or panic logs
1672 (for which there is an explicit test).
1674 PH/26 Added the /no_tell option to "control=freeze".
1676 PH/27 If a host name lookup failed very early in a connection, for example, if
1677 the IP address matched host_lookup and the reverse lookup yielded a name
1678 that did not have a forward lookup, an error message of the form "no IP
1679 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1680 could be logged. Now it outputs the IP address instead of "NULL".
1682 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1683 allows the sender and the authenticated sender to be set when
1684 submitting a message from within Exim. Since child_open_exim() is
1685 documented for local_scan(), the new function should be too.
1687 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1688 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1689 results in an empty string is now treated as unset.
1691 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1693 PH/31 Added sender_verify_fail logging option.
1695 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1696 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1697 tidied the source and removed it altogether.
1699 PH/33 When a queue run was abandoned because the load average was too high, a
1700 log line was always written; now it is written only if the queue_run log
1701 selector is set. In addition, the log line for abandonment now contains
1702 information about the queue run such as the pid. This is always present
1703 in "start" and "stop" lines but was omitted from the "abandon" line.
1705 PH/34 Omit spaces between a header name and the colon in the error message that
1706 is given when verify = headers_syntax fails (if there are lots of them,
1707 the message gets confusing).
1709 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1710 names, as there are now some PTR records that contain slashes. This check
1711 is only to protect against broken name servers that fall over on strange
1712 characters, so the fact that it applies to all lookups doesn't matter.
1714 PH/36 Now that the new test suite is complete, we can remove some of the
1715 special code in Exim that was needed for the old test suite. For example,
1716 sorting DNS records because real resolvers return them in an arbitrary
1717 order. The new test suite's fake resolver always returns records in the
1720 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1721 bounces) except when queue_only is set, to avoid logging races between
1722 the different processes.
1724 PH/38 Panic-die if .include specifies a non-absolute path.
1726 PH/39 A tweak to the "H" retry rule from its user.
1728 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1729 a label. They prevented compilation on older perls.
1731 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1732 a warning to be raised on newish perls.
1734 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1735 on queue. Changes to match documented behaviour of showing count of
1736 messages matching specified criteria.
1738 PH/40 Changed the default ident timeout from 30s to 5s.
1740 PH/41 Added support for the use of login_cap features, on those BSD systems
1741 that have them, for controlling the resources used by pipe deliveries.
1743 PH/42 The content-scanning code uses fopen() to create files in which to put
1744 message data. Previously it was not paying any attention to the mode of
1745 the files. Exim runs with umask(0) because the rest of the code creates
1746 files with open(), and sets the required mode explicitly. Thus, these
1747 files were ending up world-writeable. This was not a big issue, because,
1748 being within the spool directory, they were not world-accessible. I have
1749 created a function called modefopen, which takes an additional mode
1750 argument. It sets umask(777), creates the file, chmods it to the required
1751 mode, then resets the umask. All the relevant calls to fopen() in the
1752 content scanning code have been changed to use this function.
1754 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1755 to 24 hours. This avoids potential overflow problems when processing G
1756 and H retry rules. I suspect nobody ever tinkers with this value.
1758 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1760 PH/45 When the plaintext authenticator is running as a client, the server's
1761 challenges are checked to ensure they are valid base64 strings. By
1762 default, the authentication attempt is cancelled if an invalid string is
1763 received. Setting client_ignore_invalid_base64 true ignores these errors.
1764 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1765 they are received. Thus, the responses can be made to depend on the
1766 challenges. If an invalid string is ignored, an empty string is placed in
1769 PH/46 Messages that are created by the autoreply transport now contains a
1770 References: header, in accordance with RFCs 2822 and 3834.
1772 PH/47 Added authenticated_sender_force to the smtp transport.
1774 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1776 PH/49 Installed latest patch from the Sieve maintainer.
1778 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1779 was also set, the appendfile transport was unnecessarily scanning a
1780 directory of message files (e.g. for maildir delivery) to find the count
1781 of files (along with the size), even though it did not need this
1782 information. It now does the scan only if it needs to find either the
1783 size of the count of files.
1785 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1787 PH/52 Two bugs concerned with error handling when the smtp transport is
1790 (i) Exim was not creating retry information for temporary errors given
1791 for individual recipients after the DATA command when the smtp transport
1792 was used in LMTP mode. This meant that they could be retried too
1793 frequently, and not timed out correctly.
1795 (ii) Exim was setting the flag that allows error details to be returned
1796 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1797 recipients that were returned after the DATA command.
1799 PH/53 This is related to PH/52, but is more general: for any failing address,
1800 when detailed error information was permitted to be returned to the
1801 sender, but the error was temporary, then after the final timeout, only
1802 "retry timeout exceeded" was returned. Now it returns the full error as
1803 well as "retry timeout exceeded".
1805 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1806 do this, and (what is worse) MTAs that accept it.
1808 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1809 will now be deprecated.
1811 PH/56 New os.c-cygwin from the Cygwin maintainer.
1813 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1814 formats (previously only available in exim formats via -bpr, -bpru,
1815 and -bpra. Now also available in native and exiqgrep formats)
1817 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1818 with very large, slow to parse queues
1820 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1822 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1824 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1825 responses to authentication challenges, though it was showing the
1826 challenges; (ii) I've removed the CR characters from the debug output for
1829 PH/58 Allow for the insertion of a newline as well as a space when a string
1830 is turned into more than one encoded-word during RFC 2047 encoding. The
1831 Sieve code now uses this.
1833 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1834 data_4xx, lost_connection, tls_required.
1836 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1837 message was being sent as an SMTP response.
1839 PH/61 Add -l and -k options to exicyclog.
1841 PH/62 When verifying, if an address was redirected to one new address, so that
1842 verification continued, and the new address failed or deferred after
1843 having set something in $address_data, the value of $address_data was not
1844 passed back to the ACL. This was different to the case when no
1845 redirection occurred. The value is now passed back in both cases.
1847 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1848 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1849 use login_cap.h, so on its own it isn't the distinguishing feature. The
1850 new name refers directly to the setclassresources() function.
1852 PH/65 Added configuration files for NetBSD3.
1854 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1856 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1857 is preferred over IPv4.
1859 PH/68 The bounce_return_message and bounce_return_body options were not being
1860 honoured for bounces generated during the reception of non-SMTP messages.
1861 In particular, this applied to messages rejected by the ACL. This bug has
1862 been fixed. However, if bounce_return_message is true and bounce_return_
1863 body is false, the headers that are returned for a non-SMTP message
1864 include only those that have been read before the error was detected.
1865 (In the case of an ACL rejection, they have all been read.)
1867 PH/69 The HTML version of the specification is now built in a directory called
1868 spec_html instead of spec.html, because the latter looks like a path with
1869 a MIME-type, and this confuses some software.
1871 PH/70 Catch two compiler warnings in sieve.c.
1873 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1874 function verify_get_ident() calls ip_connect() to connect a socket, but
1875 if the "connect()" function timed out, ip_connect() used to close the
1876 socket. However, verify_get_ident() also closes the socket later, and in
1877 between Exim writes to the log, which may get opened at this point. When
1878 the socket was closed in ip_connect(), the log could get the same file
1879 descriptor number as the socket. This naturally causes chaos. The fix is
1880 not to close the socket in ip_connect(); the socket should be closed by
1881 the function that creates it. There was only one place in the code where
1882 this was missing, in the iplookup router, which I don't think anybody now
1883 uses, but I've fixed it anyway.
1885 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1886 well as to direct DNS lookups. Otherwise the handling of names in host
1887 lists is inconsistent and therefore confusing.
1893 PH/01 Two changes to the default runtime configuration:
1895 (1) Move the checks for relay_from_hosts and authenticated clients from
1896 after to before the (commented out) DNS black list checks.
1898 (2) Add control=submission to the relay_from_hosts and authenticated
1899 clients checks, on the grounds that messages accepted by these
1900 statements are most likely to be submissions.
1902 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1904 (1) Generate an error if the third argument for the ${prvs expansion is
1907 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1910 (3) Reset the variables that are obtained from the first argument of
1911 ${prvscheck and used in the second argument before leaving the code,
1912 because their memory is reclaimed, so using them afterwards may do
1915 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1916 one (it's much easier than Tom thought :-).
1918 (5) Because of (4), we can now allow for the use of $prvscheck_result
1919 inside the third argument.
1921 PH/03 For some reason, the default setting of PATH when running a command from
1922 a pipe transport was just "/usr/bin". I have changed it to
1925 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1926 anything to be listed in the output from -bV.
1928 PH/05 When a filter generated an autoreply, the entire To: header line was
1929 quoted in the delivery log line, like this:
1931 => >A.N.Other <ano@some.domain> <original@ddress> ...
1933 This has been changed so that it extracts the operative address. There
1934 may be more than one such address. If so, they are comma-separated, like
1937 => >ano@some.domain,ona@other.domain <original@ddress> ...
1939 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1940 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1941 not being looked up in the rDNS to get a host name, Exim was showing the
1942 IP address twice in Received: lines, even though the IP addresses were
1943 identical. For example:
1945 Received: from [1.2.3.4] (helo=[1.2.3.4])
1947 However, if the real host name was known, it was omitting the HELO data
1948 if it matched the actual IP address. This has been tidied up so that it
1949 doesn't show the same IP address twice.
1951 PH/07 When both +timestamp and +memory debugging was on, the value given by
1952 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1953 called by the debug printing, thereby overwriting the timestamp buffer.
1954 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1956 PH/08 When the original message was included in an autoreply transport, it
1957 always said "this is a copy of the message, including all the headers",
1958 even if body_only or headers_only was set. It now gives an appropriate
1961 PH/09 Applied a patch from the Sieve maintainer which:
1963 o fixes some comments
1964 o adds the (disabled) notify extension core
1965 o adds some debug output for the result of if/elsif tests
1966 o points to the current vacation draft in the documentation
1967 and documents the missing references header update
1971 o fixes a bug in processing the envelope test (when testing
1972 multiple envelope elements, the last element determinted the
1975 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1976 Electronic Mail") by including:
1978 Auto-submitted: auto-generated
1980 in the messages that it generates (bounce messages and others, such as
1981 warnings). In the case of bounce messages for non-SMTP mesages, there was
1982 also a typo: it was using "Auto_submitted" (underscore instead of
1983 hyphen). Since every message generated by Exim is necessarily in response
1984 to another message, thes have all been changed to:
1986 Auto-Submitted: auto-replied
1988 in accordance with these statements in the RFC:
1990 The auto-replied keyword:
1992 - SHOULD be used on messages sent in direct response to another
1993 message by an automatic process,
1995 - MUST NOT be used on manually-generated messages,
1997 - MAY be used on Delivery Status Notifications (DSNs) and Message
1998 Disposition Notifications (MDNs),
2000 - MUST NOT be used on messages generated by automatic or periodic
2001 processes, except for messages which are automatic responses to
2004 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2005 to the default Received: header definition.
2007 PH/12 Added log selector acl_warn_skipped (default on).
2009 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2010 variables because (a) they are in the wrong storage pool and (b) even if
2011 they were copied, it wouldn't work properly because of the caching.
2013 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2014 checking when decoding. Apparently there are clients that generate
2015 overlong encoded strings. Why am I not surprised?
2017 PH/15 If the first argument of "${if match_address" was not empty, but did not
2018 contain an "@" character, Exim crashed. Now it writes a panic log message
2019 and treats the condition as false.
2021 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2023 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2024 extension "envelope-auth". The code is finished and in agreement with
2025 other implementations, but there is no documentation so far and in fact,
2026 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2027 not changing the active code.
2029 Print executed "if" and "elsif" statements when debugging is used. This
2030 helps a great deal to understand what a filter does.
2032 Document more things not specified clearly in RFC3028. I had all this
2033 sorted out, when out of a sudden new issues came to my mind. Oops."
2035 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2038 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2039 canonical form (as documented). However, after a host name lookup from
2040 the IP address, check_host() was doing a simple string comparison with
2041 addresses acquired from the DNS when checking that the found name did
2042 have the original IP as one of its addresses. Since any found IPv6
2043 addresses are likely to be in abbreviated form, the comparison could
2044 fail. Luckily, there already exists a function for doing the comparison
2045 by converting both addresses to binary, so now that is used instead of
2046 the text comparison.
2048 PH/20 There was another similar case to PH/19, when a complete host name was
2049 given in a host list; looking up its IP address could give an abbreviated
2050 form, whereas the current host's name might or might not be abbreviated.
2051 The same fix has been applied.
2057 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2058 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2061 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2062 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2064 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2066 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2067 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2068 Also, for "Auto-Submitted", treat anything other than "no" as
2069 non-personal, in accordance with RFC 3834. (Previously it treated
2070 anything starting "auto-" as non-personal.)
2072 TF/01 The control=submission/name=... option had a problem with syntax
2073 errors if the name included a slash character. The /name= option
2074 now slurps the rest of the string, so it can include any characters
2075 but it must come last in the list of options (after /sender_retain
2078 PH/05 Some modifications to the interface to the fake nameserver for the new
2086 TK/01 Added the "success_on_redirect" address verification option. See
2087 NewStuff for rationale and an example.
2089 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2091 PH/02 Patch to exigrep to allow it to work on syslog lines.
2093 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2094 fread() to skip over the body file's header line, because in Cygwin the
2095 header line is locked and is inaccessible.
2097 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2098 co-exist for some time) to make it clear that it is the Exim ID that is
2099 referenced, not the Message-ID: header line.
2101 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2102 string_format() function, because snprintf() does not exist on all
2105 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2106 ${stat: expansion item.
2108 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2109 protocol synchronization error", to keep the pedants happy.
2111 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2112 well as for IRIX systems, when gcc is being used. See the host.c source
2115 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2117 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2120 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2121 transport and to the smtp transport in LMTP mode.
2123 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2125 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2126 run a filter in a subprocess. This could lead to confusion in subsequent
2127 lookups in the parent process. There should also be a search_tidyup() at
2128 the end of the subprocess.
2130 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2131 only if the host matched helo_try_verify_hosts, which caused the
2132 verification to occur when the EHLO/HELO command was issued. The ACL just
2133 tested the remembered result. Now, if a previous verification attempt has
2134 not happened, "verify = helo" does it there and then.
2136 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2138 TK/03 Fix log output including CR from clamd.
2140 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2141 exist provoked a memory error which could cause a segfault.
2143 PH/15 Installed PCRE 6.2
2145 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2147 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2148 of the problem. Specifically, suggested +O2 rather than +O1 for the
2151 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2153 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2154 record, the port was not correctly logged when the outgoing_port log
2155 selector was set (it logged the transort's default port).
2157 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2158 fallback_hosts, and "hosts" in the smtp transport.
2160 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2161 host errors such as "Connection refused".
2163 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2164 authentication with radiusclient 0.4.9:
2166 - Error returned from rc_read_config was caught wrongly
2167 - Username/password not passed on to radius server due to wrong length.
2169 The presumption is that some radiusclient API changes for 4.51/PH/17
2170 were not taken care of correctly. The code is still untested by me (my
2171 Linux distribution still has 0.3.2 of radiusclient), but it was
2172 contributed by a Radius user.
2174 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2175 expanding the "port" option of the smtp transport.
2177 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2178 while reading a MIME header. Thanks to Tom Hughes for a patch.
2180 PH/24 Include config.h inside local_scan.h so that configuration settings are
2183 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2184 that in an ACL for RCPT (for example), you can examine exactly what was
2187 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2188 commands, but it was not correctly comparing the address with the actual
2189 client host address. Thus, it would show the EHLO address in Received:
2190 header lines when this was not necessary.
2192 PH/27 Added the % operator to ${eval:}.
2194 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2195 it should be ignoring failures (because with -C, for example, it has lost
2196 privilege). It wasn't ignoring creation failures other than "already
2199 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2202 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2203 because an input line was too long, either on its own, or by virtue of
2204 too many continuations, the temporary file was not being removed, and the
2205 return code was incorrect.
2207 PH/31 Missing "BOOL" in function definition in filtertest.c.
2209 PH/32 Applied Sieve patches from the maintainer.
2211 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2213 PH/33 Added "verify = not_blind".
2215 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2216 Local/Makefile (with some defaults set). These are used in built scripts
2217 such as exicyclog, but they have never been used in the exim_install
2218 script (though there are many overriding facilities there). I have
2219 arranged that the exim_install script now takes note of these two
2222 PH/35 Installed configuration files for Dragonfly.
2224 PH/36 When a locally submitted message by a trusted user did not contain a
2225 From: header, and the sender address was obtained from -f or from an SMTP
2226 MAIL command, and the trusted user did not use -F to supply a sender
2227 name, $originator_name was incorrectly used when constructing a From:
2228 header. Furthermore, $originator_name was used for submission mode
2229 messages from external hosts without From: headers in a similar way,
2230 which is clearly wrong.
2232 PH/37 Added control=suppress_local_fixups.
2234 PH/38 When log_selector = +received_sender was set, and the addition of the
2235 sender made the log line's construction buffer exactly full, or one byte
2236 less than full, an overflow happened when the terminating "\n" was
2239 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2240 when the result of a list match is failure because a DNS lookup failed.
2242 PH/40 RM_COMMAND is now used in the building process.
2244 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2245 the "build-* directories that it finds.
2247 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2248 address in a domain literal was a prefix of an interface address.
2250 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2251 when verifying a sender address, unless rewrite_headers is false.
2253 PH/44 Wrote a long comment about why errors_to addresses are verified as
2254 recipients, not senders.
2256 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2257 the ratelimit ACL was added.
2259 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2261 PH/46 Added extra information about PostgreSQL errors to the error string.
2263 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2264 avoiding the need to install special zones in a real server. This is
2265 backwards compatible; if it can't find the fake resolver, it drops back.
2266 Thus, both old and new test suites can be run.
2268 TF/02 Added util/ratelimit.pl
2270 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2271 clock is set back in time.
2273 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2274 Candler <B.Candler@pobox.com>.
2276 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2277 OK for addresses that are the result of redirections.
2279 PH/48 A number of further additions for the benefit of the new test suite,
2280 including a fake gethostbyname() that interfaces to the fake DNS resolver
2283 TF/06 The fix for widen_domains has also been applied to qualify_single and
2284 search_parents which are the other dnslookup options that can cause
2287 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2290 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2292 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2293 service" if it encountered an MX record with an empty target hostname.
2294 The message is now "an MX or SRV record indicated no SMTP service".
2296 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2297 if the DNS of the sending site is misconfigured. This is quite a
2298 common situation. This change restores the behaviour of treating a
2299 helo verification defer as a failure.
2301 PH/51 If self=fail was set on a router, the bounce message did not include the
2302 actual error message.
2308 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2310 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2311 command itself ended in error, the underlying message about the transport
2312 filter timeout was being overwritten with the pipe command error. Now the
2313 underlying error message should be appended to the second error message.
2315 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2317 PH/02 Reduce the amount of output that "make" produces by default. Full output
2318 can still be requested.
2320 PH/03 The warning log line about a condition test deferring for a "warn" verb
2321 was being output only once per connection, rather than after each
2322 occurrence (because it was using the same function as for successful
2323 "warn" verbs). This seems wrong, so I have changed it.
2325 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2326 it should not have, which might have caused a crash in the right
2327 circumstances, but probably never did.
2329 PH/04 Installed a modified version of Tony Finch's patch to make submission
2330 mode fix the return path as well as the Sender: header line, and to
2331 add a /name= option so that you can make the user's friendly name appear
2334 TF/03 Added the control = fakedefer ACL modifier.
2336 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2337 Mark Lowes for thorough testing.
2339 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2341 TK/03 Merged latest SRS patch from Miles Wilton.
2343 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2344 with the definition in sysexits.h (which is #included earlier).
2345 Fortunately, Exim does not actually use EX_OK. The code used to try to
2346 preserve the sysexits.h value, by assumimg that macro definitions were
2347 scanned for macro replacements. I have been disabused of this notion,
2348 so now the code just undefines EX_OK before #including unistd.h.
2350 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2351 in the smtp transport. When a block could not be written in a single
2352 write() function, the timeout was being re-applied to each part-write.
2353 This seems wrong - if the receiver was accepting one byte at a time it
2354 would take for ever. The timeout is now adjusted when this happens. It
2355 doesn't have to be particularly precise.
2357 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2358 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2360 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2361 without a domain if the check to <postmaster@domain> fails.
2363 SC/01 Eximstats: added -xls and the ability to specify output files
2364 (patch written by Frank Heydlauf).
2366 SC/02 Eximstats: use FileHandles for outputing results.
2368 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2370 SC/04 Eximstats: fixed display of large numbers with -nvr option
2372 SC/05 Eximstats: fixed merging of reports with empty tables.
2374 SC/06 Eximstats: added the -include_original_destination flag
2376 SC/07 Eximstats: removed tabs and trailing whitespace.
2378 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2380 TK/06 MBOX spool code: Add real "From " MBOX separator line
2381 so the .eml file is really in mbox format (even though
2382 most programs do not really care). Patch from Alex Miller.
2384 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2385 The latter is generated from $received_to and is only set if the
2386 message has one envelope recipient. SA can use these headers,
2387 obviously out-of-the-box. Patch from Alex Miller.
2389 PH/08 The ${def test on a variable was returning false if the variable's
2390 value was "0", contrary to what the specification has always said!
2391 The result should be true unless the variable is empty.
2393 PH/09 The syntax error of a character other than { following "${if
2394 def:variable_name" (after optional whitespace) was not being diagnosed.
2395 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2396 accidental colon was present, for example, could give incorrect results.
2398 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2399 result is used (not including appendfile, where other changes are about
2402 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2403 This involved changing a lot of size variables from int to off_t. It
2404 should work with maildirs and everything.
2406 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2407 spamd dying while we are connected to it.
2409 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2412 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2413 function for BDB 4.3.
2415 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2417 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2418 the sources for PCRE have been re-arranged and more files are now
2421 PH/15 The code I had for printing potentially long long variables in PH/11
2422 above was not the best (it lost precision). The length of off_t variables
2423 is now inspected at build time, and an appropriate printing format (%ld
2424 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2425 to be "long long int" or "long int". This is needed for the internal
2426 formatting function string_vformat().
2428 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2429 the configuration file to be ":syslog", then the script "guesses" where
2430 the logs files are, rather than using the compiled in default. In our
2431 case the guess is not the same as the compiled default, so the script
2432 suddenly stopped working when I started to use syslog. The patch checks
2433 to see if log_file_path is "". If so, it attempts to read it from exim
2434 with no configuration file to get the compiled in version, before it
2435 falls back to the previous guessing code."
2437 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2438 implementing BATV in an Exim configuration. See NewStuff for the gory
2441 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2442 Makefile that are specific to HP-UX.
2444 PH/18 If the "use_postmaster" option was set for a recipient callout together
2445 with the "random" option, the postmaster address was used as the MAIL
2446 FROM address for the random test, but not for the subsequent recipient
2447 test. It is now used for both.
2449 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2450 patch removes a few documentation additions to RFC 3028, because the
2451 latest draft now contains them. It adds the new en;ascii-case comparator
2452 and a new error check for 8bit text in MIME parts. Comparator and
2453 require names are now matched exactly. I enabled the subaddress
2454 extension, but it is not well tested yet (read: it works for me)."
2456 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2457 rework some of the code of TK/09 above to avoid the hardwired use of
2458 "%lld" and "long long". Replaced the call to snprintf() with a call to
2461 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2462 records point to non-existent hosts", "retry timeout exceeded", and
2463 "retry time not reached for any host after a long failure period".
2465 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2466 experimental DomainKeys support:
2468 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2469 (2) On an error such as an illegally used "control", the wrong name for
2470 the control was given.
2472 These problems did NOT occur unless DomainKeys support was compiled.
2474 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2476 PH/24 Added ${if match_ip condition.
2478 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2479 databases so that it will be absolutely obvious if a crash occurs in the
2480 DB library. This is a regular occurrence (often caused by mis-matched
2483 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2484 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2485 picked up on a user's system that detects such things. There doesn't seem
2486 to be a gcc warning option for this - only an attribute that has to be
2487 put on the function's prototype. It seems that in Fedora Core 4 they have
2488 set this on a number of new functions. No doubt there will be more in due
2491 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2492 specify a transport. However, if an address that was verified by such a
2493 router was the subject of a callout, Exim crashed because it tried to
2494 read the rcpt_include_affixes from the non-existent transport. Now it
2495 just assumes that the setting of that option is false. This bug was
2496 introduced by 4.51/PH/31.
2498 PH/28 Changed -d+all to exclude +memory, because that information is very
2499 rarely of interest, but it makes the output a lot bigger. People tend to
2500 do -d+all out of habit.
2502 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2503 code in os-type was giving problems when libc.so lives in lib64, like on
2506 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2507 aren't the modern standard, and it seems that some systems' include files
2508 don't always have them. Exim was already checking for some of the newer
2509 ones like T_AAAA, and defining it itself. I've added checks for all the
2510 record types that Exim uses.
2512 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2513 not automatically generating a new one, as it is supposed to. This
2514 prevented TLS from working. If the file did exist, but contained invalid
2515 data, a new version was generated, as expected. It was only the case of a
2516 non-existent file that was broken.
2518 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2519 with a change in libdomainkeys > 0.64.
2521 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2522 from DNS. If the selector record carries the flag, it now has
2523 precedence over the domain-wide flag.
2525 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2527 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2528 the use of a port name that isn't defined in /etc/services) occurred, the
2529 message was deferred as in a normal delivery, and thus remained on the
2530 spool, instead of being failed because of the mua_wrapper setting. This
2531 is now fixed, and I tidied up some of the mua_wrapper messages at the
2534 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2535 the messages in a hash of arrays rather than using individual hashes.
2536 This is a bit cleaner and results in dramatic memory savings, albeit
2537 at a slight CPU cost.
2539 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2540 as requested by Marc Sherman.
2542 SC/10 Eximstats: added histograms for user specified patterns as requested
2545 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2547 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2548 fopen() in the content-scanning modules that did not already have it.
2554 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2555 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2557 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2559 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2561 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2562 to test Sieve filters that use "vacation".
2564 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2565 that changes the way the GnuTLS parameters are stored in the cache file.
2566 The new format can be generated externally. For backward compatibility,
2567 if the data in the cache doesn't make sense, Exim assumes it has read an
2568 old-format file, and it generates new data and writes a new file. This
2569 means that you can't go back to an older release without removing the
2572 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2573 work if there are any delivery delays because "one_time" forces the
2574 parent to be marked "delivered", so its unseen clone is never tried
2575 again. For this reason, Exim now forbids the simultaneous setting of
2578 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2579 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2580 entry if you want to know the details. The fix, however, neglected to
2581 consider the case when local delivery batching is involved. The test for
2582 "previously delivered" was not happening when checking to see if an
2583 address could be batched with a previous (undelivered) one; under
2584 certain circumstances this could lead to multiple deliveries to the same
2587 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2588 in its include files, and this causes problems building Exim.
2590 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2591 header_syntax) but Exim was just ignoring anything given after a slash.
2592 In particular, this caused confusion with an attempt to use "verify =
2593 reverse_host_lookup/defer_ok". An error is now given when options are
2594 supplied for verify items that do not have them. (Maybe reverse_host_
2595 lookup should have a defer_ok option, but that's a different point.)
2597 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2598 defined by RFC 821) to 2048, because there were problems with some AUTH
2599 commands, and RFC 1869 says the size should be increased for extended
2600 SMTP commands that take arguments.
2602 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2605 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2606 "unknown" error; now it says that the functionality isn't in the binary.
2608 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2609 an address' error message when a string expansion fails (syntax or
2610 whatever). Otherwise the password may appear in the log. Following change
2611 PH/42 below, there is no longer a chance of it appearing in a bounce
2614 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2616 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2617 of its addresses was ever tried. (Bugzilla bug #2.)
2619 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2620 the result incorrectly in the debug output. (It correctly added a newline
2621 to what was transported.)
2623 TF/01 Added $received_time.
2625 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2626 commented out examples of how to interface to a virus scanner and to
2627 SpamAssassin. Also added commented examples of av_scanner and
2628 spamd_address settings.
2630 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2631 and controls are allowed in which ACLs. There were a couple of minor
2632 errors. Some of the entries in the conditions table (which is a table of
2633 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2634 negation of where the condition IS allowed.
2636 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2638 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2639 header file does not have a version number, so I've had to invent a new
2640 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2641 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2642 radiusclient), but it was contributed by a Radius user.
2644 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2645 files or directories, for OpenSSL.
2647 PH/19 When an Exim process that is running as root has to create an Exim log
2648 file, it does so in a subprocess that runs as exim:exim so as to get the
2649 ownership right at creation (otherwise, other Exim processes might see
2650 the file with the wrong ownership). There was no test for failure of this
2651 fork() call, which would lead to the process getting stuck as it waited
2652 for a non-existent subprocess. Forks do occasionally fail when resources
2653 run out. I reviewed all the other calls to fork(); they all seem to check
2656 PH/20 When checking for unexpected SMTP input at connect time (before writing
2657 the banner), Exim was not dealing correctly with a non-positive return
2658 from the read() function. If the client had disconnected by this time,
2659 the result was a log entry for a synchronization error with an empty
2660 string after "input=" when read() returned zero. If read() returned -1
2661 (an event I could not check), uninitialized data bytes were printed.
2662 There were reports of junk text (parts of files, etc) appearing after
2665 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2667 PH/22 Added support for macro redefinition, and (re)definition in between
2668 driver and ACL definitions.
2670 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2671 forgetting to use the resulting value; it was using the unexpanded value.
2673 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2674 hadn't been configured. The fix is from Juergen Kreileder, who
2675 understands it better than I do:
2677 "Here's what I see happening with three configured cyrus_sasl
2678 authenticators configured (plain, login, cram-md5):
2680 On startup auth_cyrus_sasl_init() gets called for each of these.
2681 This means three calls to sasl_listmech() without a specified mech_list.
2682 => SASL tests which mechs of all available mechs actually work
2683 => three warnings about OTP not working
2684 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2686 With the patch, sasl_listmech() also gets called three times. But now
2687 SASL's mech_list option is set to the server_mech specified in the the
2688 authenticator. Or in other words, the answer from sasl_listmech()
2689 gets limited to just the mech you're testing for (which is different
2691 => the return list contains just 'plain' or 'login', 'cram-md5' or
2692 nothing depending on the value of ob->server_mech.
2694 I've just tested the patch: Authentication still works fine,
2695 unavailable mechs specified in the exim configuration are still
2696 caught, and the auth.log warnings about OTP are gone."
2698 PH/25 When debugging is enabled, the contents of the command line are added
2699 to the debugging output, even when log_selector=+arguments is not
2702 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2703 answer is "GNU", and only if the return is "GNU/something" is the answer
2706 PH/27 $acl_verify_message is now set immediately after the failure of a
2707 verification in an ACL, and so is available in subsequent modifiers. In
2708 particular, the message can be preserved by coding like this:
2710 warn !verify = sender
2711 set acl_m0 = $acl_verify_message
2713 Previously, $acl_verify_message was set only while expanding "message"
2714 and "log_message" when a very denied access.
2716 PH/28 Modified OS/os.c-Linux with
2718 -#ifndef OS_LOAD_AVERAGE
2719 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2721 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2722 nomenclature these days.)
2724 PH/29 Installed patch from the Sieve maintainer that adds the options
2725 sieve_useraddress and sieve_subaddress to the redirect router.
2727 PH/30 In these circumstances:
2728 . Two addresses routed to the same list of hosts;
2729 . First host does not offer TLS;
2730 . First host accepts first address;
2731 . First host gives temporary error to second address;
2732 . Second host offers TLS and a TLS session is established;
2733 . Second host accepts second address.
2734 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2735 and peerdn, if requested) that were in fact used only for the second
2738 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2739 attention to any local part prefix or suffix that was matched by the
2740 router that accepted the address. It now behaves in the same way as it
2741 does for delivery: the affixes are removed from the local part unless
2742 rcpt_include_affixes is set on the transport.
2744 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2745 timeout during the DATA phase of an incoming message.
2747 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2748 applied a patch sanctioned by the Sieve maintainer.
2750 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2751 the uid or gid is negative. A case of a negative gid caused this to be
2752 noticed. The fix allows for either to be negative.
2754 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2755 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2758 PH/36 The change PH/12 above was broken. Fixed it.
2760 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2761 the grounds that routing the same address twice would always produce the
2762 same answer. This might have been true once, but it is certainly no
2763 longer true now. Routing a child address may depend on the previous
2764 routing that produced that child. Some complicated redirection strategies
2765 went wrong when messages had multiple recipients, and made Exim's
2766 behaviour dependent on the order in which the addresses were given.
2768 I have moved the duplicate checking until after the routing is complete.
2769 Exim scans the addresses that are assigned to local and remote
2770 transports, and removes any duplicates. This means that more work will be
2771 done, as duplicates will always all be routed, but duplicates are
2772 presumably rare, so I don't expect this is of any significance.
2774 For deliveries to pipes, files, and autoreplies, the duplicate checking
2775 still happens during the routing process, since they are not going to be
2778 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2779 It corrects a timeout issue with spamd. This is Ian's comment: "The
2780 background is that sometimes spamd either never reads data from a
2781 connection it has accepted, or it never writes response data. The exiscan
2782 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2783 blindly assumes that writes won't block so it may never time out."
2785 PH/39 Allow G after quota size as well as K and M.
2787 PH/40 The value set for $authenticated_id in an authenticator may not contain
2788 binary zeroes or newlines because the value is written to log lines and
2789 to spool files. There was no check on this. Now the value is run through
2790 the string_printing() function so that such characters are converted to
2791 printable escape sequences.
2793 PH/41 $message_linecount is a new variable that contains the total number of
2794 lines in the message. Compare $body_linecount, which is the count for the
2797 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2798 bounce and delay warning messages, except in certain special cases, which
2801 (a) An SMTP error message from a remote host;
2802 (b) A message specified in a :fail: redirection;
2803 (c) A message specified in a "fail" command in a system filter;
2804 (d) A message specified in a FAIL return from the queryprogram router;
2805 (e) A message specified by the cannot_route_message router option.
2807 In these cases only, Exim does include the error details in bounce and
2808 warning messages. There are also a few cases where bland messages such
2809 as "unrouteable address" or "local delivery error" are given.
2811 PH/43 $value is now also set for the "else" part of a ${run expansion.
2813 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2814 being worked on, but at least Exim now implements the latest version to
2817 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2818 process to complete was treated as a delivery failure, a timeout while
2819 writing the message to the pipe was logged, but erroneously treated as a
2820 successful delivery. Such timeouts include transport filter timeouts. For
2821 consistency with the overall process timeout, these timeouts are now
2822 treated as errors, giving rise to delivery failures by default. However,
2823 there is now a new Boolean option for the pipe transport called
2824 timeout_defer, which, if set TRUE, converts the failures into defers for
2825 both kinds of timeout. A transport filter timeout is now identified in
2828 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2829 systems where "make" and "gmake" are different, calling "gmake" at top
2830 level broke things. I've arranged for the value of $(MAKE) to be passed
2831 from the Makefile to this script so that it can call the same version of
2835 A note about Exim versions 4.44 and 4.50
2836 ----------------------------------------
2838 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2839 changes of various kinds. As a consequence, a big documentation update was
2840 needed. This delayed the release for rather longer than seemed good, especially
2841 in the light of a couple of (minor) security issues. Therefore, the changes
2842 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2843 release. So 4.44 and 4.50 are in effect two different branches that both start
2846 I have left the 4.50 change log unchanged; it contains all the changes since
2847 4.43. The change log for 4.44 is below; many of its items are identical to
2848 those for 4.50. This seems to be the most sensible way to preserve the
2849 historical information.
2855 1. Minor wording change to the doc/README.SIEVE file.
2857 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2858 computation of the current number of files was incorrect.
2860 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2861 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2864 4. Give more explanation in the error message when the command for a transport
2865 filter fails to execute.
2867 5. There are several places where Exim runs a non-Exim command in a
2868 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2869 was being done only for the command run by the queryprogram router. It is
2870 now done for all such subprocesses. The other cases are: ${run, transport
2871 filters, and the commands run by the lmtp and pipe transports.
2873 6. Added CONFIGURE_GROUP build-time option.
2875 7. Some older OS have a limit of 256 on the maximum number of file
2876 descriptors. Exim was using setrlimit() to set 1000 as a large value
2877 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2878 systems. I've change it so that if it can't get 1000, it tries for 256.
2880 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2881 was an oversight, and furthermore, ever since the addition of extra
2882 controls (e.g. 4.43/32), the checks on when to allow different forms of
2883 "control" were broken. There should now be diagnostics for all cases when a
2884 control that does not make sense is encountered.
2886 9. Added the /retain_sender option to "control=submission".
2888 10. $recipients is now available in the predata ACL (oversight).
2890 11. Tidy the search cache before the fork to do a delivery from a message
2891 received from the command line. Otherwise the child will trigger a lookup
2892 failure and thereby defer the delivery if it tries to use (for example) a
2893 cached ldap connection that the parent has called unbind on.
2895 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2896 of $address_data from the recipient verification was clobbered by the
2897 sender verification.
2899 13. The value of address_data from a sender verification is now available in
2900 $sender_address_data in subsequent conditions in the ACL statement.
2902 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2904 15. Added a new option "connect=<time>" to callout options, to set a different
2907 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2908 was its contents. (It was OK if the option was not defined at all.)
2910 17. A "Completed" log line is now written for messages that are removed from
2911 the spool by the -Mrm option.
2913 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2914 information about exactly what failed.
2916 19. Added -dd to debug only the daemon process.
2918 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2919 handles timeouts, both on the server side and network timeouts. Renamed the
2920 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2922 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2923 It is now set to "smtps".
2925 22. $host_address is now set to the target address during the checking of
2926 ignore_target_hosts.
2928 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2929 being passed; this would have caused $sender_host_name to have been used if
2930 matching the list had actually called for a host name (not very likely,
2931 since this list is usually IP addresses). A host name is now passed as
2934 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2935 code that specifies a non-restarting handler (typically sigaction() in
2936 modern systems) in an attempt to fix a rare and obscure crash bug.
2938 25. Narrowed the window for a race in the daemon that could cause it to ignore
2939 SIGCHLD signals. This is not a major problem, because they are used only to
2940 wake it up if nothing else does.
2942 26. A malformed maildirsize file could cause Exim to calculate negative values
2943 for the mailbox size or file count. Odd effects could occur as a result.
2944 The maildirsize information is now recalculated if the size or filecount
2947 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2948 support for a long time. Removed HAVE_SYS_VFS_H.
2950 28. Installed the latest version of exipick from John Jetmore.
2952 29. In an address list, if the pattern was not a regular expression, an empty
2953 subject address (from a bounce message) matched only if the pattern was an
2954 empty string. Non-empty patterns were not even tested. This was the wrong
2955 because it is perfectly reasonable to use an empty address as part of a
2956 database query. An empty address is now tested by patterns that are
2957 lookups. However, all the other forms of pattern expect the subject to
2958 contain a local part and a domain, and therefore, for them, an empty
2959 address still always fails if the pattern is not itself empty.
2961 30. Exim went into a mad DNS loop when attempting to do a callout where the
2962 host was specified on an smtp transport, and looking it up yielded more
2963 than one IP address.
2965 31. Re-factored the code for checking spool and log partition space into a
2966 function that finds that data and another that does the check. The former
2967 is then used to implement four new variables: $spool_space, $log_space,
2968 $spool_inodes, and $log_inodes.
2970 32. The RFC2047 encoding function was originally intended for short strings
2971 such as real names; it was not keeping to the 75-character limit for
2972 encoded words that the RFC imposes. It now respects the limit, and
2973 generates multiple encoded words if necessary. To be on the safe side, I
2974 have increased the buffer size for the ${rfc2047: expansion operator from
2977 33. It is now permitted to omit both strings after an "if" condition; if the
2978 condition is true, the result is "true". As before, when the second string
2979 is omitted, a false condition yields an empty string. This makes it less
2980 cumbersome to write custom ACL and router conditions.
2982 34. Failure to deliver a bounce message always caused it to be frozen, even if
2983 there was an errors_to setting on the router. The errors_to setting is now
2986 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2987 canonical form (fully expanded) before being placed in
2988 $sender_host_address.
2990 36. The table in the code that translates DNS record types into text (T_A to
2991 "A" for instance) was missing entries for NS and CNAME. It is just possible
2992 that this could have caused confusion if both these types were looked up
2993 for the same domain, because the text type is used as part of Exim's
2994 per-process caching. But the chance of anyone hitting this buglet seems
2997 37. The dnsdb lookup has been extended in a number of ways.
2999 (1) There is a new type, "zns", which walks up the domain tree until it
3000 finds some nameserver records. It should be used with care.
3002 (2) There is a new type, "mxh", which is like "mx" except that it returns
3003 just the host names, not the priorities.
3005 (3) It is now possible to give a list of domains (or IP addresses) to be
3006 looked up. The behaviour when one of the lookups defers can be
3007 controlled by a keyword.
3009 (4) It is now possible to specify the separator character for use when
3010 multiple records are returned.
3012 38. The dnslists ACL condition has been extended: it is now possible to supply
3013 a list of IP addresses and/or domains to be looked up in a particular DNS
3016 39. Added log_selector=+queue_time_overall.
3018 40. When running the queue in the test harness, wait just a tad after forking a
3019 delivery process, to get repeatability of debugging output.
3021 41. Include certificate and key file names in error message when GnuTLS fails
3022 to set them up, because the GnuTLS error message doesn't include the name
3023 of the failing file when there is a problem reading it.
3025 42. Allow both -bf and -bF in the same test run.
3027 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3029 44. The "Exiscan patch" is now merged into the mainline Exim source.
3031 45. Sometimes the final signoff response after QUIT could fail to get
3032 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3033 < 0 before doing a fflush(). This bug looks as though it goes back to the
3034 introduction of TLS in release 3.20, but "sometimes" must have been rare
3035 because the tests only now provoked it.
3037 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3038 (this can affect the format of dates).
3040 47. exim_tidydb, when checking for the continued existence of a message for
3041 which it has found a message-specific retry record, was not finding
3042 messages that were in split spool directories. Consequently, it was
3043 deleting retry records that should have stayed in existence.
3045 48. Steve fixed some bugs in eximstats.
3047 49. The SPA authentication driver was not abandoning authentication and moving
3048 on to the next authenticator when an expansion was forced to fail,
3049 contradicting the general specification for all authenticators. Instead it
3050 was generating a temporary error. It now behaves as specified.
3052 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3053 (the order specifies the preference for clients). The order is now AES256,
3054 AES128, 3DES, ARCFOUR128.
3056 51. Small patch to Sieve code - explicitly set From: when generating an
3059 52. Exim crashed if a remote delivery caused a very long error message to be
3060 recorded - for instance if somebody sent an entire SpamAssassin report back
3061 as a large number of 550 error lines. This bug was coincidentally fixed by
3062 increasing the size of one of Exim's internal buffers (big_buffer) that
3063 happened as part of the Exiscan merge. However, to be on the safe side, I
3064 have made the code more robust (and fixed the comments that describe what
3067 53. Now that there can be additional text after "Completed" in log lines (if
3068 the queue_time_overall log selector is set), a one-byte patch to exigrep
3069 was needed to allow it to recognize "Completed" as not the last thing in
3072 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3073 patch that reportedly fixes this has been added. I am not expert enough to
3074 create a test for it. This is what the patch creator wrote:
3076 "I found a little strange behaviour of ldap code when working with
3077 Windows 2003 AD Domain, where users was placed in more than one
3078 Organization Units. When I tried to give exim partial DN, the exit code
3079 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3080 But simultaneously result of request was absolutely normal ldap result,
3081 so I produce this patch..."
3083 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3084 REFERENCE, so I have modified the code to exclude the patch when that macro
3087 55. Some experimental protocols are using DNS PTR records for new purposes. The
3088 keys for these records are domain names, not reversed IP addresses. The
3089 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3090 leaves it alone. Component reversal etc. now happens only for IP addresses.
3093 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3095 57. Double the size of the debug message buffer (to 2048) so that more of very
3096 long debug lines gets shown.
3098 58. The exicyclog utility now does better if the number of log files to keep
3099 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3101 59. Two changes related to the smtp_active_hostname option:
3103 (1) $smtp_active_hostname is now available as a variable.
3104 (2) The default for smtp_banner uses $smtp_active_hostname instead
3105 of $primary_hostname.
3107 60. The host_aton() function is supposed to be passed a string that is known
3108 to be a valid IP address. However, in the case of IPv6 addresses, it was
3109 not checking this. This is a hostage to fortune. Exim now panics and dies
3110 if the condition is not met. A case was found where this could be provoked
3111 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3112 components; fortuitously, this particular loophole had already been fixed
3113 by change 4.50/55 above.
3115 If there are any other similar loopholes, the new check in host_aton()
3116 itself should stop them being exploited. The report I received stated that
3117 data on the command line could provoke the exploit when Exim was running as
3118 exim, but did not say which command line option was involved. All I could
3119 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3120 running as the user.
3123 61. There was a buffer overflow vulnerability in the SPA authentication code
3124 (which came originally from the Samba project). I have added a test to the
3125 spa_base64_to_bits() function which I hope fixes it.
3128 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3129 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3131 63. The daemon start-up calls getloadavg() while still root for those OS that
3132 need the first call to be done as root, but it missed one case: when
3133 deliver_queue_load_max is set with deliver_drop_privilege. This is
3134 necessary for the benefit of the queue runner, because there is no re-exec
3135 when deliver_drop_privilege is set.
3137 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3138 This has been fixed.
3140 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3141 was in use, was not putting the data itself into the right store pool;
3142 consequently, it could be overwritten for a subsequent message in the same
3143 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3146 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3148 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3149 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3150 the function were treating the return as a boolean value, which happened to
3151 work because 0=false and not-0=true, but is not correct code.
3153 68. The host_aton() function was not handling scoped IPv6 addresses (those
3154 with, for example, "%eth0" on the end) correctly.
3156 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3157 negated items (that is, ~something) in unsigned ints. Some compilers
3158 apparently mutter when there is no cast.
3160 70. If an address verification called from an ACL failed, and did not produce a
3161 user-specific message (i.e. there was only a "system" message), nothing was
3162 put in $acl_verify_message. In this situation, it now puts the system
3165 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3166 session; change 4.31/43 added the unwanted input to the log line - except
3167 that it did not do this in the start of session case. It now does.
3169 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3170 This is wrong and can cause the other end to generate a synchronization
3171 error if it is another Exim or anything else that does the synchronization
3172 check. A QUIT command is no longer sent after a timeout.
3174 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3175 during host lookups.
3177 74. The defer_ok option of callout verification was not working if it was used
3178 when verifying addresses in header lines, that is, for this case:
3180 verify = header_sender/callout=defer_ok
3182 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3183 those file descriptors could be used for SMTP connections. If anything
3184 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3185 could be sent to the SMTP client, causing chaos. The daemon now opens
3186 stdin, stdout, and stderr to /dev/null when it puts itself into the
3189 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3190 log by default. The user can override this with suitable Perl magic.
3192 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3193 the log message when discard triggers, was not working for the DATA ACL or
3194 for the non-SMTP ACL.
3196 78. Error message wording change in sieve.c.
3198 79. If smtp_accept_max_per_host was set, the number of connections could be
3199 restricted to fewer than expected, because the daemon was trying to set up
3200 a new connection before checking whether the processes handling previous
3201 connections had finished. The check for completed processes is now done
3202 earlier. On busy systems, this bug wouldn't be noticed because something
3203 else would have woken the daemon, and it would have reaped the completed
3206 80. If a message was submitted locally by a user whose login name contained one
3207 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3208 It caused a spool format error. I have fixed the spool reading code. A
3209 related problem was that the "from" clause in the Received: line became
3210 illegal because of the space(s). It is now covered by ${quote_local_part.
3212 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3215 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3217 83. Patch from Sieve maintainer to fix unterminated string problem in
3218 "vacation" handling.
3220 84. Some minor changes to the Linux configuration files to help with other
3221 OS variants using glibc.
3223 85. One more patch for Sieve to update vacation handling to latest spec.
3226 ----------------------------------------------------
3227 See the note above about the 4.44 and 4.50 releases.
3228 ----------------------------------------------------
3234 1. Change 4.43/35 introduced a bug that caused file counts to be
3235 incorrectly computed when quota_filecount was set in an appendfile
3238 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3239 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3242 3. Give more explanation in the error message when the command for a transport
3243 filter fails to execute.
3245 4. There are several places where Exim runs a non-Exim command in a
3246 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3247 was being done only for the command run by the queryprogram router. It is
3248 now done for all such subprocesses. The other cases are: ${run, transport
3249 filters, and the commands run by the lmtp and pipe transports.
3251 5. Some older OS have a limit of 256 on the maximum number of file
3252 descriptors. Exim was using setrlimit() to set 1000 as a large value
3253 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3254 systems. I've change it so that if it can't get 1000, it tries for 256.
3256 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3257 was an oversight, and furthermore, ever since the addition of extra
3258 controls (e.g. 4.43/32), the checks on when to allow different forms of
3259 "control" were broken. There should now be diagnostics for all cases when a
3260 control that does not make sense is encountered.
3262 7. $recipients is now available in the predata ACL (oversight).
3264 8. Tidy the search cache before the fork to do a delivery from a message
3265 received from the command line. Otherwise the child will trigger a lookup
3266 failure and thereby defer the delivery if it tries to use (for example) a
3267 cached ldap connection that the parent has called unbind on.
3269 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3270 of $address_data from the recipient verification was clobbered by the
3271 sender verification.
3273 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3274 was its contents. (It was OK if the option was not defined at all.)
3276 11. A "Completed" log line is now written for messages that are removed from
3277 the spool by the -Mrm option.
3279 12. $host_address is now set to the target address during the checking of
3280 ignore_target_hosts.
3282 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3283 being passed; this would have caused $sender_host_name to have been used if
3284 matching the list had actually called for a host name (not very likely,
3285 since this list is usually IP addresses). A host name is now passed as
3288 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3289 code that specifies a non-restarting handler (typically sigaction() in
3290 modern systems) in an attempt to fix a rare and obscure crash bug.
3292 15. Narrowed the window for a race in the daemon that could cause it to ignore
3293 SIGCHLD signals. This is not a major problem, because they are used only to
3294 wake it up if nothing else does.
3296 16. A malformed maildirsize file could cause Exim to calculate negative values
3297 for the mailbox size or file count. Odd effects could occur as a result.
3298 The maildirsize information is now recalculated if the size or filecount
3301 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3302 support for a long time. Removed HAVE_SYS_VFS_H.
3304 18. Updated exipick to current release from John Jetmore.
3306 19. Allow an empty sender to be matched against a lookup in an address list.
3307 Previously the only cases considered were a regular expression, or an
3310 20. Exim went into a mad DNS lookup loop when doing a callout where the
3311 host was specified on the transport, if the DNS lookup yielded more than
3314 21. The RFC2047 encoding function was originally intended for short strings
3315 such as real names; it was not keeping to the 75-character limit for
3316 encoded words that the RFC imposes. It now respects the limit, and
3317 generates multiple encoded words if necessary. To be on the safe side, I
3318 have increased the buffer size for the ${rfc2047: expansion operator from
3321 22. Failure to deliver a bounce message always caused it to be frozen, even if
3322 there was an errors_to setting on the router. The errors_to setting is now
3325 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3326 canonical form (fully expanded) before being placed in
3327 $sender_host_address.
3329 24. Updated eximstats to version 1.33
3331 25. Include certificate and key file names in error message when GnuTLS fails
3332 to set them up, because the GnuTLS error message doesn't include the name
3333 of the failing file when there is a problem reading it.
3335 26. Expand error message when OpenSSL has problems setting up cert/key files.
3338 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3339 (this can affect the format of dates).
3341 28. exim_tidydb, when checking for the continued existence of a message for
3342 which it has found a message-specific retry record, was not finding
3343 messages that were in split spool directories. Consequently, it was
3344 deleting retry records that should have stayed in existence.
3346 29. eximstats updated to version 1.35
3347 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3348 1.35 - bugfix such that pie charts by volume are generated correctly
3350 30. The SPA authentication driver was not abandoning authentication and moving
3351 on to the next authenticator when an expansion was forced to fail,
3352 contradicting the general specification for all authenticators. Instead it
3353 was generating a temporary error. It now behaves as specified.
3355 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3356 (the order specifies the preference for clients). The order is now AES256,
3357 AES128, 3DES, ARCFOUR128.
3359 31. Small patch to Sieve code - explicitly set From: when generating an
3362 32. Exim crashed if a remote delivery caused a very long error message to be
3363 recorded - for instance if somebody sent an entire SpamAssassin report back
3364 as a large number of 550 error lines. This bug was coincidentally fixed by
3365 increasing the size of one of Exim's internal buffers (big_buffer) that
3366 happened as part of the Exiscan merge. However, to be on the safe side, I
3367 have made the code more robust (and fixed the comments that describe what
3370 33. Some experimental protocols are using DNS PTR records for new purposes. The
3371 keys for these records are domain names, not reversed IP addresses. The
3372 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3373 leaves it alone. Component reversal etc. now happens only for IP addresses.
3376 34. The host_aton() function is supposed to be passed a string that is known
3377 to be a valid IP address. However, in the case of IPv6 addresses, it was
3378 not checking this. This is a hostage to fortune. Exim now panics and dies
3379 if the condition is not met. A case was found where this could be provoked
3380 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3381 components; fortuitously, this particular loophole had already been fixed
3382 by change 4.50/55 or 4.44/33 above.
3384 If there are any other similar loopholes, the new check in host_aton()
3385 itself should stop them being exploited. The report I received stated that
3386 data on the command line could provoke the exploit when Exim was running as
3387 exim, but did not say which command line option was involved. All I could
3388 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3389 running as the user.
3392 35. There was a buffer overflow vulnerability in the SPA authentication code
3393 (which came originally from the Samba project). I have added a test to the
3394 spa_base64_to_bits() function which I hope fixes it.
3397 36. The daemon start-up calls getloadavg() while still root for those OS that
3398 need the first call to be done as root, but it missed one case: when
3399 deliver_queue_load_max is set with deliver_drop_privilege. This is
3400 necessary for the benefit of the queue runner, because there is no re-exec
3401 when deliver_drop_privilege is set.
3403 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3404 was in use, was not putting the data itself into the right store pool;
3405 consequently, it could be overwritten for a subsequent message in the same
3406 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3409 38. Sometimes the final signoff response after QUIT could fail to get
3410 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3411 < 0 before doing a fflush(). This bug looks as though it goes back to the
3412 introduction of TLS in release 3.20, but "sometimes" must have been rare
3413 because the tests only now provoked it.
3419 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3420 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3421 Now it returns an integer. A number of calls were still expecting a T/F
3422 return. Fortuitously, in all cases, the tests worked in OK situations,
3423 which is the norm. However, things would have gone wrong on any write
3424 failures on the smtp file descriptor. This function is used when sending
3425 messages over SMTP and also when doing verify callouts.
3427 2. When Exim is called to do synchronous delivery of a locally submitted
3428 message (the -odf or -odi options), it no longer closes stderr before doing
3431 3. Implemented the mua_wrapper option.
3433 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3435 5. Implemented the functions header_remove(), header_testname(),
3436 header_add_at_position(), and receive_remove_recipient(), and exported them
3439 6. If an ACL "warn" statement specified the addition of headers, Exim already
3440 inserted X-ACL-Warn: at the start if there was no header name. However, it
3441 was not making this test for the second and subsequent header lines if
3442 there were newlines in the string. This meant that an invalid header could
3443 be inserted if Exim was badly configured.
3445 7. Allow an ACL "warn" statement to add header lines at the start or after all
3446 the Received: headers, as well as at the end.
3448 8. Added the rcpt_4xx retry error code.
3450 9. Added postmaster_mailfrom=xxx to callout verification option.
3452 10. Added mailfrom=xxxx to the callout verification option, for verify=
3455 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3456 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3458 12. Inserted some casts to stop certain compilers warning when using pointer
3459 differences as field lengths or precisions in printf-type calls (mostly
3460 affecting debugging statements).
3462 13. Added optional readline() support for -be (dynamically loaded).
3464 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3465 same clock tick as a message's arrival, so that its received time was the
3466 same as the "first fail" time on the retry record, and that message
3467 remained on the queue past the ultimate address timeout, every queue runner
3468 would try a delivery (because it was past the ultimate address timeout) but
3469 after another failure, the ultimate address timeout, which should have then
3470 bounced the address, did not kick in. This was a "< instead of <=" error;
3471 in most cases the first failure would have been in the next clock tick
3472 after the received time, and all would be well.
3474 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3475 being recognized when the domain list was tested by the match_domain
3476 condition in an expansion string.
3478 16. Added the ${str2b64: operator.
3480 17. Exim was always calling setrlimit() to set a large limit for the number of
3481 processes, without checking whether the existing limit was already
3482 adequate. (It did check for the limit on file descriptors.) Furthermore,
3483 errors from getrlimit() and setrlimit() were being ignored. Now they are
3484 logged to the main and panic logs, but Exim does carry on, to try to do its
3485 job under whatever limits there are.
3487 18. Imported PCRE 5.0.
3489 19. Trivial typo in log message " temporarily refused connection" (the leading
3492 20. If the log selector return_path_on_delivery was set and an address was
3493 redirected to /dev/null, the delivery process crashed because it assumed
3494 that a return path would always be set for a "successful" delivery. In this
3495 case, the whole delivery is bypassed as an optimization, and therefore no
3498 21. Internal re-arrangement: the function for sending a challenge and reading
3499 a response while authentication was assuming a zero-terminated challenge
3500 string. It's now changed to take a pointer and a length, to allow for
3501 binary data in such strings.
3503 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3505 23. Exim was not respecting finduser_retries when seeking the login of the
3506 uid under which it was called; it was always trying 10 times. (The default
3507 setting of finduser_retries is zero.) Also, it was sleeping after the final
3508 failure, which is pointless.
3510 24. Implemented tls_on_connect_ports.
3512 25. Implemented acl_smtp_predata.
3514 26. If the domain in control=submission is set empty, Exim assumes that the
3515 authenticated id is a complete email address when it generates From: or
3516 Sender: header lines.
3518 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3519 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3520 chown and chgrp in /bin and hostname in /usr/bin.
3522 28. Exim was keeping the "process log" file open after each use, just as it
3523 does for the main log. This opens the possibility of it remaining open for
3524 long periods when the USR1 signal hits a daemon. Occasional processlog
3525 errors were reported, that could have been caused by this. Anyway, it seems
3526 much more sensible not to leave this file open at all, so that is what now
3529 29. The long-running daemon process does not normally write to the log once it
3530 has entered its main loop, and it closes the log before doing so. This is
3531 so that log files can straightforwardly be renamed and moved. However,
3532 there are a couple of unusual error situations where the daemon does write
3533 log entries, and I had neglected to close the log afterwards.
3535 30. The text of an SMTP error response that was received during a remote
3536 delivery was being truncated at 512 bytes. This is too short for some of
3537 the long messages that one sometimes sees. I've increased the limit to
3540 31. It is now possible to make retry rules that apply only when a message has a
3541 specific sender, in particular, an empty sender.
3543 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3544 it possible to be selective about when SMTP synchronization is enforced.
3546 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3548 32. Implemented hosts_connection_nolog.
3550 33. Added an ACL for QUIT.
3552 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3555 35. Added mailbox_size and mailbox_filecount to appendfile.
3557 36. Added control = no_multiline_responses to ACLs.
3559 37. There was a bug in the logic of the code that waits for the clock to tick
3560 in the case where the clock went backwards by a substantial amount such
3561 that the microsecond fraction of "now" was more than the microsecond
3562 fraction of "then" (but the whole seconds number was less).
3564 38. Added support for the libradius Radius client library this is found on
3565 FreeBSD (previously only the radiusclient library was supported).
3571 1. When certain lookups returned multiple values in the form name=value, the
3572 quoting of the values was not always being done properly. Specifically:
3573 (a) If the value started with a double quote, but contained no whitespace,
3575 (b) If the value contained whitespace other than a space character (i.e.
3576 tabs or newlines or carriage returns) it was not quoted.
3577 This fix has been applied to the mysql and pgsql lookups by writing a
3578 separate quoting function and calling it from the lookup code. The fix
3579 should probably also be applied to nisplus, ibase and oracle lookups, but
3580 since I cannot test any of those, I have not disturbed their existing code.
3582 2. A hit in the callout cache for a specific address caused a log line with no
3583 reason for rejecting RCPT. Now it says "Previous (cached) callout
3584 verification failure".
3586 3. There was an off-by-one bug in the queryprogram router. An over-long
3587 return line was truncated at 256 instead of 255 characters, thereby
3588 overflowing its buffer with the terminating zero. As well as fixing this, I
3589 have increased the buffer size to 1024 (and made a note to document this).
3591 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3592 when Exim is waiting for an SMTP response from a remote server, Exim
3593 restarts its select() call on the socket, thereby resetting its timeout.
3594 This is not a problem when such interrupts are rare. Somebody set up a cron
3595 job to run exiwhat every 2 minutes, which is less than the normal select()
3596 timeout (5 or 10 minutes). This meant that the select() timeout never
3597 kicked in because it was always reset. I have fixed this by comparing the
3598 time when an interrupt arrives with the time at the start of the first call
3599 to select(). If more time than the timeout has elapsed, the interrupt is
3600 treated as a timeout.
3602 5. Some internal re-factoring in preparation for the addition of Sieve
3603 extensions (by MH). In particular, the "personal" test is moved to a
3604 separate function, and given an option for scanning Cc: and Bcc: (which is
3605 not set for Exim filters).
3607 6. When Exim created an email address using the login of the caller as the
3608 local part (e.g. when creating a From: or Sender: header line), it was not
3609 quoting the local part when it contained special characters such as @.
3611 7. Installed new OpenBSD configuration files.
3613 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3614 try to make them clearer.
3616 9. Callout options, other than the timeout value, were being ignored when
3617 verifying sender addresses in header lines. For example, when using
3619 verify = header_sender/callout=no_cache
3621 the cache was (incorrectly) being used.
3623 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3624 only the Cygwin environment.
3626 11. When return_path_on_delivery was set as a log selector, if different remote
3627 addresses in the same message used different return paths and parallel
3628 remote delivery occurred, the wrong values would sometimes be logged.
3629 (Whenever a remote delivery process finished, the return path value from
3630 the most recently started remote delivery process was logged.)
3632 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3633 lines when AUTH and/or TLS are in use. This is the "received protocol"
3634 field. Exim used to use "asmtp" for authenticated SMTP, without any
3635 indication (in the protocol name) for TLS use. Now it follows the RFC and
3636 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3637 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3638 names appear in log lines as well as in Received: header lines.
3640 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3641 extensions, and comparison tests, and to fix some bugs.
3643 14. Changes to the "personal" filter test:
3645 (1) The test was buggy in that it was just doing the equivalent of
3646 "contains" tests on header lines. For example, if a user's address was
3647 anne@some.where, the "personal" test would incorrectly be true for
3649 To: susanne@some.where
3651 This test is now done by extracting each address from the header in turn,
3652 and checking the entire address. Other tests that are part of "personal"
3653 are now done using regular expressions (for example, to check local parts
3654 of addresses in From: header lines).
3656 (2) The list of non-personal local parts in From: addresses has been
3657 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3658 taken from the Sieve specification recommendations.
3660 (3) If the message contains any header line starting with "List-" it is
3661 treated as non-personal.
3663 (4) The test for "circular" in the Subject: header line has been removed
3664 because it now seems ill-conceived.
3666 15. Minor typos in src/EDITME comments corrected.
3668 16. Installed latest exipick from John Jetmore.
3670 17. If headers_add on a router specified a text string that was too long for
3671 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3672 of string_sprintf() is now avoided.
3674 18. $message_body_size was not set (it was always zero) when running the DATA
3675 ACL and the local_scan() function.
3677 19. For the "mail" command in an Exim filter, no default was being set for
3678 the once_repeat time, causing a random time value to be used if "once" was
3679 specified. (If the value happened to be <= 0, no repeat happened.) The
3680 default is now 0s, meaning "never repeat". The "vacation" command was OK
3681 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3682 (I found it when inspecting the code).
3684 20. There is now an overall timeout for performing a callout verification. It
3685 defaults to 4 times the callout timeout, which applies to individual SMTP
3686 commands during the callout. The overall timeout applies when there is more
3687 than one host that can be tried. The timeout is checked before trying the
3688 next host. This prevents very long delays if there are a large number of
3689 hosts and all are timing out (e.g. when the network connections are timing
3690 out). The value of the overall timeout can be changed by specifying an
3691 additional sub-option for "callout", called "maxwait". For example:
3693 verify = sender/callout=5s,maxwait=20s
3695 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3696 to the end before writing, but this should make it even safer).
3698 22. Exim was forgetting that it had advertised PIPELINING for the second and
3699 subsequent messages on an SMTP connection. It was also not resetting its
3700 memory on STARTTLS and an internal HELO.
3702 23. When Exim logs an SMTP synchronization error within a session, it now
3703 records whether PIPELINING has been advertised or not.
3705 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3706 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3707 rather than long int.
3709 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3711 26. Added the never_mail option to autoreply.
3717 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3718 crash if the getsockname() call failed; this can happen if a connection is
3719 closed very soon after it is established. The problem was simply in the
3720 order in which certain operations were done, causing Exim to try to write
3721 to the SMTP stream before it had set up the file descriptor. The bug has
3722 been fixed by making things happen in the correct order.
3728 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3729 before the connection was closed, thus losing the rejection response.
3731 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3732 some early Solaris releases, but causes trouble in current releases where
3733 socklen_t is defined.
3735 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3738 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3741 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3742 that releasing the top of it at the end releases what was used for sub-
3743 expansions (unless the block got too big). However, discard this block if
3744 the first thing is a variable or header, so that we can use its block when
3745 it is dynamic (useful for very large $message_headers, for example).
3747 6. Lookups now cache *every* query, not just the most recent. A new, separate
3748 store pool is used for this. It can be recovered when all lookup caches are
3749 flushed. Lookups now release memory at the end of their result strings.
3750 This has involved some general refactoring of the lookup sources.
3752 7. Some code has been added to the store_xxx() functions to reduce the amount
3753 of flapping under certain conditions.
3755 8. log_incoming_interface used to affect only the <= reception log lines. Now
3756 it causes the local interface and port to be added to several more SMTP log
3757 lines, for example "SMTP connection from", and rejection lines.
3759 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3761 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3763 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3764 could be overwritten at the end of the current message (or the start of a
3765 new message if it was set in a HELO ACL). The value is now preserved for
3766 the duration of the SMTP connection.
3768 12. If a transport had a headers_rewrite setting, and a matching header line
3769 contained an unqualified address, that address was qualified, even if it
3770 did not match any rewriting rules. The underlying bug was that the values
3771 of the flags that permit the existence of unqualified sender and recipient
3772 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3773 non-local messages, and by -bnq for local messages) were not being
3774 preserved with the message after it was received.
3776 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3777 "next input=" as part of the text comprising the host identity instead of
3778 the correct text. The code was using the same buffer for two different
3779 strings. However, depending on which order the printing function evaluated
3780 its arguments, the bug did not always show up. Under Linux, for example, my
3781 test suite worked just fine.
3783 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3784 doesn't work with some older versions of Perl. It has been changed to "my",
3785 which in any case is probably the better facility to use.
3787 15. A really picky compiler found some instances of statements for creating
3788 error messages that either had too many or two few arguments for the format
3791 16. The size of the buffer for calls to the DNS resolver has been increased
3792 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3793 for addresses that have a lot of PTR records. This alleviates a problem; it
3794 does not fully solve it.
3796 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3797 buffer now truncates the list and logs the incident, which is the same
3798 action as happens when Exim is looking up a host name and its aliases.
3799 Previously in this situation something unpredictable would happen;
3800 sometimes it was "internal error: store_reset failed".
3802 18. If a server dropped the connection unexpectedly when an Exim client was
3803 using GnuTLS and trying to read a response, the client delivery process
3804 crashed while trying to generate an error log message.
3806 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3807 string, for example:
3809 warn message = H1: something\nH2: something
3811 the text was added as a single header line from Exim's point of view
3812 though it ended up OK in the delivered message. However, searching for the
3813 second and subsequent header lines using $h_h2: did not work. This has been
3814 fixed. Similarly, if a system filter added multiple headers in this way,
3815 the routers could not see them.
3817 20. Expanded the error message when iplsearch is called with an invalid key to
3818 suggest using net-iplsearch in a host list.
3820 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3821 ACLs are no longer actually imposed (and a message to that effect is
3824 22. If a "gecos" field in a passwd entry contained escaped characters, in
3825 particular, if it contained a \" sequence, Exim got it wrong when building
3826 a From: or a Sender: header from that name. A second bug also caused
3827 incorrect handling when an unquoted " was present following a character
3828 that needed quoting.
3830 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3831 was not being matched caselessly.
3833 24. Arranged for all hyphens in the exim.8 source to be escaped with
3836 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3837 properly. Recipient callout cache records were still being keyed to include
3838 the sender, even when use_sender was set false. This led to far more
3839 callouts that were necessary. The sender is no longer included in the key
3840 when use_sender is false.
3842 26. Added "control = submission" modifier to ACLs.
3844 27. Added the ${base62d: operator to decode base 62 numbers.
3846 28. dnsdb lookups can now access SRV records.
3848 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3849 the configuration file.
3851 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3852 (-v) mode. This makes the output for a verbose queue run more intelligible.
3854 31. Added a use_postmaster feature to recipient callouts.
3856 32. Added the $body_zerocount variable, containing the number of binary zero
3857 bytes in the message body.
3859 33. The time of last modification of the "new" subdirectory is now used as the
3860 "mailbox time last read" when there is a quota error for a maildir
3863 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3865 35. Added +ignore_unknown as a special item in host lists.
3867 36. Code for decoding IPv6 addresses in host lists is now included, even if
3868 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3869 address was recognized as an IP address, but was then not correctly decoded
3870 into binary, causing unexpected and incorrect effects when compared with
3877 1. Very minor rewording of debugging text in manualroute to say "list of
3878 hosts" instead of "hostlist".
3880 2. If verify=header_syntax was set, and a header line with an unqualified
3881 address (no domain) and a large number of spaces between the end of the
3882 name and the colon was received, the reception process suffered a buffer
3883 overflow, and (when I tested it) crashed. This was caused by some obsolete
3884 code that should have been removed. The fix is to remove it!
3886 3. When running in the test harness, delay a bit after writing a bounce
3887 message to get a bit more predictability in the log output.
3889 4. Added a call to search_tidyup() just before forking a reception process. In
3890 theory, someone could use a lookup in the expansion of smtp_accept_max_
3891 per_host which, without the tidyup, could leave open a database connection.
3893 5. Added the variables $recipient_data and $sender_data which get set from a
3894 lookup success in an ACL "recipients" or "senders" condition, or a router
3895 "senders" option, similar to $domain_data and $local_part_data.
3897 6. Moved the writing of debug_print from before to after the "senders" test
3900 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3901 problems for message scanning, either using a data ACL, or using
3902 local_scan() because the Received: header was not generated till after they
3903 were called (in order to set the time as the time of reception completion).
3904 I have revised the way this works. The header is now generated after the
3905 body is received, but before the ACL or local_scan() are called. After they
3906 are run, the timestamp in the header is updated.
3912 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3913 before starting a queue runner without re-exec. This happened only when
3914 deliver_drop_privilege was set or when the Exim user was set to root. The
3915 effect of the bug was that timeouts during subsequent deliveries caused
3916 crashes instead of being properly handled. The handler is now left at its
3917 default (and expected) setting.
3919 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3920 message, again when deliver_drop_privilege is set or Exim is run as root.
3921 The bug described in (1) was not present in this case, but the tidying up
3922 of the other signals was missing. I have made the two cases consistent.
3924 3. The ignore_target_hosts setting on a manualroute router was being ignored
3925 for hosts that were looked up using the /MX notation.
3927 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3930 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3931 operated on the sender address. After changing the $sender_address to <>
3932 for the sender address verify, Exim was re-instated it as the original
3933 (before rewriting) address, but remembering that it had rewritten it, so it
3934 wasn't rewriting it again. This bug also had the effect of breaking the
3935 sender address verification caching when the sender address was rewritten.
3937 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3938 This has been changed so that if the ip literal address matches
3939 ignore_target_hosts, the router declines.
3941 7. Added expansion conditions match_domain, match_address, and match_local_
3942 part (NOT match_host).
3944 8. The placeholder for the Received: header didn't have a length field set.
3946 9. Added code to Exim itself and to exim_lock to test for a specific race
3947 condition that could lead to file corruption when using MBX delivery. The
3948 issue is with the lockfile that is created in /tmp. If this file is removed
3949 after a process has opened it but before that process has acquired a lock,
3950 there is the potential for a second process to recreate the file and also
3951 acquire a lock. This could lead to two Exim processes writing to the file
3952 at the same time. The added code performs the same test as UW imapd; it
3953 checks after acquiring the lock that its file descriptor still refers to
3954 the same named file.
3956 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3957 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3960 11. Added the smtp_active_hostname option. If used, this will typically be made
3961 to depend on the incoming interface address. Because $interface_address is
3962 not set up until the daemon has forked a reception process, error responses
3963 that can happen earlier (such as "too many connections") no longer contain
3966 12. If an expansion in a condition on a "warn" statement fails because a lookup
3967 defers, the "warn" statement is abandoned, and the next ACL statement is
3968 processed. Previously this caused the whole ACL to be aborted.
3970 13. Added the iplsearch lookup type.
3972 14. Added ident_timeout as a log selector.
3974 15. Added tls_certificate_verified as a log selector.
3976 16. Added a global option tls_require_ciphers (compare the smtp transport
3977 option of the same name). This controls incoming TLS connections.
3979 17. I finally figured out how to make tls_require_ciphers do a similar thing
3980 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3981 before starting the TLS session.
3983 18. Tabs are now shown as \t in -bP output.
3985 19. If the log selector return_path_on_delivery was set, Exim crashed when
3986 bouncing a message because it had too many Received: header lines.
3988 20. If two routers both had headers_remove settings, and the first one included
3989 a superfluous trailing colon, the final name in the first list and the
3990 first name in the second list were incorrectly joined into one item (with a
3991 colon in the middle).
3997 1. Added -C and -D options to the exinext utility, mainly to make it easier
3998 to include in the automated testing, but these could be helpful when
3999 multiple configurations are in use.
4001 2. The exinext utility was not formatting the output nicely when there was
4002 an alternate port involved in the retry record key, nor when there was a
4003 message id as well (for retries that were specific to a specific message
4004 and a specific host). It was also confused by IPv6 addresses, because of
4005 the additional colons they contain. I have fixed the IPv4 problem, and
4006 patched it up to do a reasonable job for IPv6.
4008 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4009 delivery, the log line now contains "pipelined" if PIPELINING was used.
4011 4. An SMTP transport process used to panic and die if the bind() call to set
4012 an explicit outgoing interface failed. This has been changed; it is now
4013 treated in the same way as a connect() failure.
4015 5. A reference to $sender_host_name in the part of a conditional expansion
4016 that was being skipped was still causing a DNS lookup. This no longer
4019 6. The def: expansion condition was not recognizing references to header lines
4020 that used bh_ and bheader_.
4022 7. Added the _cache feature to named lists.
4024 8. The code for checking quota_filecount in the appendfile transport was
4025 allowing one more file than it should have been.
4027 9. For compatibility with Sendmail, the command line option
4035 and sets the incoming protocol and host name (for trusted callers). The
4036 host name and its colon can be omitted when only the protocol is to be set.
4037 Note the Exim already has two private options, -pd and -ps, that refer to
4038 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4039 "s", but I don't think that's a major issue.
4041 10. A number of refactoring changes to the code, none of which should affect
4044 (a) The number of logging options was getting close to filling up the
4045 32-bit word that was used as a bit map. I have split them into two classes:
4046 those that are passed in the argument to log_write(), and those that are
4047 only ever tested independently outside of that function. These are now in
4048 separate 32-bit words, so there is plenty of room for expansion again.
4049 There is no change in the user interface or the logging behaviour.
4051 (b) When building, for example, log lines, the code previously used a
4052 macro that called string_cat() twice, in order to add two strings. This is
4053 not really sufficiently general. Furthermore, there was one instance where
4054 it was actually wrong because one of the argument was used twice, and in
4055 one call a function was used. (As it happened, calling the function twice
4056 did not affect the overall behaviour.) The macro has been replaced by a
4057 function that can join an arbitrary number of extra strings onto a growing
4060 (c) The code for expansion conditions now uses a table and a binary chop
4061 instead of a serial search (which was left over from when there were very
4062 few conditions). Also, it now recognizes conditions like "pam" even when
4063 the relevant support is not compiled in: a suitably worded error message is
4064 given if an attempt is made to use such a condition.
4066 11. Added ${time_interval:xxxxx}.
4068 12. A bug was causing one of the ddress fields not to be passed back correctly
4069 from remote delivery subprocesses. The field in question was not being
4070 subsequently used, so this caused to problems in practice.
4072 13. Added new log selectors queue_time and deliver_time.
4074 14. Might have fixed a bug in maildirsizefile handling that threw up
4075 "unexpected character" debug warnings, and recalculated the data
4076 unnecessarily. In any case, I expanded the warning message to give more
4079 15. Added the message "Restricted characters in address" to the statements in
4080 the default ACL that block characters like @ and % in local parts.
4082 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4083 Three changes have been made:
4085 (a) There was a serious bug; a negative response to MAIL caused the whole
4086 recipient domain to be cached as invalid, thereby blocking all messages
4087 to all local parts at the same domain, from all senders. This bug has
4088 been fixed. The domain is no longer cached after a negative response to
4089 MAIL if the sender used is not empty.
4091 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4094 (c) A new callout option, "use_sender" has been added for people who want
4095 the modified behaviour.
4101 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4104 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4105 indeed breaks things for older releases.
4107 3. Added additional logging to the case where there is a problem reading data
4108 from a filter that is running in a subprocess using a pipe, in order to
4109 try to track down a specific problem.
4111 4. Testing facility fudge: when running in the test harness and attempting
4112 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4113 getting "No route to host". Convert this to a timeout.
4115 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4118 6. Some OS don't have socklen_t but use size_t instead. This affects the
4119 fifth argument of getsockopt() amongst other things. This is now
4120 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4121 can be set for individual OS. I have set it for SunOS5, OSF1, and
4122 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4123 some earlier ones do not.
4125 7. Change 4.30/15 was not doing the test caselessly.
4127 8. The standard form for an IPv6 address literal was being rejected by address
4128 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4129 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4130 this, as well as the form without the "IPv6" on the front (but only when
4131 address literals are enabled, of course).
4133 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4135 10. Exim crashed if a message with an empty sender address specified by -f
4136 encountered a router with an errors_to setting. This could be provoked only
4137 by a command such as
4141 where an empty string was supplied; "<>" did not hit this bug.
4143 11. Installed PCRE release 4.5.
4145 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4146 remained set. It is now erased.
4148 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4149 times from message ids (which are base 36 rather than the normal 62).
4151 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4152 were being counted as actual protocol errors, and logged if the log
4153 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4154 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4155 and DATA following a set of rejected RCPTs do not count as protocol errors.
4156 In other words, Exim assumes they were pipelined, though this may not
4157 actually be the case. Of course, in all cases the client gets an
4158 appropriate error code.
4160 15. If a lookup fails in an ACL condition, a message about the failure may
4161 be available; it is used if testing the ACL cannot continue, because most
4162 such messages specify what the cause of the deferral is. However, some
4163 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4164 that caused an old message to be retained and used if a later statement
4165 caused a defer, replacing the real cause of the deferral.
4167 16. If an IP address had so many PTR records that the DNS lookup buffer
4168 was not large enough to hold them, Exim could crash while trying to process
4169 the truncated data. It now detects and logs this case.
4171 17. Further to 4.21/58, another change has been made: if (and only if) the
4172 first line of a message (the first header line) ends with CRLF, a bare LF
4173 in a subsequent header line has a space inserted after it, so as not to
4174 terminate the header.
4176 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4177 unnecessarily, used atoi() instead of strtol(), and didn't check the
4178 termination when getting file sizes from file names by regex.
4180 19. Completely re-implemented the support for maildirsize files, in the light
4181 of a number of problems with the previous contributed implementation
4182 (4.30/29). In particular:
4184 . If the quota is zero, the maildirsize file is maintained, but no quota is
4187 . If the maildir directory does not exist, it is created before any attempt
4188 to write a maildirsize file.
4190 . The quota value in the file is just a cache; if the quota is changed in
4191 the transport, the new value overrides.
4193 . A regular expression is available for excluding directories from the
4196 20. The autoreply transport checks the characters in options that define the
4197 message's headers; it allows continued headers, but it was checking with
4198 isspace() after an embedded newline instead of explicitly looking for a
4201 21. If all the "regular" hosts to which an address was routed had passed their
4202 expiry times, and had not reached their retry times, the address was
4203 bounced, even if fallback hosts were defined. Now Exim should go on to try
4206 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4207 equivalent code in the SMTP transport. Some hosts send humungous responses
4208 to HELO/EHLO, more than 1024 it seems.
4210 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4211 gives compiler warnings in some environments. I've now done it "properly",
4214 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4215 (because of problems with the built-in one) was declared to return uschar *
4216 instead of char *, causing compiler failure.
4218 25. Fixed a file descriptor leak when processing alias/forward files.
4220 26. Fixed a minor format string issue in dbfn.c.
4222 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4224 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4225 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4228 29. When a sender address is verified, it is cached, to save repeating the test
4229 when there is more than one recipient in a message. However, when the
4230 verification involves a callout, it is possible for different callout
4231 options to be set for different recipients. It is too complicated to keep
4232 track of this in the cache, so now Exim always runs a verification when a
4233 callout is required, relying on the callout cache for the optimization.
4234 The overhead is duplication of the address routing, but this should not be
4237 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4238 to be verified with callout=postmaster, and the main callout worked but the
4239 postmaster check failed, the verification correctly failed. However, if a
4240 subsequent RCPT command asked for sender verification *without* the
4241 postmaster check, incorrect caching caused this verification also to fail,
4244 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4245 it was not caching the DNS options (qualify_single, search_parents) that
4246 were used when the lookup failed. A subsequent lookup with different
4247 options therefore always gave the same answer, though there were cases
4248 where it should not have. (Example: a "domains = !$mx_any" option on a
4249 dnslookup router: the "domains" option is always processed without any
4250 widening, but the router might have qualify_single set.) Now Exim uses the
4251 cached value only when the same options are set.
4253 32. Added John Jetmore's "exipick" utility to the distribution.
4255 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4256 than a timeout (e.g. a certificate is required, and is not provided), an
4257 Exim server now closes the connection immediately. Previously it waited for
4258 the client to close - but if the client is SSL, it seems that they each
4259 wait for each other, leading to a delay before one of them times out.
4261 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4262 maintained 0.8.x compatibility because I don't think many are using it, and
4263 it is clearly obsolete.
4265 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4268 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4269 client certificate was expired. A simple patch fixes this, though I don't
4270 understand the full logic of why the verify callback is called multiple
4273 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4274 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4275 which causes problems with some clients (such as the Certicom SSL Plus
4276 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4277 disables the coutermeasure allowing Eudora to connect."
4279 38. Exim was not checking that a write() to a log file succeeded. This could
4280 lead to Bad Things if a log got too big, in particular if it hit a file
4281 size limit. Exim now panics and dies if it cannot write to a log file, just
4282 as it does if it cannot open a log file.
4284 39. Modified OS/Makefile-Linux so that it now contains
4286 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4288 The two -D definitions ensure that Exim is compiled with large file
4289 support, which makes it possible to handle log files that are bigger than
4292 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4293 instance) a domain was checked against a named list that involved a lookup,
4294 causing $domain_data to be set, then another domain was checked against the
4295 same list, then the first domain was re-checked, the value of $domain_data
4296 after the final check could be wrong. In particular, if the second check
4297 failed, it could be set empty. This bug probably also applied to
4300 41. The strip_trailing_dot option was not being applied to the address given
4301 with the -f command-line option.
4303 42. The code for reading a message's header from the spool was incrementing
4304 $received_count, but never initializing it. This meant that the value was
4305 incorrect (doubled) while delivering a message in the same process in which
4306 it was received. In the most common configuration of Exim, this never
4307 happens - a fresh exec is done - but it can happen when
4308 deliver_drop_privilege is set.
4310 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4311 it now includes up to 150 characters of the unexpected data in the log
4314 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4315 and building data strings. The size of both of these buffers was 10 000
4316 bytes - far larger than anybody would *ever* want, thought I. Needless to
4317 say, somebody hit the limit. I have increased the maximum line length to
4318 20 000 and the maximum data length of concatenated lines to 100 000. I have
4319 also fixed two bugs, because there was no checking on these buffers. Tsk,
4320 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4321 buffer is too small.
4323 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4324 lsearch lookups. Now it does.
4326 46. When parsing a route_list item in a manualroute router, a fixed-length
4327 buffer was used for the list of hosts. I made this 1024 bytes long,
4328 thinking that nobody would ever have a list of hosts that long. Wrong.
4329 Somebody had a whole pile of complicated expansion conditions, and the
4330 string was silently truncated, leading to an expansion error. It turns out
4331 that it is easier to change to an unlimited length (owing to other changes
4332 that have happened since this code was originally written) than to build
4333 structure for giving a limitation error. The length of the item that
4334 expands into the list of hosts is now unlimited.
4336 47. The lsearch lookup could not handle data where the length of text line was
4337 more than 4095 characters. Such lines were truncated, leading to shortened
4338 data being returned. It should now handle lines of any length.
4340 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4341 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4344 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4345 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4346 no longer happens in comment lines. A list of replacements is now placed
4347 at the head of all of the source files, except those whose only change is
4348 to replace PERL_COMMAND in the very first #! line.
4350 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4351 messages on the queue, with a bottom-up merge sort, using code contributed
4352 by Michael Haardt. This should make operations like -bp somewhat faster on
4353 large queues. It won't affect queue runners, except when queue_run_in_order
4356 51. Installed eximstats 1.31 in the distribution.
4358 52. Added support for SRV lookups to the dnslookup router.
4360 53. If an ACL referred to $message_body or $message_body_end, the value was not
4361 reset for any messages that followed in the same SMTP session.
4363 54. The store-handling optimization for building very long strings was not
4364 differentiating between the different store pools. I don't think this
4365 actually made any difference in practice, but I've tidied it.
4367 55. While running the routers to verify a sender address, $sender_address
4368 was still set to the sender address. This is wrong, because when routing to
4369 send a bounce to the sender, it would be empty. Therefore, I have changed
4370 it so that, while verifying a sender address, $sender_address is set to <>.
4371 (There is no change to what happens when verifying a recipient address.)
4373 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4374 target A or AAAA records (if not already returned) without resetting the
4375 qualify_single or search_parents options of the DNS resolver. These are
4376 inappropriate in this case because the targets of MX and SRV records must
4377 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4378 target that, when qualified, matched something in the local domain. These
4379 two options are now turned off when doing these lookups.
4381 57. It seems that at least some releases of Reiserfs (which does not have the
4382 concept of a fixed number of inodes) returns zero and not -1 for the
4383 number of available inodes. This interacted badly with check_spool_inodes,
4384 which assumed that -1 was the "no such thing" setting. What I have done is
4385 to check that the total number of inodes is greater than zero before doing
4386 the test of how many are available.
4388 58. When a "warn" ACL statement has a log_message modifier, the message is
4389 remembered, and not repeated. This is to avoid a lot of repetition when a
4390 message has many recipients that cause the same warning to be written.
4391 Howewer, Exim was preserving the list of already written lines for an
4392 entire SMTP session, which doesn't seem right. The memory is now reset if a
4393 new message is started.
4395 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4396 debugging output unless log_rewrite was also set.
4398 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4399 of (int)(handle) when we know that handle contains (void *)(-1).
4401 61. The Exim daemon panic-logs an error return when it closes the incoming
4402 connection. However "connection reset by peer" seems to be common, and
4403 isn't really an error worthy of noting specially, so that particular error
4406 62. When Exim is trying to find all the local interfaces, it used to panic and
4407 die if the ioctl to get the interface flags failed. However, it seems that
4408 on at least one OS (Solaris 9) it is possible to have an interface that is
4409 included in the list of interfaces, but for which you get a failure error
4410 for this call. This happens when the interface is not "plumbed" into a
4411 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4412 failure of the "get flags" call assumes that the interface is down.
4414 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4415 makes life easier for people who are doing arithmetic on fields extracted
4416 from dates, where you often get leading zeros that should not be
4417 interpreted as octal.
4419 64. Added qualify_domain to the redirect router, to override the global
4422 65. If a pathologically long header line contained very many addresses (the
4423 report of this problem mentioned 10 000) and each of them was rewritten,
4424 Exim could use up a very large amount of memory. (It kept on making new
4425 copies of the header line as it rewrote, and never released the old ones.)
4426 At the expense of a bit more processing, the header rewriting function has
4427 been changed so that it no longer eats memory in this way.
4429 66. The generation of the Received: header has been moved from the time that a
4430 message starts to be received, to the time that it finishes. The timestamp
4431 in the Received: header should now be very close to that of the <= log
4432 line. There are two side-effects of this change:
4434 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4435 logged header lines no longer include the local Received: line, because
4436 it has not yet been created. The same applies to a copy of the message
4437 that is returned to a non-SMTP sender when a message is rejected.
4439 (b) When a filter file is tested using -bf, no additional Received: header
4440 is added to the test message. After some thought, I decided that this
4443 This change does not affect the value of $received_for. It is still set
4444 after address rewriting, but before local_scan() is called.
4446 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4448 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4449 gave an unhelpful panic error message, and a defer error. I have managed to
4450 change this behaviour so that it now rejects any supplied certificate,
4451 which seems right, as the list of acceptable certificates is empty.
4453 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4454 gave an unhelpful defer error. I have not managed to make this reject any
4455 supplied certificates, but the error message it gives is "no certificate
4456 supplied", which is not helpful.
4458 70. exigrep's output now also includes lines that are not associated with any
4459 message, but which match the given pattern. Implemented by a patch from
4460 Martin Sluka, which also tidied up the Perl a bit.
4462 71. Recipient callout verification, like sender verification, was using <> in
4463 the MAIL FROM command. This isn't really the right thing, since the actual
4464 sender may affect whether the remote host accepts the recipient or not. I
4465 have changed it to use the actual sender in the callout; this means that
4466 the cache record is now keyed on a recipient/sender pair, not just the
4467 recipient address. There doesn't seem to be a real danger of callout loops,
4468 since a callout by the remote host to check the sender would use <>.
4469 [SEE ABOVE: changed after hitting problems.]
4471 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4472 temporary errors. However, in the case of such a code being given after
4473 the end of a data transmission (i.e. after ".") Exim was failing to write
4474 a retry record for the message. (Yes, there was some broken host that was
4475 actually sending 8xx at this point.)
4477 73. An unknown lookup type in a host list could cause Exim to panic-die when
4478 the list was checked. (An example that provoked this was putting <; in the
4479 middle of a list instead of at the start.) If this happened during a DATA
4480 ACL check, a -D file could be left lying around. This kind of configuration
4481 error no longer causes Exim to die; instead it causes a defer errror. The
4482 incident is still logged to the main and panic logs.
4484 74. Buglet left over from Exim 3 conversion. The message "too many messages
4485 in one connection" was written to the rejectlog but not the mainlog, except
4486 when address rewriting (yes!) was being logged.
4488 75. Added write_rejectlog option.
4490 76. When a system filter was run not as root (that is, when system_filter_user
4491 was set), the values of the $n variables were not being returned to the
4492 main process; thus, they were not subsequently available in the $sn
4495 77. Added +return_path_on_delivery log selector.
4497 78. A connection timeout was being treated differently from recipients deferred
4498 when testing hosts_max_try with a message that was older than the host's
4499 retry timeout. (The host should not be counted, thus allowing all hosts to
4500 be tried at least once before bouncing.) This may have been the cause of an
4501 occasionally reported bug whereby a message would remain on the queue
4502 longer than the retry timeout, but would be bounced if a delivery was
4503 forced. I say "may" because I never totally pinned down the problem;
4504 setting up timeout/retry tests is difficult. See also the next item.
4506 79. The ultimate address timeout was not being applied to errors that involved
4507 a combination of host plus message (for example, a timeout on a MAIL
4508 command). When an address resolved to a number of possible hosts, and they
4509 were not all tried for each delivery (e.g. because of hosts_max_try), a
4510 message could remain on the queue longer than the retry timeout.
4512 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4515 81. Fixed an obscure SMTP outgoing bug which required at least the following
4516 conditions: (a) there was another message waiting for the same server;
4517 (b) the server returned 5xx to all RCPT commands in the first message so
4518 that the message was not completed; (c) the server dropped the connection
4519 or gave a negative response to the RSET that Exim sends to abort the
4520 transaction. The observed case was a dropped connection after DATA that had
4521 been sent in pipelining mode. That is, the server had advertised PIPELINING
4522 but was not implementing it correctly. The effect of the bug was incorrect
4523 behaviour, such as trying another host, and this could lead to a crash.
4529 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4530 and daemon.c were passed as pointers to ints; they should have been
4531 pointers to socklen_t variables (which are typically unsigned ints).
4533 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4536 3. Fixed a really odd bug that affected only the testing scheme; patching a
4537 certain fixed string in the binary changed the value of another string that
4538 happened to be identical to the end of the original first string.
4540 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4541 name", it returns that address as the IP address. On some operating
4542 systems (e.g. Solaris), it also passes back the IP address string as the
4543 "host name". However, on others (e.g. Linux), it passes back an empty
4544 string. Exim wasn't checking for this, and was changing the host name to an
4545 empty string, assuming it had been canonicized.
4547 5. Although rare, it is permitted to have more than one PTR record for a given
4548 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4549 all the names associated with an address, because they do in Solaris.
4550 However, it seems that they do not in Linux for data that comes from the
4551 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4552 I found this out when I moved to a new Linux workstation and tried to run
4553 the Exim test suite.
4555 To get round this problem I have changed the code so that it now does its
4556 own call to the DNS to look up PTR records when searching for a host name.
4557 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4558 addresses that are only in /etc/hosts are still found.
4560 This behaviour is, however, controlled by an option called host_lookup_
4561 order, which defaults to "bydns:byaddr". If people want to use the other
4562 order, or indeed, just use one or the other means of lookup, they can
4563 specify it in this variable.
4565 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4566 some operating systems, this comes back from gethostbyaddr() as an empty
4567 string, and this is what Exim used to test for. However, it seems that in
4568 other systems, "." is yielded. Exim now tests for this case too.
4570 7. The values of check_spool_space and check_log_space are now held internally
4571 as a number of kilobytes instead of an absolute number of bytes. If a
4572 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4573 kilobyte. This means that much larger values can be stored.
4575 8. Exim monitor: an attempt to get the action menu when not actually pointing
4576 at a message produces an empty menu entitled "No message selected". This
4577 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4578 no entries in it ("Shell widget menu has zero width and/or height"). So I
4579 have added a single, blank menu entry in this case.
4581 9. Added ${quote_local_part.
4583 10. MIME decoding is now applied to the contents of Subject: header lines when
4586 11. Now that a reference to $sender_host_address automatically causes a reverse
4587 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4588 host lookup before query-style lookups in lists that might use this
4589 variable. This has therefore been abolished, and the "net-" prefix is no
4590 longer necessary for query-style lookups.
4592 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4593 have been a typo for LFLAGS, so it has been changed.
4595 13. The install script calls Exim with "-C /dev/null" in order to find the
4596 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4597 to be output. Howeve, since Exim outputs its version number before the
4598 error, it didn't break the script. It just looked ugly. I fixed this by
4599 always allowing "-C /dev/null" if the caller is root.
4601 14. Ignore overlarge ACL variable number when reading spool file - insurance
4602 against a later release with more variables having written the file.
4604 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4605 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4606 this, as well as the form without the "IPv6" on the front.
4608 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4609 OS/Makefile-Darwin file.
4611 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4612 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4614 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4615 variable. However, some OS use ints for inodes. I've added cast to long int
4616 to get rid of the compiler warning.
4618 19. I had forgotten to lock out "/../" in configuration file names when
4619 ALT_CONFIG_PREFIX was set.
4621 20. Routers used for verification do not need to specify transports. However,
4622 if such a router generated a host list, and callout was configured, Exim
4623 crashed, because it could not find a port number from the (non-existent)
4624 transport. It now assumes port 25 in this circumstance.
4626 21. Added the -t option to exigrep.
4628 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4629 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4630 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4631 them is set, LOOKUP_LSEARCH is forced.
4633 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4635 24. Added sender and host information to the "rejected by local_scan()" log
4636 line; previously there was no indication of these.
4638 25. Added .include_if_exists.
4640 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4641 Linux. It turns out that not all file systems support this. Apparently some
4642 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4643 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4644 not supported on the file descriptor, is now ignored when Exim is trying to
4645 sync a directory. This applies only to Linux.
4647 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4649 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4650 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4653 29. Added support for maildirsize files from supplied patch (modified a bit).
4655 30. The use of :fail: followed by an empty string could lead Exim to respond to
4656 sender verification failures with (e.g.):
4658 550 Verification failed for <xxx>
4659 550 Sender verify failed
4661 where the first response line was missing the '-' that indicates it is not
4662 the final line of the response.
4664 31. The loop for finding the name of the user that called Exim had a hardwired
4665 limit of 10; it now uses the value of finduser_retries, which is used for
4666 all other user lookups.
4668 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4671 33. Exim was neglecting to zero errno before one call of strtol() when
4672 expanding a string and expecting an integer value. On some systems this
4673 resulted in spurious "integer overflow" errors. Also, it was casting the
4674 result into an int without checking.
4676 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4677 did not work. The code looks as if it has *never* worked, though it appears
4678 to have been documented since at least releast 1.62. I have made it work.
4680 35. The "timeout_DNS" error in retry rules, also documented since at least
4681 1.62, also never worked. As it isn't clear exactly what this means, and
4682 clearly it isn't a major issue, I have abolished the feature by treating it
4683 as "timeout", and writing a warning to the main and panic logs.
4685 36. The display of retry rules for -brt wasn't always showing the error code
4688 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4689 timeout_connect_A, timeout_connect_MX.
4691 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4692 to the empty sender.
4694 39. The daemon was not analysing the content of -oX till after it had closed
4695 stderr and disconnected from the controlling terminal. This meant that any
4696 syntax errors were only noted on the panic log, and the return code from
4697 the command was 0. By re-arranging the code a little, I've made the
4698 decoding happen first, so such errors now appear on stderr, and the return
4699 code is 1. However, the actual setting up of the sockets still happens in
4700 the disconnected process, so errors there are still only recorded on the
4703 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4704 connections (as happens on some IP stacks) was logged at start up time as
4705 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4706 it from "IPv6 and IPv4", which means that two separate sockets are being
4709 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4710 says whether AF_INET or AF_INET6 was passed as an argument.
4712 42. Exiwhat output was messed up when time zones were included in log
4715 43. Exiwhat now gives more information about the daemon's listening ports,
4716 and whether -tls-on-connect was used.
4718 44. The "port" option of the smtp transport is now expanded.
4720 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4721 silently ignored. Now an error message is written to the main and panic
4724 46. There's a new ACL modifier called "logwrite" which writes to a log file
4725 as soon as it is encountered.
4727 47. Added $local_user_uid and $local_user_gid at routing time.
4729 48. Exim crashed when trying to verify a sender address that was being
4732 49. Exim was recognizing only a space character after ".include". It now also
4733 recognizes a tab character.
4735 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4736 extracting the relevant information from the specification. The man page no
4737 longer contains scrambled data for the -d option, and I've added a section
4738 at the front about calling Exim under different names.
4740 51. Added "extra_headers" argument to the "mail" command in filter files.
4742 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4745 53. Installed eximstats 1.29.
4747 54. Added transport_filter_timeout as a generic transport option.
4749 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4750 Cc: header lines. This was required by RFC 822, but it not required by RFC
4753 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4754 incoming messages that did not have them. Now it does so only if the
4755 message originates locally, that is, if there is no associated remote host
4756 address. When Resent- header lines are present, this applies to the Resent-
4757 lines rather than the non-Resent- lines.
4759 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4760 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4762 58. Messages for configuration errors now include the name of the main
4763 configuration file - useful now that there may be more than one file in a
4764 list (.included file names were always shown).
4766 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4767 for those rare installations that do not start the daemon as root or run it
4768 setuid root. I've cut out the call to initgroups() if the daemon is not
4771 60. The Exim user and group can now be bound into the binary as text strings
4772 that are looked up at the start of Exim's processing.
4774 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4776 62. Added $mailstore_basename variable.
4778 63. Installed patch to sieve.c from Michael Haardt.
4780 64. When Exim failed to open the panic log after failing to open the main log,
4781 the original message it was trying to log was written to stderr and debug
4782 output, but if they were not available (the usual case in production), it
4783 was lost. Now it is written to syslog before the two lines that record the
4784 failures to open the logs.
4786 65. Users' Exim filters run in subprocesses under the user's uid. It is
4787 possible for a "deliver" command or an alias in a "personal" command to
4788 provoke an address rewrite. If logging of address rewriting is configured,
4789 this fails because the process is not running as root or exim. There may be
4790 a better way of dealing with this, but for the moment (because 4.30 needs
4791 to be released), I have disabled address rewrite logging when running a
4792 filter in a non-root, non-exim process.
4798 1. The buildconfig auxiliary program wasn't quoting the value set for
4799 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4800 not defined. This bug was masked in 4.22 by the effect that was fixed in
4803 2. Some messages that were rejected after a message id was allocated were
4804 shown as "incomplete" by exigrep. It no longer does this for messages that
4805 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4807 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4808 have allow_domain_literals set, the ID did not get logged in the <= line.
4809 Domain literals are now always recognized in Message-ID: header lines.
4811 4. The first argument for a ${extract expansion item is the key name or field
4812 number. Leading and trailing spaces in this item were not being ignored,
4813 causing some misleading effects.
4815 5. When deliver_drop_privilege was set, single queue runner processes started
4816 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4817 same command in the process it spins off) were not dropping privilege.
4819 6. When the daemon running as "exim" started a queue runner, it always
4820 re-executed Exim in the spun-off process. This is a waste of effort when
4821 deliver_drop_privilege is set. The new process now just calls the
4822 queue-runner function directly.
4828 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4831 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4832 ignored. Though the use of -oP was forcing the writing of a pid file, it
4833 was always written to the default place.
4835 3. If the message "no IP address found for host xxxx" is generated during
4836 incoming verification, it is now followed by identification of the incoming
4837 connection (so you can more easily find what provoked it).
4839 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4841 5. Added some features to "harden" Exim a bit more against certain attacks:
4843 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4844 be put in Local/Makefile. This is like the never_users runtime option,
4845 but it cannot be overridden. The default setting is "root".
4847 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4848 prefix string with which any file named in a -C command line option
4851 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4852 is retained for -C and -D only if the caller of Exim is root. Without
4853 it, the exim user may also use -C and -D and retain privilege.
4855 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4856 command line option is disabled.
4858 6. Macro names set by the -D option must start with an upper case letter, just
4859 like macro names defined in the configuration file.
4861 7. Added "dereference=" facility to LDAP.
4863 8. Two instances of the typo "uknown" in the source files are fixed.
4865 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4866 the Configure-Makefile script screwed up while processing it.
4868 10. Incorporated PCRE 4.4.
4870 11. The SMTP synchronization check was not operating right at the start of an
4871 SMTP session. For example, it could not catch a HELO sent before the client
4872 waited for the greeting. There is now a check for outstanding input at the
4873 point when the greeting is written. Because of the duplex, asynchronous
4874 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4875 way, but not yet received, when the check is performed.
4877 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4878 on TCP/IP sockets, because this apparently causes some broken clients to
4881 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4882 unchanged) from the Cygwin maintainer.
4884 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4885 was supported instead of testing for mbx. Effectively a typo.
4887 15. The spa authenticator server code was not checking that the input it
4888 received was valid base64.
4890 16. The debug output line for the "set" modifier in ACLs was not showing the
4891 name of the variable that was being set.
4893 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4895 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4896 lookup on its own. Something else was needed to trigger the lookup. For
4897 example, a match in host_lookup or the need for a host name in a host list.
4898 Now, if $sender_host_name is referenced and the host name has not yet been
4899 looked up, a lookup is performed. If the lookup fails, the variable remains
4900 empty, and $host_lookup_failed is set to "1".
4902 19. Added "eqi" as a case-independent comparison operator.
4904 20. The saslauthd authentication condition could segfault if neither service
4905 nor realm was specified.
4907 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4908 error message that was logged was misleading, and incoming SMTP
4909 connections were dropped. The message is now more accurate, and temporary
4910 errors are given to SMTP connections.
4912 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4913 (except RSET) with a 421 error, until QUIT is received. However, it was
4914 failing to send a response to QUIT.
4916 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4917 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4918 ACL is now run after the helo_try_verify_hosts code.
4920 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4921 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4922 case-independent, so other case variants are also recognized). Apparently
4923 some systems use these upper case variants.
4925 25. If more than two messages were waiting for the same host, and a transport
4926 filter was specified for the transport, Exim sent two messages over the
4927 same TCP/IP connection, and then failed with "socket operation on non-
4928 socket" when it tried to send the third.
4930 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4932 27. The extern definition of crypt16() in expand.c was not being excluded when
4933 the OS had its own crypt16() function.
4935 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4936 as a preferred synonym for return_size_limit, both as an option and as an
4939 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4941 30. Changed the default configuration ACL to relax the local part checking rule
4942 for addresses that are not in any local domains. For these addresses,
4943 slashes and pipe symbols are allowed within local parts, but the sequence
4944 /../ is explicitly forbidden.
4946 31. SPA server authentication was not clearing the challenge buffer before
4949 32. log_message in a "warn" ACL statement was writing to the reject log as
4950 well as to the main log, which contradicts the documentation and doesn't
4951 seem right (because no rejection is happening). So I have stopped it.
4953 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4954 However, I am unable to do any testing of this.
4956 34. Fixed an infelicity in the appendfile transport. When checking directories
4957 for a mailbox, to see if any needed to be created, it was accidentally
4958 using path names with one or more superfluous leading slashes; tracing
4959 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4961 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4962 contents are added to the log line that is written for every discarded
4963 recipient. (Previously a log_message setting was ignored.)
4965 36. The ${quote: operator now quotes the string if it is empty.
4967 37. The install script runs exim in order to find its version number. If for
4968 some reason other than non-existence or emptiness, which it checks, it
4969 could not run './exim', it was installing it with an empty version number,
4970 i.e. as "exim-". This error state is now caught, and the installation is
4973 38. An argument was missing from the function that creates an error message
4974 when Exim fails to connect to the socket for saslauthd authentication.
4975 This could cause Exim to crash, or give a corrupted message.
4977 39. Added isip, isip4, and isip6 to ${if conditions.
4979 40. The ACL variables $acl_xx are now saved with the message, and can be
4980 accessed later in routers, transports, and filters.
4982 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4983 strings in the file are not string-expanded.
4985 42. If a MAIL command specified a SIZE value that was too large to fit into an
4986 int variable, the check against message_size_limit failed. Such values are
4987 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4988 day this will have to be increased, but I don't think I want to be around
4989 when emails are that large.
4996 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4997 iconv() is not standard in FreeBSD.
4999 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5000 IPv6 enabled. The observed symptom was a segmentation fault on return from
5001 the function os_common_find_running_interfaces() in src/os.c.
5003 3. In the check_special_case() function in daemon.c I had used "errno" as an
5004 argument name, which causes warnings on some systems. This was basically a
5005 typo, since it was named "eno" in the comments!
5007 4. The code that waits for the clock to tick (at a resolution of some fraction
5008 of a second) so as to ensure message-id uniqueness was always waiting for
5009 at least one whole tick, when it could have waited for less. [This is
5010 almost certainly not relevant at current processor speeds, where it is
5011 unlikely to ever wait at all. But we try to future-proof.]
5013 5. The function that sleeps for a time interval that includes fractions of a
5014 second contained a race. It did not block SIGALRM between setting the
5015 timer, and suspending (a couple of lines later). If the interval was short
5016 and the sigsuspend() was delayed until after it had expired, the suspension
5017 never ended. On busy systems this could lead to processes getting stuck for
5020 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5021 process, before it forks any delivery processes. The open lookup caching
5022 mechanism meant that the open file or database connection was passed into
5023 the delivery process. The problem was that delivery processes always tidy
5024 up cached lookup data. This could cause a problem for the next delivery
5025 process started by the queue runner, because the external queue runner
5026 process does not know about the closure. So the next delivery process
5027 still has data in the lookup cache. In the case of a file lookup, there was
5028 no problem because closing a file descriptor in a subprocess doesn't affect
5029 the parent. However, if the lookup was caching a connection to a database,
5030 the connection was closed, and the second delivery process was likely to
5031 see errors such as "PGSQL: query failed: server closed the connection
5032 unexpectedly". The problem has been fixed by closing all cached lookups
5033 in a queue runner before running a delivery process.
5035 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5036 seem to have the "const" qualifier which it has on other OS. I've
5039 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5040 *of the same type* (client or server) with the same public name that an
5041 error should be diagnosed.
5043 9. When Exim looked up a host name for an IP address, but failed to find the
5044 original IP address when looking up the host name (a safety check), it
5045 output the message "<ip address> does not match any IP for NULL", which was
5046 confusing, to say the least. The bug was that the host name should have
5047 appeared instead of "NULL".
5049 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5050 user that is built into the binary, and the -C or -D options is used, root
5051 privilege is dropped before the configuration file is read. In addition,
5052 logging is switched to stderr instead of the normal log files. If the
5053 configuration then re-defines the Exim user, the unprivileged environment
5054 is probably not what is expected, so Exim logs a panic warning message (but
5057 However, if deliver_drop_privilege is set, the unprivileged state may well
5058 be exactly what is intended, so the warning has been cut out in that case,
5059 and Exim is allowed to try to write to its normal log files.
5065 1. smtp_return_error_details was not giving details for temporary sender
5066 or receiver verification errors.
5068 2. Diagnose a configuration error if two authenticators have the same public
5071 3. Exim used not to create the message log file for a message until the first
5072 delivery attempt. This could be confusing when incoming messages were held
5073 for policy or load reasons. The message log file is now created at the time
5074 the message is received, and an initial "Received" line is written to it.
5076 4. The automatically generated man page for command line options had a minor
5077 bug that caused no ill effects; however, a more serious problem was that
5078 the procedure for building the man page automatically didn't always
5079 operate. Consequently, release 4.20 contains an out-of-date version. This
5080 shouldn't happen again.
5082 5. When building Exim with embedded Perl support, the script that builds the
5083 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5084 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5086 6. The freeze_tell option was not being used for messages that were frozen on
5087 arrival, either by an ACL or by local_scan().
5089 7. Added the smtp_incomplete_transaction log selector.
5091 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5092 was accepting AUTH without a new EHLO.
5094 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5095 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5098 10. Logging of TCP/IP connections (when configured) now happens in the main
5099 daemon process instead of the child process, so that the TCP/IP connection
5100 count is more accurate (but it can never be perfect).
5102 11. The use of "drop" in a nested ACL was not being handled correctly in the
5103 outer ACL. Now, if condition failure induced by the nested "drop" causes
5104 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5105 or "require"), the connection is dropped.
5107 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5108 that yield "discard" can now be used with an "accept" or a "discard" verb,
5109 but an error is generated for any others (because I can't see a useful way
5110 to define what should happen).
5112 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5113 are now processed in the same way as lines in the Exim configuration file.
5114 In particular, continuation lines are supported.
5116 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5118 15. Added -ti meaning -t -i.
5120 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5121 domains, and warn by logging if others are found.
5123 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5124 returned from GIFCONF when Exim is trying to find the list of interfaces on
5125 a host. The code in os.c has been modified to copy each ifreq to an aligned
5126 structure in all cases.
5128 Also, in some cases, the returned ifreq's were being copied to a 'struct
5129 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5130 means the last couple of bytes of an IPv6 address could be chopped if the
5131 ifreq contained only a normal sockaddr (14 bytes storage).
5133 18. Named domain lists were not supported in the hosts_treat_as_local option.
5134 An entry such as +xxxx was not recognized, and was treated as a literal
5137 19. Ensure that header lines added by a DATA ACL are included in the reject log
5138 if the ACL subsequently rejects the message.
5140 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5141 MD5 (which is deprecated).
5143 21. When testing a filter file using -bf, Exim was writing a message when it
5144 took the sender from a "From " line in the message, but it was not doing so
5145 when it took $return_path from a Return-Path: header line. It now does.
5147 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5148 with a valid header line field name (a series of printing characters
5149 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5151 23. Changed "disc" in the source to "disk" to conform to the documentation and
5152 the book and for uniformity.
5154 24. Ignore Sendmail's -Ooption=value command line item.
5156 25. When execve() failed while trying to run a command in a pipe transport,
5157 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5158 could be confused with a return value of 69 from the command itself. This
5159 has been changed to 127, the value the shell returns if it is asked to run
5160 a non-existent command. The wording for the related log line suggests a
5161 non-existent command as the problem.
5163 26. If received_header_text expands to an empty string, do not add a Received:
5164 header line to the message. (Well, it adds a token one on the spool, but
5165 marks it "old" so that it doesn't get used or transmitted.)
5167 27. Installed eximstats 1.28 (addition of -nt option).
5169 28. There was no check for failure on the call to getsockname() in the daemon
5170 code. This can fail if there is a shortage of resources on the system, with
5171 ENOMEM, for example. A temporary error is now given on failure.
5173 29. Contrary to the C standard, it seems that in some environments, the
5174 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5175 program. Exim now does this explicitly; it affects the formatting of
5176 timestamps using strftime().
5178 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5179 complaints. I've now initialized all the variables, to avoid this.
5181 32. Header lines added by a system filter were not being "seen" during
5182 transport-time rewrites.
5184 33. The info_callback() function passed to OpenSSL is set up with type void
5185 (*)(SSL *, int, int), as described somewhere. However, when calling the
5186 function (actually a macro) that sets it up, the type void(*)() is
5187 expected. I've put in a cast to prevent warnings from picky compilers.
5189 34. If a DNS black list lookup found a CNAME record, but there were no A
5190 records associated with the domain it pointed at, Exim crashed.
5192 35. If a DNS black list lookup returned more than one A record, Exim ignored
5193 all but the first. It now scans all returned addresses if a particular IP
5194 value is being sought. In this situation, the contents of the
5195 $dnslist_value variable are a list of all the addresses, separated by a
5198 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5199 to accept a host name and all its aliases if the forward lookup for any of
5200 them yielded the IP address of the incoming connection. Now it accepts only
5201 those names whose forward lookup yields the correct IP address. Any other
5202 names are discarded. This closes a loophole whereby a rogue DNS
5203 administrator could create reverse DNS records to break through a
5204 wildcarded host restriction in an ACL.
5206 37. If a user filter or a system filter that ran in a subprocess used any of
5207 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5208 the wrong values were passed to the pipe command ($thisaddress had the
5209 value of $0, $0 had the value of $1, etc). This bug was introduced by
5210 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5212 38. Improved the line breaking for long SMTP error messages from ACLs.
5213 Previously, if there was no break point between 40 and 75 characters, Exim
5214 left the rest of the message alone. Two changes have been made: (a) I've
5215 reduced the minimum length to 35 characters; (b) if it can't find a break
5216 point between 35 and 75 characters, it looks ahead and uses the first one
5217 that it finds. This may give the occasional overlong line, but at least the
5218 remaining text gets split now.
5220 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5221 file descriptors might be low, and that setting 1000 would always raise it.
5222 It turns out that in some environments, the limit is already over 1000 and
5223 that lowering it causes trouble. So now Exim takes care not to decrease it.
5225 40. When delivering a message, the value of $return_path is set to $sender_
5226 address at the start of routing (routers may change the value). By an
5227 oversight, this default was not being set up when an address was tested by
5228 -bt or -bv, which affected the outcome if any router or filter referred to
5231 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5232 log only when "message" or "log_message" are set. However, if one of the
5233 conditions was an address verification, or a call to a nested ACL, the
5234 messages generated by the underlying test were being passed through. This
5235 no longer happens. The underlying message is available in $acl_verify_
5236 message for both "message" and "log_message" expansions, so it can be
5237 passed through if needed.
5239 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5240 new expansion $bh_ to give the encoded byte string without charset
5241 translation. Translation happens only if iconv() is available; HAVE_ICONV
5242 indicates this at build time. HEADERS_CHARSET gives the charset to
5243 translate to; headers_charset can change it in the configuration, and
5244 "headers charset" can change it in an individual filter file.
5246 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5247 that creates RFC 2047 encoded "words" labels them as that charset instead
5248 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5249 expansion operator; (ii) when Exim creates a From: line for a local
5250 message; (iii) when a header line is rewritten to include a "phrase" part.
5252 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5253 buggy, causing it to skip the first lines of messages whose message ID
5254 ended in 'D'. This would not have bitten before Exim release 4.14, because
5255 message IDs were unlikely to end in 'D' before then. The effect was to have
5256 incorrect size information for certain domains.
5258 45. #include "config.h" was missing at the start of the crypt16.c module. This
5259 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5262 46. If there was a timeout during a "random" callout check, Exim treated it as
5263 a failure of the random address, and carried on sending RSET and the real
5264 address. If the delay was just some slowness somewhere, the response to the
5265 original RCPT would be taken as a response to RSET and so on, causing
5266 mayhem of various kinds.
5268 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5269 when I implemented it. It didn't allow for the fact that some option values
5270 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5271 the right test for positive values.
5273 48. Domain names in DNS records are case-independent. Exim always looks them up
5274 in lower case. Some resolvers return domain names in exactly the case they
5275 appear in the zone file, that is, they may contain uppercase letters. Not
5276 all resolvers do this - some return always lower case. Exim was treating a
5277 change of case by a resolver as a change of domain, similar to a widening
5278 of a domain abbreviation. This triggered its re-routing code and so it was
5279 trying to route what was effectively the same domain again. This normally
5280 caused routing to fail (because the router wouldn't handle the domain
5281 twice). Now Exim checks for this case specially, and just changes the
5282 casing of the domain that it ultimately uses when it transmits the message
5285 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5288 50. If a filter generated a file delivery with a non-absolute name (possible if
5289 no home directory exists for the router), the forbid_file option was not
5292 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5293 the existing equality matching.
5295 52. Exim was using ints instead of ino_t variables in some places where it was
5296 dealing with inode numbers.
5298 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5299 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5300 called TMPDIR, and if it finds it is different, it changes its value.
5302 54. The smtp_printf() function is now made available to local_scan() so
5303 additional output lines can be written before returning. There is also an
5304 smtp_fflush() function to enable the detection of a dropped connection.
5305 The variables smtp_input and smtp_batched_input are exported to
5308 55. Changed the default runtime configuration: the message "Unknown user"
5309 has been removed from the ACL, and instead placed on the localuser router,
5310 using the cannot_route_message feature. This means that any verification
5311 failures that generate their own messages won't get overridden. Similarly,
5312 the "Unrouteable address" message that was in the ACL for unverifiable
5313 relay addresses has also been removed.
5315 56. Added hosts_avoid_esmtp to the smtp transport.
5317 57. The exicyclog script was not checking for the esoteric option
5318 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5319 will work only if exicyclog is run under the appropriate euid.
5321 58. Following a discussion on the list, the rules by which Exim recognises line
5322 endings on incoming messages have been changed. The -dropcr and drop_cr
5323 options are now no-ops, retained only for backwards compatibility. The
5324 following line terminators are recognized: LF CRLF CR. However, special
5325 processing applies to CR:
5327 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5328 nor a local message in the state where . is a terminator.
5330 (ii) If a bare CR is encountered in a header line, an extra space is added
5331 after the line terminator so as not to end the header. The reasoning
5332 behind this is that bare CRs in header lines are most likely either
5333 to be mistakes, or people trying to play silly games.
5335 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5336 was being incorrectly given as 18 bytes larger than it should have been.
5337 This is a VOB (very old bug).
5339 60. This may never have affected anything current, but just in case it has:
5340 When the local host is found other than at the start of a list of hosts,
5341 the local host, those with the same MX, and any that follow, are discarded.
5342 When the list in question was part of a longer list of hosts, the following
5343 hosts (not currently being processed) were also being discarded. This no
5344 longer happens. I'm not sure if this situation could ever has previously
5347 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5350 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5351 header. This is something that is recommended in a new Internet Draft, and
5352 is something that is documented as being done by Sendmail. There are two
5353 possible values. For messages generated by the autoreply transport, Exim
5356 Auto-Submitted: auto-replied
5358 whereas for all other generated messages (e.g. bounces) it adds
5360 Auto-Submitted: auto-generated
5362 63. The "personal" condition in filters now includes a test for the
5363 Auto-Submitted: header. If it contains the string "auto-" the message it
5364 not considered personal.
5366 64. Added rcpt_include_affixes as a generic transport option.
5368 65. Added queue_only_override (default true).
5370 66. Added the syslog_duplication option.
5372 67. If what should have been the first header line of a message consisted of
5373 a space followed by a colon, Exim was mis-interpreting it as a header line.
5374 It isn't of course - it is syntactically invalid and should therefore be
5375 treated as the start of the message body. The misbehaviour could have
5376 caused a number of strange effects, including loss of data in subsequent
5377 header lines, and spool format errors.
5379 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5380 client host had authenticated. This control can now be exercised by an ACL
5381 for more flexibility.
5383 69. By default, callouts do not happen when testing with -bh. There is now a
5384 variant, -bhc, which does actually run the callout code, including
5385 consulting and updating the callout cache.
5387 70. Added support for saslauthd authentication, courtesy of Alexander
5390 71. If statvfs() failed on the spool or log directories while checking their
5391 size for availability, Exim confusingly gave the error "space shortage".
5392 Furthermore, in debugging mode it crashed with a floating point exception.
5393 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5394 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5395 serious problem, Exim now writes to the main and panic logs when this
5396 happens, with details of the failure. It then refuses to accept the
5397 incoming message, giving the message "spool directory problem" or "log
5398 directory problem" with a 421 code for SMTP messages.
5400 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5401 0, 1, and 2 exist, because some OS complain for execs without them (see
5402 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5403 descriptors. However, the code omitted to check that the open succeeded,
5404 causing mysterious errors if for some reason the permissions on /dev/null
5405 got screwed. Now Exim writes a message to the main and panic logs, and
5406 bombs out if it can't open /dev/null.
5408 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5409 interact so that it is all more flexible. It is supposed to remain
5410 backwards compatible. Also added extra_local_interfaces.
5412 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5413 to bomb out with an assertion failure - to the client this appears as a
5414 connection drop. This problem occurs in the part of the code that was taken
5415 from the Samba project. Fortunately, the assertion is in a very simple
5416 function, so I have fixed this by reproducing the function inline in the
5417 one place where it is called, and arranging for authentication to fail
5418 instead of killing the process with assert().
5420 75. The SPA client code was not working when the server requested OEM rather
5421 than Unicode encoding.
5423 76. Added code to make require_files with a specific uid setting more usable in
5424 the case where statting the file as root fails - usually a non-root-mounted
5425 NFS file system. When this happens and the failure is EACCES, Exim now
5426 forks a subprocess and does the per-uid checking as the relevant uid.
5428 77. Added process_log_path.
5430 78. If log_file_path was not explicitly set, a setting of check_log_space or
5431 check_log_inodes was ignored.
5433 79. If a space check for the spool or log partitions fails, the incident is now
5434 logged. Of course, in the latter case the data may get lost...
5436 80. Added the %p formatting code to string_format() so that it can be used to
5437 print addresses in debug_print(). Adjusted all the address printing in the
5438 debugging in store.c to use %p rather than %d.
5440 81. There was a concern that a line of code in smtp_in.c could overflow a
5441 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5442 initially expressed, the concern was not well-founded, because trailing
5443 spaces are removed early. However, if the trailing spaces were followed by
5444 a NULL, they did not get removed, so the overflow was possible. Two fixes
5447 (a) I re-wrote the offending code in a cleaner fashion.
5448 (b) If an incoming SMTP command contains a NULL character, it is rejected
5451 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5452 runs initgroups(), so that if the Exim user is in any additional groups,
5453 they will be used during message reception.
5459 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob